Comparing player level and monster level

[JackPhoenix]

I will say, since this is a CR thread, I vastly prefer 5E's implementation to 3.5's.

Agreed, though it could still be better if they've listed both component numbers of the final CR (because as it is, a hypotetical monster with offensive CR of 1/8 and defensive CR of 20 would have the same final CR as another monster with offensive CR of 20 and defensive CR of 1/8, despite clearly not being anywhere near close to similar in an encounter) or if they aknowledged different roles a monster could have in battle, like in 4e (I don't like 4e monster design, but that part was good) instead of reducing everything to offense and defense, especially considering some monster features are very dependant on circumstances.
A lone hobgoblin is identical to CR 1/8 guard with better equipment, but in a group, it can use Martial Advantage to vastly improve its damage.
An ambusher like an assassin is dangerous in the first round, but after that, its combat effectiveness drops sharply.
A controller like Enchanter doesn't have much damage to stand on its own, but it can serve as a powerful force multiplier if combined with allies (and the new version shows how stupid and boring reducing everything to just offense and defense is).

[rlc]

i forget if there’s math in here, but i keep one of my tabs open on this https://slyflourish.com/5e_encounter_building.html

[Easy e]

Somewhat off topic, but one of my favorite reasons to use Humanoid foes is that I can then give them Class levels to amp them up. This really makes their abilities less clear to the players, and allows me to more easily match them in the game.

Sure, you are facing a half-dozen Goblins, but what-if they are all Level 20 barbarians? That changes what you are facing a whole lot.

The most difficult opponents my players typically face are other adventuring parties.* <evil grin>




*= Or Law Enforcement. My players never like it when law enforcement gets involved, they see that as a fail state now. Really curbs the Murder-Hoboing.

[Alexander Atoz]

Hello

Posting again as the guy that started this thread.
I've seen several different ideas here, and there is obviously a lot to think about. I wanted to ask if anybody be interested in addressing the idea I suggested, and kicked off the thread with. Namely, I suggested that CR vs. PC could maybe be better addressed via the XP system than via a percentage of the level. Unoriginal was forced to admit that at high levels the fraction changes (which arguably means that there has to be a few transition levels where the ratio is in the middle of changing) and I feel that XP can avoid that, if we exempt level 1 and readjust the CR's above 20.
So far all I've heard is that I'm wrong. No facts were given, just the information that this is how it's actually done. Do any of you want to try to calculate what would happen if you tried to do it my way?
(I know that my way isn't the regular way. Think of it as my suggesting that I came up with a new, better [I think] way, and tell me what you think about it if you actually take time to think about it.)
I'm quoting myself here so that you don't have to go back to the beginning of the thread to see my suggestion.
Since there was debate, I'll clarify that equal strength means that if both fight each other, or if the same number of each fight each other (such as 4 vs. 4) they would each have approx. 50% chance of winning.

I've been trying to work out the relationship between monster level and player level. My inclination is that a PC's level is equivalent to approx. half the current monster's level, in XP.
That wasn't clear at all, so let's give examples.
A CR 6 monster is worth 2,300 XP. Based on my theory, a level six PC should be worth half that, or 1,150 XP, and be approx. equal to a CR 4 monster. (CR 4 = 11,000 XP). Similarly, a level 9 monster is worth 5,000 XP, so if my rule holds true a level 9 PC is equal to (or a bit stronger than) a level 6 monster. (5,000/2 = 2,500. CR 6 = 2,300).
Obviously, I'm talking about an ordinary build. If someone used multiclassing to build something incredibly strong or weak, that will throw off this rating. Also, in a fight it is often the case that a specific monster happens too be optimized against any one specific PC, or vice versa. Terrain, surprise, and everything else can also throw stuff off.
I will also mention that my rule isn't meant to include CR's above 20, as the XP there gets kind of crazy. (According strictly to the XP, a tarrasque should be equal to six CR 20 monsters, or to four if he fights them at once and you add the multiplier for groups. I really don't think that is the case. My best guess for monsters above CR 20 is that you should halve all XP above 20, at which point the tarrasque would be worth 85,000 XP. That sounds about right, but all this is another story.)

[Unoriginal]

So far all I've heard is that I'm wrong. No facts were given, just the information that this is how it's actually done.

This is verifiably untrue. We all have provided facts and calculations (and calculators, for that matter), proportionally to the provided data.

Do any of you want to try to calculate what would happen if you tried to do it my way?

*You* should be the one who try to do it your way. The person making the claim is the one who has to provide the data.

But fine, let's try it your way:

My inclination is that a PC's level is equivalent to approx. half the current monster's level, in XP.

and I feel that XP can avoid that, if we exempt level 1 and readjust the CR's above 20.

Since there was debate, I'll clarify that equal strength means that if both fight each other, or if the same number of each fight each other (such as 4 vs. 4) they would each have approx. 50% chance of winning.

So... exempting level 1, that gives us:

Level 2 =225 XP = CR 1

In other words, if your hypothesis was correct, a lvl 2 Fighter would 50% chances of winning and 50% chances of losing against a Bugbear, in an 1 vs 1 fight.

Do you stand by this assertion?

[Alexander Atoz]

I do, actually.
The fighter's HP at level 2 is 18.5, and that's assuming he's going for an INT build. (Eldritch knight, or one of the Tasha's options also does INT.) Otherwise, he has + 2 CON, bringing him to 20.5. Second wind, which he'll certainly use unless he went for two weapon fighting (which also uses a bonus action) raises this by 7.5, giving us 26 with INT build, 28 without. The bugbear has 27 HP.
AC. The fighter will take chain, (he probably won't have improved his armor by level 2) letting him tie with the bugbear at 16 AC. If he has a shield, he beats the bugbear by two points, at 18 AC.
+ 5 to hit. The bugbear only has + 4.
Damage. This needs to be combined with the fighting style, so there are several options.
Duelist. He uses a d8 weapon, which means he does 9.5 damage. (+ 2 from duelist.) A bit less than the bugbear, but his AC is two points higher (he'll be using a shield), combined with the higher chance to hit he has an extra + 3 over the bugbear to hit. (Meaning that the bugbear will have to roll 3 points higher on the dice than the player will. I plan to use this language again as we continue.)
Greater weapon fighting. Whether he uses a d12 weapon, or a 2d6 weapon, his average with just the weapon dice is close to 8.5. (with the d12 weapon, I calculate it as 8 and a third.) + 3 for STR, and he's doing a bit more than the bugbear.
2-weapon fighting. 2 attacks. 6.5 damage per attack. (He'd need a feat to use the d8 rapiers 2 handed, which he won't have unless he's variant human.) So a good bit more damage than the bugbear, although he might have to skip one attack if he needs to use second wind. (Most likely worth it, unless he's on the verge of killing the bugbear.)
Armor mastery. Not much to say. Either 7 damage, 2/3 of what the bugbear has, but 2 higher to hit, or 4.5, which is roughly 2/5 of what the bugbear does, but the bugbear needs to roll 4 higher to hit. I'd think it would cancel out, although a pain in the neck to prove.
I'm skipping archery and protection. Protection means I'd have to add a second PC (and a second bugbear) and that adds complication, and archery is going to be a pain in the neck to use. For starters, we'd have to decide how close they should start to each other, and there's no way to decide that.
The bugbear has surprise attack, but achieving surprise is situational and often doesn't happen. The fighter has action surge, which is a very strong ability that will definitely happen.
My conclusion is that the two are pretty much equal. I.E that have a more or less equal chance to win a fight between the two of them.

Just a few more items:

*You* should be the one who try to do it your way. The person making the claim is the one who has to provide the data.

How could I possibly do this? The only way I could think of is if we analyze every monster in every D&D book, and compare them to every possible PC. I was posting here to give people a chance to support or disprove it.

So far all I've heard is that I'm wrong. No facts were given, just the information that this is how it's actually done.

This is verifiably untrue. We all have provided facts and calculations (and calculators, for that matter), proportionally to the provided data.

If you could show me this, (perhaps in form of quotations?) I'd be most interested.
So far, all I've seen is a lot of discussion about ways to compare CR and player level, none of which is my way. That's fine, I don't claim ownership of the thread, and I'm happy for people to use it to discuss whatever. BUT, I didn't figure interrupting the conversation to ask what people thought about my way was horribly impolite when I'm the one who started the thread.
Thank you very much for taking the time to discuss this with me. I am actually very interested in what you have to say, and I look forward to hearing from you.
Alexander Atoz

[GeneralVryth]

How could I possibly do this? The only way I could think of is if we analyze every monster in every D&D book, and compare them to every possible PC. I was posting here to give people a chance to support or disprove it.

If you could show me this, (perhaps in form of quotations?) I'd be most interested.
So far, all I've seen is a lot of discussion about ways to compare CR and player level, none of which is my way. That's fine, I don't claim ownership of the thread, and I'm happy for people to use it to discuss whatever. BUT, I didn't figure interrupting the conversation to ask what people thought about my way was horribly impolite when I'm the one who started the thread.
Thank you very much for taking the time to discuss this with me. I am actually very interested in what you have to say, and I look forward to hearing from you.
Alexander Atoz

The way to test how well your algorithm works would be to generate some PC stat blocks, and calculate their CRs and see how well that matches up with the CR your algorithm produces. That's kind of what I did earlier in this thread to compare a couple of systems.

I also wonder if there is a more elegant way to express your algorithm So here is the numbers kind of brute forced:

CR	EXP Val	Half	Closest CR
1/8	25	12.5	0
1/4	50	25	1/8
1/2	100	50	¼
1	200	100	½
2	450	225	1
3	700	350	2
4	1,100	550	2?
5	1,800	900	3
6	2,300	1150	4
7	2,900	1450	4.5
8	3,900	1950	5
9	5,000	2500	6
10	5,900	2950	7
11	7,200	3600	8
12	8,400	4200	8
13	10,000	5000	9
14	11,500	5750	10
15	13,000	6500	10
16	15,000	7500	11
17	18,000	9000	12
18	20,000	10000	13
19	22,000	11000	14
20	25,000	12500	15

I don't see one elegant and precise pattern there, though there are runs of ones.

As for some data I added a column for your suggestion to may data from earlier:

Elven Blade	Level 3	CR 1/2	Ratio 1/6 	Step Diff 3	1/2 Val CR 2
Elven Scout	Level 4	CR 1	Ratio 1/4 	Step Diff 3	1/2 Val CR 2.5
App. Warmage	Level 4	CR 1	Ratio 1/4 	Step Diff 3	1/2 Val CR 2.5
Guardian	Level 6	CR 3	Ratio 1/2 	Step Diff 3	1/2 Val CR 4
Watcher		Level 8	CR 4	Ratio 1/2 	Step Diff 4	1/2 Val CR 5
Eld. Knight	Level 8	CR 4	Ratio 1/2 	Step Diff 4	1/2 Val CR 5
Warmage		Level 8	CR 5	Ratio 1/2+ 	Step Diff 3	1/2 Val CR 5
Envoy		Level 8	CR 4	Ratio 1/2 	Step Diff 4	1/2 Val CR 5
Panther		Level 8	CR 4	Ratio 1/2 	Step Diff 4	1/2 Val CR 5

Your approach mostly ends up guessing high, then again my data are NPC stat blocks built close to PC rules, but they had non-heroic arrays and not every PC ability so they would be end being weaker than a comparable PC. Really need more data and from more varied levels to say anything concrete.

[Unoriginal]

I do, actually.

[...]

My conclusion is that the two are pretty much equal. I.E that have a more or less equal chance to win a fight between the two of them.

Thank you for clarifying your assertion.

Do you also assert that a lvl 2 Bard has 50% chances of winning and 50% chances of losing in a 1 vs 1 fight against the Bugbear?

How could I possibly do this? The only way I could think of is if we analyze every monster in every D&D book, and compare them to every possible PC.

1) This isn't the only way. You could do as GeneralVryth said, for example. You could also create one "generic PC" for each class and compare them to a selection of monsters for each CR, or do any other way that show your hypothesis is based on reality.

2) IF the only way to demonstrate your hypothesis was to "analyze every monster in every D&D book, and compare them to every possible PC"... then that's what you would have to do in order to demonstrate your hypothesis. Otherwise it would just be a non-demonstrated, dataless idea, and not an hypothesis.

When I set out to demonstrate the chances of success of an invisible Imp doing a DEX (Stealth) check, I compared their capacities against the perception capacities of every D&D creature from CR 0 to CR 10 in the MM, Volo's and Mordenkainen's Tome of Foe.

Because hypotheses in a TTRPG require a *lot* of data to demonstrate.

Note that my work above was still judged insufficient by many.

3)Saying "how could I possibly do this?" when told you should provide data puts in question how you came up with that hypothesis in the first place.

Surely you had collect data in order to come up a new way to do that calculation, no?

I was posting here to give people a chance to support or disprove it.

And again, first you need to provide your data, with your methodology alongside it to explain which parameters you took into account and why.

Otherwise it's like you're asking *us* to do your work proving or disproving your idea.