Optimisation level definitions

[Biggus]

A recent thread made me realise that what people mean by "low op" or "high op" can vary wildly: unlike the tier system for classes, there isn't a broadly accepted definition as far as I'm aware. So I thought it'd be interesting to see what different people mean by them: please feel free to give definitions and/or examples of your own ideas. I'll start us off:

Very low op: at the bottom end, you don't optimise at all, either because you're very new to the game or because you make your build decisions based entirely on other criteria, such as how closely it fits your character concept or whether it looks cool. At the higher end, you do pay some attention to whether your abilities work mechanically but it's not a priority and is frequently overruled by other considerations. Level-appropriate challenges will often be difficult and sometimes deadly. An example might be someone playing a Yathrinshee just because they like the idea of it, and spending quite a lot of their skill points on things like Craft or Profession skills unlikely to be of any use in-game to represent what their character's life was like when they were growing up, or just to give them some interesting hobbies.

Low op: you do want your character to be good at what they do, but either it's only one of several competing priorities or you're still relatively inexperienced. If a class really sucks you might play a different one if it offers better options provided it still fits pretty closely with your concept. Most characters statted out in official material fit into this category (although it's not uncommon for them to be in the "very low" class) and this is also roughly what CR and XP tables assume the players will be. For example, a Mystic Theurge which doesn't use early-entry tricks: they can still be useful, but they could be much better.

Mid op: it's important to you that your character is good at what they do, but you're not spending hours searching through splatbooks and forums for all the most powerful options. You might be consciously avoiding cheesy abilities or you might not know how to use them to their best effect. You need to either be moderately experienced or take advice from someone who is to play at this level. You can handle level-appropriate challenges without much difficulty, but they're not laughably easy and at the lower end of this category they might sometimes still put you in real danger (at the higher end, this will be a rarity unless you're low on resources or otherwise at a significant disadvantage). An example would be a straight Wizard with some of the better metamagic feats, but who isn't using multiple metamagic reducers to dramatically shrink or entirely eliminate the spell level increase.

High op: most if not all of the decisions you make about your character are based on how powerful the abilities are and/ or how well they synergise with your other abilities. You have to be fairly expert to play at this level. Level-appropriate challenges are easy unless the DM is also optimising them pretty hard or giving them major advantages of some sort. Some of the feats and prestige classes commonly used at this level may be banned by some DMs. An example would be an Incantratrix who adds powerful metamagic to high-level spells and uses the Celerity series and similar tricks to control the action economy and make themselves incredibly difficult to kill.

Very high op: at this level, you are spending hours trawling splatbooks and forums for every possible advantage, and you have the experience to make full use of what you find. Every choice you make about your character is made solely on the basis of how powerful it is; their history and personality are chosen to fit the optimisation decisions you make*. Even at the lower end, level-appropriate challenges are absurdly easy; at the higher end, anything below Pun-Pun style theoretical optimisation may be considered fair game. The feats, classes and spells typically used are frequently banned or nerfed by DMs. An example character would probably have levels in half a dozen carefully cherry-picked prestige classes.

*I'm aware of the Stormwind Fallacy, in the original post it acknowledges that it's a generalisation and doesn't hold in extreme cases.

So...is this roughly similar to what you mean by these optimisation levels, or wildly different?

[ExLibrisMortis]

The lowest optimization level is "the name is cool". You only read the irrelevant label.

Then comes "the feat's fluff is cool". You read slightly more.

Then comes "the feat does a cool thing". Hey, you understand a feat now!

Then comes "there's no cooler feat than this". You understand all the feats, congratulations!



And I'm being only slightly joking. A huge part of "but it fits my character!" is people confusing names with crunch, or confusing classes with occupations, or similar naive fluff-crunch mappings. Yes, the feat is called Self-Sufficient. That means exactly nothing, least of all to your character.



Anyway, your optimization levels are based waaay too heavily on book diving and making choices because they're "powerful". Optimization isn't just about power, and it certainly isn't rated by "hours spent reading splats". It's about weighing your options and picking the coolest thing each time, for any given value of "cool".

[LordBlades]

The way I see it:

Very low op: knowingly or not, you're handicapping yourself and you're making objectively bad choices, like start g your character with 1 level of commoner because he was a farmer before adventuring or spending a bunch of skills and feats on background related stuff.

Low op: you understand the basics (like a fighter needs high str, a wizard needs high int etc.) and you aim to build accordingly. However you don't know or willingly ignore the finer points of optimization (like THF being better than sword&board or battlefield control being better than damage spells).

Mid op: you understand how the game works, what's good and what's bad and your characters are good at what they're supposed to be good at and bad at what they're supposed to be bad at. You however don't know or don't want to implement complicated tricks and synergies in your build.

High op: you apply a thorough understanding of game mechanics to build finely tuned characters, where most of the aspects of the build work together toward a unitary concept/goal. You often include compkex and/or surprising tricks in your builds.

Very high op: here it's all about maximum power. You try to get everything you can from your character and employ every trick in the book to do so. Whether the character ends up playable or a TO exercise is anybody's guess :)

[Troacctid]

I mean, it's a continuum. You can't just smush it into a stepwise scale.

[Crake]

Anyway, your optimization levels are based waaay too heavily on book diving and making choices because they're "powerful". Optimization isn't just about power, and it certainly isn't rated by "hours spent reading splats". It's about weighing your options and picking the coolest thing each time, for any given value of "cool".

This seems less about optimization capabilities, but rather optimization playstyles. So yes, while you might have great optimization skills, and the ability to draft together practically any sort of character concept, at any level of table optimization, you still need to know what level of optimization the table is playing at to be able to build the appropriate character, and I think that's what this thread is aiming to accomplish.


I mean, it's a continuum. You can't just smush it into a stepwise scale.

Considering the OP is talking about high and low ends of each "step" it reads far more like brackets used to describe areas along a spectrum rather than a stepwise scale. Like, the high end of low op is practically indistinguishable from the low end of mid op, each "step" bleeds into the one before and after it.

[ayvango]

Well, there is common flaw in optimization that prevails on this forum. Players prefer to optimize for single encounter (statistically averaged) instead of optimizing entire character path. So they are not trying to choose the fastest path.

If you could farm monsters without cool stuff you are still Ok. The true matter for optimization is efficiency: how fast could you farm that monsters. Money turnover has higher importance then margin. It could be more profitable to use non-optimal choice for single encounter as long as it remains efficient overall.

I'm wonder why so strange concept as WBL was adopted on the forum. You could transform wealth to XP by using expensive supplies extensively. You could transform XP to wealth with ritual of transference (1:5 to be exact). So you should estimate what would be more important for farming at each specific frame of time - XP or gold. And what would speed up you farm better and what opportunity cost would each decision take on your farming.

It's a basic for an experienced computer player. It's amazingly that forum so focused on optimization miss that part of the optimization process.

[Crake]

Well, there is common flaw in optimization that prevails on this forum. Players prefer to optimize for single encounter (statistically averaged) instead of optimizing entire character path. So they are not trying to choose the fastest path.

If you could farm monsters without cool stuff you are still Ok. The true matter for optimization is efficiency: how fast could you farm that monsters. Money turnover has higher importance then margin. It could be more profitable to use non-optimal choice for single encounter as long as it remains efficient overall.

I'm wonder why so strange concept as WBL was adopted on the forum. You could transform wealth to XP by using expensive supplies extensively. You could transform XP to wealth with ritual of transference (1:5 to be exact). So you should estimate what would be more important for farming at each specific frame of time - XP or gold. And what would speed up you farm better and what opportunity cost would each decision take on your farming.

It's a basic for an experienced computer player. It's amazingly that forum so focused on optimization miss that part of the optimization process.

It may be because levels of play are about optimizing how good your character is at their thing, while still actually playing the game? Dungeons and dragons isn't an MMORPG where you're out to farm gold, or power level, and while it could theoretically be played that way if you so chose, I think most people would agree, if you intended to play it that way, you may as well just actually go and play a video game. This might explain why you're so confused by WBL: WBL is the expected amount of money you should have from going out and adventuring. If you're not adventuring, but rather farming money and/or powerleveling as fast as you can, then you're not really playing within the expected parameters of the game.

[False God]

I think this is more or less accurate to how I see optimization.

While I agree that there is "alternative optimization" where you replace the word "power" with "coolness" the problem there becomes subjective. I know far too many people who think "coolness" means a highly ineffective goofball character who drags the party down and gets them into trouble they didn't need to get into. Those people might argue they're optimizing too under the "optimizing for cool".

The thing about "optimizing for power" is that is has very discernible, definable elements. Something is either powerful, or it isn't. If you make optimization subjective to the player, then you lose the ability to discuss optimization at all, because the guy who "optimized" for a fumbling moron who can't walk, talk or save their own butt thinks that's just as cool and optimized as the guy who built the master transmuter wizard who can literally turn enemies to mud and end whole encounters in one turn.

One of these is an example of power. Both of these are an example of "cool", depending on who you ask.

Personally, I float between mid and high tier, depending on the game. I want my characters to be highly capable at what I want them to do. But I'm not terribly concerned with being the absolute god of the game. (I've played actual gods. It's less fun than it sounds.)

[ayvango]

then you're not really playing within the expected parameters of the game.
So, pouncing shock trooper barbarian under War Master's Charge is still within expected parameters of game? Mailman sorcerer dealing above 1000 damage per round is too within expected bounds?

[LordBlades]

So, pouncing shock trooper barbarian under War Master's Charge is still within expected parameters of game? Mailman sorcerer dealing above 1000 damage per round is too within expected bounds?

While I highly doubt any of these particular builds were in any way anticipated by the designers, the ability to simply kill an opponent within a single standard action is an integral part of the game (there are quite a few Save-or-Die and No-Save-Just-Die spells in the PHB) so yes, I would say they both fail within what a character is expected to be able to do.

[HouseRules]

Usually you have

Linear. Party of X handles X enemies encounters.

Triangular (https://en.wikipedia.org/wiki/Triangular_number). Party of X handles ((X^2 + X) / 2) enemies encounters.

Quadratic. Party of X handles (X^2) enemies encounters.

Edit: Should be appropriate level encounters, not necessarily enemies.
Weighting the encounters based on encounter level is a much more math intensive thing to do.

[Max Caysey]

It seems to focus around combat... how about skills? I usually choose one aspect (healing, sneaking, diplomacy, etc) and focus on that. I do that by choosing classes, and feats that increase either the numerical value of that aspect, or the broadness of its applications... I don’t however do any loops or stat increasing cheese...

I usually call that practical optimization, because it’s not breaking the game, and there is literally no cheese. And if the aspect of the character is non-combat, then level appropriate encounters are damn hard!

[ayvango]

It seems to focus around combat... how about skills?
Use custom off-body magic items of skills. First skill upgrade would cost 2 (off-body) * skill_lvl^2 * 100, the second would cost 2 (off-body) * 0.75 (the same bonus on off-body item) * skill_lvl^2 * 100, the third and subsequent would cost 2 (off-body) * 0.5 (the same bonus stacked) * skill_lvl^2 * 100. So off-body extra charges goes to zero. And you could combine many different sources on single item: competence, insight, sacred, morale, luck, enhancement.

[Crake]

So, pouncing shock trooper barbarian under War Master's Charge is still within expected parameters of game? Mailman sorcerer dealing above 1000 damage per round is too within expected bounds?

This is talking apples and oranges. Or in the context of the metaphor, Objectives of the Game vs Methods of Achieving the Objectives.


It seems to focus around combat... how about skills? I usually choose one aspect (healing, sneaking, diplomacy, etc) and focus on that. I do that by choosing classes, and feats that increase either the numerical value of that aspect, or the broadness of its applications... I don’t however do any loops or stat increasing cheese...

I usually call that practical optimization, because it’s not breaking the game, and there is literally no cheese. And if the aspect of the character is non-combat, then level appropriate encounters are damn hard!

Generally speaking, at least in my experience, skills can be optimized without any detriment to combat capabilities, so if you're optimizing skills at the expense of combat prowess, I'd probably put you in the low optimization bracket. As a side note, I don't think combat is at any point directly mentioned in the OP, simply "challenges" or "encounters", which can be anything from convincing the king to send aid to your town, to fighting off a red dragon that's invading a city.


Use custom off-body magic items of skills. First skill upgrade would cost 2 (off-body) * skill_lvl^2 * 100, the second would cost 2 (off-body) * 0.75 (the same bonus on off-body item) * skill_lvl^2 * 100, the third and subsequent would cost 2 (off-body) * 0.5 (the same bonus stacked) * skill_lvl^2 * 100. So off-body extra charges goes to zero. And you could combine many different sources on single item: competence, insight, sacred, morale, luck, enhancement.

I don't think he was asking for help optimizing skills, but rather asking how the optimization of skills fits into the definitions.

[Mato]

There are 14,000,605 levels of optimization because the number is just as subjective as where the lines are drawn and I'm not attached to the five way split on "tiers" or the decimal system but I like the MCU.

The first 7,000,302 ranks are similar to the builds listed in the DMG or those box-play releases in that they are designed to suck and die to teach you about grim endings from an older generation that considered failure to be part of a game. Another 3,500,151 resemble the build examples for most prestigious class entries which are considerably more powerful, probably. Another 1,750,075 ranks are dedicated to ranking BBEG entries because WotC did actually try to optimize them for four-on-one fights. And the remainder is what people who play video games try to build which inadvertently can solo most encounters and often "break" things.

[Jay R]

D&D optimization is like driving on a one-lane road. Anyone going slower than you is a slug. Anyone going faster than you is a maniac.

[ExLibrisMortis]

This seems less about optimization capabilities, but rather optimization playstyles. So yes, while you might have great optimization skills, and the ability to draft together practically any sort of character concept, at any level of table optimization, you still need to know what level of optimization the table is playing at to be able to build the appropriate character, and I think that's what this thread is aiming to accomplish.
I guess "level of optimization" is used in two ways. There's "capable of efficiently building a character to any specs" (i.e. level of optimization skill of a person) and "specs expected at a given table" (i.e. typical optimization skill required to build characters for a table, being, presumably, the skill of the players at that table). I think the second is metonymical of the first.


Here's the last thread I remember on this subject: I give you TIERS OF THE OPTIMIZER! (http://www.giantitp.com/forums/showthread.php?542125-I-give-you-TIERS-OF-THE-OPTIMIZER!).

[upho]

This seems less about optimization capabilities, but rather optimization playstyles. So yes, while you might have great optimization skills, and the ability to draft together practically any sort of character concept, at any level of table optimization, you still need to know what level of optimization the table is playing at to be able to build the appropriate character, and I think that's what this thread is aiming to accomplish.So much this.

"Optimized" should not be mistaken for "powerful" if we're talking about practical character optimization (referring to the combination of mechanical build options of a PC intended to be played in a real game in a manner taking full advantage of the build's mechanics). Hence, a scale of "practical optimization" should be defined as something along the lines of:

"The degree to which a PC's mechanical build options a) reflect the character concept in question and b) provide an overall adventuring capability suitable for the game in question".

Notably, no matter how mechanically powerful a build is, the degree of optimization skill required to put it together, or whether its capabilities are strengthened or weakened by how it's played, if the mechanics poorly reflect the intended character concept or are regarded as OP or UP in the game , the build is poorly optimized. Full stop.

And as Crake mentioned, to be able to measure how well the power of a build ("overall adventuring capability" in my definition above) suits with a game, I think it could be most useful if we had more objective definitions for the general power levels of a game. Hopefully, such definitions could also help clearing up much of the confusion and misunderstandings resulting from poorly defined concepts such as "optimization levels" and "power levels". It may even help remove some related lingering misunderstandings, such as "optimizer = powergamer".

Also, it appears to me many people miss that TO is perhaps even less defined by power than PO, since TO builds may not necessarily be about maximizing power at all, but may just as well be about optimizing one isolated ability or aspect to the high heavens, regardless of how useful such optimization would be in a real game.

@ Biggus: Having defined levels of optimization doesn't really help much, as they don't necessarily say anything about power, and it's a very bad idea to confuse the two even if they often happen to correlate.

To give you an example of why: I could put together the most powerful high level martial control build possible in the game, and at least in a 3.5 game it wouldn't even be capable of matching the combat usefulness of a decently built and played same-level wizard with a similar control focus using only core options. Not to mention my build would be hopelessly left in the dust when it comes to overcoming other adventuring challenges. And this is despite the fact that my martial build requires tons of splat- and forum-trawling, while the wizard only needs core and maybe skimming through an online guide. Despite my build ignoring every character aspect except mechanical power, while the wizard has several options just for flavor. Or my build requiring me to spend many hours playtesting and perfecting tactics, while the wizard player only needs a few minutes to read up on a spell or two before a session. Or my build requiring me to have great optimization skill and talent along with several years of experience, while the wizard player at most needs a couple of years in the hobby and not being an idiot. Etc, etc.

Or IOW, my build would be considerably less powerful than the wizard despite being vastly more optimized.

[Biggus]

OK, thanks for the responses. It seems I need to clarify the intent of this thread a little. I was reading another thread in which somebody asked "are you playing an extremely low-op game?" then asked why they weren't using what to them were basic, everyday tactics, but which to me seemed mid-op tactics at least, bordering on high-op, and I realised that our definitions of "extremely low op" were very different.

(I intentionally didn't link to the thread or name the person who said it because I didn't want this to be a "am I right or are they?" argument thread, I wanted it to be a general discussion of what people mean when they use common terms like "low op" or "high op").


I mean, it's a continuum. You can't just smush it into a stepwise scale.



Considering the OP is talking about high and low ends of each "step" it reads far more like brackets used to describe areas along a spectrum rather than a stepwise scale. Like, the high end of low op is practically indistinguishable from the low end of mid op, each "step" bleeds into the one before and after it.

What Crake said. I wasn't intending to suggest that optimisation comes in steps or that everyone should use my definitions, quite the reverse; I wanted to know what other people mean when they use these terms.


It seems to focus around combat... how about skills?


I don't think combat is at any point directly mentioned in the OP, simply "challenges" or "encounters", which can be anything from convincing the king to send aid to your town, to fighting off a red dragon that's invading a city.


Fair question. Again, what Crake said. I meant the term "power" in a pretty broad sense of "being able to achieve your goals", so skills would be included in that. I maybe could have made that clearer.



Having defined levels of optimization doesn't really help much, as they don't necessarily say anything about power, and it's a very bad idea to confuse the two even if they often happen to correlate.
<snip>
Or IOW, my build would be considerably less powerful than the wizard despite being vastly more optimized.

I didn't confuse the two. Each level is intended to represent an optimisation level of any given class or concept. All the classes mentioned in the examples are ones a character starting as a Wizard could take. I did consider giving examples for other classes as well but I didn't want to push the definitions into TL;DR territory.



Personally, I float between mid and high tier, depending on the game. I want my characters to be highly capable at what I want them to do. But I'm not terribly concerned with being the absolute god of the game.

I'm the same, I actually find the game becomes less fun if you go past "high-ish op" into full-on high-op.

[Jack_Simth]

Personally, I'm fond of PFO and PFD numbers (Pit Fiend Offense and Pit Fiend Defense) when specifying optimization levels, but folks tend not to like to go to the work of figuring them out.

[Biggus]

Personally, I'm fond of PFO and PFD numbers (Pit Fiend Offense and Pit Fiend Defense) when specifying optimization levels, but folks tend not to like to go to the work of figuring them out.

I've not heard of those, how do they work?

[Jack_Simth]

I've not heard of those, how do they work?

PFO:
On an 'adventuring day', without warning, you're dropped in a field full of pit fiends, who are just standing around doing nothing (they take no actions, but are otherwise normal). You expected something dangerous, but didn't know exactly when (you've got the things up you'd normally have up when you go into a dangerous area, but not the sort of thing you'd pull up because you know you're about to bust down a door and expect combat on the other side - so Mage Armor? Sure, if it's on your list, you can start with it up; something like Blink (rounds/level): Not so much)Your deity tells you to down as many of them as you can. On average, at the end of ten rounds, how many of them are out of the 'fight' (unable to act against you, were it not for the scenario limitation on them - Dead, negative HP, Charmed, Dominated, Stoned, or similar)? Keep in mind, if you're rolling 1d20+29 against it's AC 40, that's a 50% likelihood of a hit, and thus half-damage is counted for what your weapon would otherwise deal for purposes of how much HP damage you're dealing; if you're spamming Charm Monster or something, have a 50% chance to beat their Will save, and a 50% chance to beat their SR, then you're getting 1/4 per round * 10 rounds = PFO 2.5). If you're stuck full attacking with a club and will only get 50% of the way through one Pit FIend's HP, then you've got a PFO of 0.5.

PFD:
Assuming the Pit Fiends know everything about you and act in concert, on average, how many of them would it take to take your character down in one round? E.g., if their best choice is Wish(Flesh to Stone) at Fort DC 27, and you have a +16 Fort save and SR 31, then you have a 50% chance of making the save (and no other defenses of note vs. that attack), then on average, it'll take two of them to do the job. That's your PFD in that instance.

It's annoying to sort out - what's your ideal attack pattern vs. a field of passive Pit Fiends, what's the Pit Fiends' best options vs. you, and so on - but if done right, it's a verifiable metric that spits out actual numbers that can then be used for comparison. Doesn't work well on party buff builds (e.g., the Dragonfire Inspiration Bard that adds 10d6 sonic damage to everyone else's attacks isn't measured well here), and doesn't account for skilled characters (those who's strength is not in combat, but out of it - gathering information, convincing folks to look elsewhere, finding traps, and so on - even if NOT done via skills), but it's an actual metric that is suitable for comparison.

[rel]

Low OP: This party will find the encounters of a published adventure to be a reasonable challenge.

High OP: Any character in this party could solo every encounter in a published adventure if they wanted to.

[Max Caysey]

Low OP: This party will find the encounters of a published adventure to be a reasonable challenge.

High OP: Any character in this party could solo every encounter in a published adventure if they wanted to.

I once made a cleric healer who could keep the party going through encounters that were much higher ecl, however he had no combat skills himself, so he could solo nothing. However he was high op healer 😊

[Telonius]

Very high op: at this level, you are spending hours trawling splatbooks and forums for every possible advantage, and you have the experience to make full use of what you find. Every choice you make about your character is made solely on the basis of how powerful it is; their history and personality are chosen to fit the optimisation decisions you make*. Even at the lower end, level-appropriate challenges are absurdly easy; at the higher end, anything below Pun-Pun style theoretical optimisation may be considered fair game. The feats, classes and spells typically used are frequently banned or nerfed by DMs. An example character would probably have levels in half a dozen carefully cherry-picked prestige classes.


Possibly true for melee, but not really for casters. Either way, it's more like, "feats, features, and spells carefully cherry-picked from two dozen different sourcebooks." Most of the super-powerful options really don't require massive amounts of different prestige classes or multiclassing. Prestige classes do have prereqs, and it's usually (usually, not always) tricky to fit more than two or three prestige classes on a single build and not have it be less powerful than if you'd taken a simpler option. This is particularly true for full casters, who can snap the game in two just using their native spell list. If you're Pun-Pun, yeah, you have the "I have all prestige classes!" ability, but that's way on the high end.

Personally I'd split "Very High Op" and "Theoretical Op" into separate categories. The split is kind of hard to see and define. I guess I'd say that for "very high op" you're still dealing with just a single character and how they can overcome the problems that the world throws at them. For theoretical op, you're changing the base assumptions about how the world works. Once you reach Pun-Pun or Tippyverse levels of theoretical op, I'd say you're playing an entirely different sort of game from the rest of the players.

[Anthrowhale]

PFO: ... PFD: ...

This seems nice.

Looking at the optimized core ECL 20 fighter (http://www.giantitp.com/forums/showsinglepost.php?p=22350680&postcount=1481) I made it looks like PFD is 2 (via Wish[Antimagic Field] + Full attack with Power Attack[18]). PFO is about 7 (5 rounds of full attack/Rapid Shot/Boots of Speed/Ring of Blinking with holy silver arrows, 2 rounds of reinitiating the ring of blinking, 3 rounds of Full attack/Rapid Shot/Boots of Speed/Ring of Blinking with holy non-silver arrows.)

[Biggus]

Possibly true for melee, but not really for casters. Either way, it's more like, "feats, features, and spells carefully cherry-picked from two dozen different sourcebooks." Most of the super-powerful options really don't require massive amounts of different prestige classes or multiclassing. Prestige classes do have prereqs, and it's usually (usually, not always) tricky to fit more than two or three prestige classes on a single build and not have it be less powerful than if you'd taken a simpler option. This is particularly true for full casters, who can snap the game in two just using their native spell list.

Yeah, I was exaggerating slightly. I've often seen optimised caster builds with three or more prestige classes, but I've also often seen ones with two or less. Some prestige classes have minimal prerequisites or ones which you're likely to have fulfilled anyway by a certain level, so they're often used to fill up levels, especially in classes with no class features like Cleric or Sorcerer.


Personally I'd split "Very High Op" and "Theoretical Op" into separate categories.

I did, I said "at the higher end, anything below Pun-Pun style theoretical optimisation may be considered fair game".

[King of Nowhere]

there was already a thread like this a while ago (www.giantitp.com/forums/showthread.php?542125-I-give-you-TIERS-OF-THE-OPTIMIZER!).

it went nowhere because everyone had different definitions

[Mato]

there was already a thread like this a while ago (www.giantitp.com/forums/showthread.php?542125-I-give-you-TIERS-OF-THE-OPTIMIZER!).

it went nowhere because everyone had different definitionsIsn't that the point?

[ExLibrisMortis]

there was already a thread like this a while ago (www.giantitp.com/forums/showthread.php?542125-I-give-you-TIERS-OF-THE-OPTIMIZER!).

it went nowhere because everyone had different definitions
I don't agree with your assessment. We didn't settle on a definitive system, but the thread definitely went places.

[The Glyphstone]

I'm pretty sure it boils down to:

Low-Op: Stuff less optimized than me/my gaming group.

Medium-Op: My preferred level of optimization.

High-Op: Stuff more optimized than me/my gaming group.

Theoretical Op: Pun-Pun et al.

[Biggus]

Isn't that the point?

Yes, exactly. That thread was trying to propose universal definitions: I'm trying to do the opposite, I want to know what range of definitions there are out there.

[upho]

OK, thanks for the responses. It seems I need to clarify the intent of this thread a little. I was reading another thread in which somebody asked "are you playing an extremely low-op game?" then asked why they weren't using what to them were basic, everyday tactics, but which to me seemed mid-op tactics at least, bordering on high-op, and I realised that our definitions of "extremely low op" were very different.This is because the "someone" you mention makes the common mistake of confusing optimization levels with power levels. Note that this someone is not actually asking about how much splat- and forum-trawling the average player needs to create a PC or anything similar you've describe in your optimization scale in the OP, they're simply asking how powerful the average PC is.


(I intentionally didn't link to the thread or name the person who said it because I didn't want this to be a "am I right or are they?" argument thread, I wanted it to be a general discussion of what people mean when they use common terms like "low op" or "high op").And this is an initiative I applaud, because I also definitely believe more clear definitions are needed in this area.


I didn't confuse the two. Each level is intended to represent an optimisation level of any given class or concept. All the classes mentioned in the examples are ones a character starting as a Wizard could take. I did consider giving examples for other classes as well but I didn't want to push the definitions into TL;DR territory.I believe you're completely missing my point. Let me rephrase: how does a defined measure of "optimization level" help people play the game? Why is that measure important?


I'm the same, I actually find the game becomes less fun if you go past "high-ish op" into full-on high-op.So for example a "full-on high-op" fighter based PC tends to be less fun in your games with say "high-ish op" dread necromancers and bards because it's too powerful? Really?

Note that the comment by False God you responded to was talking about the tiers, which is a measure of power, while you're talking about optimization levels in your response.


Personally, I'm fond of PFO and PFD numbers (Pit Fiend Offense and Pit Fiend Defense) when specifying optimization levels, but folks tend not to like to go to the work of figuring them out.Same game tests such as this are often the only really viable method to measure power, I agree. But again, I fail to see why this necessarily has much to do with a discussion about optimization levels.


Low OP: This party will find the encounters of a published adventure to be a reasonable challenge.

High OP: Any character in this party could solo every encounter in a published adventure if they wanted to.Right, so no 3.5 fighter or monk can be high-op, but only full casters, a few 6/9 casters and some higher level PF martials can?


I once made a cleric healer who could keep the party going through encounters that were much higher ecl, however he had no combat skills himself, so he could solo nothing. However he was high op healer 😊And this is of course a very real issue when testing the power levels of individual PCs, demanding you also have some kind of "standard fellow party members" in your tests. Which results in a lot of work...

[Gnaeus]

I mostly agree with the low op definitions and I agree it’s a scale.

The best hallmark I think of between mid and high op is that Mid op is about good ingredients, high op is about good combinations and using things in unforeseen ways. Like combining something that makes damage non-lethal with immunity to non lethal damage. The parts may not be impressive. Sanctum Spell kind of sucks. But using it to enter prestige classes multiple levels early then retraining it away once the class gives you casting which qualifies you for the class is a high op trick.

[pabelfly]

Not sure how you'd separate low and mid optimization, but I'd define high optimization as being any character that requires specialized tactics to defeat, and theoretical optimization is anything you could only discuss on a forum and your DM would laugh at you if you ever suggested using it in an actual game.

[upho]

The best hallmark I think of between mid and high op is that Mid op is about good ingredients, high op is about good combinations and using things in unforeseen ways. Like combining something that makes damage non-lethal with immunity to non lethal damage. The parts may not be impressive. Sanctum Spell kind of sucks. But using it to enter prestige classes multiple levels early then retraining it away once the class gives you casting which qualifies you for the class is a high op trick.This is pretty accurate, I think, and could probably be made into good definitions of both mid- and high-op.

I also believe mid-op is typically about using good solid build options combined with basic min-maxing, but doesn't normally include combos involving more than a couple of options or a few obviously related options (such as feat chains), and a build with very similar qualities can often be made using different options.

And I think your definition of high-op as being "about good combinations and using things in unforeseen ways" is probably the best I've heard so far, and it illustrates why high-op is typically difficult and often requires both system mastery and a lot of experience playing the game. I'd only like to add "for uniquely superior results" to that, as I believe that unlike the mechanical qualities and strengths of mid-op builds, those of high-op builds largely rely on several options with very specific and unique benefits.

(Using simple examples from PF: a mid-op charge-focused build might be a Titan Mauler barbarian pouncing with an over-sized butchering axe and a bite, while a high-op supercharger version is more likely a Bloody-Knuckled Rowdy Primalist bloodrager with a MoMS monk dip who pounces with primary natural attacks boosted by the Horn of the Criosphinx feat.)


Not sure how you'd separate low and mid optimization, but I'd define high optimization as being any character that requires specialized tactics to defeat, and theoretical optimization is anything you could only discuss on a forum and your DM would laugh at you if you ever suggested using it in an actual game.Honestly, I don't find either definition particularly good.

I think the high-op one is poor because it defines a power level rather than an optimization level. I think many common monster combat abilities/strengths and very basic default combat tactics could easily defeat for example a high-op monk or 3.5 fighter. And the TO one is poor because a TO build doesn't necessarily include broken or even questionable combos or components, nor is a TO build necessarily OP or unplayable in most games. Using a recent TO exercise from this forum as an example, I don't see why a PC ending up with a Perception bonus well above +1000 is bound to break anything or have the DM laugh at you for suggesting you play it in a real game, nor why such a PC couldn't be made into at least a decent adventurer overall.

[GrayDeath]

They seem to mostly fit in with my experience.

Although it would be best to clearly seperate a lack of skill in optimization out of the definition.

Assuming one is a D&D long time palyer and avid Forum Reader, I would (in short words) list it like this:

No OP: Dont play. Even if I have a ridicolous concept, I dont saddle my coplayers with a incapable Character. It must at elast be remotely decent at its aimed spot in the party.

Low OP: Pick a concept. Go with it. No matter if its good or not.
Example:
You always wanted to play a Monk because of (enter inspirational Char here) and in this group this can actually work. Kind of.

Medium OP: Normal Play for most people I play with. You msotly either have a decently OP Concept in mind and pick fitting fluff, or have a FLuff cocnept that does not force you to abandon good choices.
Example: Playing a Sorcerer with good but not "perfect" Spell picks (say 1 Spell per Level purely for the feel, or somesuch).
Never includes anything that entierely invalidates other concepts or breaks the game PER SE.

High OP: Minmaxing to a unfluffy degree starts here. YOu might have to read up on details regarding how you best implement your Characters Strengths, less reliant on perfected tactics than the above, overall Power and flexibility rise by say 50% OR one of them by at least 100%.
Example: Martial Initiator/Rogue/Telflammar Shadowlord with feats retrained/exchanged to get back the horrendous prerequs.
Well buolt but not "always perfectly equipped and prepared" Wizards with Planar Binding Underlings. Top of the Line Sorcerer or Druid.

Very High OP: The highest LEvel I have ever played at, and that rarely. Needs around 5 books per CHaracter at a minimum, EVERYTHING but clear rule disputes takes a backs eat for POWAAA!
Frowned upon on most tables even with "legal without a doubt" RAW CHoices (and we all know how less clear RAW arguments can go).
WIzard played to GITP Standards or close, and built with perfection over days.

Non-PunPun TO: Semi.infinite loops, RAW Interpretations that would outstink mature Stilton, cherry picking certain itnerpretations, and using all you can find to maximize your Power. Often requires multiple setting dependant things from different settings and "perect availability" of feats, creatures and Items.
Never played it.

And then of course there is Pun Pun......

[King of Nowhere]

Yes, exactly. That thread was trying to propose universal definitions: I'm trying to do the opposite, I want to know what range of definitions there are out there.

well, then. my definition is based on the martial/caster divide

very low op: monks are considered the best class

low op: martials are as strong as casters

medium op: casters are stronger, but martials meaningfully contribute.

high op: martials contribute only if the casters let them

[upho]

well, then. my definition is based on the martial/caster divide

very low op: monks are considered the best class

low op: martials are as strong as casters

medium op: casters are stronger, but martials meaningfully contribute.

high op: martials contribute only if the casters let themOut of curiosity, if you bring a martial from one of your high-op games to one of your mid-op games, would that martial be higher op than the other martials in the mid-op game?

And say you have built a martial which includes several unorthodox/unforeseen unique combos on top of some of the strongest more obvious ones, the result of you spending hundreds of hours on splat- and forum-trawling, the assistance from several of the world's most experienced players and skilled optimizers, careful refinement and thorough playtesting, all brought together with a very high degree of general system mastery on top of enormous build talent and creativity.

Do you consider your martial to be more or less optimized than my boringly standard cleric which I threw together all by myself in less than an hour to play in one of your mid-op games?
Why or why not?
Would your answers to the above questions change depending on whether the mechanics of your martial makes for a more or less capable adventurer than those of my cleric?

[Jack_Simth]

Same game tests such as this are often the only really viable method to measure power, I agree. But again, I fail to see why this necessarily has much to do with a discussion about optimization levels.Optimization and power aren't really separable. Optimization is, by definition, towards some goal - one might optimize a vehicle for speed, hauling capacity, mileage, or any number of other things - and at the end of the day, the measure of how well one optimized is how well one achieved the goal. If you're achieving goals... how is that different from power?

[PoeticallyPsyco]

EDIT: I wrote this whole thing because I got the impression that high-op was being associated with powerful choices at the expense of character concept/consistent characterization, but now that it's all written I cannot for the life of me figure out where I got that impression from. I am, however, unwilling to delete all this work, even if it is contesting a point that (apparently) no one was making. So, for your viewing pleasure:

I don't think I agree that, as optimization increases, character concept decreases.

Fluff and mechanics should always inform each other, and while that's not always the case, it's pretty easy to see in action when it does. Even at its most basic, a player optimizing is still sticking to the character concept of "the best at this thing, The Ace (https://tvtropes.org/pmwiki/pmwiki.php/Main/TheAce)". And most optimization tricks are often in character for someone that wants to improve at that thing (for another basic example, a rogue that wants to be better at killing things would logically want to learn tricks from assassins, represented by a dip in the class, even if they don't share an assassin's ideology or motivation).

Going the other way, crunch can inform fluff quite well, too. Take my character Futility (http://www.giantitp.com/forums/showsinglepost.php?p=23920484&postcount=63). Some of the coolest things about her are the fluff implied by her mechanics, which I then ran with. The most obvious of example is how RAW she gains all the powers of the [swarm] subtype, which I then fluffed as dissolving into a humanoid shaped cloud of swarm members, each carrying a piece of her original body. A less obvious one (the judge missed it) is that her mental stats come from a plant swarm; this suggests that the plant swarm is actually the mind of the monster, and since it was nonsentient before symbiosis, Futility is naive, inexperienced, petulant, and initially terrified of the world outside of her comfort zone. She's child-like, and this behavior is both her biggest weakness and, I think, one of the creepier things about the character.

Or lets go with a more common example: the Craven feat. This can go any way, really; you can ignore the fluff, choose it because of its fluff, or (like me) not choose it because you hate the fluff. But on a recent character I made, I thought about what it would say about him that despite his confident, even arrogant, airs, he took a penalty on any fear related saves. Led into some interesting characterization directions.

Let's wrap up this post before it gets any longer. Basically, it's a similar thing to the relationship between optimization and power; correlation but not necessarily causation. Just as a martial can be highly optimized but still less powerful than a low-op caster, a character can be extremely high-op but still very dedicated to concept and characterization.

[rel]

I once made a cleric healer who could keep the party going through encounters that were much higher ecl, however he had no combat skills himself, so he could solo nothing. However he was high op healer 😊

That can certainly happen, but in my experience a cleric can switch roles pretty dramatically by just changing up spell selections at dawn.



Right, so no 3.5 fighter or monk can be high-op, but only full casters, a few 6/9 casters and some higher level PF martials can?

I'm not sure what published adventures you're playing, but the ones I participated in tend to have non combat encounters that pretty much anyone can resolve and straight up fights with CR appropriate foes.

A well built fighter (say an ubercharger or dungeon crasher) should be able to sort those out without too much trouble.

I have less knowledge of monks but given my experience of published adventures, I'd be surprised if a well built monk couldn't get through most encounters solo.


No idea what pathfinder has to do with it. If anything, pathfinder reduces the power of martials at high levels of optimisation.

[upho]

Optimization and power aren't really separable. Optimization is, by definition, towards some goal - one might optimize a vehicle for speed, hauling capacity, mileage, or any number of other things - and at the end of the day, the measure of how well one optimized is how well one achieved the goal. If you're achieving goals... how is that different from power?The first key difference is that for purposes relevant in this discussion, power optimization is also a measure of change (and arguably also of the difficulty of, or effort required for, that change), not just a measure of the end result. Ignoring this difference means that a defined optimization level won't necessarily say much about the time, experience, talent and skill typically required to match a build of a certain category with that optimization level, while it very much gives the impression of doing so. Instead, power optimization level definitions should be based on a view of optimization as a process, a verb, and power as one of the noticeable results of that process, an adjective. And again, a decent illustration of this is that a fighter can be vastly more optimized than wizard, despite being less powerful.

The second and IMO often far more important key difference is that practical optimization is not just about working towards a goal defined by power, and much less a goal of simply maximizing power. Instead, as I mentioned in my first post (http://www.giantitp.com/forums/showsinglepost.php?p=24059837&postcount=18), PO is about making a build's mechanics reflect the character concept in question as accurately, interestingly and playable as possible and have those mechanics result in an overall power suitable for the game the PC is to be played in. And this of course typically adds multiple optimization parameters more or less separate from the power parameters, many of which may very well be just as vital and difficult to improve as it is to match the build's overall power to that of the game. By equating optimization with power, a majority of these optimization parameters are ignored, and the optimization level definitions are bound to be misleading or worse.

So again, the definitions which would be most useful are those of power levels. Optimization levels are obviously not just more complex and multi-faceted and therefore more difficult to properly define, but thankfully also not as important as power level definitions when it comes to mitigating the risks of PC mechanics ruining a game.

[Jack_Simth]

So again, the definitions which would be most useful are those of power levels. Optimization levels are obviously not just more complex and multi-faceted and therefore more difficult to properly define, but thankfully also not as important as power level definitions when it comes to mitigating the risks of PC mechanics ruining a game.

Ah, I get it. You've got a noticeably different definition of the word "optimization" than I do, even outside the modifiers of "low", "practical", "high", and "theoretical". Well, base definitions aren't arguable. Negotiable when both folks want to communicate clearly, but not arguable, as language is fundamentally arbitrary. I could go through point-by-point on where I disagree with you, but there's little point. I'm not going to convince you of anything (and vice versa). Oh well.

[upho]

I'm not sure what published adventures you're playing, but the ones I participated in tend to have non combat encounters that pretty much anyone can resolve and straight up fights with CR appropriate foes.

A well built fighter (say an ubercharger or dungeon crasher) should be able to sort those out without too much trouble.While overcoming a large majority of challenges in published adventures run as written won't require a more than a party of very modest power and versatility and little player experience, it's far from easy for a single PC to match such a party's capabilities, and especially beyond 10th level or so, most builds based on martial classes won't be able to do so even if optimized as high as they possibly can be. And if you really think a 3.5 fighter could solo most higher level adventures, I have to say I very much suspect you've played very few adventures designed for levels above 10th or so, are used to DMs who routinely pull the punches of the opposition or have them fight with the same lack of tactical sense you'd only expect from a random gaggle of leaderless zombies, and/or possibly grant the PCs other advantages far greater than those recommended in official guidelines.

The fact is that despite being near exclusively focused on combat, a solo high-op 3.5 ubercharger (includes Dungeon Crasher in my world) wouldn't stand much of a chance in many combat encounters of a CR above their level, and would frequently be curb-stomped so far into kingdom come it's not even fun in the many encounters preventing effective use of their charge-dependent combat style. Not to mention those where a competent higher level caster is part of the opposition. Because while an ubercharger can typically easily one-shot a single enemy by dealing tons of hp damage, on their own they're sadly incapable of adapting their combat style to the frequently changing and many varied demands of higher level combat, and thus they often won't even get a chance to put that one-shot ability to use.

Alone, the fighter simply won't have even remotely sufficient passive durability to survive focused fire and no active defenses anywhere near strong enough to compensate with (no reliable enemy detection abilities, insufficient Will, Ref, initiative and often also AC and hp, no reliable emergency escape methods, no effective means of acting out of turn to prevent enemy offense etc). And when it comes to offense, against higher level opponents the fighter would arguably be even more sorely lacking (poor targeting capabilities, no effective tools for dealing with magic defenses, insufficient mobility and ranged offense, no effective ambush abilities, no multitarget/AoE abilities anywhere near worth the action cost or other reliable means to take out more than at best a couple of enemies per round, etc).

Unfortunately also other 3.5 martials suffer from many similar serious deficiencies as solo combatants in higher levels.

And then there's of course the typically far more serious issue of the many huge gaping holes in the 3.5 fighter's tool set for overcoming challenges other than combat...


l have less knowledge of monks but given my experience of published adventures, I'd be surprised if a well built monk couldn't get through most encounters solo.Has much the same issues as the fighter has, albeit with a less abysmal tool set for non-combat challenges and no way of matching the fighter's few actually meaningful offensive capabilities in combat.

If you doubt any of this, I suggest you for example look through some of the combat encounters above CR 10 and other challenges found here (https://thetrove.net/Books/Dungeons%20and%20Dragons/AD&D%203rd%20Edition/Modules/Expedition%20to%20the%20Demonweb%20Pits.pdf) (or in whichever higher level adventures you prefer) and ask yourself how much of a chance a lone 10th level fighter or monk would have to successfully overcome all of them. (Hint: virtually none.)


No idea what pathfinder has to do with it. If anything, pathfinder reduces the power of martials at high levels of optimisation.It appears you believe PF martials are basically limited only to the options found in the CRB for some reason... Is this correct? Or could it be quite a few years have passed since you personally last played or read up on PF martials and looked into their options found outside the CRB?

Regardless, perhaps I might be able to offer a little help catching up: here (http://www.giantitp.com/forums/showsinglepost.php?p=22924595&postcount=561) (scroll down to spoiler) are (http://www.giantitp.com/forums/showsinglepost.php?p=21860394&postcount=240) three (http://www.giantitp.com/forums/showsinglepost.php?p=21556361&postcount=89) mechanically related examples (plus combo details and example opening combat round (http://www.giantitp.com/forums/showsinglepost.php?p=21556450&postcount=91)) I've happened to have made in the past which may enlighten you. (Note that they're quite a bit weaker than they could be with the content now available (or if less hastily made in the case of the first example). I'll hopefully soon find the time and energy to properly update and refine them...)

Now if your statement actually remains correct and my poor old demented brain has simply forgotten just how awesome 3.5 martials can be, I'm sure you could direct me to a 3.5 martial build which clearly shows how PF "reduces the power of martials at high levels of optimisation". At least when compared to my three little PF creations linked to above.

[AnimeTheCat]

Right, so no 3.5 fighter or monk can be high-op, but only full casters, a few 6/9 casters and some higher level PF martials can?

I'm pretty sure that a well optimized fighter or monk could solo, at least the early levels, of the Bastion of Broken Souls adventure path. I think they would probably start to struggle most around the third adventure, as anything before that is just smashing numbers together until one of the numbers remains.

That actually leads me what I think are appropriate definitions of Optimization, Performance, and Power as well as how those three interact with each other.

Optimization is how well you've utilized your resources to accomplish a specified goal. Optimization has two constraints, the goal and the resources. In this way, optimization is a fluid definition as it changes depending on that goal or the resources available.

Performance is how well a certain character performs a particular task in comparison to other characters performing the same task. Performance has two vectors, the task and the other characters. Performance is very similar to power, but is slightly different because it is a comparison within a narrow field of view.

Power is the comparison of a character's performance in ALL tasks with other character's performance in ALL tasks. The key difference is that power is a comparison of performance in every task as opposed to just a specific task.

With these three definitions out of the way, you can see where, in my mind, I've isolated optimization as simply a measure of how well the resources available have been utilized. In this way it isn't a singular measure of how good a character is, but rather a measure of how well a certain character has utilized the resources available to them. Because of this, optimization really can always be an absolute measure. How well did you utilize your resources to accomplish your specified goal. Optimization levels vary from poorly optimized to well optimized, poorly meaning that there are many options available that were not used, and well optimized meaning that there are nearly no options remaining that could improve how a character accomplishes or achieves their goal.

TL;DR: Optimization is how well you use your resources to achieve your stated goal rather than power or performance. Optimizatoin is a static measure within the confines of the stated parameters, those being the goal and the available resources, and can range from poorly optimized to well optimized.

[HouseRules]

Optimization is how well you use your resources to achieve your stated goal rather than power or performance. Optimizatoin is a static measure within the confines of the stated parameters, those being the goal and the available resources, and can range from poorly optimized to well optimized.

Integrating Performance requires that the Resource is Effective Character Level and the Optimization of that is the collection of Character Class Levels, Racial Hit Dice, and Level Adjustment that make up the character. Equipment, Feats, Skills, and other stuff are at another level.

Separating Performance requires that the Resource is a Character of a Specific Class because of some reason.

As stated by AnimeTheCat, do not confuse these two types of optimizations.
Certain Classes regardless of how optimized (2nd category) as their class could be are not optimized (1st category) because of their TIER.
Class Tier defines the optimization of the 1st category.

[Quertus]

Most people, IME, use the phrase(s) to describe *power* level, not *optimization* level. Someone who plays a random Druid may well be called as higher optimization level than someone who carefully optimized their Commoner or their Wizard or their +8 LA Vampire to match their vision for the character, and/or to match the power level of the party.


PFO:
PFD:
It's annoying to sort out - what's your ideal attack pattern vs. a field of passive Pit Fiends, what's the Pit Fiends' best options vs. you, and so on - but if done right, it's a verifiable metric that spits out actual numbers that can then be used for comparison. Doesn't work well on party buff builds (e.g., the Dragonfire Inspiration Bard that adds 10d6 sonic damage to everyone else's attacks isn't measured well here), and doesn't account for skilled characters (those who's strength is not in combat, but out of it - gathering information, convincing folks to look elsewhere, finding traps, and so on - even if NOT done via skills), but it's an actual metric that is suitable for comparison.

For a lot of my characters, it spits out "0-1” or "NI" - but, then, that's largely a factor of level. I'll have to see if I can round up enough level appropriate characters, to see if it gives meaningful numbers.

For force multiplier characters, I'd think you'd want to measure the party's numbers without and with them, and call the difference their contribution.

And, yeah, it doesn't measure things outside combat. Or even every type of combat. Heck, when I did this in 2e, I had the calculation *start* with an infinite hallway of Kobalds, and compare that score against other challenges, to get a final score.

[AnimeTheCat]

Integrating Performance requires that the Resource is Effective Character Level and the Optimization of that is the collection of Character Class Levels, Racial Hit Dice, and Level Adjustment that make up the character. Equipment, Feats, Skills, and other stuff are at another level.

Separating Performance requires that the Resource is a Character of a Specific Class because of some reason.

While that helps with a balanced comparison, that's not necessarily true. It's just a comparison, not necessarily a balanced one. If you wanted a balanced comparison, yeah you'll want to make sure that you're comparing equal level characters with the same amount of in-game available (number of feats, value of equipment) but that's really where it ends. Within the resources available to the individual making the character, you're making a comparison between two characters doing the same thing and see which one performs better. In fact, if you're comparing two rogues that have been optimized for the same goal, they should perform nominally the same. Seperating performance, if anything, only requires that there are two different characters attempting to do the same task. in this way, you can see which character performs said task better.


As stated by AnimeTheCat, do not confuse these two types of optimizations.
Certain Classes regardless of how optimized (2nd category) as their class could be are not optimized (1st category) because of their TIER.
Class Tier defines the optimization of the 1st category.

I only categorized one thing as "optimization" because optimization is only one thing. The rest are comparisons of performance in a single task or in multiple tasks. Performance in a single task or every task isn't a measure of optimization, because the measure of optimization is how well you've utilized the resources available to you in order to achieve your stated goal. Don't misrepresent me by claiming I've said something I've, verifiably, did not.

[HouseRules]

Optimization without any context means
"Minimum Resource for Maximum Performance".
Therefore, the constraint has to exist, and that means Resource.

Thus, Optimization means, for the same resource, what is the best performance (for an intended goal).
The problem is that there are many intended goals that are implied by too many people.
Implicit Intended Goal screws up more discussions than they progress the knowledge.

[PoeticallyPsyco]

If you doubt any of this, I suggest you for example look through some of the combat encounters above CR 10 and other challenges found here (https://thetrove.net/Books/Dungeons%20and%20Dragons/AD&D%203rd%20Edition/Modules/Expedition%20to%20the%20Demonweb%20Pits.pdf) (or in whichever higher level adventures you prefer) and ask yourself how much of a chance a lone 10th level fighter or monk would have to successfully overcome all of them. (Hint: virtually none.)


Actually, I just made a monk for an optimization challenge (sorry, can't provide any more details than that until the reveal) that would be fully capable of soloing all the encounters I looked at (the first three, and a random three others chosen by scrolling to a random point in the document) at the appropriate level. Monks have a pretty high optimization ceiling for martials, because there are so many ACFs for them that quite a few ended up really strong.

[upho]

Ah, I get it. You've got a noticeably different definition of the word "optimization" than I do, even outside the modifiers of "low", "practical", "high", and "theoretical". Well, base definitions aren't arguable. Negotiable when both folks want to communicate clearly, but not arguable, as language is fundamentally arbitrary. I could go through point-by-point on where I disagree with you, but there's little point. I'm not going to convince you of anything (and vice versa). Oh well.Whoa, hold your dark horses of lost causes for just a sec there! I think you grossly underestimate the Bluff DCs I'd have to beat to be regarded as a man of principles... :smallbiggrin:

To clarify, I don't really care much about the semantics here, only that no single actually separate important measure/parameter/thingy is confused with another and that whatever labels we decide to put on these things are defined clearly and distinctly enough to be easily understood and useful in practice. I'd definitely welcome any suggestions you may have on a label more suitable than "optimization" to describe the mechanical build fiddling made for primary purposes other than altering overall power level and/or "performance" levels in some more specific regard or task (as defined above by AnimeTheCat).


I'm pretty sure that a well optimized fighter or monk could solo, at least the early levels, of the Bastion of Broken Souls adventure path. I think they would probably start to struggle most around the third adventure, as anything before that is just smashing numbers together until one of the numbers remains.I'm not familiar with that AP, but you may very well be right. And there are probably a few other similar very combat focused lower level exceptions. But I'd be genuinely surprised if a fighter or monk could be made capable of reliably soloing more than maybe a couple of unusually forgiving higher level adventures. And I'm also certain several high-op builds based on the most powerful and versatile classes could in contrast be made capable of breezing through an overwhelmingly large majority of higher level adventures.


That actually leads me what I think are appropriate definitions of Optimization, Performance, and Power as well as how those three interact with each other.

....

TL;DR: Optimization is how well you use your resources to achieve your stated goal rather than power or performance. Optimizatoin is a static measure within the confines of the stated parameters, those being the goal and the available resources, and can range from poorly optimized to well optimized.Hmm... Yeah, I like this, as I fully agree that having a separate defined measure of power for specific functions could be very handy, perhaps most obviously to help a DM evaluate a party's specific strengths and weaknesses, or to help a player find out what their PC should focus on in order to best complement the other PCs in a party.

Any ideas on how to define some rough levels of performance? Come to think of it, I wonder whether it may prove very difficult to come up with such universal definitions sufficiently relevant and meaningful for all the many wildly varying functions measured, not to mention a universal scale with of set objective values is simply impossible.


Most people, IME, use the phrase(s) to describe *power* level, not *optimization* level.This is true also IME. I guess one possibly important directly related question is whether it would be worth the time and effort of trying to change the meaning of optimization level into something distinctly separate from power level also in most peoples minds. Or whether it would be better to simply accept the IMO confusing/incorrect but more widely established "optimization = power" definition and instead come up with some new label for "non-power optimization".


Someone who plays a random Druid may well be called as higher optimization level than someone who carefully optimized their Commoner or their Wizard or their +8 LA Vampire to match their vision for the character, and/or to match the power level of the party.Precisely. And I think there's another potentially serious issue with the more common "optimization = power" definition as it distorts the meaning and implied goal of Practical Optimization, reducing its meaning into nothing more than something like "the process of making a build as mechanically powerful as possible without using clearly cheesy/OP combos or questionable interpretations of RAW". Which is not just far less useful than the meaning I've mentioned in previous posts, but also risks directly hurting games by failing to include any concept of "balancing to the table" or to recognize that non-power/verisimilitude optimization ("verisim-op"?) is a thing, much less a thing which may often be far more rewarding and important than power-op.

[AnimeTheCat]

I'm not familiar with that AP, but you may very well be right. And there are probably a few other similar very combat focused lower level exceptions. But I'd be genuinely surprised if a fighter or monk could be made capable of reliably soloing more than maybe a couple of unusually forgiving higher level adventures. And I'm also certain several high-op builds based on the most powerful and versatile classes could in contrast be made capable of breezing through an overwhelmingly large majority of higher level adventures.

It's the adventures that start with Sunless Citadel and then progresses in to Forge of Fury and then The Speaker in Dreams. I feel like a well optimized fighter or Monk would likely stand a chance through the first two, but by the time they hit The Speaker in Dreams, they would probably run in to pretty major issues.


Hmm... Yeah, I like this, as I fully agree that having a separate defined measure of power for specific functions could be very handy, perhaps most obviously to help a DM evaluate a party's specific strengths and weaknesses, or to help a player find out what their PC should focus on in order to best complement the other PCs in a party.

Any ideas on how to define some rough levels of performance? Come to think of it, I wonder whether it may prove very difficult to come up with such universal definitions sufficiently relevant and meaningful for all the many wildly varying functions measured, not to mention a universal scale with of set objective values is simply impossible.

So, some things are easy. For example, it's easy to compare performance on skill interactions as those are static or opposed DCs. So there's an easy, static point of measure. This is easy, but will yield skewed results. For example, a rogue that picked up the Investigator feat will of course perform better at Gather Information and Search checks, yet Investigator is hardly what is considered an "optimized" choice. It can be, if you have the build resources to pick it up and those two skills are vital to that build's success, but in large there are usually more effective or efficient ways to achieve what you're after than static bonuses at the cost of a feat.

Another way is really, really complex though. For instance, if the stated goal is to be an investigator, compare how well the two characters can serve as investigators. This is less numeric, though it does have a numeric value to it (you're going to use skills). So, going with that investigator goal, you might think a Rogue would be an 'optimal' choice due to high skill points and great class skills, you may also think Bard as well because of the same reasons, but also Magic, and you may also think Enchanter specialist wizard. So what you would do is optimize each of those characters, taking specific care to use only the same books for each. In this way you can gauge whether your character is optimized to the task you're after simply by evaluating what options are available and how well you've utilized them and how many, if any, remain for you to use. If you can't fathom any further way to make that particular character any better at that specific task, then you have a well optimized character. Then, to compare performance, you can simply run the characters through some numbers and see how far each character can go. How often will a character with that build succeed at a certain skill DC? Is the character using magical buffs, if yes cool, if not, add that to your calculation and just make a not of how often that can occur. Specifically for things like being an "investigator" it's up to the individual doing the comparison to judge how much of an impact it is to have limited spell uses, if you're even using spell slots rather than magic items (because money is a resource after all). It's all something to be taken in to account during the comparison. The last part should probably be just comparing the numbers of each character. Who has the higher number in the investigative skills? Who can perform investigative tasks the best? Can that performance be consistent or is it reliant on successfully casting a spell on a target other than yourself? For example, if the Enchanter Wizard is reliant on Charm Person to perform well, what is their likelihood of success with that spell? That's dependent upon the targets, so targets should be taken in to consideration. If the wizard is reliant on Divination, do those divination have risk of failure (saves) or giving wrong data? What about the rogue? Does the Rogue have anything in their build with a risk of failure? Can they consistently roll a high enough bluff to beat the target's sense motive? Again, target consideration. There is such a wide breadth of questions to ask, this method of comparison is, like you said, nearly impossible to do. Too many variables.

My 'Idea" is that specific comparisons need to be made by the individual doing the comparison with specific milestones and challenges that all builds are put up against. If some specific, well though out ones are generally or universally agreed upon, then these can be used as universal milestones (clearly... circular reasoning is circular) which can be referenced during discussions. If some of these are to be decided upon, that should probably be a new thread.

[King of Nowhere]

Out of curiosity, if you bring a martial from one of your high-op games to one of your mid-op games, would that martial be higher op than the other martials in the mid-op game?


generally yes, the martial would be higher op, and it would actuallly be stronger. it's just that you can squeeze much more optimization from a caster than from a martial


And say you have built a martial which includes several unorthodox/unforeseen unique combos on top of some of the strongest more obvious ones, the result of you spending hundreds of hours on splat- and forum-trawling, the assistance from several of the world's most experienced players and skilled optimizers, careful refinement and thorough playtesting, all brought together with a very high degree of general system mastery on top of enormous build talent and creativity.

Do you consider your martial to be more or less optimized than my boringly standard cleric which I threw together all by myself in less than an hour to play in one of your mid-op games?
Why or why not?
Would your answers to the above questions change depending on whether the mechanics of your martial makes for a more or less capable adventurer than those of my cleric?

the martial would be more optimized even if the cleric was actually stronger. though if the cleric is stronger than a high-op martial, it must be at least mid-high op itself.
the reason the fighter is higher op is that it took much more effort to build it, and if a simmilar effort was used also on casters, then the caster would definitely break the game.

though my definitions tend to fall apart a bit if confronted by those questions. they work better for describing the power level of a table.

[Jack_Simth]

Whoa, hold your dark horses of lost causes for just a sec there! I think you grossly underestimate the Bluff DCs I'd have to beat to be regarded as a man of principles... :smallbiggrin:

To clarify, I don't really care much about the semantics here, only that no single actually separate important measure/parameter/thingy is confused with another and that whatever labels we decide to put on these things are defined clearly and distinctly enough to be easily understood and useful in practice. I'd definitely welcome any suggestions you may have on a label more suitable than "optimization" to describe the mechanical build fiddling made for primary purposes other than altering overall power level and/or "performance" levels in some more specific regard or task (as defined above by AnimeTheCat).We're using language differently.

"Optimizating" is building or modifying a thing to accomplish some set of goals better within a set of resources.
"Power" is the ability to accomplish some set of goals.

The way I know the words, the two aren't really separable. More skill was put into that Fighter-12, but that Wizard-12 with less player skill is trouncing him at the tests in use. Well, your class is either part of your optimization choices (in which case, Fighter's player made some non-optimization choices that are hampering things, in which case the wizard player is optimizing better), or the class selections were made for the players (in which case, the fighter's player and the wizard's player do not have the same resource pool for the comparison tests in question).

When you talk of separating optimization and power, it doesn't make sense to me. Story and character are fluff. They have only passing ties to "crunch" - the optimization. It's like you're saying there's no blue in purple. You're speaking a different language. You're using your words differently. We're using different base tokens. Those are not arguable.

[upho]

Thus, Optimization means, for the same resource, what is the best performance (for an intended goal).
The problem is that there are many intended goals that are implied by too many people.
Implicit Intended Goal screws up more discussions than they progress the knowledge.Precisely this.


Actually, I just made a monk for an optimization challenge (sorry, can't provide any more details than that until the reveal) that would be fully capable of soloing all the encounters I looked at (the first three, and a random three others chosen by scrolling to a random point in the document) at the appropriate level. Monks have a pretty high optimization ceiling for martials, because there are so many ACFs for them that quite a few ended up really strong.Cool. Give it your best shot and please let us know (or send me a PM) how your megamonk fared in the pits. Though it may be worth keeping in mind that your method for picking challenges most likely doesn't meet the criteria I set; to solo all combat encounters of a CR above 10 as a 10th level fighter or monk. And as AFAICT just from quickly having skimmed through the last parts of the adventure, I suspect there's a very high probability a monk would end up in a "kill 'em all"-scenario and having to face at least some of the higher CR opponents simultaneously.

When's the reveal of your megamonk btw?


"Optimizating" is building or modifying a thing to accomplish some set of goals better within a set of resources.
"Power" is the ability to accomplish some set of goals.This may possibly be confusing capability to optimize with the resulting level of optimization (per your definition). Otherwise I have no objections to this.

But just to be sure there are no hidden implications in the above, could you perhaps give a few different examples of "a set of goals" one might try to accomplish by "optimizing" in practice (in your language and per your definitions)?


The way I know the words, the two aren't really separable. More skill was put into that Fighter-12, but that Wizard-12 with less player skill is trouncing him at the tests in use. Well, your class is either part of your optimization choices (in which case, Fighter's player made some non-optimization choices that are hampering things, in which case the wizard player is optimizing better), or the class selections were made for the players (in which case, the fighter's player and the wizard's player do not have the same resource pool for the comparison tests in question).Going by this example and what you've said so far, it appears to me you're at the very least implying that some things are always one and the same and can only be viewed in one certain perspective and used only for one certain general purpose, even though these things can in reality be objectively different and used for other purposes. This remains true regardless of the language you prefer to use and regardless of whether you recognize the existence of these differences.

Or to use your color analogy, it doesn't matter whether you're color blind and unable to perceive any differences between red and green, they still reflect very different wavelengths of light. And they remain two different colors even though you keep insisting they're both colors and therefore "the two aren't really separable". And more importantly, if it's also highly beneficial for a certain purpose to recognize they are different colors, why not do so?


When you talk of separating optimization and power, it doesn't make sense to me.And again, I could be fine with not separating them. But again, then we need to come up with some other labels for the mechanical tweaking made for other purposes than maximizing power, along with a new well defined term replacing the relatively useless and misleading "Practical Optimization".

[Jack_Simth]

But just to be sure there are no hidden implications in the above, could you perhaps give a few different examples of "a set of goals" one might try to accomplish by "optimizing" in practice (in your language and per your definitions)?

Sure. One might optimize a Cleric for spellcasting. Put the high roll (or a high value on a point buy) into Wisdom, grab the Madness domain (http://www.d20srd.org/srd/divine/domains.htm#madnessDomain), maybe bump up the character a few age categories, get +Wis items at the first opportunity, put all level up bonuses into Wisdom, and so on.

Another might optimize a Cleric for melee combat. Pick up Divine Metamagic(Persistent Spell), extra turning, nightsticks (when available), and persist the standard trio of Divine Power, Divine Favor, and Righteous Might (when each is available, of course; other spells are also useful, but that's the "basic" trio for a DMM(Persist) build); wear full plate armor (until a Monk's Belt is better, anyway), cast Magic Vestments and Greater Magic Weapon on the armor and weapon of choice (respectively); pick up a few appropriate combat feats, and so on.

Another might optimize a Cleric for making pastries. Take full ranks in Profession(Pastry Chef), skill focus in Profession(Pasty Chef), prepare Divine Insight (or Guidance of the Avatar, if online WotC content is permitted) to use on Profession(Pastry Chef) checks, buy a masterwork oven, and so on.

The first is very powerful at spellcasting (for the level), and will do very well for any goals where spellcasting is useful (VERY common in D&D and derivatives).
The second is very powerful at melee combat (for the level), and will do very well for any goals where melee combat is useful (common in D&D and derivatives).
The third is very powerful at making pastries (for the level), and will do very well for any goals where making pastries is useful (rare in D&D and derivatives, but can come up sometimes).

If you're optimizing a thing, you're increasing the thing's ability to accomplish some set of goals (often at the expense of other things - to continue one of the requested examples above: the Madness domain power also reduces will saves, Spot checks, Listen checks, Sense Motive checks, and so on most of the time; increasing your age category increases your mental stats, but also saps Strength, Dex, and Con - making one more fragile and easier to hit; starting by going after the +Wis item means you're NOT starting by going after the +Saves items; and so on).

Meanwhile, power is the ability to accomplish goals. If you're optimizing a thing for some aspect, you're increasing that thing's power at that aspect.


Going by this example and what you've said so far, it appears to me you're at the very least implying that some things are always one and the sameNot one and the same, no, but intrinsically tied. With how human eyes work, all purples contain some blue. That does NOT mean that blue and purple are the same thing.

and can only be viewed in one certain perspective and used only for one certain general purpose, even though these things can in reality be objectively different and used for other purposes. This remains true regardless of the language you prefer to use and regardless of whether you recognize the existence of these differences.
Where do you get that from? What definition of "optimization" and "power" are you using such that they're unrelated to each other?


Or to use your color analogy, it doesn't matter whether you're color blind and unable to perceive any differences between red and green, they still reflect very different wavelengths of light. And they remain two different colors even though you keep insisting they're both colors and therefore "the two aren't really separable". And more importantly, if it's also highly beneficial for a certain purpose to recognize they are different colors, why not do so?
Well, in this case, they're not at opposite ends of a spectrum.


And again, I could be fine with not separating them. But again, then we need to come up with some other labels for the mechanical tweaking made for other purposes than maximizing power,Well, it's not optimizing if you're changing things to avoid achieving goals. It's trying to reflect fluff with mechanics.

along with a new well defined term replacing the relatively useless and misleading "Practical Optimization".If "practical optimization" means "keep your power level to that of the table you're playing at", you've got a problem in that lots of folks refer to games they're in as "high optimization" quite regularly.

[upho]

The first is very powerful at spellcasting (for the level), and will do very well for any goals where spellcasting is useful (VERY common in D&D and derivatives).
The second is very powerful at melee combat (for the level), and will do very well for any goals where melee combat is useful (common in D&D and derivatives).
The third is very powerful at making pastries (for the level), and will do very well for any goals where making pastries is useful (rare in D&D and derivatives, but can come up sometimes).

If you're optimizing a thing, you're increasing the thing's ability to accomplish some set of goals (often at the expense of other things - to continue one of the requested examples above: the Madness domain power also reduces will saves, Spot checks, Listen checks, Sense Motive checks, and so on most of the time; increasing your age category increases your mental stats, but also saps Strength, Dex, and Con - making one more fragile and easier to hit; starting by going after the +Wis item means you're NOT starting by going after the +Saves items; and so on).Seems absolutely fine to me and perfectly in line with how I view and define these things, as I've mentioned in my reply to AnimeTheCat regarding his "performance" concept/measure.

But first, I have absolutely no clue as to how you square this with your far more important and more relevant definition of overall optimization level, as you touch upon here:
The way I know the words, the two aren't really separable. More skill was put into that Fighter-12, but that Wizard-12 with less player skill is trouncing him at the tests in use. Well, your class is either part of your optimization choices (in which case, Fighter's player made some non-optimization choices that are hampering things, in which case the wizard player is optimizing better), or the class selections were made for the players (in which case, the fighter's player and the wizard's player do not have the same resource pool for the comparison tests in question).According to this, your "pastry-cleric" above isn't as optimized as your caster cleric or melee cleric, and the more relatively capable he becomes at achieving goals related to pastry-making, the less relatively capable he becomes overall in a vast majority of D&D games. In other words, it appears your definition of "optimization" leads to a paradox: the more optimized he is, the less optimized he is. Using my definitions, he instead simply becomes less powerful but not necessarily any less optimized. And this is one of the very reasons why I'd prefer to separate optimization level from power level.

Second, by extension I believe it's pretty darn useless, if not counterproductive, to insist that overall optimization must equate power simply because of the obvious fact that improving a certain mechanical aspect gives a PC a greater ability to achieve goals where that certain aspect is relevant. Because again, as you touch upon yourself, two builds equally optimized to achieve their respective different specific mechanical goals virtually always also have different overall power because of their chosen goals. So for all practical purposes of having defined universal power levels (or "optimization levels" per your definition), it's at the very least misleading to say a PC highly optimized for making pastries is powerful. IOW, the PC's mechanical pastry-making optimization may very well be important for the purpose of say reflecting parts of the PCs background and personality, but in itself it's extremely unlikely to have any impact whatsoever on any important power balance aspects of the game. And to be clear, of course pastry-making optimization obviously is optimization and obviously alters the related mechanical power also in my mind, it's just not optimization relevant when defining levels of overall power.

Third, using your definition of optimization as solely being about increasing the mechanical power to achieve a certain in-game goal, what do you call mechanical alterations made with the explicit primary goal of decreasing mechanical power? Would that be "optimization to decrease optimization"?


Well, it's not optimizing if you're changing things to avoid achieving goals. It's trying to reflect fluff with mechanics.I think "changing things to avoid achieving goals" sounds like a paradox. If you want to a avoid a certain goal and change things to do so, then obviously the avoidance itself is your goal. And why cannot "reflect fluff with mechanics" be a goal? I personally find it to often be a very important one in a game as mechanics-heavy as 3.5/PF, where a character having fluff not backed up with the proper mechanics (when/if relevant and if such mechanics exist) tend to make the character look like a joke or simply decreases verisimilitude and the possibilities of immersion.


If "practical optimization" means "keep your power level to that of the table you're playing at", you've got a problem in that lots of folks refer to games they're in as "high optimization" quite regularly.Indeed. Hence why I've said I believe this needs to be taken into account when trying to define power levels.

[ExLibrisMortis]

I suppose that, at this point, we need to introduce the concept of "vector" to describe optimization. Optimization has both magnitude (power) and direction (goal), and all that.

[AnimeTheCat]

The thing is, you dont really optimize classes or characters usually, but you normally optimize task performance. Make my character the best at [task]. Dealing damage. Battlefield control. Healing. Sneaking. It's much less common to see "make my character the best [class]". Its really because classes and characters are open ended while tasks are specific and closed ended. It's why when people ask for help building a monk the forum goes, "what about monk do you want?"

Now, usually classes have an understood but unwritten universal role that they fill. Rogue=skillmonkey/damage, fighter=damage, barbarian=damage, bard=support/skillmonkey, etc. So usually when someone says optimize my [class] we default to this, but it's entirely possible to optimize within other constraints, because we're optimizing tasks, not classes or characters.