The basic premise of 3e encounter design does not work

[Just to Browse]

So in 3rd edition we have a two premises for fights:
A level-appropriate encounter will consume 25% of party resources.
A level-appropriate encounter (for a party of 4) is equal in resources and resource-removing capability to one party member.

The original version of this is spoilered below for your viewing pleasure. It makes less sense, but also uses less math (perhaps those two are correlated :smallfrown:).

The Preface (almost as long as the rest of the post)
The assumptions you see above this paragraph are two assumptions set out by the DMG. Both have problems in their own right--the action economy is brought up a lot for (2), and sustaining through fights each day has been brought up in this very thread for (1). But those problems aren't what I'm here to discuss.

What I'm here to talk about is how these options cannot be used as rules for modeling encounters--one has to fail in order for the other to work. The following splerg is a demonstration of combat (extremely simplified) in which we strictly following rule (2), and then find that while following rule (2) it's impossible to enforce rule (1). It should be noted that with increased complexity comes increased deviation, and that the designers (who may not have been aware of this problem) make monsters to be more appropriate foes. So while it is totally possible in D&D to have combats where the opponents are CR = party level, the basics behind monster and encounter design have to be severely bent to accommodate such events.

Some other assumptions, which I sort of implicitly thought of, but didn't write down until problems started cropping up are as follows. These are for an ideal, balanced D&D game.
All creatures and encounters with the same level and CR need to have similar total resources, and similar ability to remove resources--an example of this is the wizard who has fewer HP and lower weapon damage than the warrior, but can cast a limited number of defensive spells and has magical attacks.
Options that are available to one side of a fight can also be made available to another side. If some creatures have strong crowd control, then there must be some other creature(s) of equal level with stronger single-target control. If enemies have some way of using regeneration, that must also be a player option. A party might choose not to take such an option, but it has to be at least available. This fits in with Premise (2).
1 CR X creature == 2 CR X - 2 creatures == 4 CR X - 4 creatures, etc. This is also in the DMG. A creature with half the strength (CR X - 2) has half the resources and half the resource-removal capability of something that is CR X. This is assumed so that encounters with multiple creatures are balanced (i.e. minions don't outdamage their single-creature counterparts).

The Example
Assume the average fight has a length L (measured in rounds). The average character contributes R 'resource depletion' per round. If this term sounds dumb to you, that's because it kind of is. There are a lot of things you can do in a fight--buff your party members to live longer, buff them to hit harder, cut down on enemy offenses, change the battlefield to make achieving objectives easier, etc.--but the reason I lump all of that together is because we're assuming each party member is balanced to one another (yes this isn't true in 3.x, everyone and their dog knows that already. Bear with me.), so each party member should be contributing equally on average, which is why we can simplify all useful actions down to a single quantity.

With this idea in mind, a party of P adventurers thus depletes P*R*L resources from their opponent over the fight, so the opponent must have that quantity of resources to deplete (TR). By assumption #2 (which is part of the DMG and MM, can't deny this), this means each player also has TR = P*R*L resources. Again using assumption #2, we know that the opponent in each fight must be capable (on average) of depleting R*L resources from the party as whole. That's where we see the problem...

There are P players in this party. If each of them has TR = P*R*L resources, that means the total party resources is P*TR = (P^2)*R*L, so an opponent that removes only R*L resources depletes 1/(P^2) of player resources. For a standard party of 4, this is 1/16 of the desired resources, which directly clashes with assumption #1. An equal-level enemy should on average be taking out a quarter of party resources, yet we can see that it would actually be taking out a quarter of that.

What this means is that even if there was excellent balance between the classes, and perfect monster-to-PC translation, combat wouldn't be functional because the players would barely lose any resources. I think this core problem is at least partly to blame for why the monsters have such out-of-whack CRs, or are CR'd 'appropriately' yet are totally unsuited as PCs of similar levels. The writers were trying to balance two unbalance-able concepts.

Thoughts?

EDIT: This is probably also why a lot of people feel like players breeze through their boss fights, despite ramping up numbers. In this example, you'd want to quadruple enemy resources to make it a standard fight for the PCs (though it would be a serious grindfest). A final boss made to take down all of the party would have quadruple resources and resource depletion*.

I'll do this by example because the math is easier. Assume our fight must last 4 rounds, and the average character contributes 1 damage per round (you might not be specifically dealing 1 damage per round, perhaps you buff the fighter to do +1 each round, or you charge-up for 2 damage every 2 rounds). This means that the average total damage per round is 16.

Since opposition to the party is supposed to be equal to the party, this means the enemy has 16 HP and also deals 1 damage per round, and it means each member of the party has ~16 HP.

However, if the opposition is going to deal only 1 damage per round, then it will take out 4 HP total over the course of the fight. Since each character has 16 HP, this is actually 1/16 of party resources, not 1/4.

A monster 2 CR higher is supposed to be twice as powerful (either twice as much HP or twice damage), which means it theoretically takes a creature 4 CR higher to actually challenge a party.

I think this is a big reason why monsters are so hard to translate from Team Monster to Team PC--as designers continued to write things, they realized that a CR 12 encounter couldn't be both a useful PC and a useful enemy, so they designed the monsters to be extra powerful.

If one of these assumptions is going to hold, the other has to fall. My question (other than "How did I mess up here?") is which of these assumptions do we have to scrap to solve this problem?

*At the upper and lower limits of the spectrum, this actually breaks down. A monster with 4R and 4TR would be a normal challenge for a party of level + 4, but would dispatch a low-level party without dying at all, because (on average) one player would fall per round, so total depletion would be something like 4(PR) + 3(PR) + 2(PR) + 1(PR), or 6PR less than necessary to take the guy out. An inverse case occurs for really weak mooks, in that players waste damage taking out targets with a TR < R*L, so they lose more resources in a mook-heavy fight without cleave effects.

And as a reminder: This is not about specific D&D mechanics or party sizes. This is an imaginary version of D&D where the classes are balanced and the challenges follow the assumptions listed above. Waddacku says it better than me:


The entirety of the actual details in monster design, spells, abilities, and all that stuff is completely irrelevant, because the issue at hand is the conflicting premises the CR system makes. The reality of how the game works doesn't matter whatsoever for it, because it has nothing to do with the assumptions made when setting up the CR system in the first place. The DMG does state that a 4v1 fight doesn't work as their CR rules imply it should, but the fact is the CR rules doesn't work in general either.

Consider this: CR is (aside from the level scale) system independent. You you just rip it straight out and implant it somewhere else and the only thing you need do is replace level with whatever advancement the other system is using.
However, even removed from any game system, the math of the CR system fails to actually meet its own premises.

[Flickerdart]

The basic premise of 3.5 combat doesn't hold if you're going to replace the math, yes.

[Tvtyrant]

Doesn't your math assume the enemy can only hurt one person per round, and that damage is the overriding consideration for such? Enemies have to effect multiple opponents a round to be effective, while the party frequently only fights a single foe.

This means that any math is going to be off, because they are aimed in different directions. A web spell at level 3 is wonderful for a bad guy because it can net a significant part of the party, while it is merely good for a player because so many of the hardest opponents are going to be alone and thus the massive area is wasted. Fireball is decidedly better for enemies than for the party unless the party is battling mooks constantly, because the enemies job is to hurt multiple party members a round.

[Segev]

If the party consists solely of fighters, you're correct. Fighters do their 1 hp of damage directly with no expenditure of resources.

But that's not the only expenditure of resources. The mage buffs the fighter to do 2 damage per round. The cleric heals the fighter after the fight is over. These spells are amongst the primary resources actually expended.

It's also an average of 1/4 party resources per CR-appropriate encounter. So one encounter will go as you described, while another will have some "luck" one way or the other.

In order for the lone CR-equal foe to deplete more resources, we merely need to give him expendable resources of his own. He, being outmatched 4:1, will use as many as he can to stave off death, and so will do more than the "expected" 1 hp/turn in effective contribution. This will cause more hp to be depleted from the party-of-4, and will further result in a bit more need to step up the daily consumables to mitigate, prevent, or to do more damage faster.

Still, the party-of-4 will only use a quarter of their resources because they don't NEED to go all-out. The party-of-1 needs to, and still likely will perish. (In fact, with no other reason to stay in the fight, I'd probably have such an outmatched foe try to surrender or retreat. There are plenty of ways to orchestrate otherwise, and the PCs of course can pursue, but the resources are going to be blown trying to stay ALIVE as long as possible, so more time exists to bring more to bear, and hopefully survive.)

It's one reason the idea that monsters won't blow through their full array of powers in a life-or-death fight is a bit specious: if the PCs are facing a foe that is as powerful as 4 of each of them, you can bet they're not conserving resources for later battles. They'll blow them now just to live so they can hole up and rest and recover.

[Just to Browse]

I seem to have not communicated my point. The math is identical to the desired math for 3.x, but simplified. I'm not even sure what "replaced" means in this context.

The idea is that everyone is different, but should on average contribute equally to combat. Because this is a simulation, contribution is measured in "damage per round". Yes if the mage buffs the fighter's damage output, then the fighter deals more damage, but the mage also deals less in exchange and in order for the system to be balanced that output has to be equal on average. If it's not then somebody is OP.

The math does assume the enemy can only make 1 attack per round, because that's what players can do. 1 enemy == 1 player (e.g. a fighter 1 is a CR 1 encounter). If the enemy can make multiple attacks, or gets extra resources, or deals more damage, or gets a free casting of web then that enemy is no longer a CR 1 encounter and thus should not be depleting only 25% of player resources.

[eggynack]

Why do you assume that everyone will necessarily get an attack off? Getting four guys to hit one guy is far more difficult than getting one guy to hit one out of four guys.

Additionally, what if our enemy is casting grease, color spray, or sleep instead of web? Now, he's a perfectly allowable CR 1 opponent, and he's doing more to the party than an equivalent action by the party wizard would do to him. Maybe the enemy has combat reflexes and improved trip, and maybe that gives him an advantage over the enemy fighter doing the exact same thing, because he actually does get to use more than one AoO in the same turn. Maybe, the tactics that one guy uses against four are different than the tactics that four guys use against one, even if they're of the same level, and even if they're at the same balance level.

Finally, the game system itself doesn't seem to bear out your reasoning much. If you set a low tier character, and a low tier monster (here defined as a monster using a lot of spells. Seriously, you have perfect balance as a premise, and that makes actually looking at the game through this lens tricky as hell) against each other, a lot of the time the character will have better than even odds of winning the fight. In other words, even if the CR system may be off sometimes, monsters aren't generally all that overpowered. If you were correct, all or most monsters would be overpowered, and I don't think they are.

[Firechanter]

In theory, yes, the Fighter's resource is hit points, the Cleric's and Wizard's are spell slots, the Rogue's also hit points I guess, and everyone can expend consumables (like x/day items etc).

In practice, all CRs are not created equal. A CR9 Druid is going to give you a much harder time than a CR9 Fighter. Monsters can be anywhere. I've actually toyed with the idea of giving certain classes a CR-Adjustment based on their tier. Off the top of my head, maybe something like "Character level + (4 - Tier)". But of course the same should also apply to party level calculation.

Thing is - we recently touched this in another thread - the CR system is calibrated to a game where everyone plays like a T4 or T5 class. There's a Fighter who hits things with a stick, a Rogue who spends several rounds just moving into flanking position, a Wizard who blasts until he runs out of spells, and a Healbot Cleric who swings a mace für 1d8 damage and spends his spell slots on Cure.

THIS is the actual premise of 3.5 combat. It is like this because the designers need to cater to casual gamers who are not at all into optimization, can be barely bothered to read the rules, and pick feats purely because the name sounds cool. Even the most un-optimized party is supposed to beat published adventures.
And as long as you stick to this premise, it more or less works. But more invested players behave entirely differently. For one thing, their Clerics don't waste spells on Cure and their Wizards do more efficient things than blasting. And their Melees first-strike so hard the enemy doesn't even get to hit back.
In short, they are much more efficient, and thus expend way fewer resources.

[Yahzi]

You've rediscovered the basic flaw in their design: the action economy. That was not a concept they had while they were making the game; it didn't occur to them that 4 actions were not just 4 times better than 1 action, but much, much better.

[zlefin]

that's not the basic premise of 3.5 combat; that's the premise of the CR system (which is known to have numerous issues already).

[Firechanter]

You've rediscovered the basic flaw in their design: the action economy.

I think I remember reading in an old WotC article, "Acting first is good, acting last is better".

[Brookshw]

Why yes, CRs are not representative in many cases of reasonable challenges, especially with the great many splat books out there. Nothing new here in a horribly broken system that I still love =D

[Killer Angel]

...and when the ubercharger wins initiative and one-shot the enemy, you will spend even less resources.
The theory of the CR system is only a guideline.

[Maginomicon]

I'm reminded of the tale of the Leaping Wizards (http://www.youtube.com/watch?v=9G5PjlCMlGw).

To summarize: Three 1-spell-slot nerds (GM) vs six football players (the party, half of which are clerics). A nerd wins initiative, casts sleep, knocks out 2/3 of the party.

Where's your math now? :smallbiggrin:

[Talderas]

So in 3rd edition we have a two premises for fights:
A level-appropriate encounter will consume 25% of party resources.
A level-appropriate encounter (for a party of 4) is one party member.

A level appropriate encounter is an an encounter with a challenge rating equal to the average party level. This can be a single opponent or it can be multiple opponents. Your entire assumption is based off an erroneous starting point.

[Psyren]

I highly recommend everyone drop what they're doing and read this guide. (https://docs.google.com/document/d/1nx-o8VAjhUwh3nnfzDQT-JA5eFLnN_BZJiBitGjBMDg/edit) It covers a lot of the myths and misconceptions about CR and the undervaluing of the action economy being discussed here, and provides great solutions.

[Segev]

Got a non-google docs version? Work blocks that.

[Psyren]

Got a non-google docs version? Work blocks that.

It's not mine - it was written by Alex Augunas (one of the guys behind the PF Binder conversion).

Maybe look at it on your phone until you can get home? It's really good. (Also, OotS jokes!)

[Tyndmyr]

I seem to have not communicated my point. The math is identical to the desired math for 3.x, but simplified. I'm not even sure what "replaced" means in this context.

The idea is that everyone is different, but should on average contribute equally to combat. Because this is a simulation, contribution is measured in "damage per round". Yes if the mage buffs the fighter's damage output, then the fighter deals more damage, but the mage also deals less in exchange and in order for the system to be balanced that output has to be equal on average. If it's not then somebody is OP.

The math does assume the enemy can only make 1 attack per round, because that's what players can do. 1 enemy == 1 player (e.g. a fighter 1 is a CR 1 encounter). If the enemy can make multiple attacks, or gets extra resources, or deals more damage, or gets a free casting of web then that enemy is no longer a CR 1 encounter and thus should not be depleting only 25% of player resources.

Your assumption that a CR 1 is exactly equal to a level 1 PC is a bit off.

For instance, two orcs is a CR 1 encounter, and your math is...wildly off for representing what that would be like. This isn't really an exotic encounter, either.

[Talderas]

Got a non-google docs version? Work blocks that.

My post earlier in the thread basically points out the topic of the document.


A level appropriate encounter is an an encounter with a challenge rating equal to the average party level. This can be a single opponent or it can be multiple opponents. Your entire assumption is based off an erroneous starting point.

A single CR appropriate enemy will generally be underwhelming due to action economy. More challenging encountered tend to be ones where you have a higher number of lower CR enemies which brings the action economy back to parity.

[Fax Celestis]

Because this is a simulation, contribution is measured in "damage per round".

There are far more things you can do than DPS (DPR?). Most of them don't even affect other people's DPS.

[Chronos]

Quoth Tyndmyr:

Your assumption that a CR 1 is exactly equal to a level 1 PC is a bit off.
That's not his assumption; it's one of the game's assumptions. An enemy of a standard race with levels in PC classes has a CR equal to its character level.

And to put the OP's argument a little more directly: Suppose your party of four were to face their exact mirror image party. You would expect them to then have a 50-50 chance of winning, which means that they will, on average, expend 100% of their resources on that encounter. But facing those enemies one at a time, the party is almost certain of victory, because fighting enemies one at a time is a lot easier than fighting them all at once. So the party should be expected to expend a lot less than 25% of their resources on each of the four separated equal-CR encounters.

But that's not really a problem with the CR system; it's just a problem with the assumption that you spend 25% of your resources on each encounter. After about four encounters, it's probably a good idea to rest and replenish, but that doesn't mean that your resources are all at zero. A party that's down to zero resources is literally dead. There's a world of difference between "down to zero resources" and "down on resources low enough that you want to rest".

[Talderas]

Suppose your party of four were to face their exact mirror image party.

The CR of such an encounter is APL + 4.

Edit: So a party of 4 level 6 characters fighting themselves would be facing a CR10 encounter. I would not expect them to win by only using 25% of their resources.

[Cheiromancer]

@Just to Browse: The circumstances you describe are fairly unusual. But when they hold, each doubling of the number of creatures will add +4 to the EL instead of the standard +2. Conversely, halving the number of opponents will reduce the EL by 4, if this allows the larger group to gang up on the smaller. An Encounter Level (EL) of N is a challenge which uses up 25% of a level N party's resources. Every +2 in the EL doubles the resources used; every -2 halves the resources used.

The argument is just what you gave in your post: If there is a balanced party of four level N characters (all equal in power, as in your example), then a single level N character is an EL N -4 encounter. You can calculate this by observing that 6.25% of the party's resources are expended defeating. Using the same mathematical procedure, two level N characters would be an EL N encounter, and so we determine that doubling numbers adds +4 to EL.

This applies when encounters unfold in a way that is similar to how you describe. There are some exceptions: monsters whose abilities scale with the number of opponents (e.g. area attacks); situations where the action advantage is reduced (e.g. the larger group can't gang up on the smaller group due to fighting on a spiral staircase or in a twisting tunnel). Some monsters may also be "puzzle monsters" and not all characters are effective against them.

These special considerations also hold in reverse. 16 goblins ought to use up 16 times more resources than 4 goblins (according to your method), but a single fireball might be enough to finish them all in one round; a 16 goblin encounter would then use up exactly the same amount of resources as 4 goblins would. Or maybe only 4 can engage the party at any time, or they arrive in waves or something; then quadrupling the number only uses up 4 times the resources. So adding +4 to the EL per each doubling will usually be too much. Adding +2 to the EL for doubling the numbers is a decent compromise.

Assuming that all party members can contribute equally to the defeat of the monster (as in your example), then if a single CR N monster is an adequate challenge for a party with N members, the monster has been mislabeled. Treat it as really CR N+4.

[Chronos]

Quoth Talderas:

The CR of such an encounter is APL + 4.

Edit: So a party of 4 level 6 characters fighting themselves would be facing a CR10 encounter. I would not expect them to win by only using 25% of their resources.
Right, I wouldn't expect that, either. In fact, I'd expect them to use about 100% of their resources, just like I said.


Quoth Cheiromancer:

Assuming that all party members can contribute equally to the defeat of the monster (as in your example), then if a single CR N monster is an adequate challenge for a party with N members, the monster has been mislabeled.
No, that's exactly what CR N is supposed to mean. The catch is that "adequate challenge" doesn't mean "must be at the top of your game and probably somewhat lucky or you'll get wiped out". An adequate challenge is something that you should be able to face multiple times a day, repeatedly, without anyone dying.

[prufock]

I think we all knew already that CR is not solidly rated. There's a difference between facing a level 7 druid and a level 7 fighter, for instance, but the CR is the same.

[eggynack]

I think we all knew already that CR is not solidly rated. There's a difference between facing a level 7 druid and a level 7 fighter, for instance, but the CR is the same.
More tellingly, there's a difference between facing a level one druid, with a riding dog animal companion, and facing a riding dog on its own, but the CR is the same. Hell, the druid's riding dog is generally independently better at combat than the lone riding dog, because the druid's riding dog will often have barding of some kind.

[Chronos]

I'd say that's more a problem with druids than with the CR system, though.

[eggynack]

I'd say that's more a problem with druids than with the CR system, though.
Yeah, probably. Still, it's an oddity of some kind. There're a lot of discrepancies between the power levels of different creatures of the same CR, just as there're a lot of discrepancies between the power levels of various PC's of the same level.

[Person_Man]

3.5 combat works fine, it can just be very fiddly, and the mechanisms for balancing them are not always self evident.

Earlier editions of D&D had a much stronger emphasis on exploration (the "hidden map" game) and resource management (limited hit points, spells, gp, and xp, which could be spent on different things). There was an assumption that you wouldn't or couldn't just rest to restore our spells; there was a much greater emphasis on exploration, combat was generally something you avoided when possible, xp was granted for gp instead of killing things, random encounters/wandering monsters were common (especially when you tried to make camp in or near a dungeon), many DMs would punish players for trying to rest too often (and some modules explicitly encouraged them to do so), and workarounds like Rope Trick were generally rarer or non-existent.

Part of this carried over into the initial design philosophy of 3.0. But it quickly broke down, as they added more and more ways to avoid the necessity of resource management. Designers realized that varying levels of resource management wasn't fun for many players and caused tremendous imbalance within game groups, which led directly to encounter based abilities, and eventually 4E.

But this doesn't negate the fact that an experienced DM can make combat "work" for pretty much any game group. It just takes discernment, and trial and error. There is not a single formula or rule as to how all 3.5 combats should work. And that's ok. Different people like different things out of combat, and for some of us, it's nice having a system that accommodates those different options, even if it does take more effort for them to work right.

[Flickerdart]

I seem to have not communicated my point. The math is identical to the desired math for 3.x, but simplified. I'm not even sure what "replaced" means in this context.
The party wizard casts Grease on the orc. How much damage did he deal?

[eggynack]

The party wizard casts Grease on the orc. How much damage did he deal?
It's even difficult to work out for more cut and dry cases. There's an assertion in the thesis that a buff spell is merely handing over his point of damage for the turn over to the party fighter. Either that, or the assertion is that that's how it should be. However, buff spells don't work like that at all. They stretch across several rounds, often covering an entire encounter. Thus, unless you can be assured of the precise length of an encounter, you can't perfectly balance out the damage between a fighter and a wizard.

For example, maybe the fighter is dealing one damage per turn, and the wizard ups it to 1.4 damage per turn, so the wizard is dealing more damage if the encounter lasts more than three turns, and less at one or two turns. So, do you set the power level of wizard buffs to increasing fighter damage to 1.4, or do you increase it assuming short encounter lengths, or do you decrease it, assuming long encounter lengths? Are we assuming the fighter is optimal or suboptimal when we make the 1.4 assessment? A figure like that is clearly just expressing a ratio between the fighter's old and new damage, so any change in fighter damage parameters will change the value of a buff, both due to synergies with the fighter's attacks, and due to the relative damage level compared to old damage.

Thus, even in this seemingly simple case, there're a ridiculous amount of issues. It's nice to say that we can just set things equal, and control for variables, and form a hypothesis, but the way you've described the game isn't the way the game really works at all. It'd be like a physicist saying that real life walking doesn't work, because if you set friction equal to zero then you can't get any traction. Ignoring these factors, as convenient as it might be, is just that: a convenience.

[Spuddles]

Why yes, CRs are not representative in many cases of reasonable challenges, especially with the great many splat books out there. Nothing new here in a horribly broken system that I still love =D

I find that with 3.5, the biggest imbalances are in core. There are tons of low level monsters with save or dies and permanent ability drain.

The allip, for instance. How the hell are you supposed to deal with that at level 2?


The party wizard casts Grease on the orc. How much damage did he deal?

8 damage per round.

[Psyren]

The allip, for instance. How the hell are you supposed to deal with that at level 2?

Magic Weapon oils and a trip to the local cleric afterward :smalltongue:

But yeah, Allips are pretty damn under-CR'ed.

[Just to Browse]

So my vocabulary choices appear to confuse some people, but I'm glad others are getting it. I'll address this specific post:


The party wizard casts Grease on the orc. How much damage did he deal?

Take the chance of grease successfully inflict [condition], then determine how much less damage the enemy deals based [condition] (factoring in duration, actions taken to remove it, and lack of mobility), and now the enemy inflicts that much less damage and thus removes less of the party's resources otherwise. That's one bit, but it's not damage.

To find damage inflicted, you take the increased chance to succeed (or perhaps increased damage inflicted) due to the [condition] inflicted by grease, and now take the altered chance that the players will inflict the condition (taking into account mobility, etc.) and the potentially altered amount of damage, and that is the increased damage. Yes there are ways to make that system incredibly complicated, and yes it requires predictive knowledge of all possible outcomes, but we're simplifying things.

My point is that, even if the CR-judging system worked, and even if all the classes were balanced to each other on average (yes classes are broken, we all know this), then the 3.5 combat idea would be a failure. "Level-appropriate" encounters do not do what they are supposed to do, and I believe that this inherent problem is a root cause of inter-class balance screwiness as well as monster screwiness.

I'm adding a new-vocab, more-mathy explanation in the post so people stop getting sidetracked by terms like "damage" and "HP".

[eggynack]

I understood what you were saying. It's just that the game doesn't work that way. You're making grease and stabbing each a single unit on the scale of enemy deadness. When you reach 16 units, the enemy dies. It'd be nice if the model were good for anything but a model, but it's really not. Grease isn't just a unit equivalent action. In fact, fireball isn't even a unit equivalent action. If the party wizard casts a fireball on the single enemy, presumably that's a single unit. However, if the enemy wizard casts fireball on the whole party, that is therefore four units of damage. By the end of the combat, after four fireballs, the party's resources are depleted by 25%. Fwoom, the system works perfectly. Except it doesn't. My model isn't really more or less accurate than yours, because we've simplified out all of the nuances of the system. The nuances correct for a lot of the issues you're talking about, and they don't correct for all of the issues.The game is different from the image you've constructed, and that's a problem.

[Just to Browse]

Yes, you can build complexity into a system, but the fact that there is complexity doesn't magically make the basic premise of the CR system functional again--it just obfuscates that system's failures. A grease spell provides some benefits balanced to make the wizard useful against groups of enemies of a certain size--hitting more is better than a regular attack, but sometimes you hit fewer and it's worse. The same thing goes for fireball--it's like a regular attack, but you reduce the damage based on the number of intended targets. Sometimes all the targets will get fried but other times they will make their reflex save or you'll botch the casting or they have a contingent 15' dash.

The point of removing complexity is not to presuppose some theoretical boring version of D&D, it's to take a normalized average of resource depletion over combat.

[Fax Celestis]

The point of removing complexity is not to presuppose some theoretical boring version of D&D, it's to take a normalized average of resource depletion over combat.

That's all well and good, but it doesn't work.

To crib a phrase, D&D combat is about condensing fact from the vapor of nuance. There are so many little fiddly bits (High ground! Flanking! Charge! Bless! Demoralize! Grease!) and different options and functions of the rule system that, frankly, the game as it stands is pretty close to being as boiled down as possible while still having it be both functional and recognizable.

I don't think anyone's going to argue that CR works very well: it doesn't. But sometimes it does, and sometimes we can fudge it a few points in the right direction to make it work.

[OldTrees1]

@Fax Celestis

Stop for a moment and answer this tiny question:
D&D has a premise that a 4 on 1 encounter will deplete 1/4th of the resources of the 4 and 100% of the resources of the 1. If we consider the action economy advantage, does this premise make any sense? When viewed in the abstract Just to Browse does not think this holds.

Remember this is a premise of the combat system and thus is judged prior to complex concrete examples. Such examples are supposed to conform to the premise thus the premise cannot be judged by them.

Addendum: What would have to change for the premise to hold?

[Psyren]

Addendum: What would have to change for the premise to hold?

Did you read the guide I linked?

And as Fax said, there are many other factors (spells, environment, special qualities etc.) that complicate this issue and make it difficult if not impossible to slather on a one-size-fits-all solution to designing perfect encounters.

[Novawurmson]

I highly recommend everyone drop what they're doing and read this guide. (https://docs.google.com/document/d/1nx-o8VAjhUwh3nnfzDQT-JA5eFLnN_BZJiBitGjBMDg/edit) It covers a lot of the myths and misconceptions about CR and the undervaluing of the action economy being discussed here, and provides great solutions.

Oh, man. I wish I had read this guide as a beginning DM years ago. These are all lessons I had to learn through trial and error (and trial and error and trial and error)...

[OldTrees1]

Did you read the guide I linked?

And as Fax said, there are many other factors (spells, environment, special qualities etc.) that complicate this issue and make it difficult if not impossible to slather on a one-size-fits-all solution to designing perfect encounters.

I did (and loved it).

However I am haunted by this final summary:

1. Boss Monsters are not effective encounters because the PCs have them beat in the number of actions that they can perform. This is known as the Action Economy.

How would you adjust the CR rules such that a plain 4v1 encounter has an accurately measured difficulty rather than being easier than advertised due to the action economy? (assuming no complicating factors)

[Talderas]

How would you adjust the CR rules such that a plain 4v1 encounter has an accurately measured difficulty rather than being easier than advertised due to the action economy? (assuming no complicating factors)

Well, for such encounters, it shouldn't be an encounter that consumes 20% of the party's resources. It should actually account for more. So the CR of the encounter should be around 3-4 above the party. The boss itself should be a CR equal to party level and then you populate the encounter with other monsters in order to raise that CR up. This eliminates the action economy isuse.

[OldTrees1]

@Talderas
Answer the question asked next time.

I did not ask how to change a 4v1 encounter so it was not 4v1. I asked how the CR rules should be changed so that they advertise a 4v1 encouter's difficulty accurately rather than what we have currently (4v1 is advertised as a harder encounter than it actually is)

Perhaps consider 4 enemies to be default and +2CR for doubling the number of enemies but -3CR(arbitrary guess) for halving the number of enemies?

[Fax Celestis]

Stop for a moment and answer this tiny question:
D&D has a premise that a 4 on 1 encounter will deplete 1/4th of the resources of the 4 and 100% of the resources of the 1.
False.

D&D has a premise that a party of any given ECL will deplete approximately 25% of its resources against an equally-CR'd encounter.

There are a variety of factors that play into CR, up to and including terrain, monster equipment, monster choice, and number of opponents. Monsters are a factor in determining an encounter's CR. Four CR 3 archers in an archer tower that you have to climb up from the outside could be a CR 5 or 6 encounter, simply because of the terrain advantage the archers have. Six CR 5 monsters (young black dragons, for instance) is a CR 10 encounter. They outnumber the players physically and in actions available. They breathe water, so given the correct terrain there can be terrain advantage.

4v1 is the laziest kind of encounter--and the least likely to actually do its intended purpose of providing challenge for the players.

[Talderas]

@Talderas
Answer the question asked next time.

I did answer the question. You just didn't like the answer.

1vParty encounters do not work. Full stop. Action economy will always favor the party until such a point that the sole creature can wipe the party without breaking a sweat. The only exception to this involves breaking the action economy for the single monster, in which case you've driven the game into rocket tag territory.

[OldTrees1]

@Fax Celestis
D&D assumes that a generic PC has a CR equal to their level (assuming proper gear). D&D has a premise that a party of any given ECL will deplete approximately 25% of its resources against an equally-CR'd encounter. So a 4v1 encounter (without environmental factors) is claimed to deplete 25% of the party's resources. It does not as a result of the action economy. This is a flaw in the guidelines to calculate the CR of an encounter since it breaks down in this (admittedly boring) type of encounter.

How should the CR calculating guidelines be altered so that become more accurate and thus can accurately measure the difficulty in this 4v1(no environmental factors) case?

@Talderas

1vParty encounters do not work. Full stop. Action economy will always favor the party until such a point that the sole creature can wipe the party without breaking a sweat.
False. A 1vParty encounter is still an encounter(it still expends resources) so it must have a CR. The question is what would the accurate CR be and how can we adjust the calculator so that it takes advantage of what we can learn from the action economy imbalance?

[Fax Celestis]

How should the CR calculating guidelines be altered so that become more accurate and thus can accurately measure the difficulty in this 4v1(no environmental factors) case?

The standard encounter should always be at least 4v4.

[Psyren]

How would you adjust the CR rules such that a plain 4v1 encounter has an accurately measured difficulty rather than being easier than advertised due to the action economy? (assuming no complicating factors)

It's very, very difficult to hit the sweet spot required. Either you make the lone monster tough enough to withstand 4:1 actions and therefore risk a TPK, or you make him too weak and the "boss encounter" is a joke. It's possible to hit the golden mean here but it's extremely difficult.

Which is why the guide says minions are key - they give you a lot more leeway to tweak the encounter. And really, it just makes sense for them to be there - most boss encounters involve creatures so powerful that lesser beings would venerate them or be enslaved by them, and/or being so evil that their vileness can warp reality/thin the veil around them. So even if you have, say, a lone ghost or lich or fiend or even chromatic dragon as the boss fight, it's not hard to imagine their supreme "wrongness" resulting in the creation of nearby oozes/elementals/swarms/undead, perhaps with some templates slapped on for good measure to explain their spontaneous generation.

[OldTrees1]

The standard encounter should always be at least 4v4.

Do you really want to use this distraction just to avoid finding an answer? I expected that an accurate theoretical CR could still be found even if it is not an engaging encounter design. We can learn from such improvements to the CR calculation. If you do not want to help please do not obstruct.

@Psyren
Thank you for your reply. It does not give a definitive answer but it does address the difficult question.

Rather than a boss encounter, what CR relative to the party would you give for a 4v1 or 4v2 against a rival PC or PCs? (CR=Party Level-3,-1)?

[Talderas]

False. A 1vParty encounter is still an encounter(it still expends resources) so it must have a CR. The question is what would the accurate CR be and how can we adjust the calculator so that it takes advantage of what we can learn from the action economy imbalance?

What you're not getting is that the answer is nothing will fix the problem that you think can be fixed. The nature of the game and the nature of action economy means that in any Xv1 encounter the side with numerical superiority will win without much difficulty until such time that the sole opponent is capable of wiping the floor with the party without breaking a sweat. There's no gentle curve. It's an abrupt cliff. This is true not only in D&D but basically any JRPG has the same issue. Single bosses either wipe you out or you wipe them out even if it's a slow grind.

[Fax Celestis]

Do you really want to use this distraction just to avoid finding an answer? I expected that an accurate theoretical CR could still be found even if it is not an engaging encounter design. We can learn from such improvements to the CR calculation. If you do not want to help please do not obstruct.

I am helping. Standard encounters should be 4v4. There is no other way to fairly adjudicate actions and ability power without threatening TPK or party steamroll swinginess.

While a CR listed is theoretically for a single monster, in practice, it does not function that way.

[OldTrees1]

@Talderas and Fax Celestis

Then the accurate CR adjustment should show that. Either that or the combat premise has a serious hole.

[Talderas]

@Talderas and Fax Celestis

Then the accurate CR adjustment should show that. Either that or the combat premise has a serious hole.

The combat premise of D&D is that an encounter of a CR equal to the average party level should consume roughly 25% of the party's resources. The premise itself is not wrong and in fact the DMG even admits that single monster encounters tend to be swingy and will result in a TPK or the party just runs over the monster. The DMG basically endorses that random encounters should never be single monster and that if you are going to go with a single monster you need to be depleting the party's resources, splitting the party, or other things to avoid letting them wield their full might.

[OldTrees1]

@Talderas
Are you taking the position that the DMG calculates the CR accurately but the results are swingy?
Or are you taking the position that the DMG does not calculate the CR accurately and the results are swingy?

It is obvious that 4v1 encounters will not be as precise as other encounters of their CR. But the question is what is the accurate CR from which the swinging occurs?

[Fax Celestis]

@Talderas
Are you taking the position that the DMG calculates the CR accurately but the results are swingy?
Or are you taking the position that the DMG does not calculate the CR accurately and the results are swingy?

It is obvious that 4v1 encounters will not be as precise as other encounters of their CR. But the question is what is the accurate CR from which the swinging occurs?

Neither. He's stating that the DMG states that 4v1, while mathematically and theoretically acceptable for CR, is in practice not a very good idea, and that you should probably use groups of lower-CRed enemies instead.

World's Most Interesting DM: I don't always agree with Talderas, but when I do, it's about CR.

[Talderas]

@Talderas
Are you taking the position that the DMG calculates the CR accurately but the results are swingy?
Or are you taking the position that the DMG does not calculate the CR accurately and the results are swingy?

It is obvious that 4v1 encounters will not be as precise as other encounters of their CR. But the question is what is the accurate CR from which the swinging occurs?

Neither. A 4v1 encounter is either a cakewalk or a TPK. There is potential for something in the middle but that band is so narrow it's not even worth contemplating and the adjustment for each creature is going to be variable so there's no real guidelines on how to do it. The DMG advises against single creature fights. As far as the DMG is concerned, single monster encounters aren't in scope of the challenge rating system as you must utilize functions (such as party splitting) in order to bring them into the scope. They are, consequentally, exceptions.

[Maginomicon]

Neither. He's stating that the DMG states that 4v1, while mathematically and theoretically acceptable for CR, is in practice not a very good idea, and that you should probably use groups of lower-CRed enemies instead.
Interestingly, the Bell Curve Rolls variant explicitly states that if you're encountering four or more of the same creature, you should drop each one's CR by 1.


Many monsters—especially low-CR monsters encountered in groups—rely heavily on a lucky shot to damage PCs. When rolling 3d6, those lucky shots are fewer and farther between. In a fair fight when everyone rolls a 10, the PCs should win almost every time. The bell curve variant adheres more tightly to that average (which is the reason behind the reduction in CR for monsters encountered in groups).

[truncated]

Any time creatures are encountered in groups of four or more, reduce their CR by 1. For example, a single troll is CR 5, and two trolls are CR 5 each (and thus a EL 7 encounter). But four trolls are only CR 4 each (making a EL 8 encounter).

Monsters with fractional CRs move down to the next lowest fraction when encountered in groups of four or more; the goblin (ordinarily CR 1/2) becomes CR 1/3, for example.

[Karnith]

Neither. He's stating that the DMG states that 4v1, while mathematically and theoretically acceptable for CR, is in practice not a very good idea, and that you should probably use groups of lower-CRed enemies instead.
For reference, this is on page 49 of the Dungeon Master's Guide, in the aptly-titled "Single Monster Encounters" section, though other parts of the book deal with the issue as well. And it does say "encounters with single monsters can be very 'swingy,'" verbatim.

[DR27]

This is true not only in D&D but basically any JRPG has the same issue. Single bosses either wipe you out or you wipe them out even if it's a slow grind.This. It's a universal problem for situations where the boss is outnumbered and can't act more often. I really like how the angry DM outlined the problem (http://angrydm.com/2010/04/the-dd-boss-fight-part-1/), even if it is for 4th ed - this still holds true for 3.5:
1. Most Solos Do Not Act Often Enough
2. Solos Lump Most Of Their Actions Together
3. Solo Fights Are Static
4. Solos Are Disproportionately Affected By Conditions


His solution ended up being to homebrew the "video game" three-act boss monster, allowing for a more interesting, varied fight over a series of blog posts. I know that's not what a lot of people want. But that's what you probably have to do when your solo monster is so outgunned on actions and doesn't have a way to split the party and pick them off one by one.

What does this mean? The premise of combat costing 25% of resources breaks down when you open your monster manual and say "This Ettin should be a perfect encounter for my four 6th level PC's." That Ettin gets swarmed, entangled, and then whittled down (if not one-shotted by a charger). However, if your party is ambushed by a hidden CR-appropriate dragon casting a chained dimensional anchor and quickened wall to split the party? Well, then not so much.

It ends up being easier to throw a song of the white raven crusader with minions at them and letting go of the idea that a single enemy should be taking 25% of your party's resources in an encounter.

[OldTrees1]

Since you two are adamant in your advise not to answer the question, I will take heed of your wisdom and leave the question unanswered.

Such a shame though that wisdom was used to block knowledge.

[Fax Celestis]

Since you two are adamant in your advise not to answer the question, I will take heed of your wisdom and leave the question unanswered.

Such a shame though that wisdom was used to block knowledge.

I am not sure what you are talking about: we did answer the question. The game knows that CR doesn't work for single-monster encounters and specifically states "Don't do it" within the bounds of its core material, so your asking "how do we re-calibrate CR to work for single-monster encounters" is...well, not productive. The game doesn't need to work for that level: single-monster encounters don't work, and don't work in pretty much any system. Even video games generally make boss encounters to be two or three monsters instead of just one.

[Talderas]

Such a shame though that wisdom was used to block knowledge.

Please, don't try to take some high road. We're point out the major initial flaws with what you're seeking. That doesn't even start to touch to ancillary flaws that would be caused by making an such changes to the CR system. There's far more to the CR system than just trying to find a challenge appropriate encounter and once you do that you need to start altering other things which are dependent on it, such as treasure tables and other sundry issues.

The fact of the matter is you're seeking a solution for a problem which is recommended to be avoided without seeming to grasp an understand of what causes the problems and how the whole problem leads to a more or less binary outcome. The problem itself is not with the CR system but with the combat system so to actually fix it you need to somehow fix the combat system in a way that limits the effect of the action economy which essentially would require an entire rewriting of combat for 3.5 D&D.

That which you are attempting to seek is also best suited for the homebrew forum and not this one.

[DR27]

Since you two are adamant in your advise not to answer the question, I will take heed of your wisdom and leave the question unanswered.

Such a shame though that wisdom was used to block knowledge.Don't be a troll, they answered you.

[Eldariel]

Best thing to do is to just throw CR down the drain and focus on creating interesting encounters that you balance by numbers rather than by an ephemeral, dysfunctional system. Party power level varies wildly from party to party as does encounter difficulty even with the same party depending on the environment.

CR cannot even begin to hope to account for even half of all the relevant features when a Fighter 6 can either do 1d8+4 damage a turn or threaten to one-shot a PC depending on the build (and it can be either on the PCs' side or against them).


But yeah, CR is borked. It's also unnecessary, at best a guideline for new DMs. Experienced DM has no trouble making beatable but hard, beatable easy and unbeatable encounters at will off the numbers without relying on CR.

[Just to Browse]

To crib a phrase, D&D combat is about condensing fact from the vapor of nuance. There are so many little fiddly bits (High ground! Flanking! Charge! Bless! Demoralize! Grease!) and different options and functions of the rule system that, frankly, the game as it stands is pretty close to being as boiled down as possible while still having it be both functional and recognizable.Like I said in the post you quoted, complexity does not obviate the core premise. Every option the monster has, the player also has, and vice versa. That's why Characters of level X are supposed to (supposed to, this is theory) be a standard challenge.


I don't think anyone's going to argue that CR works very well: it doesn't. But sometimes it does, and sometimes we can fudge it a few points in the right direction to make it work.It's not that CR 'works very well', it's that CR is predicated on 2 beliefs that are not compatible. If (1), then ~(2). If (2), then ~(1). Adding complexity to a system makes it so that edge cases in a simple system become more common, but the average combat fails to deliver on what is promised in 3.5.

Another important thing people seem to be missing: This is not about action economy, this is about total output. A combat with a creature with RD resource depletion per round and RP resources available could just as easily be a 2 creatures with RD/2 and RP/2, or three creatures with RD/3 and RP/3, and the same problem would arise. The problem is that the party's total RD is 4x the standard encounter's RD, and their total RP is 4x the standard encounter's RP.

Also, there are totally ways to do 1v4 fights that aren't cakewalks or roflstomps (just like how it's possible to do Army v 4 fights), it just requires more limited design. That is a very interesting topic in itself, but I'd rather it not go here. If people want to talk about that, we can start another thread. I could talk all day about how to properly design boss monsters for a 1v4 fight.

[Flickerdart]

the average combat fails to deliver on what is promised in 3.5.

The average combat is not identical PCs fighting a fifth identical PC and all of them using straightforward single-target damage or effects that directly affect straightforward single-target damage. We've already gone over the reasons why.

[Just to Browse]

The average combat is not identical PCs fighting a fifth identical PC and all of them using straightforward single-target damage or effects that directly affect straightforward single-target damage. We've already gone over the reasons why.

Yes, and I've already repeated how it's not single-target damage and how the effects are available to both sides and how it doesn't need to be identical because this is an average.

[eggynack]

Yes, and I've already repeated how it's not single-target damage and how the effects are available to both sides and how it doesn't need to be identical because this is an average.
Yes, but your claims partially hinge on the tactics available to either side. Your logic only holds if the actions available to each member of the party are equal in power to the actions available to the enemy. Just setting up your example in a different way, what if every action by the enemy did one damage to each party member, instead of one damage to one party member? For example, a perfectly valid line of attack for the party wizard is shooting a solid fog at the enemy, and that's a theoretically one damage action. An equally valid line of attack for the enemy wizard is shooting a solid fog at the party, and in that case, the damage is equal to the number of enveloped party members. In other words, it's more than one damage.

You say that it's not about single target damage, but the party actually doesn't have any multi-target damage, which is a problem. If the scenario as I presented comes to pass, then the enemy wizard will end up depleting 25% of the party's resources, because his spells have four times the potency. Maybe it'll be less than that by a bit, because he can't hit the whole party all the time, but it'll be higher than how you've presented it.

[Platymus Pus]

I don't agree with 2 due to a single person very well being able to wipe parties out with proper preparation, spells, and an ambush.

1 seems to be an exaggeration.
25% of resources? Just how much of that is easily recoverable?

[Just to Browse]

Yes, but your claims partially hinge on the tactics available to either side. Your logic only holds if the actions available to each member of the party are equal in power to the actions available to the enemy. Just setting up your example in a different way, what if every action by the enemy did one damage to each party member, instead of one damage to one party member? For example, a perfectly valid line of attack for the party wizard is shooting a solid fog at the enemy, and that's a theoretically one damage action. An equally valid line of attack for the enemy wizard is shooting a solid fog at the party, and in that case, the damage is equal to the number of enveloped party members. In other words, it's more than one damage.If the enemy has an attack that depletes resources from 4 people, it will deplete 1/4 the quantity of resources. If someone uses crowd control, then (just like I explained for grease above) then it either wipes out resources by allowing the rest of the party to attack more advantageously, or wipes out resources by allowing the rest of the party to attack more. It's not like solid fog contributes to an ether layer of Combat Things that change combat without interacting with the mechanics--that's a specifically synergistic ability. Your example of a wizard is four times stronger than an equal-CR encounter, and is thus not a valid subject.


You say that it's not about single target damage, but the party actually doesn't have any multi-target damage, which is a problem. If the scenario as I presented comes to pass, then the enemy wizard will end up depleting 25% of the party's resources, because his spells have four times the potency. Maybe it'll be less than that by a bit, because he can't hit the whole party all the time, but it'll be higher than how you've presented it.If his spells have four times the potency, he is either a) overpowered and thus problematic in an entirely different way, or b) he's not an equal-CR encounter.


I don't agree with 2 due to a single person very well being able to wipe parties out with proper preparation, spells, and an ambush.

1 seems to be an exaggeration.
25% of resources? Just how much of that is easily recoverable?

It doesn't matter whether you agree with #2 and under what circumstances you think it applies--that's a statement in the DMG, and it's what 3.x monster design is predicated on.

The 'resources' are your daily expendables, so [Quest Consumables]/[Days per Quest], [HP]/[Fights between full rests] + [HP per day], etc.

[eggynack]

If the enemy has an attack that depletes resources from 4 people, it will deplete 1/4 the quantity of resources. If someone uses crowd control, then (just like I explained for grease above) then it either wipes out resources by allowing the rest of the party to attack more advantageously, or wipes out resources by allowing the rest of the party to attack more. It's not like solid fog contributes to an ether layer of Combat Things that change combat without interacting with the mechanics--that's a specifically synergistic ability. Your example of a wizard is four times stronger than an equal-CR encounter, and is thus not a valid subject.
A solid fog cast by a party member causes one enemy to be trapped. A solid fog cast by an enemy causes four party members to be trapped. It takes the enemy the same amount of resources to leave the fog as it takes each individual party member, so you're depleting four times the resources as an enemy. Also, how is my wizard four times stronger than an equal CR encounter? I'm assuming a seventh level party, and a seventh level wizard. You can do it with color spray instead, if you like. The enemy causes four times the stunned rounds as the party does.


If his spells have four times the potency, he is either a) overpowered and thus problematic in an entirely different way, or b) he's not an equal-CR encounter.
His spells have four times the potency because he's using spells with an area of effect. A fireball cast on four party members deals four times the damage as a fireball cast on one bad guy. The party is at level 5, and the enemy is at level 5. I don't see the issue.

Edit: A solid fog also brings divide and conquer effects into play. If the enemy can limit the number of party members he has to engage at one time, then the party loses their action economy advantage. The party, by contrast, is physically incapable of dividing the enemy in the same manner, because it's just the one guy (unless they're literally dividing him with a sword or something).

Double edit: I just came up with another issue. The party wizard has to reserve his spell slot usage to some extent, because of future requirements. The enemy wizard, by contrast, is under no such constraint. Thus, the enemy is dealing his single resource damage every turn, while the party wizard has to deal his only periodically, or else deal less resource damage with lower level spells.

[Just to Browse]

A solid fog cast by a party member causes one enemy to be trapped. A solid fog cast by an enemy causes four party members to be trapped. It takes the enemy the same amount of resources to leave the fog as it takes each individual party member, so you're depleting four times the resources as an enemy. Also, how is my wizard four times stronger than an equal CR encounter? I'm assuming a seventh level party, and a seventh level wizard. You can do it with color spray instead, if you like. The enemy causes four times the stunned rounds as the party does.Before I write this, let me preface it with the following: my example is not about a level 7 party of wizards with ultimate crowd control. My example is about a normalized party removing resources from an opponent.

With that out of the way, if the opponent has access to solid fog (a multi-target crowd control), then the PCs need access to something with four times as much crowd control, or twice as much crowd control and a comparable amount of damage or buff effect. A party of level 7 wizards against a level 7 enemy can see solid fog just as readily as they could see polymorph or black tentacles. And since every spell is effectively capable of ending the fight on its own, my example still stands.


His spells have four times the potency because he's using spells with an area of effect. A fireball cast on four party members deals four times the damage as a fireball cast on one bad guy. The party is at level 5, and the enemy is at level 5. I don't see the issue.You're assuming the fireball is dealing as much damage as a single-target spell, which makes no sense because then everyone would just use fireball and that makes fireball overpowered. Since fireball is made to hit multiple people, it deals less damage than a single-target spell.


Edit: A solid fog also brings divide and conquer effects into play. If the enemy can limit the number of party members he has to engage at one time, then the party loses their action economy advantage. The party, by contrast, is physically incapable of dividing the enemy in the same manner, because it's just the one guy (unless they're literally dividing him with a sword or something).I... yes... and the players can instead use a comparably useful spell that isn't solid fog on the enemy and achieve their mob tactics. This is the texas sharpshooter fallacy.

[eggynack]

Before I write this, let me preface it with the following: my example is not about a level 7 party of wizards with ultimate crowd control. My example is about a normalized party removing resources from an opponent.

With that out of the way, if the opponent has access to solid fog (a multi-target crowd control), then the PCs need access to something with four times as much crowd control, or twice as much crowd control and a comparable amount of damage or buff effect. A party of level 7 wizards against a level 7 enemy can see solid fog just as readily as they could see polymorph or black tentacles. And since every spell is effectively capable of ending the fight on its own, my example still stands.
What would the PC's have access to with any of those things? Solid fog is a great spell. Black tentacles would actually work for my argument just as well, if not better, because it can damage party members. You actually have to name the spells you're using here. That's why just creating perfectly smooth head models doesn't really work that well. I really don't see how what you've said here proves your argument at all.



You're assuming the fireball is dealing as much damage as a single-target spell, which makes no sense because then everyone would just use fireball and that makes fireball overpowered. Since fireball is made to hit multiple people, it deals less damage than a single-target spell.
What single target third level spell deals four times as much, or even twice as much damage as a fireball? Fireball isn't overpowered, beca us use it just deals damage, but it's pretty good at just dealing damage. It's doubly good against a party because you can be assured that the fireball will be used against four characters. It's a little metagamey, but the spell is iconic enough that it's reasonable that a villain would have it.



I... yes... and the players can instead use a comparably useful spell that isn't solid fog on the enemy and achieve their mob tactics. This is the texas sharpshooter fallacy.
Which ones? Seriously, stop citing hypotheticals at me, and name some actual thing that you're using here. The game we're playing isn't a hypothetical, so featureless resource removal models can't give us a complete picture. They can give us a model, and a model is useful, but it's still just a model.

[Platymus Pus]

It doesn't matter whether you agree with #2 and under what circumstances you think it applies--that's a statement in the DMG, and it's what 3.x monster design is predicated on.


Yea, monster design.
You said another party member.

A level-appropriate encounter (for a party of 4) is one party member.
And I can say this could be true at some instances even if they are all the same level with the numbers being 10 to 1 simply due to optimization along with the fact that the single person doesn't have to hold back in the least.
They could screw the party over in a single surprise round, and then wipe them out in a single move.

For example... You have a surprise round due to being unaware of the single PC(easy enough.); the PC also rolls a 70 in initiative

The PC uses this round to stop time for 4 rounds as an immediate action, you can't stop it.
The PC then uses this time to set down 8 delayed fireballs in those 4 rounds.
They also use those rounds to completely get out of the parties sight using 4 move actions.

The fireballs all go off at once on the party.
160d6 to everyone(960 max dmg 400-500 avrg). oh +8 as well, they are actually delayed iceballs. If anyone has delay they might survive, but they have to make 8 checks to come out unscathed in a single round.
Improved Evasion means half dmg even if failed. (half dmg could mean death here)
(yes there are lots of counters to a meager fireball even if delayed)
That was the surprise round without any of the extra dmg stuff thrown in that could make it even worse for the party aside from simple spell prep aka buffs.

Now the real round happens, guess who most likely has highest initiative due to prep and having an already high ini.
It happens again, "time" is stopped as an immediate action for 4 rounds and there is nothing you can do about it, no saves or anything.
The pc gets away after killing a great deal of party members or flat out killing them all.

A 1 versus many situation can be bad.

[TuggyNE]

Can we drop the 1vParty comparisons, please? Those are relatively useless, for the reasons Fax and Talderas have so ably demonstrated.

Instead, consider a 4v4 encounter. A level 6 party would face 4 CR 2 enemies, as one example. If those enemies can, over the course of ~4 rounds, consume ~20% of PC resources (partly by chewing away HP, healing spells, etc, and partly by requiring spells, item charges, etc to be killed), then that means they must each do 5% of one PC's resources per round. The PCs, meanwhile, must use up between ~50-90% of their enemies' resources (depending on how much of their spell selection or item charges the enemies blow to stay alive/attack PCs/whatever before they die) in that same time, so at least 12.5% of a given enemy's resources per round per PC, and maybe as much as 22.5% or more.

From this, we see that PC resources, and ability to consume enemy resources, must scale at the rate of at least 2.5 times, and as much as 4.5 times, every 4 levels. Not all of that is offensive power, though; simply gaining almost triple HP by virtue of acquiring 4 more levels (in this particular case) obviously accomplishes a lot!

The chief flaw in this analysis, of course, and also in the CR system, is that "resources" is rather awfully abstract to express spell slots/pp, daily item charges, consumable charges, HP, daily abilities, and so forth. (It's especially bad when all-day buffs can be enough to deal with the majority of foes, effectively amortizing the same resource expenditure across almost any number of enemies.)

[Just to Browse]

Yea, monster design.
You said another party member. A character of level X == a monster of CR X. If this ends up not being true in some place, then something is wrong with the design of that monster or character, not the CR system.


And I can say this could be true at some instances even if they are all the same level with the numbers being 10 to 1 simply due to optimization along with the fact that the single person doesn't have to hold back in the least.
They could screw the party over in a single surprise round, and then wipe them out in a single move.

[snip]

You're entire argument is predicated on anecdotal examples and the idea that classes are unbalanced. So address those:
The classes are assumed to be balanced. Do not assume that one is more optimized for contribution than another.

Any anecdote pointing in one direction can just as easily point in another. Do not use specifics to disprove a general point.

EDIT: OK, I'm in with tuggyNE here.

If the players in the party (who all have 16 "resources" and remove 1 quantity of "resources" per round) face an encounter of 4 members, those members need to have 1/4 the resource removal and 1/4 the total resources to remove.

So it's 16/1, 16/1, 16/1, 16/1 versus 4/.25, 4/.25, 4/.25, 4/.25. The players are capable entirely removing one enemy in a single round, but it takes the opposing party 16 rounds of constant aggression before they even manage to take out a single PC. This is actually a great example of the inverse-boss problem, where the players are killing the enemies off so quickly that they don't even lose as many resources as they should. Total enemy resource removal is (4 + 3 + 2 + 1) (.25) = 2.5, out of 64. So basically nothing.

[theIrkin]

So in 3rd edition we have a two premises for fights:
A level-appropriate encounter will consume 25% of party resources.
A level-appropriate encounter (for a party of 4) is one party member.

I originally posted a simple version, but people appear to be getting caught up on my word choice, so I'm putting a different explanation down. The original one is spoilered at the end for your viewing pleasure.

Assume the average fight has a length L (measured in rounds). The average character contributes R 'resource depletion' per round. If this term sounds dumb to you, that's because it kind of is. There are a lot of things you can do in a fight--buff your party members to live longer, buff them to hit harder, cut down on enemy offenses, change the battlefield to make achieving objectives easier, etc.--but the reason I lump all of that together is because we're assuming each party member is balanced to one another (yes this isn't true in 3.x, everyone and their dog knows that already. Bear with me.), so each party member should be contributing equally on average, which is why we can simplify all useful actions down to a single quantity.

With this idea in mind, a party of P adventurers thus depletes P*R*L resources from an opponent, so the opponent must have that quantity of resources to deplete (TR). By assumption #2 (which is part of the DMG and MM, can't deny this), this means each player also has TR = P*R*L resources. Again using assumption #2, we know that the opponent in each fight must be capable (on average) of depleting R*L resources from the party as whole. That's where we see the problem...

There are P players in this party. If each of them has TR = P*R*L resources, that means the total party resources is P*TR = (P^2)*R*L, so an opponent that removes only R*L resources depletes 1/(P^2) of player resources. For a standard party of 4, this is 1/16 of the desired resources, which directly clashes with assumption #1. An equal-level enemy should on average be taking out a quarter of party resources, yet we can see that it would actually be taking out a quarter of that.

What this means is that even if there was excellent balance between the classes, and perfect monster-to-PC translation, combat wouldn't be functional because the players would barely lose any resources. I think this core problem is at least partly to blame for why the monsters have such out-of-whack CRs, or are CR'd 'appropriately' yet are totally unsuited as PCs of similar levels. The writers were trying to balance two unbalance-able concepts.

Thoughts?

EDIT: This is probably also why a lot of people feel like players breeze through their boss fights, despite ramping up numbers. In this example, you'd want to quadruple enemy resources to make it a standard fight for the PCs (though it would be a serious grindfest). A final boss made to take down all of the party would have quadruple resources and resource depletion*.

I'll do this by example because the math is easier. Assume our fight must last 4 rounds, and the average character contributes 1 damage per round (you might not be specifically dealing 1 damage per round, perhaps you buff the fighter to do +1 each round, or you charge-up for 2 damage every 2 rounds). This means that the average total damage per round is 16.

Since opposition to the party is supposed to be equal to the party, this means the enemy has 16 HP and also deals 1 damage per round, and it means each member of the party has ~16 HP.

However, if the opposition is going to deal only 1 damage per round, then it will take out 4 HP total over the course of the fight. Since each character has 16 HP, this is actually 1/16 of party resources, not 1/4.

A monster 2 CR higher is supposed to be twice as powerful (either twice as much HP or twice damage), which means it theoretically takes a creature 4 CR higher to actually challenge a party.

I think this is a big reason why monsters are so hard to translate from Team Monster to Team PC--as designers continued to write things, they realized that a CR 12 encounter couldn't be both a useful PC and a useful enemy, so they designed the monsters to be extra powerful.

If one of these assumptions is going to hold, the other has to fall. My question (other than "How did I mess up here?") is which of these assumptions do we have to scrap to solve this problem?

*At the upper and lower limits of the spectrum, this actually breaks down. A monster with 4R and 4TR would be a normal challenge for a party of level + 4, but would dispatch a low-level party without dying at all, because (on average) one player would fall per round, so total depletion would be something like 4(PR) + 3(PR) + 2(PR) + 1(PR), or 6PR less than necessary to take the guy out. An inverse case occurs for really weak mooks, in that players waste damage taking out targets with a TR < R*L, so they lose more resources in a mook-heavy fight without cleave effects.

Please explain why TR = P*TR. That's a recursive loop... logic fail

[Just to Browse]

Please explain why TR = P*TR. That's a recursive loop... logic fail

I'm not sure where you're looking, but I assume it's in my explanation of fixing?

Going off that, it's because the equation is TRmonster = P*TRplayers. It's one of the ways to "balance" the fight so the monster actually removed 1/4 of party resources.

[Platymus Pus]

A character of level X == a monster of CR X. If this ends up not being true in some place, then something is wrong with the design of that monster or character, not the CR system.



You're entire argument is predicated on anecdotal examples and the idea that classes are unbalanced. So address those:
The classes are assumed to be balanced. Do not assume that one is more optimized for contribution than another.

Any anecdote pointing in one direction can just as easily point in another. Do not use specifics to disprove a general point.
Classes are never balanced some are more optimized for contribution than others. It's called variety and it's a fact.
Assuming something doesn't make it true.

Do not ignore specifics like they don't effect your very broad point, because they do.

A PC can do whatever they want, the CR won't matter for the encounter because they aren't a monster. They could be CR+100 and decide to fight the party if they wished. You can't apply the same rule of a monster to a Player character, because it's just that a PC. So you can't say party member.

You have no point here unless it's a NPC.

25% of resources per encounter isn't even that challenging anyway going by what most of said resources are especially and most especially with a "balanced" party having so many ways around it.


as for the discussion on CR, yes it's broken.

They’re basic humanoids, so I have to build them with class levels. The rule that says “an nth-level NPC is a CR n monster”… well, let’s just say that the rule isn’t beyond reproach. It’s true of some classes within some level ranges, but it’s simply not accurate as a general rule. I don’t think any designer will tell you with a straight face that a 1st-level NPC wizard is a good challenge for four 1st-level PCs. (Better hope the NPC gets that sleep spell off, huh?) So my low-level drow have 1 point of CR vanish into thin air, and they lose more oomph because they’re built with class levels.
http://www.d20source.com/2006/07/challenge-rating-equals-level-maybe-not

[Bogardan_Mage]

If the players in the party (who all have 16 "resources" and remove 1 quantity of "resources" per round) face an encounter of 4 members, those members need to have 1/4 the resource removal and 1/4 the total resources to remove.
Not necessarily, and your example demonstrates why. Once the PCs kill off one of their enemies, the amount of resource removal remaining at the enemies' disposal is reduced. To be a level-appropriate encounter the enemies must be able to reduce the party's resources by 25% over the whole encounter, by your own definition. As you demonstrated that they cannot do so, they are not a level-appropriate encounter.

[Just to Browse]

Classes are never balanced some are more optimized for contribution than others. It's called variety and it's a fact.
Assuming something doesn't make it true. Variety != Balance. Classes can contribute differently against different opponents across different rounds, for different positionings, but the average contribution of each class must be equal. That is D&D's assumption, that is my assumption. It's not something you can debate.


Do not ignore specifics like they don't effect your very broad point, because they do.http://i.imgur.com/PkJBa3a.gif

They totally don't. Because for every situation you can think of going one way, I can think up another going the opposite. The point is that on average these contributions are equal. If it's possible for the enemy to get surprise and one-shot the players, then those players should be able to do the same because the CR is equal to their level.


A PC can do whatever they want, the CR won't matter for the encounter because they aren't a monster. They could be CR+100 and decide to fight the party if they wished. You can't apply the same rule of a monster to a Player character, because it's just that a PC. So you can't say party member.The core assumption is that the enemy is the same level as the party. Yes a player can go kill himself on a CR 100 monster, but this is not the issue at hand. You need to accept the premises of this discussion before you start throwing your opinions in there.


[snip]This is also not the issue at hand.


Not necessarily, and your example demonstrates why. Once the PCs kill off one of their enemies, the amount of resource removal remaining at the enemies' disposal is reduced. To be a level-appropriate encounter the enemies must be able to reduce the party's resources by 25% over the whole encounter, by your own definition. As you demonstrated that they cannot do so, they are not a level-appropriate encounter.

Indeed you are correct, and that's the reason I didn't want to do 4v4 CR=level fights. They ruin the math oh-so-perfectly. In my above post, I identified this as the other side of the Boss Fight problem (let's call this the Minion Fight problem). In boss fights, the opposition eliminates PCs so quickly they can't possibly contribute, but in Minion fights the players eliminate minions so quickly that two things occur:
If the minions have perfectly equal resources to a player hit, then players are capable of removing each minion every round, reducing the average resource reduction and making the combat easier.
However if the minions have fewer resources than the minimum amount that players remove, the players end up not killing the minions as fast as they should and it makes combat harder.

As I said before, the Minion and Boss Fight problems are on the edge of the combat spectrum and should both be relatively rare. It's also better to discuss them in another thread because you can say a whole lot on that topic alone. Again I'm more than willing to throw down another thread on that specific topic, let's just keep it away from here.

[Bogardan_Mage]

Indeed you are correct, and that's the reason I didn't want to do 4v4 CR=level fights. They ruin the math oh-so-perfectly. In my above post, I identified this as the other side of the Boss Fight problem (let's call this the Minion Fight problem). In boss fights, the opposition eliminates PCs so quickly they can't possibly contribute, but in Minion fights the players eliminate minions so quickly that two things occur:
If the minions have perfectly equal resources to a player hit, then players are capable of removing each minion every round, reducing the average resource reduction and making the combat easier.
However if the minions have fewer resources than the minimum amount that players remove, the players end up not killing the minions as fast as they should and it makes combat harder.

As I said before, the Minion and Boss Fight problems are on the edge of the combat spectrum and should both be relatively rare. It's also better to discuss them in another thread because you can say a whole lot on that topic alone. Again I'm more than willing to throw down another thread on that specific topic, let's just keep it away from here.
Why? The topic is "The basic premise of 3.5 combat does not work" not "...does not work in one particular instance". And not to beat a dead horse but as others have mentioned, a particular instance that is specifically advised against by the DMG.

The fact is that your example is not a level appropriate encounter, even if you increase (I assume you meant that, rather than "fewer") the number of resources of the enemies. If the enemies survive all four rounds (either because they have more resources or because the players helpfully attack them one-on-one each) they will have drained not one quarter but one sixteenth of the party's total resources. You've assumed that four characters equals one character with four times each stat, which is not an assumption the game makes in its CR calculations, and that's where the "minion problem" is coming from. Give them the actual ability to deplete 25% of the party's resources and they will actually be a level appropriate challenge.

If you still don't believe that's on topic I have another observation that applies to both scenarios: You've equated "resources" with HP alone, while the game actually has a broad range of "resources" which are not interchangeable. The average player will drain some of his own resources merely by attempting to drain the enemy's resources (the average is important, because while a Fighter can dish out damage at no cost, the Wizard has to spend spells. Assuming everyone is a Fighter isn't going to give an accurate model of the game). So whether they're fighting a boss or his minions, the players are going to be losing resources as they attack. Sure, the enemies are too, but it works out to the players losing comparatively more because the boss is just one guy to their four (they will lose four times as many as he does assuming that everyone performs the same action, by definition) and the minions are much weaker (assuming it turns out that blindly dividing everything by four is accurate).

[NichG]

The fallacy here is the assumption that the CR system is actually integral to 3.5's combat system. But really, 'combat in 3.5' is a separate issue from 'encounter design in 3.5', and the two should not be conflated.

Basically, the idea of a CR that is somehow rigorous is unfounded, even 'in the average'. The debate has sort of been focused around this idea of 'well, there must be some 'correct' CR value that we could assign to any given character/monster that would appropriately create this 25% resource consumption metric we want' but that isn't true.

There are a couple issues. One is that the average is poorly defined because it involves things like player skill, which the system doesn't try to model. So you have to go to data from outside the system in order to analyze the system, and that data isn't even stationary (since players improve with time). So this is just messy.

The other is that, even within the system, synergistic effects are a fairly dominant contribution. In effect, a monster 'on its own', paired with another of the same type, paired with another of a different type, paired with terrain, etc all will have very different difficulties. Averaging over those differences is clumsy, because the standard deviation is much larger than the mean and so the mean doesn't characterize the distribution of outcomes very well.

None of this however prevents an actual combat from working well though. It just means that you can't use a single number and a simple addition formula to accurately determine the outcome of the combat independent of the party composition/etc.

[Just to Browse]

The fallacy here is the assumption that the CR system is actually integral to 3.5's combat system. But really, 'combat in 3.5' is a separate issue from 'encounter design in 3.5', and the two should not be conflated.Er... OK, I think this is just a problem with word choice. Just replace "combat" with "encounter design", and the statement holds.


There are a couple issues. One is that the average is poorly defined because it involves things like player skill, which the system doesn't try to model. So you have to go to data from outside the system in order to analyze the system, and that data isn't even stationary (since players improve with time). So this is just messy.System mastery is explicitly called out in D&D. In the DMG, it says that if your players demonstrate system mastery (read: high skill), then you need to throw higher-CR challenges against them. All a game needs to do is presuppose a skill level, then tell the DM to compensate for skill level.


The other is that, even within the system, synergistic effects are a fairly dominant contribution. In effect, a monster 'on its own', paired with another of the same type, paired with another of a different type, paired with terrain, etc all will have very different difficulties. Averaging over those differences is clumsy, because the standard deviation is much larger than the mean and so the mean doesn't characterize the distribution of outcomes very well.I'm not going to debate with you what the standard deviation of resource depletion is, but if synergistic effects are available to team monster then they're available to team PC, and vice versa. A monster in a favorable condition can just as likely be in an unfavorable condition, so assuming that there is some 'synergy' modifier that completely changes CR does not make sense. Monsters in favorable conditions will be more difficult, and the encounter should be a higher-CR to reflect that.

Like I said before, both sides get the same options. Adding complexity does not somehow change the core math--it just makes it complicated.


Why? The topic is "The basic premise of 3.5 combat does not work" not "...does not work in one particular instance". And not to beat a dead horse but as others have mentioned, a particular instance that is specifically advised against by the DMG.No, it's not a particular instant. It's that sometimes the CR system appears to work despite being designed to fail, but most of the time the CR system fails because the CR system was designed in such a way that it has to fail.


The fact is that your example is not a level appropriate encounter, even if you increase (I assume you meant that, rather than "fewer") the number of resources of the enemies. If the enemies survive all four rounds (either because they have more resources or because the players helpfully attack them one-on-one each) they will have drained not one quarter but one sixteenth of the party's total resources. You've assumed that four characters equals one character with four times each stat, which is not an assumption the game makes in its CR calculations, and that's where the "minion problem" is coming from. Give them the actual ability to deplete 25% of the party's resources and they will actually be a level appropriate challenge.I invite you to read the original post. That is the entire point. Either the opponents drain 25% of player resources and are not equal-CR by D&D standards, or they are equal-CR and instead drain 1/16. That is what this thread is all about. The designers made that assumption, and that is why the system doesn't work.


Resources [snip]Those resources are equatable in that every single one will be used as defensive material (stopping you from dropping) or offensive material (to drop enemies), and the goal of the fight is to have non-zero resources when the enemy has zero.

If a wizard uses a spell, he's used one of his limited resources and now has less, which means his attack should be relatively more effective than a simple at-will attack. This is the basic theory that demonstrates why at-will casting is weaker than vancian casting, and it can still be modeled in the example with no problem. The wizard simply pays some resource cost (call it 1) to remove extra resources from the enemy (call it 2). This adds tactical depth, and the increase in horizontal power is something you need to take into account, but the math remains relatively the same.

[Waddacku]

The entirety of the actual details in monster design, spells, abilities, and all that stuff is completely irrelevant, because the issue at hand is the conflicting premises the CR system makes. The reality of how the game works doesn't matter whatsoever for it, because it has nothing to do with the assumptions made when setting up the CR system in the first place. The DMG does state that a 4v1 fight doesn't work as their CR rules imply it should, but the fact is the CR rules doesn't work in general either.

Consider this: CR is (aside from the level scale) system independent. You you just rip it straight out and implant it somewhere else and the only thing you need do is replace level with whatever advancement the other system is using.
However, even removed from any game system, the math of the CR system fails to actually meet its own premises.

[zlefin]

long thread; in which post is your proposed fix to the system?

[TuggyNE]

long thread; in which post is your proposed fix to the system?

Who said anything about a fix? Getting the facts straight is worth a thread on its own.

[skyth]

Your 25% per encounter number is not accurate. The idea is that the party can survive 4 encounters before needing to regroup/rest. If every encounter drained 25% of the party resources, then the party would, on average, not survive 4 encounters.

If you assume that the party stops at 50% resources left, each encounter should drain 12.5% of resources. If you assume that the party actually stops at 75% resources, each encounter drains 6.25% (IE 1/16) of total resources.

Something else to consider...A 3rd level character does not do twice the resource depletion of a 1st level character. There is a diminishing return on the amount of resources that will be depleted by a character as they go up in levels.

[BlasphemousSlug]

@Just to Browse

I tried very very hard to read through the entire thread, but found that it is too much for me. That you managed to read, and reply to many comments and people who "just aren't getting it" is astounding to me, and I applaud your amicability.

I think the stumbling block that most people are having to trying to understand what you are saying is that you a) named the thread what you did and b) didn't highlight (enough) on the incongruity of the two statements

Thanks for the info though, it was a very interesting read. I noticed that traditional characters never really seemed to fit the challenge rating they were supposed to be given, but I never knew how bad it was. I only wish that I had decided to do the math earlier, you know, before all those TPKs... :smallredface:

[zlefin]

Who said anything about a fix? Getting the facts straight is worth a thread on its own.

but this isn't getting the facts straight; it's demonstrating that there's issues with the CR system; but we already long knew that the CR system has issues; and we knew what those issues were.
So without adding a fix i'm not sure what it really adds to the discourse.

That's without getting into issues the model used herein has.

[Psyren]

long thread; in which post is your proposed fix to the system?

The "fix" is to design your encounters properly - I shared a useful guide (https://docs.google.com/document/d/1nx-o8VAjhUwh3nnfzDQT-JA5eFLnN_BZJiBitGjBMDg/edit) on how to do that.

[Platymus Pus]

They totally don't. Because for every situation you can think of going one way, I can think up another going the opposite. The point is that on average these contributions are equal. If it's possible for the enemy to get surprise and one-shot the players, then those players should be able to do the same because the CR is equal to their level.

Except it's blatantly not fair to the players then.
Who wants to make a new character for every single encounter.


There are more encounters than there are party members.
So no it's not equal at all.
You're only using a single encounter working on the logic of depletion for your example that's why it doesn't work.

It also may not be possible for players to do the aforementioned due to other factors despite being equal CR as a surprise round has other factors despite being a part of combat.

Surprise

When a combat starts, if you are not aware of your opponents and they are aware of you, you’re surprised.
Determining Awareness

Sometimes all the combatants on a side are aware of their opponents, sometimes none are, and sometimes only some of them are. Sometimes a few combatants on each side are aware and the other combatants on each side are unaware.

Determining awareness may call for Listen checks, Spot checks, or other checks.
The Surprise Round

If some but not all of the combatants are aware of their opponents, a surprise round happens before regular rounds begin. Any combatants aware of the opponents can act in the surprise round, so they roll for initiative. In initiative order (highest to lowest), combatants who started the battle aware of their opponents each take a standard action during the surprise round. You can also take free actions during the surprise round. If no one or everyone is surprised, no surprise round occurs.
Unaware Combatants

Combatants who are unaware at the start of battle don’t get to act in the surprise round. Unaware combatants are flat-footed because they have not acted yet, so they lose any Dexterity bonus to AC.



Are there rules for CR being effected regarding surprise rounds?
Because I can't find any.


Another reason CR doesn't work as is because the dice are blatantly chaotic.
The dice rolling bad for the party and rolling good for the monster can result in a loss of a party member screwing with the balance of the CR for the encounter. Then it goes downhill from there.

On the other hand the party rolling good and the monster rolling bad ends up with a dead monster.
The monsters don't actually have to worry about depletion because of this making things skewed.
The party moves on until they run into another monster and the scenario happens again.

That's ignoring that it's actually possible to go into the negatives with resources and still fight.



Because for every situation you can think of going one way, I can think up another going the opposite.
That's not the point, once a character dies it expends all of their resources and possibly their parties on top of it to revive said character. Expending resources to regain them.

The monsters don't have to.






The core assumption is that the enemy is the same level as the party. Yes a player can go kill himself on a CR 100 monster, but this is not the issue at hand. You need to accept the premises of this discussion before you start throwing your opinions in there.


Quit saying assumption. It's the mistake the system makes.
It's assuming levels themselves matter for combat that much. They don't.
Levels measure base power in regards to nothing else, it's also why assuming a balanced character is a mistake as levels by themselves don't think about what a character gains.

Also note that with enough optimization, it's generally possible to go up a tier, and if played poorly you can easily drop a few tiers, but this is a general averaging, assuming that everyone in the party is playing with roughly the same skill and optimization level. As a rule, parties function best when everyone in the party is within 2 Tiers of each other (so a party that's all Tier 2-4 is generally fine, and so is a party that's all Tier 3-5, but a party that has Tier 1 and Tier 5s in it may have issues).

The Tier System

Tier 1: Capable of doing absolutely everything, often better than classes that specialize in that thing. Often capable of solving encounters with a single mechanical ability and little thought from the player. Has world changing powers at high levels. These guys, if played well, can break a campaign and can be very hard to challenge without extreme DM fiat, especially if Tier 3s and below are in the party.

Examples: Wizard, Cleric, Druid, Archivist, Artificer, Erudite

Tier 2: Has as much raw power as the Tier 1 classes, but can't pull off nearly as many tricks, and while the class itself is capable of anything, no one build can actually do nearly as much as the Tier 1 classes. Still potencially campaign smashers by using the right abilities, but at the same time are more predictable and can't always have the right tool for the job. If the Tier 1 classes are countries with 10,000 nuclear weapons in their arsenal, these guys are countries with 10 nukes. Still dangerous and world shattering, but not in quite so many ways. Note that the Tier 2 classes are often less flexible than Tier 3 classes... it's just that their incredible potential power overwhelms their lack in flexibility.

Examples: Sorcerer, Favored Soul, Psion, Binder (with access to online vestiges)

Tier 3: Capable of doing one thing quite well, while still being useful when that one thing is inappropriate, or capable of doing all things, but not as well as classes that specialize in that area. Occasionally has a mechanical ability that can solve an encounter, but this is relatively rare and easy to deal with. Challenging such a character takes some thought from the DM, but isn't too difficult. Will outshine any Tier 5s in the party much of the time.

Examples: Beguiler, Dread Necromancer, Crusader, Bard, Swordsage, Binder (without access to the summon monster vestige), Wildshape Varient Ranger, Duskblade, Factotum, Warblade, Psionic Warrior

Tier 4: Capable of doing one thing quite well, but often useless when encounters require other areas of expertise, or capable of doing many things to a reasonable degree of competance without truly shining. Rarely has any abilities that can outright handle an encounter unless that encounter plays directly to the class's main strength. DMs may sometimes need to work to make sure Tier 4s can contribue to an encounter, as their abilities may sometimes leave them useless. Won't outshine anyone except Tier 6s except in specific circumstances that play to their strengths. Cannot compete effectively with Tier 1s that are played well.

Examples: Rogue, Barbarian, Warlock, Warmage, Scout, Ranger, Hexblade, Adept, Spellthief, Marshal, Fighter (Dungeoncrasher Variant)

Tier 5: Capable of doing only one thing, and not necessarily all that well, or so unfocused that they have trouble mastering anything, and in many types of encounters the character cannot contribute. In some cases, can do one thing very well, but that one thing is very often not needed. Has trouble shining in any encounter unless the rest of the party is weak in that situation and the encounter matches their strengths. DMs may have to work to avoid the player feeling that their character is worthless unless the entire party is Tier 4 and below. Characters in this tier will often feel like one trick ponies if they do well, or just feel like they have no tricks at all if they build the class poorly.

Examples: Fighter, Monk, CA Ninja, Healer, Swashbuckler, Rokugan Ninja, Soulknife, Expert, OA Samurai, Paladin, Knight

Tier 6: Not even capable of shining in their own area of expertise. DMs will need to work hard to make encounters that this sort of character can contribute in with their mechanical abilities. Will often feel worthless unless the character is seriously powergamed beyond belief, and even then won't be terribly impressive. Needs to fight enemies of lower than normal CR. Class is often completely unsynergized or with almost no abilities of merit. Avoid allowing PCs to play these characters.

Examples: CW Samurai, Aristocrat, Warrior, Commoner

And then there's the Truenamer, which is just broken (as in, the class was improperly made and doesn't function appropriately).

http://www.brilliantgameologists.com/boards/?topic=1002.0
There is a party of 4 level 1 commoners.
Are you going to use that broken CR formula when 4 house cats can kill the party?(1/4 x 4 = CR 1)
That can kill the party of commoners, with a surprise round even more so.
add a 5th housecat and suddenly it goes from 50% to 15% survival rate.
That lasts all the way to 14 housecats somehow.
http://www.d20srd.org/extras/d20encountercalculator/

Now I'm not sure about you, but I don't think that many classes would consider a housecat a real threat if there was a 1 1/4 CR's worth of them.





Well, officially there is a formula, and you can automate it with this Encounter Calculator:

http://www.penpaperpixel.org/tools/d20encountercalculator.htm

For your example, that's a CR14 encounter against an average PL of 10.5, so Very Difficult.

Unofficially, CR is guesswork at the best of times. If CR is unreliable, sometimes you have to go into the grit and manually compare your party's damage output/survivability to various monster stat blocks before you choose them for an encounter.

The fact that each party is good at different things and worse at others, combined with the fact that WotC can't judge a creature's CR for their life (MMII, I'm looking at you), combined with the fact that their formula doesn't work (10 CR 1 creatures will not challenge your average ECL 10 Party, no matter what the EL Calculator says) turns CR evaluation into a game of "guess and check", except with more TPKs.

EDIT: Ninja'd

Vorpal Tribble made up a pretty nifty system to calculate the correct CR of any creature. You can find it here:
http://www.giantitp.com/forums/showthread.php?t=192329

Well, first you look at the monster or monsters, then you look at the capabilities of the party. Then you close your eyes, spin in a circle until you fall down, and count the spots floating in your vision. This is called "the WotC method."

Or you can follow Ciccio's link, but that's not nearly as fun.

http://www.giantitp.com/forums/showthread.php?t=208308

It's been discussed to death. The system is majorly flawed people have come up with better systems than it.
http://www.giantitp.com/forums/showthread.php?t=143809

[TuggyNE]

Except it's blatantly not fair to the players then.
Who wants to make a new character for every single encounter.

The point is symmetry of CR, not that Just to Browse was recommending such a TPK-monster. Your rejoinder does not follow.


There are more encounters than there are party members.
So no it's not equal at all.
You're only using a single encounter working on the logic of depletion for your example that's why it doesn't work.

You can, if you like, track it out by encounters per day (4, generally), but it makes the math slightly longer. Doesn't change the results any though.


There is a party of 4 level 1 commoners.
Are you going to use that broken CR formula when 4 house cats can kill the party?(1/4 x 4 = CR 1)
That can kill the party of commoners, with a surprise round even more so.
add a 5th housecat and suddenly it goes from 50% to 15% survival rate.
That lasts all the way to 14 housecats somehow.

That's not what "challenging" means at all. "Challenging", and the 50%/15% numbers, are references to the DMG's guide for how often a given encounter should come up. Namely, a normal equal-CR encounter should make up 50% of encounters, and a Very Difficult encounter should be 15% of them. "Challenging", though, is the stated 20-25% of resources level of difficulty, not a 50/50 toss-up.

[Bogardan_Mage]

No, it's not a particular instant. It's that sometimes the CR system appears to work despite being designed to fail, but most of the time the CR system fails because the CR system was designed in such a way that it has to fail.
I don't think we understand one another, since I was talking about your apparent belief that only this one particular scenario of 4 against 1 is on topic in this thread. You don't seem to believe that 4 against 4 is any less broken in the system, so why should it be off topic when the topic is the flaws in the system?


I invite you to read the original post. That is the entire point. Either the opponents drain 25% of player resources and are not equal-CR by D&D standards, or they are equal-CR and instead drain 1/16. That is what this thread is all about. The designers made that assumption, and that is why the system doesn't work.
No, you made an assumption that an equal-CR encounter of four characters will together drain a quarter of the resources of any one of the characters (or, equivalently, a singular equal-CR opponent). That is not borne out by any game mechanics or by your other abstract assumptions, and I think you'll find that that is what you are missing.


Those resources are equatable in that every single one will be used as defensive material (stopping you from dropping) or offensive material (to drop enemies), and the goal of the fight is to have non-zero resources when the enemy has zero.

If a wizard uses a spell, he's used one of his limited resources and now has less, which means his attack should be relatively more effective than a simple at-will attack. This is the basic theory that demonstrates why at-will casting is weaker than vancian casting, and it can still be modeled in the example with no problem. The wizard simply pays some resource cost (call it 1) to remove extra resources from the enemy (call it 2). This adds tactical depth, and the increase in horizontal power is something you need to take into account, but the math remains relatively the same.
The math would remain the same if you weren't double counting by dividing the "resources drained" figure by 4 in the minions fight. If we assume, however, that some of those "resources drained" represent the characters spending spells or single-use items or other offensive resources, there's no reason to assume they would suddenly used one quarter of those when confronted with 4 weaker opponents than when confronted with a single equal-level opponent. If they were, for example by withholding higher level spell slots in anticipation of tougher battles, then the resources they drained from the enemy would be less. Either way dividing both figures by 4 results in you double counting the effect, and does not give you an accurate model of the CR system.

[Platymus Pus]

You can, if you like, track it out by encounters per day (4, generally), but it makes the math slightly longer. Doesn't change the results any though.




I'm pointing out the first encounter can go far beyond what the game thinks will be used resource wise due to monster variety at CR levels.
It can also go below what it thinks will happen. This shows CR is flawed in accurately predicting things.

That's not what "challenging" means at all. "Challenging", and the 50%/15% numbers, are references to the DMG's guide for how often a given encounter should come up. Namely, a normal equal-CR encounter should make up 50% of encounters, and a Very Difficult encounter should be 15% of them. "Challenging", though, is the stated 20-25% of resources level of difficulty, not a 50/50 toss-up.
Yes I can read.
http://www.d20srd.org/extras/d20encountercalculator/

This is how many encounters of the indicated difficulty an adventure should have. See "Difficulty" on page 49 of the Dungeon Master's Guide for details.

The first encounter likely would result in a TPK with housecats going beyond the 50% resource that the calc thinks would happen in that encounter with normal commoners.

It doesn't account for the fact commoners can all be killed in a single surprise round by housecats or any round after and it isn't even considering that said commoners are having massive trouble with something considered "equal".

It has flaws that the GM has no choice but to balance despite the action economy also being equal. 4 housecats vs 4 commoners. Add an extra cat and all of a sudden the chances of winning goes dramatically down for the commoners despite the CR only going up by 1/4th

Who in their right mind would throw something that can one shot your party members for 50% of all encounters? CR thinks it's fair.

[TuggyNE]

Yes I can read.
http://www.d20srd.org/extras/d20encountercalculator/


The first encounter likely would result in a TPK with housecats going beyond the 50% resource that the calc thinks would happen in that encounter with normal commoners.

I'm not sure how to say this politely, but… the encounter calculator very specifically does not say in any form that the encounter as specified will consume 50% of resources. Nor that it will have a 50% chance of killing the characters, or anything else like that.

It says, quite clearly and distinctly, "% of total9" and "9 This is how many encounters of the indicated difficulty an adventure should have. See "Difficulty" on page 49 of the Dungeon Master's Guide for details."


It doesn't account for the fact commoners can all be killed in a single surprise round by housecats or any round after and it isn't even considering that said commoners are having massive trouble with something considered "equal".

It has flaws that the GM has no choice but to balance despite the action economy also being equal. 4 housecats vs 4 commoners. Add an extra cat and all of a sudden the chances of winning goes dramatically down for the commoners despite the CR only going up by 1/4th

Who in their right mind would throw something that can one shot your party members for 50% of all encounters? CR thinks it's fair.

Commoners are not PC classes, and as mentioned, there is no reason at all to assume that CR actually considers four housecats to have a 50% chance of killing half the party. Even half the "worst-NPC-class-ever" party.

[Platymus Pus]

I'm not sure how to say this politely, but… the encounter calculator very specifically does not say in any form that the encounter as specified will consume 50% of resources. Nor that it will have a 50% chance of killing the characters, or anything else like that.

It says, quite clearly and distinctly, "% of total9" and "9 This is how many encounters of the indicated difficulty an adventure should have. See "Difficulty" on page 49 of the Dungeon Master's Guide for details."



Commoners are not PC classes, and as mentioned, there is no reason at all to assume that CR actually considers four housecats to have a 50% chance of killing half the party. Even half the "worst-NPC-class-ever" party.
Commoners can be played as a PC, people do it all the time despite it being a NPC class.
Besides commoners have a CR of .5 while a house cat has a CR of 1/4.
Doesn't work anyway.

The 50% encounter rate is based on resources taken due to difficulty based on what CR is supposed to do.

1. A level-appropriate encounter will consume 25% of the party's resources.
That has already been established and what is considered more level appropriate than something with a 50% encounter rate according to CR.
I'm presenting something where there is a 50% encounter rate with something that is supposed to take 25%, but instead takes much more.
25-100%. 100% is a dead party, the 25% is the loss of a member since it's such low play.

I also said there would be a higher chance than 50% of the cats killing the party yet they have an encounter rate of 50%. I'm showing an imbalance.
The chances of dying are actually higher than the encounter rate recommended as an average encounter.

[TuggyNE]

Commoners can be played as a PC, people do it all the time despite it being a NPC class.

But the system is not designed around that! Criticizing it for being "broken" when you're introducing weird random factors that it doesn't attempt to account for is foolish.


Besides commoners have a CR of .5 while a house cat has a CR of 1/4.
Doesn't work anyway.

And there you have it.


100% is a dead party, the 25% is the loss of a member since it's such low play.

Hmm. And you're sure it's not, say, one member at 1 or 2 HP, another two several HP down, and the fourth untouched?


I also said there would be a higher chance than 50% of the cats killing the party yet they have an encounter rate of 50%. I'm showing an imbalance.
The chances of dying are actually higher than the encounter rate recommended as an average encounter.

I still don't get why you're weirdly fixated on the 50%/50% figures. The numbers are unrelated to each other, and off-hand comparisons like "The chances of dying are actually higher than the encounter rate recommended as an average encounter" are basically meaningless except for news-station-quality talking points. What's meaningful is to consider the overall chance of dying/TPKing over a series of recommended encounters, and determine how often a new party will need to be constructed.

Skipping the commoners, let's try Rogue/Wizard/Fighter/Cleric for a minute against those same four cats. At low-op, the Cleric (14 Str, 13 Dex, 12 Con, 10 Int, 15 Wis, 8 Cha; +2 attack for 1d8+2, 17 AC with chainmail and light shield, 9 HP) is healbotting whenever not bashing a cat with a morningstar; 3 orisons and 3 first-level spells to spend over the day. The Rogue (14 Str, 15 Dex, 13 Con, 12 Int, 10 Wis, 8 Cha; +2 attack for 1d6+1d6+2 or 1d6+1d6, 16 AC with chain shirt, 7 HP) is trying to flank (or might use a shortbow in the surprise/opening round against flat-footed cats at AC 12). The Wizard (10 Str, 13 Dex, 14 Con, 15 Int, 12 Wis, 8 Cha; +1 attack for 1d8, 15 AC with mage armor, 6 HP) could use color spray except that's not low-op, so just magic missile for 1d4+1 auto damage. The Fighter (15 Str, 12 Dex, 14 Con, 13 Int, 10 Wis, 8 Cha; +4 attack for 2d6+3, 16 AC with chainmail, 12 HP) Weapon Focus: Greatswords his way to victory and glory.

The cats have slightly less than a 50% chance to hit on their claws for most of the characters, and a considerably worse chance on their bite, for about 1 or 2 damage per round per cat average; uniformly attacking the Wizard in the surprise round and following gives them a good chance of dropping him into negatives on the first round, but the Cleric can heal him back up with one spell. Meanwhile, (almost) any hit by any character will drop them, so assuming the Rogue goes before at least two cats and acts in the surprise round, there's half of them already dying, and the other two follow soon after. Net expenditure one to three spell slots and maybe some ammo; little or no chance of death, much less TPK.

[Just to Browse]

Quick responses, then the length one:

Waddaku: Thank you, that's actually a great explanation for complexity. If you don't mind, I'd like to quote you in the OP.

zlefin: To use an allegory, I think this is more of a "fighters are bad" thread than a "let's fix the fighter" thread. The reason I'm not adding a proposed fix is because this is a new (I think) revelation for the community, and we need to expose this before we go on to argue over solutions. I mean, looking at the chaos generated by just writing what I wrote, I think adding arguments over what the "best" fixes would bog down the thread to incomprehensibility.

skyth: That's not my assumption, that's an assumption in the DMG. It does have problems unrelated to my post, but those problems are not what I'm debating.

BlasphemousSlug: Agreed. I think I'm going to change those things now. Better understanding is always appreciated.

Platymus Plus: I no longer understand how the things you are saying relate to my argument at all, and frustrating tuggyNE is not going to help us here. Could you re-explain why you think this system does not accurately model the basic structure of combat?

Long response:


I don't think we understand one another, since I was talking about your apparent belief that only this one particular scenario of 4 against 1 is on topic in this thread. You don't seem to believe that 4 against 4 is any less broken in the system, so why should it be off topic when the topic is the flaws in the system?This is what I believe:

In a fight where any creature can remove a number of resources equal to or greater than any opponent's current resources with a single turn of action at least one turn before the projected last round of the fight, and in the process of such an action is capable of directing all that resource-removal at one target, then the system breaks down. If you fight a boss monster with a 25%-removal that hits 4 people at a time, an APL + 4 fight works out fine. If you have players with a 25%-removal that hits 4 targets at a time, a 4v4 at equal CR works out fine. But if you make all the removal single-target (which we need for convenience in the simulation), then you run into the Minion Fight problem and Boss Fight problem.


No, you made an assumption that an equal-CR encounter of four characters will together drain a quarter of the resources of any one of the characters (or, equivalently, a singular equal-CR opponent). That is not borne out by any game mechanics or by your other abstract assumptions, and I think you'll find that that is what you are missing.I took the two assumptions out of the DMG, and showed how they cannot work together. A creature of CR X == 4 creatures CR X - 4. Alternatively, a creature with resources X and removal Y == 4 creatures with resources X/4 and removal Y/4. That's not just borne out in game mechanics, it's what challenges are designed around.

Think about it this way: If I instead created an encounter designed backwards in order to remove 25% of total party resources, you would notice that the total CR of the encounter does not equal party level, indicating the system is broken. I just did this the other way around (proving an equal-CR opponent does not remove 25% of resources), because math is hard.


The math would remain the same if you weren't double counting by dividing the "resources drained" figure by 4 in the minions fight. If we assume, however, that some of those "resources drained" represent the characters spending spells or single-use items or other offensive resources, there's no reason to assume they would suddenly used one quarter of those when confronted with 4 weaker opponents than when confronted with a single equal-level opponent. If they were, for example by withholding higher level spell slots in anticipation of tougher battles, then the resources they drained from the enemy would be less. Either way dividing both figures by 4 results in you double counting the effect, and does not give you an accurate model of the CR system.I'm really confused by this statement. You're saying that against weak enemies, players might not use 25% of their resources because they could decide not to? Well, I'll counter by saying that players might decide to use more than 25% because they decide to do so. The assumption is that the players will burn as much as necessary, and not more, and that "as much as necessary" has to equal the amount that the minions would be draining in the first place.

And no, dividing both numbers by 4 is necessary, because 1 CR X == 4 CR X - 4. If I let the minions keep the same damage, they'd be doing 4/round, whereas their equivalent 1-creature encounter would be doing 1/round, so the CR's would not be balanced.

[Platymus Pus]

But the system is not designed around that! Criticizing it for being "broken" when you're introducing weird random factors that it doesn't attempt to account for is foolish.



And there you have it.



Hmm. And you're sure it's not, say, one member at 1 or 2 HP, another two several HP down, and the fourth untouched?

DnD is practically nothing but weird random factors.
It's why CR doesn't work fully as intended it's not complicated enough.

Have what? A commoner having 2/4 , but actually being lower in every other area that matters in level 1 combat besides 2 more hp compared to a 1/4 who has a 50% of doing that amount of hp? Yea, we have that.

Do you want to actually have a real battle to see how that'd actually go a majority of the time? Because the % of it actually being %25 through those means is low. There is a percentage for the percentage.



Skipping the commoners, let's try Rogue/Wizard/Fighter/Cleric for a minute against those same four cats. At low-op, the Cleric (14 Str, 13 Dex, 12 Con, 10 Int, 15 Wis, 8 Cha; +2 attack for 1d8+2, 17 AC with chainmail and light shield, 9 HP) is healbotting whenever not bashing a cat with a morningstar; 3 orisons and 3 first-level spells to spend over the day. The Rogue (14 Str, 15 Dex, 13 Con, 12 Int, 10 Wis, 8 Cha; +2 attack for 1d6+1d6+2 or 1d6+1d6, 16 AC with chain shirt, 7 HP) is trying to flank (or might use a shortbow in the surprise/opening round against flat-footed cats at AC 12). The Wizard (10 Str, 13 Dex, 14 Con, 15 Int, 12 Wis, 8 Cha; +1 attack for 1d8, 15 AC with mage armor, 6 HP) could use color spray except that's not low-op, so just magic missile for 1d4+1 auto damage. The Fighter (15 Str, 12 Dex, 14 Con, 13 Int, 10 Wis, 8 Cha; +4 attack for 2d6+3, 16 AC with chainmail, 12 HP) Weapon Focus: Greatswords his way to victory and glory.

The cats have slightly less than a 50% chance to hit on their claws for most of the characters, and a considerably worse chance on their bite, for about 1 or 2 damage per round per cat average; uniformly attacking the Wizard in the surprise round and following gives them a good chance of dropping him into negatives on the first round, but the Cleric can heal him back up with one spell. Meanwhile, (almost) any hit by any character will drop them, so assuming the Rogue goes before at least two cats and acts in the surprise round, there's half of them already dying, and the other two follow soon after. Net expenditure one to three spell slots and maybe some ammo; little or no chance of death, much less TPK.


The cats are so little of a challenge to the standard party it won't take 25% of the resources they have despite it being CR 1.

[zlefin]

I see, well, you can believe that if you like; but it's just not true. These aren't problems that people are unaware of; this is not a revelation; they've been known for ages and discussed for just as long.

[eggynack]

I see, well, you can believe that if you like; but it's just not true. These aren't problems that people are unaware of; this is not a revelation; they've been known for ages and discussed for just as long.
True enough. My primary dispute is with this smooth and sterile model as a source for explanation. It can only ever be imperfect, when a far simpler explanation would go twice as far. Basically, the whole point is that the action economy matters. Four characters against one significantly more powerful enemy will have the odds tilted towards the four characters, simply because they get more actions. Having encounters with more enemies helps out a lot, but the CR system is intrinsically marred by the creators of the game not really understanding what gives a monster (or character) power. That's basically it in a nut shell. There's no need for expressing everything a character is in terms of "resource points" and trying to determine how every action affects that stockpile.

[TuggyNE]

DnD is practically nothing but weird random factors.

Possibly. "The PCs are all using NPC classes!" is not a design goal, though, so it can be ignored.


Have what? A commoner having 2/4 , but actually being lower in every other area that matters in level 1 combat besides 2 more hp compared to a 1/4 who has a 50% of doing that amount of hp? Yea, we have that.

No, more like the fact that a Commoner 1 is CR 1/2, while a PC class at level 1 is CR 1, which shows that all-Commoner parties are, of course, irrelevant to CR calculations in the general case: the system goes out of its way to tell you that they do not count the same.


The cats are so little of a challenge to the standard party it won't take 25% of the resources they have despite it being CR 1.

Make up your mind: are you arguing that a level 1 party gets slaughtered by EL 1 encounters because of the system's design, or are you arguing that a particular EL 1 encounter is too weak and isn't exactly the right challenge? Because those assertions do not work well together.

[Bogardan_Mage]

I took the two assumptions out of the DMG, and showed how they cannot work together. A creature of CR X == 4 creatures CR X - 4. Alternatively, a creature with resources X and removal Y == 4 creatures with resources X/4 and removal Y/4. That's not just borne out in game mechanics, it's what challenges are designed around.

Think about it this way: If I instead created an encounter designed backwards in order to remove 25% of total party resources, you would notice that the total CR of the encounter does not equal party level, indicating the system is broken. I just did this the other way around (proving an equal-CR opponent does not remove 25% of resources), because math is hard.
Right. What you did was demonstrate that given certain assumptions about class balance and so forth, an ostensibly level-appropriate 4 on 1 fight does not in fact deplete as many resources as it ought to. You then divided the same figures by 4 to make a 4 on 4 fight and acted surprised that the same problem appeared. Of course it did, you're using the same figures! You're not deriving the CR-4 stats from anything in the game, you're deriving them from the figures that you have already demonstrated to be faulty. Find an independent method of deriving these stats or they are useless.


I'm really confused by this statement. You're saying that against weak enemies, players might not use 25% of their resources because they could decide not to? Well, I'll counter by saying that players might decide to use more than 25% because they decide to do so. The assumption is that the players will burn as much as necessary, and not more, and that "as much as necessary" has to equal the amount that the minions would be draining in the first place.
No, you are saying that, by claiming that the minions drain a quarter as many resources as the boss by virtue of being lower level. I'm saying that the resources drained will not go down at the same rate as the resources available.


And no, dividing both numbers by 4 is necessary, because 1 CR X == 4 CR X - 4. If I let the minions keep the same damage, they'd be doing 4/round, whereas their equivalent 1-creature encounter would be doing 1/round, so the CR's would not be balanced.
Except the resource depletion is not solely a property of the enemy, which is my entire point. The players are spending resources every time they attack. If you're claiming that this is taken into account in the resource depletion stat, then that stat cannot vary so greatly based solely on the level of the enemy.

[Just to Browse]

I see, well, you can believe that if you like; but it's just not true. These aren't problems that people are unaware of; this is not a revelation; they've been known for ages and discussed for just as long.

Er... people already knew that a CR = level encounter is designed to drain 1/16 of player resources instead of 1/4? Well, it's at least news to me and I've been on D&D forums for a while, so if it's news to me then I'm sure it's news to at least someone else. In fact, BlasphemousSlug even admitted such a thing.


Right. What you did was demonstrate that given certain assumptions about class balance and so forth, an ostensibly level-appropriate 4 on 1 fight does not in fact deplete as many resources as it ought to. You then divided the same figures by 4 to make a 4 on 4 fight and acted surprised that the same problem appeared. Of course it did, you're using the same figures! You're not deriving the CR-4 stats from anything in the game, you're deriving them from the figures that you have already demonstrated to be faulty. Find an independent method of deriving these stats or they are useless.Ah, I see. You're assuming I'm begging the question, then? Let me try to re-explain.

I demonstrated at the beginning that CR = level depletes 1/16 the resources.
People claim that's a fault of 1v4 combats, and a 4v4 should work better.
I use the CR rules (where a monster with R and TR == 4 monsters with 1/4 R and 1/4 TR) to demonstrate that the math is the same.
You and I get into a confusing debate.

I actually did do this independently. Those monsters aren't designed so that they have 1/4 R and 1/4 TR, they are designed so they are the same CR encounter, and the math just so happens to work out such that they have 1/4 R and 1/4 TR, which proves my point that the number of combatants doesn't change whether or not the CR system works. It's simple enough to look like it's a circular derivation, but it's not (at least not anymore than proving 1 = 4*.25).


No, you are saying that, by claiming that the minions drain a quarter as many resources as the boss by virtue of being lower level. I'm saying that the resources drained will not go down at the same rate as the resources available.These are rules dictated by math. If the the resource depletion is anything but R/P and total resources are anything but TR/P, the math is borked and one encounter is stronger/weaker than the other. You have to make this assumption or the encounters aren't balanced.


Except the resource depletion is not solely a property of the enemy, which is my entire point. The players are spending resources every time they attack. If you're claiming that this is taken into account in the resource depletion stat, then that stat cannot vary so greatly based solely on the level of the enemy.It can entirely vary based on the enemy, for reasons you stated in your last post. If the enemy is a big monster, then the player will use abilities that deplete lots of their own resource in exchange for depleting lots of enemy resources. If the enemy is a minion, the player will use abilities that deplete very little of their own resource in exchange for depleting very little enemy resources without any waste.

For example: If enemy [1] removes X resources over the encounter and has Y resources of its own, then player [alpha] can use an ability that removes X of his own resources in order to remove Y resources from the monster. If the monster actually has Y + Z resources and removes X + W of the player's, then the player can remove X + W of his own resources to get rid of Y + Z of the monster. That is how spells and such must be set up, and if they aren't set up like that then they're overpowered or underpowered, and we're not concerned with that.

[eggynack]

Er... people already knew that a CR = level encounter is designed to drain 1/16 of player resources instead of 1/4? Well, it's at least news to me and I've been on D&D forums for a while, so if it's news to me then I'm sure it's news to at least someone else. In fact, BlasphemousSlug even admitted such a thing.
No, primarily because it doesn't. A single enemy against a party of four, with all of them at the same level, will just drain less than 1/4 of the party's resources. It won't hit the party by some exact 1/16 amount, because your measurement method leaves out a pretty huge number of factors.

You assume a lot of things that aren't true, like the idea that four party members will be equally capable of bringing their force to bear against a single enemy as that enemy is capable of bringing his force to bear against a member of the party, or the idea that resource depletion stays constant whether you're attacking a single target or a group. There are also more basic factors, like the fact that a party wizard is draining his own resources with every attack, while an enemy wizard only exists for one encounter, and can therefore ignore that resource expenditure.

These things all change the numbers in various complicated ways, and it all adds up to your numbers being flawed. The simpler answer, that the action economy causes one monster against four PC's to be a shaky proposition at best, is also a better answer.

[Just to Browse]

No, primarily because it doesn't. A single enemy against a party of four, with all of them at the same level, will just drain less than 1/4 of the party's resources. It won't hit the party by some exact 1/16 amount, because your measurement method leaves out a pretty huge number of factors. You keep saying "no it doesn't happen because it doesn't". That does not work--I showed you the math, you need to demonstrate that the math doesn't work or show some new math. And of course it's not exactly 1/16--try looking at the OP and doing ctrl-f for the term "on average".


You assume a lot of things that aren't true, like the idea that four party members will be equally capable of bringing their force to bear against a single enemy as that enemy is capable of bringing his force to bear against a member of the party, or the idea that resource depletion stays constant whether you're attacking a single target or a group. There are also more basic factors, like the fact that a party wizard is draining his own resources with every attack, while an enemy wizard only exists for one encounter, and can therefore ignore that resource expenditure. The first problem has already been discussed--if you're using limited resources, then your abilities need to do more to compensate for the lost resources. The second problem does not matter, because the players are supposed to use up all enemy resources, so it's actually a bad thing if the wizard can't use all his spells (and if you're going to metagame that hard, then the wizard will certainly be spending them on hella buffs. Please don't try to debate "the wizard might NOT use his spell slots!" because that also fails against the premises of the argument).


These things all change the numbers in various complicated ways, and it all adds up to your numbers being flawed. The simpler answer, that the action economy causes one monster against four PC's to be a shaky proposition at best, is also a better answer. Action Economy != More Resource Removal. I've done the math for this twice.

Complexity != Balance changes. If one side can get a benefit, then the other side can get a benefit.

[Bogardan_Mage]

Ah, I see. You're assuming I'm begging the question, then? Let me try to re-explain.

I demonstrated at the beginning that CR = level depletes 1/16 the resources.
People claim that's a fault of 1v4 combats, and a 4v4 should work better.
I use the CR rules (where a monster with R and TR == 4 monsters with 1/4 R and 1/4 TR) to demonstrate that the math is the same.
You and I get into a confusing debate.

I actually did do this independently. Those monsters aren't designed so that they have 1/4 R and 1/4 TR, they are designed so they are the same CR encounter, and the math just so happens to work out such that they have 1/4 R and 1/4 TR, which proves my point that the number of combatants doesn't change whether or not the CR system works. It's simple enough to look like it's a circular derivation, but it's not (at least not anymore than proving 1 = 4*.25).

These are rules dictated by math. If the the resource depletion is anything but R/P and total resources are anything but TR/P, the math is borked and one encounter is stronger/weaker than the other. You have to make this assumption or the encounters aren't balanced.

It can entirely vary based on the enemy, for reasons you stated in your last post. If the enemy is a big monster, then the player will use abilities that deplete lots of their own resource in exchange for depleting lots of enemy resources. If the enemy is a minion, the player will use abilities that deplete very little of their own resource in exchange for depleting very little enemy resources without any waste.

For example: If enemy [1] removes X resources over the encounter and has Y resources of its own, then player [alpha] can use an ability that removes X of his own resources in order to remove Y resources from the monster. If the monster actually has Y + Z resources and removes X + W of the player's, then the player can remove X + W of his own resources to get rid of Y + Z of the monster. That is how spells and such must be set up, and if they aren't set up like that then they're overpowered or underpowered, and we're not concerned with that.
If they are set up like that then they're unbalanced for the reasons you've demonstrated! I'd like to bring up the area of effect thing again, because your argument there was similar. The boss cannot do more damage to the party total than the party can do to the boss individually. You claim this is balanced, but it's not. If the resource draining functions on all enemies, and it's becoming increasingly clear that it must, then the problem doesn't exist. The players drain 1 resource each from all of their enemies (even though it's just one guy), and the boss drains 1 resource from each of his enemies (1 from each player). Net loss, 1 quarter of the total party's resources. It also works for the 1/4 minion scenario; they die a lot quicker but end up depleting 25% of the party's resources all the same.

I really think you'll find this is the solution. The CR system's problem is that it assumes every party member can engage every opponent at the same time.

[skyth]

Some things to consider...An NPC of equivalent level to the PC party will use 1/5 of the part resources, not 1/4.

Another thing is the disparate use of resources. In D&D, encounters can be defeated without all of their resources being used. One missed save or a crit could mean all HP depleted without any other resources being used. Combat does not last long enough to allow all spell slots/charges/arrows to be used by the enemy before their HP are depleted.

Take, for example, A fight of a level 3 party against a level 3 cleric with a wand of shield of faith. The enemy cleric will not use all his spell slots in the fight. Even taking into account some pre-combat buffing (Which assumes he has time and motivation to pre-buff), he will have some spell slots to heal himself after the fight along with some to use to heal himself after combat and to give himself flexibility. There is no way that he will use all 50 charges on the wand, not all his crossbow bolts. The potions of cure lights wounds he has on himself will heal himself slower than the party can do damage to him. There is almost no way that the party will reduce the resources of the cleric to 0 before overcoming the challenge that he poses. Even if he runs away, the PC's get the reward for overcoming the challenge that he poses.

[Thunndarr]

1. A level-appropriate encounter will consume 25% of party resources.
2. A level-appropriate encounter (for a party of 4) is equal in resources and resource-removing capability to one party member.

It really seems like people are not grasping why these two statements don't mesh. For the record, I think the game designers want #1 to be true, but they didn't do the math and figure out that #2 doesn't follow. I'll use a simple example.

Let's assume the party is 4x4hp fighters who can each do 1 dmg per round.

According to premise 2, their opposition would be one 4hp fighter, doing 1 dmg per round.

The fight would break down with the party doing a total of 4 dmg, killing the opponent fighter, and the party would take 1 damage out of their total 16hp, or 1/16th their resources.

To amend premise #2 to mesh with #1, the opponent needs to have a combination of resources and resource depletion 4 times what a single party member has.

In the example above, the 4x4hp 1dmg/round fighter party would find an appropriate challenge in a single 16 hp 1dmg/round fighter (4xresources), or 8 hp 2dmg/round (2x resources and 2x resource depletion), or 4hp 4dmg/round. (4x resource depletion.)

Assuming simultaneous combat, in each case the opposing fighter is able to kill 1 of the party's fighters, or 25% of resources.

[eggynack]

You keep saying "no it doesn't happen because it doesn't". That does not work--I showed you the math, you need to demonstrate that the math doesn't work or show some new math. And of course it's not exactly 1/16--try looking at the OP and doing ctrl-f for the term "on average".
I'm saying that the number is greater than 1/16, not that the number is distributed evenly around 1/16. I'm showing you a number of variables that would shift the resource expenditure upwards in general, which is about as good as I can do, and it'd be pretty stupid for me to present new math, given that I'm saying that mathing around all of the complexities of the game doesn't make all that much sense. I might consider running some for the 4 v 4 fight at some point though.



The first problem has already been discussed--if you're using limited resources, then your abilities need to do more to compensate for the lost resources.
How do you get your abilities to do more, despite your abilities being nigh on equivalent to the enemy's? Imagine it as a four on one pure fighter match, for the sake of simplicity. Three of the fighters are able to surround the enemy fighter, dealing their point of resource damage, but one of them just doesn't have the movement to pull it off. Meanwhile, given the two groups being within striking range of each other, the enemy fighter will always be able to hit the enemy for his one point of resource damage. Now the enemy fighter is dealing one point to the enemy's three in that round, and the party has no way to recoup that loss. By my way of looking at it, there's no real way that this situation could go the other way, with the party on the winning end of it, so it can only push the number up.



The second problem does not matter, because the players are supposed to use up all enemy resources, so it's actually a bad thing if the wizard can't use all his spells (and if you're going to metagame that hard, then the wizard will certainly be spending them on hella buffs. Please don't try to debate "the wizard might NOT use his spell slots!" because that also fails against the premises of the argument).
That's the thing of it though. The party doesn't need to use up the enemy's spell slots, and the enemy doesn't much care if he does. The enemy wizard will always be able to cast nothing but their highest level spells, which means that he'll be able to deplete the enemy's resources more efficiently than the party wizard can deplete his. The enemy wizard can also cast spells in every round, while a party wizard usually wouldn't want to do that. The same is true for expendable items. The enemy character can use all of his potions, scrolls, and use/day items in this one fight, under the assumption that this will be the last fight he ever has. The party has no such luxury.



Action Economy != More Resource Removal. I've done the math for this twice.
Have you proven that these lower power enemy actions are actually at 1/4 the resource removal, and that they have 1/4 the total resources? I don't think you have, and I don't think you could.



Complexity != Balance changes. If one side can get a benefit, then the other side can get a benefit.
This is absolutely untrue. There are advantages accessible to a single enemy fighting a party of four that the party has no access to. You can't just assume that contribution by each side's characters will remain symmetrical when the enemy for each side is completely different. For example, you asserted the existence of some kind of quadruple damage single target fireball, when such a thing clearly doesn't exist at third level spells. You can't just say that things should be a certain way, and that that clears up all issues, because things are what they are. We're playing a game where an area of effect ability against a group probably deals more resource damage than a single target ability against a single target will, and that's how it should be. Otherwise, what would be the point in casting AoE spells at all?

[Scow2]

The big problem here is an underlying math issue.

A CR=APL encounter isn't 25% of the party's resources: It's only 20%. The party should be near-exhausted after four CR 4 encounters, with a fifth likely to wipe them.

A lot of monsters are badly-CRed.


Most monsters have a way of handling multiple attackers, depending on role. Lurker-types (Such as rogues) can isolate-and-spike, partially negating its enemy's greater action economy (They have to spend rounds getting into position to strike), to deal 20% damage to a party's resources, and having a 50/50 chance of beating a lone party member with the survivor having most of their own resources depleted (Or none, with it being a case of "Rocket Tag"). Controllers can likewise isolate a party, and shift things to its favor long enough to drain 20% of a party's resources. Brutes can generally tough out a few rounds, and deal either herd-hitting attacks, or massive spike damage to a single target.

20% of a party's resources isn't "Kill one person, leave the others unscathed" - it's more "Drain a few HP and spell slots from everyone."

[Platymus Pus]

Make up your mind: are you arguing that a level 1 party gets slaughtered by EL 1 encounters because of the system's design, or are you arguing that a particular EL 1 encounter is too weak and isn't exactly the right challenge? Because those assertions do not work well together.

Both, because that is how the system is designed.

[Just to Browse]

I'm saying that the number is greater than 1/16, not that the number is distributed evenly around 1/16. I'm showing you a number of variables that would shift the resource expenditure upwards in general, which is about as good as I can do, and it'd be pretty stupid for me to present new math, given that I'm saying that mathing around all of the complexities of the game doesn't make all that much sense. I might consider running some for the 4 v 4 fight at some point though. So you're saying there is something special that affects all CR = level fights independent of how many combatants are on either side and what resources are available to them, and that factor is called "a number of variables"? This just does not work.


How do you get your abilities to do more, despite your abilities being nigh on equivalent to the enemy's? Imagine it as a four on one pure fighter match, for the sake of simplicity. Three of the fighters are able to surround the enemy fighter, dealing their point of resource damage, but one of them just doesn't have the movement to pull it off. Meanwhile, given the two groups being within striking range of each other, the enemy fighter will always be able to hit the enemy for his one point of resource damage. Now the enemy fighter is dealing one point to the enemy's three in that round, and the party has no way to recoup that loss. By my way of looking at it, there's no real way that this situation could go the other way, with the party on the winning end of it, so it can only push the number up.They're not doing more--there are four players on one side and the equivalent of one player on the other side. And yes, if we pre-suppose this is the standard unbalanced D&D 3.5 where every fighter can only move 30' and none of them are going to charge and literally none of them start closer than 30' from the enemy and they decide to go on turn one and none of them have a ranged attack and none of them have a maneuver... etc.

In that specific edge case, you will see a potential 1-damage deficit. But in some other cases the party will have surprise, or will garner a flanking bonus (because surrounding people is difficult), or will otherwise do something that makes swarm tactics viable, and now you have edge cases on either side of the spectrum which is why the simulation is taken on average. Heck, this could even happen for a swarm of 16 combatants whose total CR == player level, and they can't all swarm the players because reasons, so now the players have it slightly easier. The same problem actually affects both sides and is irrespective of CR!


That's the thing of it though. The party doesn't need to use up the enemy's spell slots, and the enemy doesn't much care if he does. The enemy wizard will always be able to cast nothing but their highest level spells, which means that he'll be able to deplete the enemy's resources more efficiently than the party wizard can deplete his. The enemy wizard can also cast spells in every round, while a party wizard usually wouldn't want to do that. The same is true for expendable items. The enemy character can use all of his potions, scrolls, and use/day items in this one fight, under the assumption that this will be the last fight he ever has. The party has no such luxury.Yes, the party will be using 1/4 of their per-day slots and 1/4 of their per-day items if this is supposed to be taxing 1/4 of their per-day resources. Having the enemy blow all their spells and per-day items isn't just supposed to happen, it's the determining basis by which encounters are designed. If the enemy wizard is not casting all his spells he is not using all his resources and the fight is too easy.


Have you proven that these lower power enemy actions are actually at 1/4 the resource removal, and that they have 1/4 the total resources? I don't think you have, and I don't think you could.Um, yes. In order for challenges to be equivalent, they should have the same number of resources and the same resource-removing capabilities. So four enemies need R/4 and TR/4. This is not some difficult proof.


This is absolutely untrue. There are advantages accessible to a single enemy fighting a party of four that the party has no access to. You can't just assume that contribution by each side's characters will remain symmetrical when the enemy for each side is completely different. For example, you asserted the existence of some kind of quadruple damage single target fireball, when such a thing clearly doesn't exist at third level spells. You can't just say that things should be a certain way, and that that clears up all issues, because things are what they are. We're playing a game where an area of effect ability against a group probably deals more resource damage than a single target ability against a single target will, and that's how it should be. Otherwise, what would be the point in casting AoE spells at all?Yes, I can, because like I said in the beginning of this post, this is not a debate about how D&D currently works. We already know D&D is broken--it is broken because HD inflation is terrible, it is broken because traps have CR's (what?), it is broken because wizards are better than fighters, it is broken because monsters are not appropriately CR'd. That is not the point. The point is that even if the game were balanced, even if we had abilities that were useful for all situations available as options, and even if we had an internal consistency that didn't make one target better than the other, the system still would not work.

And yes, I can assume that there are specific tactics that are good for the 1 in a 1v4 and good for the 4 in a 1v4, because if one side gets it then the game is broken unless the other side gets it. I don't understand how you think fireball is useless unless it gets to do the same exact damage as any single-target spell. That makes no sense.

[zlefin]

you keep using the word broken; but that's too strong a word for some of those things; though it's apropos for the others.

Also, if those other things were balanced, the system WOULD work; because it would have been designed with much higher levels of rigor; so they would have made sure it worked as intended, or changed the system itself to one that is accurate.

I still think your time would be better spent finding solutions than arguing about a mathematical abstraction which has plenty of issues of its own.

But if you really REALLY want to argue it; then i'm willing to go full logic on it.

[Scow2]

Yes, the party will be using 1/4 of their per-day slots and 1/4 of their per-day items if this is supposed to be taxing 1/4 of their per-day resources. Having the enemy blow all their spells and per-day items isn't just supposed to happen, it's the determining basis by which encounters are designed. If the enemy wizard is not casting all his spells he is not using all his resources and the fight is too easy.Actually, the number of spells a caster has do take the action economy into account. However, they give the caster more options to respond to the threat they face, and they can spend a spell every round without fear of having to conserve their resources. It's not "They only work if they cast all their spells", it's "They work if they cast a spell every round, instead of holding back."


Um, yes. In order for challenges to be equivalent, they should have the same number of resources and the same resource-removing capabilities. So four enemies need R/4 and TR/4. This is not some difficult proof.Fortunately, the challenges aren't equivalent. The deck is stacked against a CR = APL monster. It needs 1/4 the combined resources of the party to work, and a way to have its resources negate the drawbacks of its poor action economy.


Yes, I can, because like I said in the beginning of this post, this is not a debate about how D&D currently works. We already know D&D is broken--it is broken because HD inflation is terrible, it is broken because traps have CR's (what?), it is broken because wizards are better than fighters, it is broken because monsters are not appropriately CR'd. That is not the point. The point is that even if the game were balanced, even if we had abilities that were useful for all situations available as options, and even if we had an internal consistency that didn't make one target better than the other, the system still would not work.Traps have CR based on the threat they can pose to a party, based on difficulty on finding and disabling them. I'm not sure how you think that's nonsense - A CR 5 trap has a chance of absolutely wasting a level 1 party because they cannot find nor disarm it, and end up dying when they trigger it.

That aside - while trying to throw out the 'bad' in CR calculation, you're also throwing out the fiddly 'good' as well that makes it all largely balance out.


And yes, I can assume that there are specific tactics that are good for the 1 in a 1v4 and good for the 4 in a 1v4, because if one side gets it then the game is broken unless the other side gets it. I don't understand how you think fireball is useless unless it gets to do the same exact damage as any single-target spell. That makes no sense.PC/NPC asymmetry is not indicative of a broken game. If anything, this debacle shows that the symmetry is what breaks it.

[Brookshw]

I think the players won the arms race. I don't even glance at CR when I plan encounters, stat blocks are more informative on if the party would be challenged.

[Just to Browse]

zlefin: "If the system were balanced, it would be balanced" is not an argument I'm trying to make. Yes, if the designers realized their two base assumptions conflicted and fixed them, then their base assumptions would no longer conflict, but this thread is about the fact that those assumptions do conflict (which you seem to agree with me on? I'm not sure), and since people still remain unconvinced I'm going to spend time making sure to convince them before I throw yet another monkey wrench into the mix. If you think a thread would be better dedicated to writing up solutions, I invite you to start one and populate it with a few ideas, and I'll be sure to post.

Brookshw: Spoilered for size (couldn't find a smaller one in 30sec)
http://th03.deviantart.net/fs70/PRE/f/2011/288/7/a/i_know_that_feel_bro_by_rober_raik-d4cxn5a.png
Since around my third campaign session, I just stopped trying to use CRs and eyeballed my encounters. The reason I got thinking about this thread is because I wanted to know why my unbuffed CR=lvl encounters with NPCs always got stomped so fast, but monsters tended to do better.


Actually, the number of spells a caster has do take the action economy into account. However, they give the caster more options to respond to the threat they face, and they can spend a spell every round without fear of having to conserve their resources. It's not "They only work if they cast all their spells", it's "They work if they cast a spell every round, instead of holding back."So I must preface these responses once again with: This is not about a real D&D 3.x party, this is about a theoretically-balanced D&D party that follows the assumptions in the OP. This specific post was addressing someone who said casters can be OP because they spend all their daily resources in a fight and that's not fair while the player's can only spend 1/4. My response is that this is how the game is designed. Also, if a caster has excess spells, then he can spend them on buffs and resource replenishing outside of combat. That's the basis of spells like energy resistance and vigor.


Fortunately, the challenges aren't equivalent. The deck is stacked against a CR = APL monster. It needs 1/4 the combined resources of the party to work, and a way to have its resources negate the drawbacks of its poor action economy.Again, this would be valid in D&D 3.x as it currently is, but since we're talking about a theoretical version of D&D that actually follows the premises outlined in D&D (classes are equal, 1 CR X == 1 CR X - 2, etc.), this is not valid. We all know that a CR 4 encounter is generally a cakewalk for a level 4 party, but I'm making the case that this is not about action economy but about what you can do with that action economy. If a monster can take a standard action to make 4 attacks, it has the same action economy but is much more effective.


[snip]Those are minor examples about why D&D is broken. I'd rather we didn't spend time nitpicking them.


PC/NPC asymmetry is not indicative of a broken game. If anything, this debacle shows that the symmetry is what breaks it. This is not PC/NPC symmetry, this is Big Group/Small Group symmetry. If the players get an advantage in 1v4 always, then they will have the disadvantage in Army v 4 always, and if that's true then the game isn't very good but my point holds anyways.

[Waddacku]

Waddaku: Thank you, that's actually a great explanation for complexity. If you don't mind, I'd like to quote you in the OP.

By all means, do so if you feel it'd be useful.

[zlefin]

zlefin: "If the system were balanced, it would be balanced" is not an argument I'm trying to make. Yes, if the designers realized their two base assumptions conflicted and fixed them, then their base assumptions would no longer conflict, but this thread is about the fact that those assumptions do conflict (which you seem to agree with me on? I'm not sure), and since people still remain unconvinced I'm going to spend time making sure to convince them before I throw yet another monkey wrench into the mix. If you think a thread would be better dedicated to writing up solutions, I invite you to start one and populate it with a few ideas, and I'll be sure to post.


My position is this:
1. the original CR system has a number of flaws and issues; we know of them in detail.
2. the stuff you're pointing out isn't new information.
3. there are flaws in your model/arguments which makes it inaccurate and reduces its utility.

Its not that people are unconvinced that there are flaws in the system; it's that they are not convinced that your math is a good model either, and if you use inaccurate math to try to make a new system, then it will work poorly.
So people are mostly arguing over issue 3 with you at this point.

minor note: you changed the subtitle on the original post, but not the thread title.

[Just to Browse]

My position is this:
1. the original CR system has a number of flaws and issues; we know of them in detail.
2. the stuff you're pointing out isn't new information.
3. there are flaws in your model/arguments which makes it inaccurate and reduces its utility.

[snip]I addressed 1 and 2 already in my previous responses to you, so those are resolved.

I know you think there are flaws in the example, but so far your only your only concerns have been addressed as being designed to exist in the system (wizards go nova), do not change the result (Flanking, movement requirements, solid fog), or are specific to the current broken version of 3.x and not the theoretical balanced version that is assumed in the OP.


minor note: you changed the subtitle on the original post, but not the thread title.Are you sure? It looks like the title is changed for me... is it still the same for you?

[zlefin]

hmmm, upon further review; the thread title does appear to have changed properly; though oddly, the thread subtitle has not for most of the posts.

As for the other issues, I will have to review; though i'm more looking at fixing the system, perhaps I shall start a thread on that.

[Scow2]

hmmm, upon further review; the thread title does appear to have changed properly; though oddly, the thread subtitle has not for most of the posts.

That's because the title of any post is independent from the thread topic, but it's automatically "Re:(Thread topic name)" by default.

[skyth]

I know you think there are flaws in the example, but so far your only your only concerns have been addressed as being designed to exist in the system (wizards go nova), do not change the result (Flanking, movement requirements, solid fog), or are specific to the current broken version of 3.x and not the theoretical balanced version that is assumed in the OP.

No, the flaws with the hypothesis in the OP is two-fold. The first is that 1/5 of resources are used in an average encounter, not 1/4. The other is that you do not need to completely deplete the challenge's resources to overcome it. In fact, it is usually impossible to completely deplete their resources.

[Just to Browse]

No, the flaws with the hypothesis in the OP is two-fold. The first is that 1/5 of resources are used in an average encounter, not 1/4.

Thank you for pointing this out, because it brings up an interesting point. So I went looked in the DMG to confirm what I thought was true and found this passage, which holds both assumptions in it:


Challenging: Most encounters seriously threaten at least one member of the group in some way. These are challenging encounters, about equal in Encounter Level to the party level. The average adventuring group should be able to handle four challenging encounters before they run low on spells, hit points, and other resources. If an encounter doesn’t cost the PCs some significant portion of their resources, it’s not challenging.

This is where both assumptions are, and in it I find that both of us are wrong. A group of players needs to use up X amount of their daily resources, where X*5 >= 100%. So the range of resources used in a fight is between 1/4 and 1/5.

With this new information, we can still support the original point because 1/16 < 1/5 < 1/4. The way combat is set up still doesn't follow with the revised premise.


The other is that you do not need to completely deplete the challenge's resources to overcome it. In fact, it is usually impossible to completely deplete their resources.This is not a problem at all. If you have to remove fewer of an opponent's resources, then you're losing less of your own and thus the number is smaller than 1/16, which means the premise of encounter design still doesn't work.

[Scow2]

This is not a problem at all. If you have to remove fewer of an opponent's resources, then you're losing less of your own and thus the number is smaller than 1/16, which means the premise of encounter design still doesn't work.

Actually, most monsters are overloaded on resources... and the math works out to 1/5th, not 1/16th, even with the broken action economy against them.

[Just to Browse]

Actually, most monsters are overloaded on resources... and the math works out to 1/5th, not 1/16th, even with the broken action economy against them.


This is not about a real D&D 3.x party, this is about a theoretically-balanced D&D party that follows the assumptions in the OP.


A level-appropriate encounter (for a party of 4) is equal in resources and resource-removing capability to one party member.


All creatures and encounters with the same level and CR need to have similar total resources, and similar ability to remove resources--an example of this is the wizard who has fewer HP and lower weapon damage than the warrior, but can cast a limited number of defensive spells and has magical attacks.

No, the monsters are not overloaded, because if they have TR resources then each player does to, or if they have 4*TR then so do all the players.

[Scow2]

No, the monsters are not overloaded, because if they have TR resources then each player does to, or if they have 4*TR then so do all the players. Most player characters are also overloaded on resources: They're more likely to run out of HP before they run out of scrolls/spell slots/Wonderous Item uses/Wand Charges/feather tokens/specialized arrows/etc in any given combat, unless they enter combat in a depleted state.

[Just to Browse]

Most player characters are also overloaded on resources: They're more likely to run out of HP before they run out of scrolls/spell slots/Wonderous Item uses/Wand Charges/feather tokens/specialized arrows/etc in any given combat, unless they enter combat in a depleted state.

Scow2, I need you to read the following quoted section three times, and then read it another three times. It's really annoying to repeat myself over and over.


This is not about a real D&D 3.x party, this is about a theoretically-balanced D&D party that follows the assumptions in the OP.

Heck, I'll even bite on part of that: Some of those items are called Quest Items (tokens, enchanted arrows, special wondrous items) and I called out in this very thread that we're concerned with daily resources, which are determined by [Quest Resources] / [Resources per Quest]. So if the players stumble into a bad encounter and die with extra quest resources, that's OK because the system is designed that way.

[eggynack]

This is not about a real D&D 3.x party, this is about a theoretically-balanced D&D party that follows the assumptions in the OP.
Perhaps, but then the title might need to become something like, "The basic premise of theoretically balanced D&D that follows my assumption doesn't work," which is fine, I guess. It just doesn't tell us much.

[TuggyNE]

Perhaps, but then the title might need to become something like, "The basic premise of theoretically balanced D&D that follows my assumption doesn't work," which is fine, I guess. It just doesn't tell us much.

It actually does, sort of. It means (if true) that not only is CR plagued by the difficulty of determining a single number for any given monster, the variability of classes, and the quirks of situational abilities and tactics and environment, but also that even the most fundamental equations are wholly unsound and the system can only ever produce vaguely accurate results by pure coincidence, rather than, as some might assume, that the system can produce vaguely accurate results simply by compensating for those various factors. Put another way, not only is it incomplete in its factors, but it is wrong in the factors that it does include, and is essentially composed entirely of thoroughly-inaccurate fudging at every stage.

I haven't had the gumption to revisit the calculations since my last post, so I'm not sure that this is actually the case, but those are the implications if it is.

[Just to Browse]

Perhaps, but then the title might need to become something like, "The basic premise of theoretically balanced D&D that follows my assumption doesn't work," which is fine, I guess. It just doesn't tell us much.

The basic premise of theoretically balanced D&D is the same as the basic premise of regular D&D. The DMG spells out the basic premise, and since the basic premise fails in our case, then that assumption can be extended to any balanced version of the game.

EDIT: Tuggyne you're far better at saying these things than I am.

[TuggyNE]

EDIT: Tuggyne you're far better at saying these things than I am.

Why thank you! :smallsmile:

I've also gone back through the thread trying to find the post where you laid out the reason for assuming a CR APL-2 enemy does half damage and has half resources (and therefore that a CR APL-4 enemy does quarter damage and has quarter reserves). It seems to have been stated only in the OP, as part of assumption 3. Unfortunately, it does not actually follow; there is no need to assume that particular relation — no need to assume that splitting an encounter into minions must or should cause them to do equal or less total resource damage than the solo, or to have equal or less total resources. And it is that assumption that is in fact problematic for this thread's thesis.

There's another consideration I don't think has been properly brought up yet, and it is the second crucial point: characters (in general, in theory) expend resources not only when attacked by enemies, but when attacking, and thus enemies can force resource expenditure not merely when they attack a PC, but when they are themselves attacked or killed. Magic missile is the most obvious case of this, but Barbarian rage or Paladin smite are also valid examples.

Let's take a few examples, round by round, calculated in Excel. In every case, all characters are assumed to possess only single-target limited-use attacks, and to expend resources evenly upon being attacked. Both sides focus their fire. (The resource expenditure code is not perfect, and notably it can cause multiple enemies to be affected by the same PCs' efforts in rounds where one dies. This is not significant on the scale of these simulations.)
{table=head]Character|Spent|Depleted|Resources|1|2|3|4|5|6|7|8 |9|10|11
Party 1|0.25|0.5|16|14.75|13.5|12.25|11.25|10.25|9.25|8. 5|7.75|7.25|6.75|6.5
Party 2|0.25|0.5|16|15.75|15.5|15.25|15|14.75|14.5|14.25 |14|13.75|13.5|13.25
Party 3|0.25|0.5|16|15.75|15.5|15.25|15|14.75|14.5|14.25 |14|13.75|13.5|13.25
Party 4|0.25|0.5|16|15.75|15.5|15.25|15|14.75|14.5|14.25 |14|13.75|13.5|13.25
Party total|0|0||97%|94%|91%|88%|85%|82%|80%|78%|76%|74% |72%
Enemy 1|0.25|0.25|8|5.75|3.5|1.25|-1|-1.25|-1.5|-1.75|-2|-2.25|-2.5|-2.75
Enemy 2|0.25|0.25|8|7.75|7.5|7.25|5|2.75|0.5|-1.75|-2|-2.25|-2.5|-2.75
Enemy 3|0.25|0.25|8|7.75|7.5|7.25|7|6.75|6.5|4.25|2|-0.25|-0.5|-0.75
Enemy 4|0.25|0.25|8|7.75|7.5|7.25|7|6.75|6.5|6.25|6|3.75 |1.5|-0.75[/table]

Very basic. Also, very slow; ~11 rounds to take the foes down. Within a smallish margin of the desired (75-80%) resources left.

{table=head]Character|Spent|Depleted|Resources|1|2|3|4|5
Party 1|0.5|1|16|13.5|11.5|10|9|8.5
Party 2|0.5|1|16|15.5|15|14.5|14|13.5
Party 3|0.5|1|16|15.5|15|14.5|14|13.5
Party 4|0.5|1|16|15.5|15|14.5|14|13.5
Party total|0|0||94%|88%|84%|80%|77%
Enemy 1|0.5|0.5|8|3.5|-1|-1.5|-2|-2.5
Enemy 2|0.5|0.5|8|7.5|3|-1.5|-2|-2.5
Enemy 3|0.5|0.5|8|7.5|7|2.5|-2|-2.5
Enemy 4|0.5|0.5|8|7.5|7|6.5|2|-2.5[/table]

Party resources at the end are even closer to the right range, and things go faster. Mind you, I did nothing but multiply all the spent and depleted numbers by 2.

{table=head]Character|Spent|Depleted|Resources|1|2|3|4|5
Party 1|0.05|1.25|16|13.95|12.4|11.35|10.8|10.75
Party 2|0.95|1|16|15.05|14.1|13.15|12.2|11.25
Party 3|0.15|1.25|16|15.85|15.7|15.55|15.4|15.25
Party 4|0.85|0.5|16|15.15|14.3|13.45|12.6|11.75
Party total|0|0||94%|88%|84%|80%|77%
Enemy 1|0.5|0.5|8|3.5|-1|-1.5|-2|-2.5
Enemy 2|0.5|0.5|8|7.5|3|-1.5|-2|-2.5
Enemy 3|0.5|0.5|8|7.5|7|2.5|-2|-2.5
Enemy 4|0.5|0.5|8|7.5|7|6.5|2|-2.5[/table]

Once more with feeling. This time, the PCs spend different amounts on attacking, and do different amounts of "damage", but because the total spent/depleted numbers are the same, the outcome is identical.

Two major things still to do: work out a model that semi-sort-of handles multi-target attacks, the most obvious unknown quantity; compare the numbers required for (8) CR APL-6 enemies to (the 4) CR APL-4 enemies, and cross-reference with at least one more data point to establish the scaling rate. But that'll probably have to wait until tomorrow.

[zlefin]

why do you use a constant for resource depletion?

consider the following character:
has 7 resource.
abilities: once a day, deplete 4 resource on one target. at will, deplete 1 resource on one target.


by my count, if 4 of those character face 4 encounters of 1 each of that character the results go something like this:
all fights last one round. The party uses 3 at wills and one daily to kill the monster in one round, dealing 7 resource in damage. The monster uses its one daily to hit the party in that round, costing the party 4 resource.

after 4 fights, the party would have lost 16 out of its 28 resource.


Now, this does not address your issue directly; as the daily power should be considered a resource in itself, and it is not done so here.
However this scenario does seem interesting to me, your comment upon it?


also, looking at your page 1 argument; there are different conclusions possible than the one you make.
The simplest is that premise 2 is false, and is also not a core premise required for the system to work. Rather, the CR system is designed to satisfy 1, and is produced in an ad hoc fashion so that it does so; they guessed premise 2, but it's an ancillary premise and as such, it being wrong doesn't prevent you from designing good encounters in general.

[Just to Browse]

I would consider having access to a daily power that removes 4 resources as a resource in itself, and you could equate the character (in this world) to one with 11 resources (7 + 4) and an at-will that removes 1/turn. Also, the players would probably be better off using two dailies per fight to remove the monster (but that causes excess wasted damage, which is a problem entirely in its own right) as it would cut down on the number of turns monsters would get against them.

I had the pleasure of looking (1) and (2) up in the DMG, so they can't be false. EL == Party Level for a fight that removes 1/5 to 1/4 of party daily resources. I think the best way to tackle the problem is to determine which of those two premises is more important (#2 IMO) and base the encounter rules around that.

[skyth]

And you just pointed out another problem with the model...That you overcommit resources to a problem. To heal 4 points of damage with a cure light wounds spell, you are likely to use resources for no benefit (The extra 9-13 it could have healed).

Same with offensively...Your fireball does 30 points of damage. The monster only had 10 hp left. You had to overuse resources to accomplish the task.

Btw, the DMG specifies that a challenging encounter should us 20% of party resources, not 20-25% like you keep on saying.

[zlefin]

I would consider having access to a daily power that removes 4 resources as a resource in itself, and you could equate the character (in this world) to one with 11 resources (7 + 4) and an at-will that removes 1/turn. Also, the players would probably be better off using two dailies per fight to remove the monster (but that causes excess wasted damage, which is a problem entirely in its own right) as it would cut down on the number of turns monsters would get against them.

I had the pleasure of looking (1) and (2) up in the DMG, so they can't be false. EL == Party Level for a fight that removes 1/5 to 1/4 of party daily resources. I think the best way to tackle the problem is to determine which of those two premises is more important (#2 IMO) and base the encounter rules around that.

when I look at the math of saying they have 11 resources and just the at-will, it yields different results for the amount of resources they have left.

For my example, let's replace resources with HP. The party starts with 28 total; and 4 daily uses (one per person).
After 4 encounters, the party has 12 hp, and 0 daily uses. That looks to me like a small amount of resources. Do you get different results for remaining hp/dailies?

if you wish to use an alternate combat strategy, please spell out its round by round choices and effects.

[NichG]

It seems to me that if you want to keep with this simplified model but also take into account some of the factors that people have brought up, the easiest way to do it is to institute a per-round 'cost to fight' independent of what you're facing (but realistically, dependent on class).

This models the wizard/cleric/druid expending spell-slots to kill the enemy, buff, etc.

I still think this model is too far from the realities of actual monster design to be useful for implementing any sort of 'fixes' though. I certainly can't look at, e.g., a beholder and say what its 'resource count' is as an enemy or a PC, much less say what the 'average resource count' should be for 'the average Lv7 wizard' or 'the average Lv7 rogue'.

Perhaps rather than trying to design encounters in isolation that satisfy a mathematical goal about resource depletion, it would be better to devise a way that lets you basically measure resource depletion in progress for a given party, so you can basically adjust future encounters dynamically to take that into account?

Basically, a measure of the relative value of 1hp to a Lv3 spell slot to a 500gp item to other stuff in the game that's expendable.

[eggynack]

Nifty stuff.
We may have just hit the point where you can just kinda arbitrarily argue both sides of the issue. I mean, yeah, that's basically about the point I was making. If you assume that this encounter with several lower level enemies is going to consume the same resources as an encounter with one equal level enemy, well of course it's going to do that. I get that you can step out of the actual numbers of the game to some extent when you're considering equal CR opponents, but you can't just draw broad comparisons between two enemies of different levels without proving that they're actually part of the system. I can't even be at all sure that the 1/4 resource figure was intended as a part of the game, so it can't be considered a basic premise.

[Just to Browse]

And you just pointed out another problem with the model...That you overcommit resources to a problem. To heal 4 points of damage with a cure light wounds spell, you are likely to use resources for no benefit (The extra 9-13 it could have healed).This is a problem that occurs for both sides (and is even compensated by the <25% resource-use concept), so it doesn't change the CR problem.


Btw, the DMG specifies that a challenging encounter should us 20% of party resources, not 20-25% like you keep on saying.Let's play a game I like to call Find That Page Number! I found my reference, now you find yours.

On the topic of weak opponents: The reason you have to assume that opponents of 1/2 the strength have 1/2 the RT and 1/2 the R is that encounters don't make sense otherwise. If opponents that are half as strong (and show up twice as often) somehow get more of either, then they are a stronger encounter. This isn't even begging the question--you just cannot have minions be stronger or weaker the same way boss monsters can't be stronger or weaker. If you do, you're ignoring the most important premise of the exercise.

IMPORTANT THINGS NOW


when I look at the math of saying they have 11 resources and just the at-will, it yields different results for the amount of resources they have left.

For my example, let's replace resources with HP. The party starts with 28 total; and 4 daily uses (one per person).
After 4 encounters, the party has 12 hp, and 0 daily uses. That looks to me like a small amount of resources. Do you get different results for remaining hp/dailies?All right, this is definitely an interesting idea.

Let's look at how this differs from the basic example:
The monster has HP such that unless the players have an ability removing 4 RT at once, there is no way to not have excess damage in the encounter.
The players have this daily ability--whose weight in resources in unknown (I guessed 4).
Combats aren't lasting 4 rounds, and couldn't even last 4 rounds if the monster had 11 HP and the players were wailing on it.

After doing a lot of math poorly, I've come to the following conclusions:

Wasted Attacks
From a party's perspective, these don't matter at all. When extra players are hitting a dude it really doesn't change total resources lost. If attacks are wasted for the monster, it's a huge deal though. 4 wasted player attacks = 1 monster wasted attack, effectively, so I figure these can balance out over the course of the encounters.

How long combat lasts
There should be an "ideal combat length" where everyone can only remove R and all combatants have RT = P*R*L, but I don't know how to apply this without somehow calculating daily costs... which is where my problem arises.

Calculating Daily Costs
It appears that daily costs cannot be balanced between players and monsters equally--they need to cost more for monsters than players because the players have the power of P*R (instead of R) per-round removal, and P*RT (instead of RT) daily resources.

Example: The enemy deals R and has RT, same as each player. For a player, dealing 4R is like spending 3 rounds attacking. In each round, that players would lose R/P which means the cost of using a 4R ability in the standard party should only be 3/4 RT. This is a reasonably low number and makes sense in the paradigm of 1/16-total-resource-loss for a fight that we previously calculated up top.

However, from the monster's perspective, a daily ability that deals 4R is like getting 3 rounds to attack which means the monster circumvents P*R*3 damage, so it should cost him as much. For the standard example, this means the cost is 12, or 16x more expensive than for the player.

This means dailies do change the fundamental workings of combat, and there may be a way to balance CR around them. I apologize for wasting so many people's time instead of doing this myself (thank you zlefin). Now I have no doubt that no one at WotC thought this way (considering the incredible minority that get any dailies at all), but the fact that CR can work if you give monsters the appropriate number of limited abilities means that the DMG's 2 central premises can somehow be balanced if you include a third premise regarding the amount of resources consumed by limited-use-abilities.

So if we just assume that there is some number and magnitude of per-day abilities that make CR=level opponents worth 20-25% of the player's resources (it's more than a 4R attack, I can tell you that), the challenge is now to calculate that. And god I don't have a clue where to start.

[eggynack]

On the topic of weak opponents: The reason you have to assume that opponents of 1/2 the strength have 1/2 the RT and 1/2 the R is that encounters don't make sense otherwise. If opponents that are half as strong (and show up twice as often) somehow get more of either, then they are a stronger encounter. This isn't even begging the question--you just cannot have minions be stronger or weaker the same way boss monsters can't be stronger or weaker. If you do, you're ignoring the most important premise of the exercise.
Lemme see if I can explain this one. A 4 v 4 encounter, where the CR of the encounter is equal to the party's level, has its constituent monsters at a certain level of power. You might want that power level to be exactly 1/4, but unlike in the other case, where you can just claim aberrant data in both directions, here you actually have to prove it. For example, four CR 6 enemies will make a CR 10 encounter, if I'm not mistaken. If I am, just replace the numbers as you wish. You have to prove that a CR 6 enemy is about 1/4 as powerful as a CR 10 enemy. That's basically the whole thing in a nutshell.

The complicated part of it is, here's a place where the game's actual balance issues become problematic for your premise. I've gotta figure that a 10th level monk probably won't be four times as powerful as a 6th level monk, due to monks having a pretty low power growth rate. You can pretty much do the same for other melee classes, especially when the low level character has just finished his feat chain, and is thus acquiring variety rather than scale. On the other end of it, you have a wizard. Those guys have something like exponential growth, so it's very possible that a 10th level wizard will be more than four times as powerful as a 6th level wizard. We're talking two spell levels. The fact that this aspect of imbalance, the linear growth of fighters vs. the exponential growth of wizards, is a likely premise of the game, means that I'm not sure you can just ignore it.

Anyway, the point I'm trying to make is that you're saying that each member of that enemy party will be a fourth as powerful as the single big enemy. I don't think that you're correct, so it's a thing you really need to prove. You can say that you'd like the game to be like that, or that the game should be like that, but your assertions partially hinge on the idea that the game is like that, and I disagree.

[TuggyNE]

It seems to me that if you want to keep with this simplified model but also take into account some of the factors that people have brought up, the easiest way to do it is to institute a per-round 'cost to fight' independent of what you're facing (but realistically, dependent on class).

This models the wizard/cleric/druid expending spell-slots to kill the enemy, buff, etc.

So basically my "(Resources) Spent" column in the tables above.


Perhaps rather than trying to design encounters in isolation that satisfy a mathematical goal about resource depletion, it would be better to devise a way that lets you basically measure resource depletion in progress for a given party, so you can basically adjust future encounters dynamically to take that into account?

Basically, a measure of the relative value of 1hp to a Lv3 spell slot to a 500gp item to other stuff in the game that's expendable.

This might be rather intriguing.


If you assume that this encounter with several lower level enemies is going to consume the same resources as an encounter with one equal level enemy, well of course it's going to do that.

The crucial thing is really that "1/4 overall strength" is not the same as "resource depletion/spent/available are each individually 1/4", and there is no reason to assume this equivalence at all.


On the topic of weak opponents: The reason you have to assume that opponents of 1/2 the strength have 1/2 the RT and 1/2 the R is that encounters don't make sense otherwise. If opponents that are half as strong (and show up twice as often) somehow get more of either, then they are a stronger encounter. This isn't even begging the question--you just cannot have minions be stronger or weaker the same way boss monsters can't be stronger or weaker. If you do, you're ignoring the most important premise of the exercise.

I'm still not following. Mathematically, about all we know is that two monsters of level X-2 consume (approximately) the same resources from the PCs as one monster of level X, and have approximately the same chance of killing one or more PCs. What we do not know is whether this halving applies to all their stats individually, and in fact the nature of stat synergy very strongly argues that it should not: the strength of a monster is much closer to the product of all its stats, not the sum. That is, doubling or halving offensive resource depletion, on its own, can result in a doubling or halving of resources consumed!

In other words, I think you're layering an assumption onto the system that the system itself actually relies on not having, and then criticizing the system for having it. This assumption is that all stats (or at any rate resource depletion and resource pool) scale linearly with overall strength. Instead, ideal monsters scale those stats logarithmically.

[skyth]

This is a problem that occurs for both sides (and is even compensated by the <25% resource-use concept), so it doesn't change the CR problem.

Wrong. This is a problem that does not affect the monsters at all. They will not heal unless low on hit points. Also, theoretically they will not kill a PC. (They may drop one to negative HP, but the party will use resources to restore that PC). Also considering the attrition nature of the system, combat loss groupings should be taking place.

(CLG is a term from Battletech which basically means that attrition damage will roughly equally damage each individual unit until at the end, all the units drop in rapid succession).

Only the PC's will likely over-heal or over-damage. Dropping a monster to negative HP is different from dropping a PC to negative HP.




Let's play a game I like to call Find That Page Number! I found my reference, now you find yours.



'An encounter with an Encounter Level (EL) equal to the PCs' level is one that should expend about 20% of their resources' (DMG Pg 49)

[NichG]

So basically my "(Resources) Spent" column in the tables above.


Ah, thats what that meant! Okay, yes, the same as your tables then. I didn't quite get what you were doing the first time through.

[zlefin]

there's more factors than just dailies. PCs in particular have a budget for consumables. Consumables vary somewhat in their use patterns; but of interest here would be high power consumables which generate a high effect on a per action basis but are costly in gold. These might not ever be used in the course of a day of ordinary fights against ordinary encounters; they're saved for things like boss fights or when you get in trouble. Since the monsters tend to be the ones in trouble, they will be using any of these that are suitable at a much higher frequency than players will.

As to fixing the CR system, I have a variety of thoughts and ways to do that; though I'd want to review what current fixes are out there.
But the first steps of a fix are to more deeply analyze how the current system works.
One basic test series of tests would compare a level X party vs a single level Y character; to see what kind of drains occur in practice (there is of course huge variance based on which classes and builds are used).
I suspect that for a party of level X, a character of level X+1 is more likely to be around the intended level of an appropriate ordinary encounter than X is; testing to see if that is so, and how other relative levels fare, should provide useful info.

Ideally you want a large database of played encounters to mine.
If one really wanted to, one could look through all the games on this and other boards and make note of all encounters and their outcomes; but I'm certainly not bored enough to do that :P

[Big Fau]

Challenging: Most encounters seriously threaten at least one member of the group in some way. These are challenging encounters, about equal in Encounter Level to the party level. The average adventuring group should be able to handle four challenging encounters before they run low on spells, hit points, and other resources. If an encounter doesn’t cost the PCs some significant portion of their resources, it’s not challenging.

If the party's Wizard stops an enemy from outright murdering the party's Fighter with a single spell, the encounter did not seriously threaten the Wizard in the least but challenged the Fighter. That's encounter #1 for the day. Encounters 2, 3, and 4 should focus on other members of the party, such as the Rogue or the fore-mentioned Wizard. If those encounters challenge the relevant party member but not the rest then job well done, those encounters were appropriate CRs for the party. If not, then either bad luck happened (one way or another) or the challenge wasn't appropriately CR'ed.

According to the DMG, this is how it was supposed to work. The action economy and dice often don't allow this to happen. An encounter that seriously threatens all X of the party members isn't supposed to be 1/4 of the encounters they are supposed to face that day, it's 4/4 of the encounters they should face. An encounter that only slightly challenges all 4 members of the party is 1/4, possibly 1/3 or 1/5.

Sadly the CR system and Spellcasting are so screwed up that adjudicating what the appropriate challenge would be damn-near impossible. What a Wizard finds challenging would be utterly unbeatable by a Fighter or Rogue, but a Cleric might stand a better chance. What the Fighter and Rogue find challenging would usually be a pushover for a Wizard, or possibly be a Wizard.

Optimization is also a crucial factor. The S&B Fighter and Fireball Wizard are going to struggle if you throw Druids with Greenbound Summoning or DMM Persist Clerics at them, while the Shadowcraft Mage Wizard is going to walk all over your horde of zombies.

[TuggyNE]

Ah, thats what that meant! Okay, yes, the same as your tables then. I didn't quite get what you were doing the first time through.

Yeah, I was in a bit of a hurry last night, so wasn't as clear as I could have been. :smallsigh:

However! Here's another table, with sundry additions and corrections; most notably, PC overkill is assumed to be nil (rather than maximal; neither is quite right, but nil is closer), AoEs are taken into account in an idealized sort of way, and healing is accounted for by way of lossy resource transferral. Most happily of all, the result is precisely the 80% figure one would wish, to within experimental error (various possible minor off-by-one errors in the calculation and the inevitable messiness of actual combats).
{table=head]Character|Resources Spent|Target Depletion|AoE|Heals|Resource Pool|1|2|3|4|5
Party 1|0.1|1.5|FALSE|FALSE|16|14.3|13.1|12.4|12.2|12.5
Party 2|1|0.5|TRUE|FALSE|16|15|14|13|12|11
Party 3|0.1|1.75|FALSE|FALSE|16|15.9|15.8|15.7|15.6|15.5
Party 4|0.8|0.25|FALSE|TRUE|16|15.2|14.4|13.6|12.8|12
Party total|0|0||||94%|90%|85%|82%|80%
Enemy 1|0.5|0.5|||8|3.5|0|0|0|0
Enemy 2|0.5|0.5|||8|7|5|0|0|0
Enemy 3|0.5|0.5|||8|7|6|5.5|0|0
Enemy 4|0.5|0.5|||8|7|6|5|5|0[/table]

[Just to Browse]

eggynack: I've been saying the same thing at you over and over, so this is the last time because I seriously hate wasting time and space. Encounters of CR = level must be designed so that they are all the same difficulty. If your set-up, whether it has harder terrain, more dangerous minions, stronger single monsters, or random lightning bolts, makes the encounter use up more resources for 1 party or make 1 party remove fewer resources then it is not the same difficulty. Minion fights need to be just as draining as equal-difficulty single monster fights because that's how difficulty works.


Wrong. This is a problem that does not affect the monsters at all. They will not heal unless low on hit points. Also, theoretically they will not kill a PC. (They may drop one to negative HP, but the party will use resources to restore that PC). Also considering the attrition nature of the system, combat loss groupings should be taking place.Spells that grant regeneration and healing over time won't overheal? Monsters with enormous healing powers do not exist? Monsters that cast true strike aren't real? Players with sacrificial minions do not exist? Hmmmmmm.... And if we are talking D&D 3.x here, dropping a PC to -X HP is X - 1 more damage than necessary from keeping them out of the fight because healing in combat is a Bad Idea. Even if it were, then an intelligently-played monster would just kill the PC to stop it from standing up and multiplying the PC action economy by 133%.


'An encounter with an Encounter Level (EL) equal to the PCs' level is one that should expend about 20% of their resources' (DMG Pg 49)Thank goodness for citations! 1/5 < 1/16.


there's more factors than just dailies. PCs in particular have a budget for consumables. Consumables vary somewhat in their use patterns; but of interest here would be high power consumables which generate a high effect on a per action basis but are costly in gold. These might not ever be used in the course of a day of ordinary fights against ordinary encounters; they're saved for things like boss fights or when you get in trouble. Since the monsters tend to be the ones in trouble, they will be using any of these that are suitable at a much higher frequency than players will.Consumables are limited resources and are budgeted over the course of quests, which is composed of days and/or encounters, so the two equate. If you are using the same currency system for consumables as you are for out-of-combat resources (living places) and permanent boosts (+1 weapons), your system is borked already. Conservation of consumables runs prey to the same problem as per-day, just over a longer time period. Whatever the solution for dailies, there is a similar solution for consumables.


[snip]I think that if you're fixing the CR system, you need to work from the ground up and just redesign your challenges, especially because (as people have said over and over) there are a lot of things that can happen in a fight, and it's easier to write challenges for cut-and-dry encounters, then add factors that make it harder/easier, than it is to figure out what the current broken easiness and hardness things are.

[eggynack]

Thank goodness for citations! 1/5 < 1/16.
Indeed, 1/5 is less than 1/16, but it is also about equal to TuggyNE's numbers. So, y'know, the game works perfectly all the time forever, with absolutely no room for error. Huzzah indeed.

edit:
eggynack: I've been saying the same thing at you over and over, so this is the last time because I seriously hate wasting time and space. Encounters of CR = level must be designed so that they are all the same difficulty. If your set-up, whether it has harder terrain, more dangerous minions, stronger single monsters, or random lightning bolts, makes the encounter use up more resources for 1 party or make 1 party remove fewer resources then it is not the same difficulty. Minion fights need to be just as draining as equal-difficulty single monster fights because that's how difficulty works.
But they're not designed that way. You're arguing for premises that you want to be true, while ignoring the premises that actually exist in the game. Minion fights aren't just as draining as fights against one guy, because the action economy exists, and is a relevant factor. The basic premise of encounter design that you're espousing isn't one that's a fundamental aspect of the game, so you can't just assume it and move on. At some point, you have to look at the things you're saying, and assess whether they do anything to reflect real game states. I don't think they do, if they ever did.

[Just to Browse]

Indeed, 1/5 is less than 1/16, but it is also about equal to TuggyNE's numbers. So, y'know, the game works perfectly all the time forever, with absolutely no room for error. Huzzah indeed.

Those tables assume very specific resource costs and target depletion. So if the system is based around using your own resources and an enemy's resources in that specific manner (which it cannot be, because different classes are supposed to have different resource management schemes) then it would make sense. Heck, it appears the enemies are designed to act differently (with very different ratios) from the players, which violates premise #2. But I'm kind of OK with that at this point.

[eggynack]

Those tables assume very specific resource costs and target depletion. So if the system is based around using your own resources and an enemy's resources in that specific manner (which it cannot be, because different classes are supposed to have different resource management schemes) then it would make sense. Heck, it appears the enemies are designed to act differently (with very different ratios) from the players, which violates premise #2. But I'm kind of OK with that at this point.
There're a bunch of different versions of the thing. I think that the enemies are acting differently than the players at some kinda ratio, such that they're worse off than the higher level players. In any case, the point of his tables likely has far less to do with specific depletion values, and far more to do with the fact that the PC resources are being depleted at all. An enemy wizard casting four top level spells during an encounter hasn't pushed himself anywhere closer to death, and neither has a party wizard. However, the wizard has expended long term daily resources, which complicates things. That logic holds true for expendables of all kinds.

I've gotta say though, this whole premise of wizards and fighters being balanced throws things off in some odd ways. Like, which wizard is the fighter balanced against? Is he balanced against the wizard at the start of the day, at the end of the day, or neither because a core premise of the game is that wizards are more powerful than fighters later on? The reason I'm mentioning that is that your numbers seem to universally be geared towards beginning of the day fights. If the wizard is starting off the day equal to the fighter, then in later fights he should be reducing enemy's resources by less than one/turn. If he's hitting equality with the fighter later in the day, then a given enemy wizard's ability to deplete party resources should be greater than that of any given party member who is not a wizard, and should be so at every point in the day, because enemy wizards tend to enter fights fresh. It's an odd thing.

[TuggyNE]

Those tables assume very specific resource costs and target depletion.

They... actually do not. They require only one basic constraint, I think, and that is that a normal 4x4 encounter is composed of enemies with half the total resource depletion and half the total resource pool that the party has, total. (Even AoE and healing abilities do not change this math, apparently, at least not enough to be obviously wrong.) This constraint, along with the insight NichG and I came up with independently that attacking uses resources just like being attacked does, is the highlight of the exercise.

If you like, I can run a bunch more examples of this, varying the pool sizes between enemies, changing up their attacks, and so on. Maybe adding enemy AoEs, daily abilities, and healing. (Some of that would be tricky, but eh.)


So if the system is based around using your own resources and an enemy's resources in that specific manner (which it cannot be, because different classes are supposed to have different resource management schemes) then it would make sense. Heck, it appears the enemies are designed to act differently (with very different ratios) from the players, which violates premise #2. But I'm kind of OK with that at this point.

Initially I stuffed all the classes in as just "spends 0.5 per turn, deals 0.5 to enemies" and had the enemies uniformly at 0.25/0.25. That worked fine. Then I shifted it so that different characters had different resource expenditures (0.05, 0.1, 0.95, 0.85, 0.15, etc etc etc, keeping only the total the same), as well as changing their resource depletion damage, and that also worked fine. So neither of these problems applies to the model; all that is required is to assume that players and monsters both advance at the rate of * Sqrt(2) in resources and resource depletion every 2 levels. (This is the same as the constraint mentioned earlier, just rephrased.)

[skyth]

Spells that grant regeneration and healing over time won't overheal?


In an encounter, if a monster has these going, likely it will be after they've already been damaged and won't keep up with damage being dealt to it and thus it won't overheal. For a PC party, they exist after the fight and thus the resource depletion is more.



Monsters with enormous healing powers do not exist? Monsters that cast true strike aren't real? Players with sacrificial minions do not exist?

All irrelevant as the enormous healing powers won't over-heal in combat. I'm not sure what you're getting at via true strike.

Of course, make up your mind...Are we talking about generalities or specific situations? You constantly talk about this being a general situation, but now bring up very specific, rarely encountered things to argue against something? Can't have your cake and eat it too.


Hmmmmmm.... And if we are talking D&D 3.x here, dropping a PC to -X HP is X - 1 more damage than necessary from keeping them out of the fight because healing in combat is a Bad Idea. Even if it were, then an intelligently-played monster would just kill the PC to stop it from standing up and multiplying the PC action economy by 133%.

In combat killing downed enemies generally doesn't happen. But even if that happens, that makes the point even better. It takes more resources to revive a PC from the dead than to just cure them. This is not an issue with the encounter. There is absolutely NO difference for an encounter to be dead or at -1 HP.

And the thing you are forgetting, with a PC, dropping them to x-1 negative uses up more party resources than dropping them to -x hp. After the fight, more healing is needed to bring them back up. This is not true of the encounter. The difference if a monster is brought to -1 and -9 HP is a lot different than a PC being brought to -1 or -9 HP.

PC's have LONG TERM resource issues. Encounters do not. This leads to encounters being able to nova and thus deplete PC resources better, and it forces PC's to over-use resources in an inefficient way. This is where your entire argument breaks down.

[Eldariel]

PC's have LONG TERM resource issues. Encounters do not. This leads to encounters being able to nova and thus deplete PC resources better, and it forces PC's to over-use resources in an inefficient way. This is where your entire argument breaks down.

Depends on the type of enemy to be honest. Some recurring enemies will run into the same problem as the PCs, especially if the PCs force extended defense of the same location, employ mooks of their own or whatever.

[eggynack]

In an encounter, if a monster has these going, likely it will be after they've already been damaged and won't keep up with damage being dealt to it and thus it won't overheal. For a PC party, they exist after the fight and thus the resource depletion is more.
If you factor regeneration into the power level of the enemy, which you should, any encounter round in which the enemy isn't healing is one in which he is effectively overhealing. It makes no sense to only count rounds where the enemy is healing one HP, and not count rounds where the enemy is healing zero HP.

[ahenobarbi]

CR system has numerous problems of whatever system tries to describe complex reality with a single number. You are right that if one side gets 1/4th of defensive & offensive power of the other it will manage to "spend" much less than 1/4th of the other side resources. That's why generally monsters that have both CR and ECL CR is much lower.

[skyth]

Depends on the type of enemy to be honest. Some recurring enemies will run into the same problem as the PCs, especially if the PCs force extended defense of the same location, employ mooks of their own or whatever.

First off, pc's using mooks is a very rare situation. Second, a reoccurring villian will not have to face the pc's multiple times in fhe same day. On different days they don't have to look to replace consumables...the dm just provides

[Eldariel]

First off, pc's using mooks is a very rare situation. Second, a reoccurring villian will not have to face the pc's multiple times in fhe same day. On different days they don't have to look to replace consumables...the dm just provides

Depends. D&D is designed in a way that midlevel games involve a lot of minions. Animate Dead, Planar Binding, Simulacrum, Leadership, Craft Construct, hirelings, there are plenty of simple, obvious ways of generating extra hands with a little effort. Some cost more than others. OD&D dungeons often involved using hirelings especially on low levels to decrease the chance of death.

Does every group make heavy use of underlings? Probably not. Do most use make heavy use of underlings? Why, I don't think either of us is qualified to comment on this. Your comment sounds very much like "This is how I play", which is fine but you gave it as an answer to a general issue with a comment that's not generalizable. Which makes it quite useless for the scope of the discussion here.


And again, why would a DM provide his enemies with effectively limitless resources? How is the Last Alliance assembled by the PCs supposed to force Sauron out of his tower if not by depleting his armies and sieging his fortress first? Sauron can't have infinite resources or the whole system makes no sense.

And in another world, why could Irenicus resupply his consumables and spell slots if PCs have specifically dimension locked the whole continent and he's barely escaped in his hiding place? Why shouldn't he be forced to make do with what he's got, as long as they're not giving him a moment to rest?

You seem to have your mind fixed on a particular playstyle even though the game is built for and is capable of supporting countless different ways of playing. As such, I find it hard to apply your statements very broadly. Sorry.

[eggynack]

And again, why would a DM provide his enemies with effectively limitless resources? How is the Last Alliance assembled by the PCs supposed to force Sauron out of his tower if not by depleting his armies and sieging his fortress first? Sauron can't have infinite resources or the whole system makes no sense..
Your point about possible slot expenditure on an enemy wizard is potentially valid, but this point seems irrelevant. As long as you engage with each enemy set on an encounter by encounter basis, you can discuss the resources of each of those encounters on a micro scale. For example the enemy boss could send a barbarian at you, and if the party kills it, that'd diminish the total resources of the boss. However, the wizard that he sends afterwards will be just as refreshed as the barbarian, and if we assume that there are four encounters in a day, the particular remaining resources of a dungeon has no bearing on the amount of fighting you have to do.

[Eldariel]

Your point about possible slot expenditure on an enemy wizard is potentially valid, but this point seems irrelevant. As long as you engage with each enemy set on an encounter by encounter basis, you can discuss the resources of each of those encounters on a micro scale. For example the enemy boss could send a barbarian at you, and if the party kills it, that'd diminish the total resources of the boss. However, the wizard that he sends afterwards will be just as refreshed as the barbarian, and if we assume that there are four encounters in a day, the particular remaining resources of a dungeon has no bearing on the amount of fighting you have to do.

But there's certainly a difference in the BBEG fight if you TP in and face him with the Barb, the Wizard and himself at the same time vs. facing the Barb, then the Wizard, then the BBEG. The two mooks are effectively resources he's spent and thus he's gotten weaker (as with all the items/spells he's used to empower his mooks for the battle).

[NichG]

But there's certainly a difference in the BBEG fight if you TP in and face him with the Barb, the Wizard and himself at the same time vs. facing the Barb, then the Wizard, then the BBEG. The two mooks are effectively resources he's spent and thus he's gotten weaker (as with all the items/spells he's used to empower his mooks for the battle).

In practice, because of action economy things, this isn't how it works. Its kind of dumb but the BBEG can pretty much always win if he depopulates his lair and has them all with him at all times, because fighting all four encounters of the dungeon in the same room is much harder than fighting them as four separate events.

Basically, for meta-game reasons the villain can't actually make full use of their resources.

[eggynack]

But there's certainly a difference in the BBEG fight if you TP in and face him with the Barb, the Wizard and himself at the same time vs. facing the Barb, then the Wizard, then the BBEG. The two mooks are effectively resources he's spent and thus he's gotten weaker (as with all the items/spells he's used to empower his mooks for the battle).
I agree with that, but it's a different kind of resource. The BBEG loses his wizard resource when the wizard engages the party in battle, but the wizard has no effective loss in resources on a round to round basis if he dies at the end of the fight. This thing you're discussing is a resource, but it's a very different kind of resource than the one we're discussing.

[Eldariel]

I agree with that, but it's a different kind of resource. The BBEG loses his wizard resource when the wizard engages the party in battle, but the wizard has no effective loss in resources on a round to round basis if he dies at the end of the fight. This thing you're discussing is a resource, but it's a very different kind of resource than the one we're discussing.

Wouldn't e.g. a bought Riding Dog or a Leadershipped Cohort be effectively a character resource?

[eggynack]

Wouldn't e.g. a bought Riding Dog or a Leadershipped Cohort be effectively a character resource?
It feels a lot like you'd be better off just sticking the cohort in a new spot. Like, it's not the druid spending riding dog resources, or the druid spending druid resources, even if that may be how it works from a build perspective. Instead, it would be the riding dog spending riding dog resources. He has his own pile of HP, actions, and if it's a leadershipped guy with daily resources, daily resources. This definitely makes things act wonky in terms of putting down equal CR encounters, but that's actually just generally true of this kind of effect.

This is even more true from the perspective of the BBEG. Traditionally, the BBEG doesn't actually spend build resources to get henchmen at his beck and call, so there's no particular reason to call it a character resource in that case, while it may make more sense for PC's, who do need to spend build resources for these effects.

[Scow2]

It feels a lot like you'd be better off just sticking the cohort in a new spot. Like, it's not the druid spending riding dog resources, or the druid spending druid resources, even if that may be how it works from a build perspective. Instead, it would be the riding dog spending riding dog resources. He has his own pile of HP, actions, and if it's a leadershipped guy with daily resources, daily resources. This definitely makes things act wonky in terms of putting down equal CR encounters, but that's actually just generally true of this kind of effect.

This is even more true from the perspective of the BBEG. Traditionally, the BBEG doesn't actually spend build resources to get henchmen at his beck and call, so there's no particular reason to call it a character resource in that case, while it may make more sense for PC's, who do need to spend build resources for these effects.I'm not sure what I or you are arguing, but my insights:

It is the Druid spending Druid resources when his Riding Dog companion gets hurt, or his Summoned Nature Allies get summoned and subsequently destroyed, or a Cleric's called allies or Wizard's bound minions or Bard's Horny Barbarian's or Anyone's Leadership minions/cohorts get taken out. Adding NPCs independent of Character Sheet-given resources give them their own resources, and increase the effective CR of a party. Most BBEGs have fiat powers, and thus their "minions" are not actually CR-based resources, and instead additions to a BBEG's resources.

An encounter that ends early (Such as a Villain running away) doesn't have the same resource cost as one fought to completion (Annihilation of one side or the other).

It could be argued that non-'summoned' or Leadership-exempt allies count similarly, as they are partially fiat-based. The problem with 3e's CR is that it considers all summons and similar allies "Already figured into the CR of an enemy."


Of the OP's two assertations ("CR = APL encounters are supposed to take 20% of a party's resources" and "the APL = CR of an average member"), only the first is actually true. The second is merely a rough/idealized approximation, not a core assumption any more than "A Rakshasa is a CR 10 monster" is a core assumption.

[eggynack]

I'm not sure what I or you are arguing, but my insights:

It is the Druid spending Druid resources when his Riding Dog companion gets hurt, or his Summoned Nature Allies get summoned and subsequently destroyed, or a Cleric's called allies or Wizard's bound minions or Bard's Horny Barbarian's or Anyone's Leadership minions/cohorts get taken out.
I was mostly just arguing from the perspective of tactical resource allocation, which fits into the realm of encounter design, rather than from the perspective of strategic resource allocation, which is more macro and game design related. On a given round, the druid doesn't actually spend anything to have his riding dog attack. It'd be functionally equivalent to have a party with four normal party members and one player who made a dog character for no reason.

On the other end of things, you have bound and summoned minions. The way I figure it, a minion that you bring to the table only counts against your tactical resources if you spend those resources within that day. If you bind a minion one day, and use it to assault the BBEG on the next, what has the wizard really lost from a tactical perspective? The answer is not much, and it's part of the reason that these abilities are so powerful. On the other hand, if you summon a minion, you're generally paying an in combat action cost, as well as a daily spell cost.

A big problem with this issue is that the game's CR system is obviously imperfect, as you and others have mentioned. You can say that the animal companion, and the bound minion, and the cohort all represent a loss of build resources, and that means that you can count those factors into the normal CR, and everything will be fine. However, that's not actually how the game works. In reality, a first level druid is just two separate CR 1 characters. A first level druid without an animal companion is still a full caster with a more than healthy amount of battlefield control on his list, and a riding dog without a druid friend is still a CR 1 encounter, and deservedly so. That first level druid has two completely separate stacks of resources, actions, and resource expenditures in a turn. It's all still obviously a CR 1 character per the game rules, but in every way that's important to encounter design, it might as well be two.