Determining Racial Build Costs

[Moxxmix]

This is a fresh attempt at determining racial build costs, on the assumption that WotC uses a strict system to balance races against each other.

Coronoides on rpg.net did this same sort of work back when 5E first came out, and has an elaborate system based on it. However it's problematic in many ways, and even a cursory attempt at creating a cost system yields results very different from his. So I'm redoing things from scratch.

I made a couple posts in the Feat vs Darkvision thread (p5-6 in this thread (http://www.giantitp.com/forums/showthread.php?552468-Is-a-Feat-really-better-than-Darkvision)) regarding this attempt, but since there have been suggestions on alternate ways to approach the problem, I'm creating a separate thread so as not to go off topic too far in that other thread.

[JNAProductions]

This is a fresh attempt at determining racial build costs, on the assumption that WotC uses a strict system to balance races against each other.

Coronoides on rpg.net did this same sort of work back when 5E first came out, and has an elaborate system based on it. However it's problematic in many ways, and even a cursory attempt at creating a cost system yields results very different from his. So I'm redoing things from scratch.

I made a couple posts in the Feat vs Darkvision thread (p5-6 in this thread (http://www.giantitp.com/forums/showthread.php?552468-Is-a-Feat-really-better-than-Darkvision)) regarding this attempt, but since there have been suggestions on alternate ways to approach the problem, I'm creating a separate thread so as not to go off topic too far in that other thread.

I bolded an issue. That's a mighty big assumption to make.

That being said, the races are relatively balanced against each other.

Although, I've had discussions with one of my friends. One of the things is that, if we rate every races' power on a scale of 1-100, the VHuman is probably merely 60 or 65, while other races can hit 100. Yet, VHuman is an oft-suggested race, and I agree with that, even over some of the 100 races. Why? Because, let's take Mountain Dwarf. +4 total stats, Darkvision, Poison resistance/Advantage on saves against Poison, and Medium armor. That's pretty good, but it's difficult to find a single class or build that can use it ALL. +2 Strength is only helpful to Strength based classes, which pretty much universally have Medium Armor, making that go to waste. In addition, Darkvision is best when you're being stealthy-if you're not sneaking, you might as well have a torch. None of the rest of the Mountain Dwarf supports that. Whereas a VHuman, since it's entirely mutable, always sees 100% use for whatever your build needs.

Or, in brief-abilities X, Y, and Z might be worth 5 points apiece. But if they aren't synergistic, the total value of a race is not 15, it's less.

[Moxxmix]

Preliminary work on finding a cost for Stat+2:

Baseline: Stat +1 = 4 points. Everything is relative to this value.

Size Medium is the default, and given a value of 0.

Speed 30 is the default, and given a value of 0.

Common Language is the default, and given a value of 0.

Assumption: Minimum cost of a feature is 1 point.

Assumption: Extra language is 1 point.


Due to additive nature, stat +2 cannot be lower than 8 (2x stat +1).

Due to elf stats, stat +2 cannot be higher than 12.

Due to dwarf stats, if speed 25 is -1, Stat +2 cannot be higher than 9. If speed 25 is -2, Stat +2 cannot be higher than 10.


Basic limits at this early stage indicates Stat+2 must be between 8 and 10 points.


You might also try redoing your math under the assumption that +2 to a stat is roughly equal in value to one feat, because we have reason to believe the designers tried to make that true. (Because ASIs can be converted to/from feats.)

I looked at that. In order for a Feat provided by a race choice to be worth a Stat+2 (at the max value of 10 points), a Skill Proficiency must be worth 6 points. However Lizardfolk put an upper bounds on Skill Proficiency cost of 2 points if Stat+2 is 10 points. If Stat+2 is 8 points, Skill Proficiency has an upper limit of 3 points.

Thus a Feat can't have a cost equal to a Stat+2 feature. Because of the other bounds, it should be at least 3 points higher than Stat+2, so has a lower bound of 11 points.

[E’Tallitnics]

You should know that during the 2.5 years of play testing the overwhelming feedback to, “should we balance the races/classes,” was, “NO!”.

We wanted races/classes that were more or less equal, but not so much that they started to homogenize into too much ‘sameness’.

Not that you can’t find what you’re looking for! Just that it might not be where you expect to find it….

[MaxWilson]

You should start with a couple of assumptions:

Feats and ASIs (+2 to any stat) are equivalent in value. Note: ASIs and +2 to a fixed stat, like +2 to Dex, need not be equivalent in value to each other. Flexibility has value.

Misty Step once per short rest is roughly as valuable as any of the elven subrace packages, per the 5E DMG guidance.

-Max

[Moxxmix]

I bolded an issue. That's a mighty big assumption to make.
Given that Mike Mearls, in his recent talks about building subclasses, has explicitly talked about the spreadsheets they use to balance out the classes, I don't find it to be a very big assumption that they would do the same for races.


That being said, the races are relatively balanced against each other.

Although, I've had discussions with one of my friends. One of the things is that, if we rate every races' power on a scale of 1-100, the VHuman is probably merely 60 or 65, while other races can hit 100. Yet, VHuman is an oft-suggested race, and I agree with that, even over some of the 100 races. Why? Because, let's take Mountain Dwarf. +4 total stats, Darkvision, Poison resistance/Advantage on saves against Poison, and Medium armor. That's pretty good, but it's difficult to find a single class or build that can use it ALL. +2 Strength is only helpful to Strength based classes, which pretty much universally have Medium Armor, making that go to waste. In addition, Darkvision is best when you're being stealthy-if you're not sneaking, you might as well have a torch. None of the rest of the Mountain Dwarf supports that. Whereas a VHuman, since it's entirely mutable, always sees 100% use for whatever your build needs.

Or, in brief-abilities X, Y, and Z might be worth 5 points apiece. But if they aren't synergistic, the total value of a race is not 15, it's less.

But, as a caveat to my own earlier statement, Mike has also noted that just because it's mathematically balanced doesn't mean it's functionally balanced. Features that each make sense individually may not be something that can actually be used together when applied to various classes. So there's still an art to making a balanced race, even after you work out the mathematical boundaries.

Or put another way, just because you have a polygon with 6 sides, doesn't mean you have a cube.


You should start with a couple of assumptions:

Feats and ASIs (+2 to any stat) are equivalent in value. Note: ASIs and +2 to a fixed stat, like +2 to Dex, need not be equivalent in value to each other. Flexibility has value.

Misty Step once per short rest is roughly as valuable as any of the elven subrace packages, per the 5E DMG guidance.

-Max

Umm. I can't assume the first point. I can test it, but I can't assume it.

As for flexibility having value (and thus +1 to any two stats of your choice, vs +1 to Dex and +1 to Cha, or whatever) does make sense. I'll have to see if there's a way to consider that.

[MaxWilson]

Umm. I can't assume the first point. I can test it, but I can't assume it.

Hmmm. I'd have said the reverse was true: you can make assumptions, but you cannot really validate the assumptions you make except to the extent that they contradict each other. For example, you can't falsify your theory that races are spreadsheet-balanced against each other. You can assume it and explore the implications of that assumption; but you can't really test it.

I'm just saying that you should number among your assumptions one of the few things we know to be a design constraint of 5E: ASIs and feats are of equivalent value in the designers' eyes.

If you come up with a system that says that feats are roughly as valuable as +4 to a stat, that's a pretty good sign that the system you've derived probably doesn't match the real system used by the designers.

[thoroughlyS]

Musicus (http://zappyman2.wixsite.com/musicushomebrew/homebrew-races) did some work on this early on in 5E's history, and I've seen it referenced many times because of that. However, I prefer Detect Balance (http://www.giantitp.com/forums/showthread.php?501272-Detect-Balance-an-Improved-Scale-for-Measuring-5e-Races), because it is more granular, and take things like non-synergistic abilities and flexibility into account. I still have some issues with some of the costs, but perhaps this thread could become another continuation?

[Moxxmix]

Musicus (http://zappyman2.wixsite.com/musicushomebrew/homebrew-races) did some work on this early on in 5E's history, and I've seen it referenced many times because of that. However, I prefer Detect Balance (http://www.giantitp.com/forums/showthread.php?501272-Detect-Balance-an-Improved-Scale-for-Measuring-5e-Races), because it is more granular, and take things like non-synergistic abilities and flexibility into account. I still have some issues with some of the costs, but perhaps this thread could become another continuation?

I've seen the Musicus data, but hadn't seen Detect Balance. Interestingly, the Detect Balance scaling is very close to what I came up with, including many of the same result values (but not all of them). It seems like a very solid starting point, and I'll double check what I end up with against it.