(Removed)

[LudicSavant]

Edit: It has been pointed out to me that there are some cases where the suggested change does impact some of the upper tier races in ways I didn't properly anticipate, and as such I no longer stand by the suggestion I made.

[Seerow]

Yes if you pick a race that has a penalty to your primary stat, you are going to end up behind. Let us count the people who are surprised. *crickets*



On the other hand, let us look at a different example. A pair of archers, one human, one elf. They are going to go for Dex first (up to 16), then Strength.

Elf ends up with 16 dex (6 points) and 16 strength (10 points), with 2 points leftover to spare. Which he goes ahead and uses to boost above the initial target of 16 dex to 17 dex, because he has nothing better to do with it.

Human ends up with 16 dex (10 points) and 15 strength (8 points). Looks like the elf is ahead!


If you want to push to an 18 it's even better. Because now the Elf has 18 dex (10 points) and 15 str (8 points), while the Human has 18 dex (16 points) and 10 str (2 points).



Humans don't "cheat" at point buy. They are just good all around. They are easily adaptable to any concept, while other races to be effective need to play to their strengths, but when they do, they come out ahead of the human stat-wise.

[nedz]

Racial modifiers promote racial stereotyping and penalise playing against type — more news at eleven.

Suppose your Wizards wanted 18 Str and made Int secondary (and yes I have played such a character, well he was Sorcerer but no matter)

Half Orc: 18 Str and 13 Int
Human: 18 Str and 10 Int

[LudicSavant]

Yes if you pick a race that has a penalty to your primary stat, you are going to end up behind. Let us count the people who are surprised. *crickets* I think you missed the point of the article. It's not that the half-orc was worse. It's that it's absolutely and completely worse, with no tradeoffs (even bad tradeoffs), simply because of the way point buy math works. And that the issue is easily resolvable. The point isn't that the Human is just better in general (even though it certainly is), it's that the human ends up with more strength despite investing the same amount of points, even though the Half-Orc is getting a bonus to Strength.


On the other hand, let us look at a different example. A pair of archers, one human, one elf. They are going to go for Dex first (up to 16), then Strength.

Elf ends up with 16 dex (6 points) and 16 strength (10 points), with 2 points leftover to spare. Which he goes ahead and uses to boost above the initial target of 16 dex to 17 dex, because he has nothing better to do with it.

Human ends up with 16 dex (10 points) and 15 strength (8 points). Looks like the elf is ahead!


If you want to push to an 18 it's even better. Because now the Elf has 18 dex (10 points) and 15 str (8 points), while the Human has 18 dex (16 points) and 10 str (2 points). Again, I think you may have missed the point of the article. The point is that applying racial penalties after point buy causes absurd skyrocketing costs for equal investments that outweighs what the penalty would normally do (e.g. in rolling) if you invest at a point when point buy costs scale. Your example doesn't involve any racial penalties being applied in that situation, so the issue does not occur in your example.

To put it another way, the half-orc is actually paying the cost of an 18 intelligence to get to 16, and this doesn't cause the half-orc's intelligence to change relative to a competitior... it causes all other statistics to suffer in a very disproportionate way, since that took up a whopping 6 extra points.

[Deophaun]

I think you missed the point of the article. It's not that the half-orc was worse. It's that it's absolutely and completely worse, with no tradeoffs (even bad tradeoffs), simply because of the way point buy math works.
No, there is a tradeoff: An orc sacrifices being a smart wizard in favor of being an awesome barbarian.

I have to ask: with this system, do you also apply stat bonuses before point-buy?

[LudicSavant]

No, there is a tradeoff: An orc sacrifices being a smart wizard in favor of being an awesome barbarian. No, they do not. This is what I'm talking about when I say the issue runs counter to intuitions. The half-orc and the human have the same intelligence, but the half-orc has 3 less strength.


I have to ask: with this system, do you also apply stat bonuses before point-buy?

No.

[Deophaun]

No, they do not. This is what I'm talking about when I say the issue runs counter to intuitions. The half-orc and the human have the same intelligence, but the half-orc has 3 less strength.
Because you've intentionally gamed the system to make the least effective use of your half-orc's racials. Yawn.

No.
Ah, so the half-orc is cheating at point buy.

[Seerow]

I think you missed the point of the article. It's not that the half-orc was worse. It's that it's absolutely and completely worse, with no tradeoffs (even bad tradeoffs), simply because of the way point buy math works. And that the issue is easily resolvable.

Except they aren't completely worse. They were completely worse in your extremely limited scenario because you specifically highlighted the character trying to prioritize a stat that they have a penalty to. Of course that is worse. Nobody ever questioned that. On the other hand, as I showed, when you highlight a situation where the penalty doesn't come into play, and you are focusing on the stat where you get a bonus, you come out significantly ahead.

Let's look at another example using the Half-Orc that even brings in the stat penalty. You want to make a standard tripping Fighter. This means you need a high strength, and at least 13int for Combat Expertise. Same 18 points to go around as before

Human: 13int (5 points) 17str (13 points) = 18

Half-Orc: 13int (8 points) 18 str (10 points) = 18.

Even when needing to boost a stat he gets a penalty to, the half-orc comes out ahead in his primary stat, strength.

Humans are a great race, but it's not because of the lack of stat bonuses/penalties. That lack just means you can use them for any concept you want. What makes them stand out is the extra feat, which they get in place of all the minor features other races get that nobody really cares about.

Similarly, Half-Orcs are a bad race. But it's not just because of their stat distribution (though them getting a penalty to two stats instead of one because developers overvalued strength is a sore point), it's because on top of that they get literally nothing but darkvision.

[LudicSavant]

I think the tongue-in-cheek title may have been confusing, so I changed it.


Except they aren't completely worse. Yes, the character is in fact completely worse. The fact that a different character would not be completely worse is missing the point.


Humans are a great race, but it's not because of the lack of stat bonuses/penalties. Never said that that was what made humans great.


Similarly, Half-Orcs are a bad race. But it's not just because of their stat distribution (though them getting a penalty to two stats instead of one because developers overvalued strength is a sore point), it's because on top of that they get literally nothing but darkvision.

Never said that that was what made half-orcs bad.

The topic is about why things like multiclassing penalties and applying racial penalties after point buy are crippling for character diversity, not an analysis of which races are better and why. I could just as easily have made the comparison between any two races as long as one had a racial penalty and another did not. I am highlighting an element of 3.5e's core design issues that apply universally... not just to any two specific races.

[Seerow]

I think the tongue-in-cheek title may have been confusing, so I changed it.

Changing the title really doesn't change that the only thing you proved is that trying to play a race with a penalty to your primary stat is going to be a bad idea. Basically your whole rant boils down to other races aren't as flexible in the archetypes they can represent as a human... which is something nobody has ever disagreed with, and kind of goes without saying.



The topic is about why things like multiclassing penalties and applying racial penalties after point buy are crippling for character diversity, not an analysis of which races are better and why.

And what you are describing is the game working as intended.

Humans are supposed to be the ultra flexible "I can do anything race". Other races then by definition are intended to be more limited in the concepts they support. The entire point of favored classes and stat penalties is to put up signs saying "hey this race isn't supposed to focus on these things, and one who tries is going to be worse at it than average".

[LudicSavant]

Changing the title really doesn't change that the only thing you proved is that trying to play a race with a penalty to your primary stat is going to be a bad idea. Basically your whole rant boils down to other races aren't as flexible in the archetypes they can represent as a human... which is something nobody has ever disagreed with, and kind of goes without saying.

And a way that you can help alleviate this issue, similar to how removing favored classes / multiclassing penalties helps alleviate the issue.

[Jormengand]

Solution: don't be an orc wizard.

Humans are the most versatile race, so it shouldn't surprise you that they're better at - le gasp - being versatile than an orc is.

[Seerow]

And a way that you can help alleviate this issue, similar to how removing multiclassing helps alleviate the issue.

In which case you should be posting over on homebrew if you want people to be talking about your proposed houserule rather than the 'problem' within the system you are describing.

[Ashtagon]

Seriously, this wasn't even news a decade ago back when 3e had just come out. Orcs make lousy wizards, by design. Orcs make decent barbarians, again, by design.

You're basically telling us something that was intentionally built into the system from the ground up.

[OldTrees1]

Before getting too sidetracked, would someone please analyse the "racial modifiers before point buy" and see what Cons emerge to offset the Pros the OP sees?

For instance
Orc Wizard (Int > Con) 16 Int, 14 Con
Before: 18 points
After: 22 points

Orc Brute (Str > Con) 16 Str, 14 Con
Before: 12 points
After: 10 points

Orc Warrior (Str > Int > Con) 16 Str, 14 Int, 12 Con
Before: 18 points
After: 18 points

Orc Warrior (Str > Int > Con) 18 Str, 14 Int, 12 Con
Before: 26 points
After: 20 points

Orc Warrior (Str > Int > Con) 16 Str, 16 Int, 12 Con
Before: 22 points
After: 24 points

Edit: Forgot the +2 Con bonus

[nedz]

Opening with an appeal to authority and then patronising everyone doesn't go down well — more news at eleven.

[Renen]

This man's on a roll with some guides! Go go!

[eggynack]

As has been noted, you've designed a character that's intended to be excellent at two things, strength and intelligence, and you've given them a race that makes them better at one of those things and worse at the other. Your race is acting at cross purposes with one of the goals of your character, so the character finds conflict. Why are you beginning with setting your intelligence score if your primary goal is high intelligence, and why are you picking half-orc when there isn't a separate incentive beyond stats for doing so (like substitution levels)? It does your argument a disservice to pick only the most extreme and illogical scenario.

Let's look at something a bit more realistic, as something of an expansion of seerow's efforts. Thus, instead of using the crappy half-orc, we'll use the far more useful water orc. This race gets +4 strength, +2 constitution, and -2 intelligence as its key stats for the situation. Under your weird 18 point buy situation, your human has 16 intelligence and 15 strength. Our "flabby" water orc, using the same starting assignment, has 14 intelligence, 19 strength, and 10 constitution, where that latter stat is likely to be extremely relevant. Yes, we needed to drop intelligence some, but isn't that the whole point of using a low intelligence race, that you'd wind up with lower intelligence and higher strength than the moderate humans?

Let's make it even more realistic, actually. 18 point buy isn't really a thing, after all, and both 10 and 8 constitution are unrealistic on a wizard intended for battle. Thus, let's set point buy at 25, and because it's actually a generally agreed upon stat assignment for wizards, let's set constitution at 14 (see treantmonk's guide to wizards, which is inflexible with that assignment at the cost of all else). Now the human has, say, 14 strength, 14 constitution, and 17 intelligence for a rather normal strength setup, and the water orc has, say, 20 strength, 15 constitution, and 14 intelligence.

Again, our water orc will be significantly better at grappling, and capable of casting, and it will be in that fashion because that's what the character was designed to do. You wanted a grapplemancer, and that's what you get. A character capable of out-strengthing even the strongest humans, backed up by magical force, and such a setup wouldn't even be possible on a human. The trick is, if you want to do a concept, you do it right. You think that the racial system doesn't let you pursue weird concepts, but that's because you refused to go all in, picking a below average race, and continuing to set your race based on what you think a wizard should or must be. You put intelligence first because you cared more about casting ability than grappling ability, and you paid the price for it, but the orc wizard will be grappling opponents capably.

[Seerow]

Before getting too sidetracked, would someone please analyse the "racial modifiers before point buy" and see what Cons emerge to offset the Pros the OP sees?

It basically makes the penalties less harsh or anyone who wants to get above a 12 in the stat.

Stat | Before | After
12 | 6 | 6
13 | 8 | 7
14 | 10 | 8
15 | 13 | 10
16 | 16 | 12
17 | na | 15
18 | na | 18


So using the examples from earlier in the thread:
Half-Orc Gish: 16int (12 points) 16 str (6 points)
Half-Orc Fighter: 13int (7 points) 18 str (10 points) [1 point left over]


Basically it makes sure that a race with a stat change is going to come out ahead in stats anytime their boosted stat is desirable, regardless of the penalty.

[LudicSavant]

It basically makes the penalties less harsh or anyone who wants to get above a 12 in the stat.

Stat | Before | After
12 | 6 | 6
13 | 8 | 7
14 | 10 | 8
15 | 13 | 10
16 | 16 | 12
17 | na | 15
18 | na | 18

So using the examples from earlier in the thread:
Half-Orc Gish: 16int (12 points) 16 str (6 points)
Half-Orc Fighter: 13int (7 points) 18 str (10 points) [1 point left over]

Basically it makes sure that a race with a stat change is going to come out ahead in stats anytime their boosted stat is desirable, regardless of the penalty.

Pretty much. The intention is to lower the weight of "trap" options while not really impacting the optimal options on a basic systemic level. Note that 4e's designers caught on to this and implemented it by simply removing racial penalties entirely (not that they didn't futz up the math in many other areas...).

[OldTrees1]

It basically makes the penalties less harsh or anyone who wants to get above a 12 in the stat.

Stat | Before | After
12 | 6 | 6
13 | 8 | 7
14 | 10 | 8
15 | 13 | 10
16 | 16 | 12
17 | na | 15
18 | na | 18


So using the examples from earlier in the thread:
Half-Orc Gish: 16int (12 points) 16 str (6 points)
Half-Orc Fighter: 13int (7 points) 18 str (10 points) [1 point left over]


Basically it makes sure that a race with a stat change is going to come out ahead in stats anytime their boosted stat is desirable, regardless of the penalty.

+4
score | mod before | mod after
14 | 2 | 2 (equal costs here and lower)
15 | 4 | 3
16 | 6 | 4
17 | 9 | 5
18 | 12 | 6
19 | na | 8
20 | na | 10
21 | na | 13
22 | na | 16

+2
score | mod before | mod after
14 | 4 | 4 (equal costs here and lower)
15 | 6 | 5
16 | 8 | 6
17 | 11 | 8
18 | 14 | 10
19 | na | 13
20 | na | 16

-2
score | mod before | mod after
12 | 6 | 6 (equal costs here and lower)
13 | 7 | 8
14 | 8 | 10
15 | 10 | 13
16 | 12 | 16
17 | 15 | na
18 | 18 | na

Cons:
Using racial modifiers before point buy lowers the maximum for boosted stats (just as it raises the maximum for penalized stats)
Using racial modifiers before point buy makes boosted stats more expensive (just as it makes penalized stats less expensive)

[Seerow]

To be clear, I was showing the chart for -2, with before being the current costs, and after being the new costs with the proposed change.

[georgie_leech]

No opinion one way or the other, but it seems the discussion on what it does to pointbuy misses that OP is saying that Racial Penalties are applied before and not bonuses. You can still have a STR 20 Half-orc.

[OldTrees1]

No opinion one way or the other, but it seems the discussion on what it does to pointbuy misses that OP is saying that Racial Penalties are applied before and not bonuses. You can still have a STR 20 Half-orc.

Oh. I did miss that.

[Chronos]

The whole point of racial ability modifiers is that some races are supposed to be better at some things than others. You can't fix that, because it's already fixed. Any attempted "fix" will just break things.

[Troacctid]

To be honest, when I saw the thread title, I was expecting it to be about the complete opposite phenomenon--how Orcs often end up with higher mental stats than Humans, because they can get the same Strength score with a lower point-buy investment and use their savings to improve what would otherwise be dump stats.

[Mando Knight]

The whole point of racial ability modifiers is that some races are supposed to be better at some things than others. You can't fix that, because it's already fixed. Any attempted "fix" will just break things.

The other thing is that the intent of point-buy is to roughly mimic a normalization of rolling dice: scores closer to 10 cost less because you're much more likely to roll them, while much higher scores cost a lot more because they're appropriately harder to roll. The Half-Orc wizard with 16 Int has to spend a lot more points because it has to buy up to 18 first, which is much rarer than a 16.

Now, if you decide that you want 14 Intelligence and 16 Strength, then the Half-Orc and Human break even: the Half-Orc buys to 16 Int and 14 Strength, and the Human to 14 Int, 16 Str, and the more you prioritize Strength over Intelligence, the better the Half-Orc gets, as intended. If you decide to start a character with 18 Strength and then assign mental stats with whatever's left over, then the Half-Orc comes out as far ahead as the Human did in the original case study (where the goal was 16 Int and then spend the rest on physical stats).

[LudicSavant]

The other thing is that the intent of point-buy is to roughly mimic a normalization of rolling dice: scores closer to 10 cost less because you're much more likely to roll them, while much higher scores cost a lot more because they're appropriately harder to roll. The Half-Orc wizard with 16 Int has to spend a lot more points because it has to buy up to 18 first, which is much rarer than a 16.

Now, if you decide that you want 14 Intelligence and 16 Strength, then the Half-Orc and Human break even: the Half-Orc buys to 16 Int and 14 Strength, and the Human to 14 Int, 16 Str, and the more you prioritize Strength over Intelligence, the better the Half-Orc gets, as intended. If you decide to start a character with 18 Strength and then assign mental stats with whatever's left over, then the Half-Orc comes out as far ahead as the Human did in the original case study (where the goal was 16 Int and then spend the rest on physical stats). Yep. The math works out completely differently with rolling. It is just as impossible to have the "literally no tradeoff" situation with rolling as it is to have it with point buy when you apply penalties before costs rather than after.


To be honest, when I saw the thread title, I was expecting it to be about the complete opposite phenomenon--how Orcs often end up with higher mental stats than Humans, because they can get the same Strength score with a lower point-buy investment and use their savings to improve what would otherwise be dump stats.

True! Half-Orcs also "cheat" at point buy, because a 0 vs a +2 has the same relationship as a -2 vs 0. The irregularity is a core aspect of the underlying mathematical system. Take case:

Prioritize Strength, then Int:

Human = 18 Strength (16 point cost), 10 Int (2 point cost), total 18 spent
Half-Orc = 18 Strength (10 point cost), 14 Int (8 point cost), total 18 spent

So, the end result character effectively has racial adjustments of +0 Strength /+4 Intelligence.

You can fix this irregularity too (in basically the same way), but I didn't want to mess with the more optimal cases.

[georgie_leech]

The whole point of racial ability modifiers is that some races are supposed to be better at some things than others. You can't fix that, because it's already fixed. Any attempted "fix" will just break things.

On the other hand, I'd expect a Half-Orc Wizard to be stronger than their Human counterpart. Is there something that would make it so that the naturally stronger race has a more difficult time staying fit while studying?

[eggynack]

On the other hand, I'd expect a Half-Orc Wizard to be stronger than their Human counterpart. Is there something that would make it so that the naturally stronger race has a more difficult time staying fit while studying?
A half-orc wizard is stronger than their human counterpart. They're also less intelligent. If you try to act against the intelligence reduction, then you're going to sacrifice the strength advantage and then some. I mean, really, basic doings, start with a 16 in strength and intelligence on a human, keep the points right where they are, and the half-orc will have an 18 in strength and a 14 in intelligence.

[georgie_leech]

A half-orc wizard is stronger than their human counterpart. They're also less intelligent. If you try to act against the intelligence reduction, then you're going to sacrifice the strength advantage and then some. I mean, really, basic doings, start with a 16 in strength and intelligence on a human, keep the points right where they are, and the half-orc will have an 18 in strength and a 14 in intelligence.

I suppose my point is, why is it that when you try to shore up a weakness, your talents are neglected to the point that an average person is better at it than you are?

[eggynack]

I suppose my point is, why is it that when you try to shore up a weakness, your talents are neglected to the point that an average person is better at it than you are?
Because getting a high score is universally difficult, and on a half-orc, a 16 intelligence is a high score. If a human wizard tries to put an 18 in intelligence, then an average person will often be better at their less important stats than they are. The same holds true here. This really just seems like the game working the way it's supposed to to me.

[Troacctid]

I suppose my point is, why is it that when you try to shore up a weakness, your talents are neglected to the point that an average person is better at it than you are?

An average person has 10 or 11. No half-orc will ever have less than a 10 in Strength if you use point buy--the minimum is 8, and they have a +2 racial bonus. So the answer is they won't be worse than an average person, they'll be the same.

[Curmudgeon]

One is a half-orc. She spends 16 points for a 16 in Intelligence, and puts her remaining points into Strength. End result? 16 int, 12 str.
18 points spent.

The other is a human. She buys up to 16 Intelligence, then puts her remaining points into Strength. End result? 16 int, 15 str.
18 points spent.

You probably notice something weird is going on by now.
Yes, I noticed the weirdness: you're purposely obfuscating the actual rules with your examples (emphasis added in RED). Point buy always has the same costs, and is finished before racial modifiers are applied, but you're treating those later racial modifiers as part of the point buy process.

Please stop.

[LudicSavant]

Point buy always has the same costs, and is finished before racial modifiers are applied, but you're treating those later racial modifiers as part of the point buy process. Uhm, what? No. If a half-orc character wants a 16 intelligence, they pay 6 more points than a human character does. That is not an "obfuscation," that is a fact of the math. The cost for the same stat is not the same, and the effect of the inflated costs is to lower other stats.

[Curmudgeon]

Uhm, what? No. If a half-orc wants a 16 intelligence, they pay 6 more points than a Human does.
And if you roll dice instead, it's nigh-impossible to get that 16 INT as a Half-Orc. Point buy is fairer in that it makes working against racial modifiers to such an extent a guaranteed success instead of likely impossible.

So what are you complaining about?

[atemu1234]

Solution: don't be an orc wizard.

Humans are the most versatile race, so it shouldn't surprise you that they're better at - le gasp - being versatile than an orc is.

I agree quite firmly with Jormengand on this.

[LudicSavant]

I'm seeing a lot of responses like "just don't roleplay an orc wizard. You should totally be punished if you do."

The effect of rules like multiclassing XP penalties, favored classes, and inflated point buy costs is mostly to make suboptimal things more suboptimal, which serves to constrict the range of viable concept options. If that's what you want, then great, I guess. *Shrug*

[Flickerdart]

I'm seeing a lot of responses like "just don't roleplay an orc wizard. You should totally be punished if you do."
You can totally roleplay an orc wizard. As long as you don't play an orc wizard, you're fine.

[Zanos]

Pretty much. The intention is to lower the weight of "trap" options while not really impacting the optimal options on a basic systemic level. Note that 4e's designers caught on to this and implemented it by simply removing racial penalties entirely (not that they didn't futz up the math in many other areas...).
I think calling them "trap" options is a bit dishonest. It says right in the Wizard description that intelligence is important and strength is not, so picking a race that has a penalty to the important stat and a bonus to the unimportant isn't a trap, it's just picking an inherently poor option.

3.5 does have a lot of trap options(Two Weapon Fighting, Tower Shields), but "Races with intelligence penalties make poor wizards" is not a trap option, because the book says that they do.

Personally, I have no problem with racial stat adjustments pigeonholing certain races. Elves tend towards archery and finesse because they have a dex bonus and a con hit, Orcs are good melee combatant because they get physical stat bumps, but their lower mental capacity causes their societies to be more primitive and not be particularly good at arcane magic, Dwarves are hardy, etc.

I guess overall whether or not you think this is okay is whether or not you have a problem with racial stereotypes being mechanically enforced, which I really don't.

[Drelua]

I'm definitely going to have to agree that this isn't a problem. The fact is, orcs are supposed to be dumb and wizards need to be smart to be, well, wizards. If you want to play a dumb race on a class that needs to be smart, great, have fun. If you think you should be just as good at magic as a person who's naturally more inclined to that sort of thing, then that's honestly a bit ridiculous.

This sounds to me like someone complaining that short people should be able to be just as good at basketball as tall people. That's just not how basketball works. If you train against a disadvantage, it doesn't just go away. Being naturally ill-suited for something makes you bad at that thing, that's how it works.

Simply put, this is a case of the rules supporting the fluff. The fluff says orcs aren't very smart or good at magic, and the rules say the same thing. That's just good game design.

[eggynack]

I'm seeing a lot of responses like "just don't roleplay an orc wizard. You should totally be punished if you do."

The effect of rules like multiclassing XP penalties, favored classes, and inflated point buy costs is mostly to make suboptimal things more suboptimal, which serves to constrict the range of viable concept options. If that's what you want, then great, I guess. *Shrug*
The -2 to intelligence is supposed to be a downside, and it's a downside you're reducing by doing this. Downsides restrict options by design. Multiclassing XP penalties aren't bad because they're a downside, or even because they restrict options, but because they're a downside that apply only to the weakest characters. By contrast, there are plenty of perfectly optimal characters out there that make use of races with ability score penalties, and this rule aids those significantly.

Take the classic gray elf wizard, for example, now capable of boosting constitution to around normal while gaining the advantage of bumps to their other primary stats and substitution levels to boot, or hell, even the desert half-orc druid, now able to lift up a tertiary stat to a reasonable place while doing usual awesome desert half-orc druid things. This even opens up the jermalaine as a more viable druid option, as the drops to constitution, intelligence, and charisma are no longer nearly as important, thus making a +6 wisdom race even better. You've decided on one of the absolute worst character designs possible for your example, and you got a bad character out of it (shock and awe), but not all characters who would make use of this rule are half-orc wizards. By contrast, most of the things you'd do with a lifting of the multiclassing penalty rules are worse than things you could do without them, and the design on them is intrinsically silly. There is little parity here.

[LudicSavant]

I guess overall whether or not you think this is okay is whether or not you have a problem with racial stereotypes being mechanically enforced, which I really don't.

I guess this is why I'm a bit surprised by some of the reactions here. Everyplace else I've brought this up public opinion seemed to be going in the opposite direction. They wanted their orc wizards and dwarf ninjas and so forth, and thought that enforcing racial stereotypes was a downside for the game's design. And I tend to agree with them. Let me see if I can explain why.

Actually, I think one of the best places I've seen this explained was another post by FrankTrollman. I'll see if I can dig it up... here:



Well, if you want to render race irrelevant to your options and your ability at what options you have, you say that powerful archmages can pick "orc".

If you want it to matter, then you pick orc first and then have any of the things orcs are capable of qualifying for open.

Sigh. Space Jam!

Look, if there's even one Chinese man in 1.3 billion people who can play in the NBA as a giant, then you are willing to concede that this is something a Chinese man is "capable" of qualifying for. The problem is that you're then reversing yourself when talking about other scenarios where the population isn't even a billion.

Or to put it another way:

Being 3 standard deviations away from the mean of a demographic group you belong to does not make you be of a specific power level. Being a crazy tall pygmy is insufficient to be a basketball superstar. Height is measured against other NBA superstars when determining power. Being crazy tall as a Chinese man puts Yao Ming into the 7th standard deviation. Having a Kenyan come out that tall would him in like the 4th or 5th standard deviation. It would be more likely for "a Kenyan" chosen at random to have the height and athleticism to be an NBA superstar than to for "a Chinese guy" to be. After all, there are only four Chinese NBA players, and Kenya fields more than that out of a population that is 1/38th the size.

Truly your chances of becoming an NBA superstar as a man from China are laughably small compared to the Kenyans who are per capita literally a hundred times more likely to succeed at Basketball. But see, that doesn't matter. At all. A Space Jam character isn't selected at random from the people of his home nation or even the people of the world. He's not even selected at random from people in the Basketball industry or even people in the NBA. He's a specific, exceptional, main character, NBA superstar. He's not chosen randomly at all. He's constructed by the player and is a unique individual who by definition is one of the twelve most exceptional basketball players on the planet.

Yes, less than 1% of 1% of 1% of 1% of Yao Ming's people can be like Yao Ming. But that means **** all for the purposes of making a Space Jam character. His height stat is balanced for a character in his power level (over nine thousand), and the fact that he's Chinese makes the height stat chosen "extremely rare." But not unbalanced. If Yao Ming's player had chosen to make him Kenyan instead, his height would be astoundingly less exceptional. Like literally a hundred or a thousand times less exceptional. But you don't create your character based on how exceptional they are, but by how powerful they are. Where people fall on demographic bell curves doesn't matter - the only thing that matters is where people fall on power scales.

-Frank

I'd like to raise 3 points here:

1) Levels are supposed to measure power level. How "rare" it is for someone to be at that power level doesn't matter... a Sorcerer might not have had to work as hard as a Wizard, but his level isn't about that. Demographics stats in the D&D sourcebooks say some classes are rarer than others, but they don't take taxes for that. Superman is not a lower level character than Batman. Ideally, two characters of the same level would be close to each other in power. Clearly this isn't the case in 3.5e, but the closer it is to being the case, the better.

2) The average orc is not a relevant demographic for "epic adventurer heroes who are the stuff of legends" any more than "average Chinese man" is a relevant demographic for "Space Jam." If I was making a Space Jam game, I would like people to comfortably be able to play as Yao Ming. If I'm making an epic fantasy adventure game, I would like people to be able to comfortably play as a legendary exception. Epic fantasy is not a genre about average people. Your PC is a specific, exceptional main character.

3) I believe that there is a very real demand for playing Orc Wizards and the like, and that many of these people really do feel like their friends are kicking them in the shins for even suggesting it... even if that demographic hasn't been posting much in this thread. I have seen countless threads pop up over the years complaining about it, or even with people feeling like they've been bullied by their parties about it to the point that they want to leave the game. If there is no such desire in your group, then hey, whatever. As for me, I want more of this: https://www.pinterest.com/pin/307089268315271751/

[Vhaidara]

I'm seeing a lot of responses like "just don't roleplay an orc wizard. You should totally be punished if you do."

No, you aren't being punished. Being punished is making something, then having a penalty added at the end. You go to make an Orc Wizard, you know from the outset that you WILL be weaker than a wizard of a better suited race. That's the point of penalties: you aren't as good at some things.

You treat a 16 Int orc as the equivalent of a 16 Int human. It isn't. a 16 Int orc is comparable to an 18 Int human: The highest possible starting spot.

[eggynack]

1) Levels are supposed to measure power level. How "rare" it is for someone to be at that power level doesn't matter... a Sorcerer might not have had to work as hard as a Wizard, but his level isn't about that. Demographics stats in the D&D sourcebooks say some classes are rarer than others, but they don't take taxes for that. Superman is not a lower level character than Batman. Ideally, two characters of the same level would be close to each other in power. Clearly this isn't the case in 3.5e, but the closer it is to being the case, the better.
Race isn't a significant factor of power level, and to the extent that it is, it's more like choosing a bad feat than like choosing superman over batman. Maybe make that two bad feats, if you're choosing an actively bad race.


2) The average orc is not a relevant demographic for "epic adventurer heroes who are the stuff of legends" any more than "average Chinese man" is a relevant demographic for "Space Jam." If I was making a Space Jam game, I would like people to comfortably be able to play as Yao Ming. If I'm making an epic fantasy adventure game, I would like people to be able to comfortably play as a legendary exception. Epic fantasy is not a genre about average people. Your PC is a specific, exceptional main character.
Race in game is supposed to be a bigger thing than race in reality. It's not just about averages, but about actual fundamental differences in nature.


3) I believe that there is a very real demand for playing Orc Wizards and the like, and that many of these people really do feel like their friends are kicking them in the shins for even suggesting it... even if that demographic hasn't been posting much in this thread. I have seen countless threads pop up over the years complaining about it, or even with people feeling like they've been bullied by their parties about it to the point that they want to leave the game. If there is no such desire in your group, then hey, whatever.
But orcs can be wizards. It's two intelligence, not the end of the world. A wizard can start with eleven intelligence and only be behind in spell level at three. This isn't as big of a deal as you think it is.

[Zanos]

I guess this is why I'm a bit surprised by some of the reactions here. Everyplace else I've brought this up public opinion seemed to be going in the opposite direction. They wanted their orc wizards and dwarf ninjas and so forth, and thought that enforcing racial stereotypes was a downside for the game's design. And I tend to agree with them. Let me see if I can explain why.
I'm actually a bit confused by this. Are you using this as a specific example? Dwarves have no ability score penalty and aren't really bad rogues. Unless you're suggesting that each race be equally good at everything, which seems to remove the point of selecting a race as a mechanical option(not that there's anything wrong with that, but it's a more major change than what you suggested in your OP.)

1) While I agree with your point that two characters of similar level should, ideally, be of similar strength, that is rarely the case. Even if you throw out race as making some people better Wizards, a Wizard that spent all his feats on exotic weapon proficiency feats probably isn't as powerful as a Wizard who spent all his feats on things relating to magic. Every option isn't equally good for every class.

2) Yes, the PCs are special. A 16 int Orc is one of the smartest Orcs in existence. He isn't as intelligent as a 20 int Gray Elf who is one of the smartest members of one of the smartest races in existence. I see no problem with the smart guy of the smart guy race being smarter than the smart guy of the dumb guy race. You may also have noticed that people running directly against the grain has caused, in my opinion anyway, a lot of really bad characters. I've had issues with players who needed their character to be super special beyond being a PC, and it usually didn't end well(Drizzt Syndrome, heyo.)

3) I don't have a problem with people selecting less than optimal race/class combos in my games, but I expect them to be able to contribute to the party, both as a DM and a Player. 2-4 int isn't going to make a Wizard that was otherwise played quite well suddenly terrible, and I've never complained about someone's build unless they were worthless in combat. Like the transmuter who's only prepared spells were ALL animate rope.

[Vortenger]

I'm a bit surprised by the tone of response in this thread.

[OldTrees1]

I'm a bit surprised by the tone of response in this thread.

Yeah the tone is surprising. The conclusion is less so once we looked into the math (it makes a penalty so much less significant than a bonus that it is problematic). But I would have expected the intent of the suggestion to be warmly received.

[LudicSavant]

It makes a penalty so much less significant than a bonus that it is problematic.

That seems like a potentially productive line of inquiry.

Can you give a specific example of it becoming problematic?

When I think problematic, I would think "outclassing the better LA 0 races in the roles they're best at" or "noticeably buffing them at the roles they're best at." And when I think of better LA 0 races, I think of, say...

- Strongheart halfings (like small Humans)
- Azurin (like Humans but with bonus essentia)
- Silverbrow humans (like Humans but with the dragonblood subtype)
- Whisper Gnomes
- Elans
- Warforged
- Warforged Scout
- Lesser Tiefling
- Lesser Aasimar
- Dragonborn of Bahamut template applied to something appropriate (like Lesser Aasimar)
- Deep Dwarves
- Grey Elves
- Star Elves
- Water Orcs
- Air Goblins
- Water Halflings
- Mongrelfolk
- Hadozee
- Darfellans (their bites get iterative attacks from a good base attack bonus)
- Hellbred (both Mind Hellbred and Body Hellbred are good)
- Spellscales

[OldTrees1]

That seems like a potentially productive line of inquiry. Can you give a specific example of it becoming problematic?

When I think problematic, I would think "outclassing the better LA 0 races in the roles they're best at." And when I think of better LA 0 races, I think of, say...

Well it was highlighted that the same problem seen with 16 Int Humans and Orcs is reversed with 16 Str Humans and Orcs. Your solution weakens the impact of the penalties (or bonuses if applied to the bonuses). If we apply your idea to both penalties and bonuses then we just decreased the point of racial ability modifiers. If we apply your idea to only one, then we are showing favoritism towards either Half Orcs(races with modifiers) or Humans(races without modifiers) respectively.

Now this problem occurs prior to looking at any specific race, but I follow the philosophy of "Don't use a broken mechanic to fix an extant problem".

PS: When I think problem, my mind first thinks of a Math Textbook or a Puzzle book. So I do not imply any particular scale or attitude when I use that word.

[LudicSavant]

If we apply your idea to only one, then we are showing favoritism towards either Half Orcs(races with modifiers) or Humans(races without modifiers) respectively.

Can you give an actual, specific, practical example of it being problematic from a balance perspective, though? Like, a case where the approach creates an option that is actually better than the old optimal options?

The intention is to impact the "skewness (http://en.wikipedia.org/wiki/Skewness)" of D&D's options in a positive manner without really changing the powerful side much. If you can show me an example where the optimal options change meaningfully, I will revise my position.

You talk about "favoritism" against humans, and I absolutely understand where you're coming from. It really does make bonuses more impactful than penalties (although penalties were already less impactful than bonuses, since they would get tossed into dump stats by optimal configurations unless we're talking about something nobody wants to dump like Constitution). But are the optimal configurations really changing much? Note that people often talk about how, say, Hobgoblins would probably be okay at LA+0 as written. Or that Lesser Aasimar already have two bonuses with no penalties, in addition to some passably decent racial features.


If we apply your idea to both penalties and bonuses then we just decreased the point of racial ability modifiers.

If I was remaking the game from the ground up, I'd honestly probably do something like this. It's an increasingly popular approach in game design to undermine differences in racial statistics in favor of other ways of making races unique (See things like Guild Wars 2, for instance), or simply put races on the same resource buying tracks with different bases, partly because of some of the points I referenced above. I haven't really tried to apply such a system to D&D 3.5e, however.

[OldTrees1]

Can you give an actual, specific, practical example of it being problematic from a balance perspective, though? Like, a case where the approach creates an option that is actually better than the old optimal options?

The intention is to impact the "skewness (http://en.wikipedia.org/wiki/Skewness)" of D&D's options in a positive manner without really changing the powerful side much. If you can show me an example where the optimal options change meaningfully, I will revise my position.

You talk about "favoritism" against humans, and I absolutely understand where you're coming from, but I daresay I suspect that humans will remain just fine in the scheme of things.

Can you name 2 races that are balanced with respect to the same goal despite one having relevant racial ability modifiers(bonuses or penalties) and the other not having relevant racial modifiers? Remember the 2 races are balanced despite this.

If you can show me those 2, then I would use those 2 as specific examples. For now, I will call them +/-, and 0 respectively.

Both races are aiming at a 16 in the stat +/- has the relevant ability modifier in.
When balanced, this cost 0 10 points and cost +/- either 6 or 16 points.
With the change the cost to +/- changes to 8 or 12 points.
If the races in question were balanced, they now are less balanced.

So this change is throwing imbalance after imbalance at the hopes that it balances out. I personally consider that bad game design. I am much more inclined to fix existing imbalance itself instead.

[LudicSavant]

I could suggest a fix that was more complete rather than a stopgap band-aid, I suppose. I picked the one I did for brevity rather than completeness of solution. In retrospect I am not sure why I thought that was a good idea, especially since my original draft has a better fix. Maybe I thought it was too long because I subconsciously fear TL;DR responses? :smallconfused:

[OldTrees1]

I could suggest a fix that was more complete rather than a stopgap band-aid, I suppose.

Math is quirky, there probably is some solution as simple as this that does not have any problems. A Higher number of points per character seems to approach the desired result.

[LudicSavant]

A Higher number of points per character seems to approach the desired result.

I am not sure I see how.

[eggynack]

Can you give an actual, specific, practical example of it being problematic from a balance perspective, though? Like, a case where the approach creates an option that is actually better than the old optimal options?
I already gave you several. Gray elf on wizard, and desert half-orc on druid are already highly optimal options, and this would make them better. Jermalaine is still probably worse than anthropomorphic bat, but I don't think that's exactly helping your argument, and it too was optimal prior to this, and significantly moreso afterwards. And actually, while I'm at it, I'll add shifter druid to the list too. Great race, weighed down by some bad ability score adjustments. There are probably others I'm missing, but those seem to top the list given they're high powered options attached to the highest powered classes in the game. Powerful races with constitution penalties, like gray elf, are probably a good place to look, because everyone loves constitution.

[LudicSavant]

I already gave you several. Gray elf on wizard, and desert half-orc on druid Thanks, didn't see these before. And I wasn't aware of Jermalaines.


Powerful races with constitution penalties, like gray elf, are probably a good place to look, because everyone loves constitution. Yeah, that figures. That does seem to be an exception to the tendency to simply dump the stat with a penalty completely. I shall revise my position, then. Looks like I mucked this one up.

[OldTrees1]

I am not sure I see how.

Higher point buy -> The further each character goes into the realm of diminishing returns -> The smaller the difference in benefit compared to the total package. -> Breaking stereotypes is easier.

Aka if race X is 4 points more efficient at point buy than race Y for concept Z, it is a bigger deal at 4/25 than at 4/36.

Likewise
Higher point buy -> -2 races are getting 16s to the +2 race's 20 rather than 12s to 16s. -> Since 16 is good enough even if bigger is better ... -> Breaking stereotypes is easier.

[LudicSavant]

High point buy -> The further each character goes into the realm of diminishing returns -> The smaller the difference in benefit compared to the total package.

Aka if race X is 4 points more efficient at point buy than race Y for concept Z, it is a bigger deal at 4/25 than at 4/36.

Drowning the issue, eh? Hardly seems elegant, though. It seems like it would qualify for some of the complaints you voiced about the idea of reducing the impact of both bonuses and penalties.

[OldTrees1]

Drowning the issue, eh? Hardly seems elegant, though.

Kinda, yeah. I agree there. However I see it as a happy side effect of balancing SAD vs MAD.
But to avoid too much drowning, I try to enable 16s without giving enough points to max out multiple stats. So racial modifiers still are relevant, but the baseline is high enough for Orc Shamans.

Still, there is probably an even better solution waiting to be found.

[LudicSavant]

Kinda, yeah. I agree there. However I see it as a happy side effect of balancing SAD vs MAD.
But to avoid too much drowning, I try to enable 16s without giving enough points to max out multiple stats. So racial modifiers still are relevant, but the baseline is high enough for Orc Shamans.

Still, there is probably an even better solution waiting to be found.

Actually, I wonder. Would it really be a bad idea if both penalties and bonuses applied before point buy, even though it simply serves to make stat adjustments less impactful? In that case, neither bonuses nor penalties would suffer the odd scaling issue. It would make very high stats (higher than 18) notably more difficult, but consider that this would hurt SAD more than it would hurt MAD. On the other hand, I can see the argument being that you wouldn't want to wall out those 22 Strength orc berserkers.

Nah. It wouldn't work.

[eggynack]

Drowning the issue, eh? Hardly seems elegant, though. It seems like it would qualify for some of the complaints you voiced about the idea of reducing the impact of both bonuses and penalties.
Penalties, yes, but bonuses less so. The biggest benefit of bonuses tends to be that they give you a higher top end, allowing for stuff like 20 int wizards, 22 strength barbarians, and 24 wisdom druids. The bonuses are also relevant when they apply to a secondary stat, but it's somewhat lesser. Notably, high point buy tends to favor low tier classes some, both because they tend to be MAD, and because a small benefit tends to be a bit less relevant on a high power character. You shouldn't necessarily push massive point buys, because at some point you're talking something akin to all 18's, which would indeed lead to a drowning problem, but above average tends to be pretty helpful.

[OldTrees1]

Actually, I wonder. Would it really be a bad idea if both penalties and bonuses applied before point buy, even though it simply serves to make stat adjustments less impactful? In that case, neither bonuses nor penalties would suffer the odd scaling issue. It would make very high stats (higher than 18) notably more difficult, but consider that this would hurt SAD more than it would hurt MAD. On the other hand, I can see the argument being that you wouldn't want to wall out those 22 Strength orc berserkers.

It would dilute racial differences. Based on a comment of yours, I think you would like this. However I like the racial differences since it makes them feel more different despite being surmountable.

[rockdeworld]

Edit: So apparently I just summed up what everyone else posted except for one or two points.
LudicSavant, this article is bad. It looks like you made one observation about the game and decided to write it up in an official-looking post. I can see just how bad because I've written articles like it before. Please do more research before posting an article to inform other people how the game works.

Here are the problems:
1. You claim point buy hurts races with racial penalties but don't compare it to any standard. Assuming a standard of dice rolling with an elite array average, your same example would result in (assuming 14 str/15 int) an orc with 16 str/13 int and a human with 14 str/15 int. Even though the orc loses a +1 int mod compared to the human, he gains a +1 str mod - exactly what the bonus/penalty system was meant to do. So the claim should be that relative to the dice-rolling system, point buy is worse with racial penalties. But even that approach would be obly half-true, because the point of a point-buy system is that it trades the potential highs and statistical balance of dice rolls for the security of knowing what the result will be. The reason I prefer point buy (and competitions use it) is that I know my character's stats will adhere to some baseline of power, rather than being potentially god-like or not-worth-playing. And racial penalties don't take away from that, so in that sense they rally don't cheat the system.
2. You compared a human with a half-orc and said that the half-orc got nothing in exchange for its racial penalties, which is totally false, because you didn't consider the rest of the compairson. If what you suggested were the obly considerations, no one would ever want to play a kobold for mechanical strength, because of their stat penalties. But kobolds mechanically make some of the best sorcerers, despite their stat penalties. The dull comparison includes (1) humans also get a bonus feat and an extra skill point, plus access to human-only feats, but (2) half-orcs get darkvision and half-orc-only feats, all of ehich are applicable to a melee character like the one in your example. So racial penalties don't suck for point-buy with no advantage, they're just one factor in a bundle that help determine a race's mechanical strengths and weaknesses. (A race might still have no advantage, but it's not based solely on stat penalties.)
3. Your point was that if you want to play a class/race combination where the race has a penalty to a primary stat of the class, you shouldn't use point buy. But instead of suggesting an official variant that already deals with this problem (dice rolling), you offered a homebrew solution that you claimed was better without any analysis to back it up. Yes, races with penalties have their penalties reduced under your system. How does that affect game balance? That darkvision/feat access suddenly seem a lot better when my stats are going to be almost the same regardless of my race. (This question is hypothetical. If half-orcs are still worse, we can just find a race that isn't.) Another example: grey elves have 2 racial stat penalties. But they're considered the best race for a wizard in core. Your house-rule makes them even more powerful and widens the gap with other races, rather than closing it (as I guess was your intention). The fact that this type of exception wasn't addressed in your post undermines its validity.
4. And how does your home-brew rule compare to dice-rolling? Does it result in higher average stats? The same?
5. I never saw what your goal actually was (possibly I missed it). Do you want to bring races closer together in mechanical power? Or do you just want everyone to reduce the effect of racial penalties in their games? This article strikes me as the latter, and I can guarantee it will fail that goal, because we as people don't do things just because we're told. This could've been an interesting debate thread (do we want to reduce the effect of racial penalties in this way?), but by framing this as a problem and providing a 'solution', you've killed the possibility of that by making everyone who disagrees with you look like they support a bad way of gaming (despite still not defining why it's bad).

My point is that your article is doing more bad than good right now. By all means write more articles, but do the research before posting them, and include the results in your post. We don't need more half-baked ideas about how the game should work - we get enough from WotC itself.

Edit: please excuse typos, as I wrote this up on a phone.

[Feint's End]

@rockdeworld: sometimes it lays to look back at the thread if you write a wall of text :smallamused:

On topic (the resolved one): I think it's pretty wrapped up now but I like the suggestion of simply using higher point buys. They actually fix a lot of the problems as long as they aren't arbitrarily high. My personal sweet spot is somewhere between 36 and 40 (I like to find a way of adding some odd numbers though so it gets less predictable).

[atemu1234]

@rockdeworld: sometimes it lays to look back at the thread if you write a wall of text :smallamused:

On topic (the resolved one): I think it's pretty wrapped up now but I like the suggestion of simply using higher point buys. They actually fix a lot of the problems as long as they aren't arbitrarily high. My personal sweet spot is somewhere between 36 and 40 (I like to find a way of adding some odd numbers though so it gets less predictable).

My IRL players use 36 point-buy.

[Chronos]

Applying racial modifiers before point buy almost always means that the racial modifiers effectively don't exist, and the one case where they do end up existing is when one of the modified scores is a dump stat. In which case it unambiguously hurts you if the dump stat is the one you get a bonus to, because you can't dump it as far, and unambiguously helps you if the dump stat is one you get a penalty to, because you can dump it deeper. But how far you can dump a stat is, in this case, the only effect of having the racial modifiers at all.

[Vhaidara]

Oh, and preapplying makes half-orcs fundamentally worse than everyone else, no matter what they do.
No modifier starting line: 8, 8, 8, 8, 8, 8
+2/-2 starting line (normal races): 10, 8, 8, 8, 8, 6
+2/-2/-2 starting line (half-orc): 10, 8, 8, 8, 6, 6

Basically, they now just come out 2 points behind on any point buy.

[daemonaetea]

I think the simplest fix to the issue you have is to just allow players to play a race without taking any of the stat adjustments, positive or negative. Simply put, allow them to have the other racial features, just not the stats, if they so choose. That gets around your problem without having to worry about adjusting point buy at all, yet allowing for flexibility of stat choice if they wish to play a concept that the stat changes naturally work against. A DM I played with before used this, and it seemed reasonable to me.

[eggynack]

I think the simplest fix to the issue you have is to just allow players to play a race without taking any of the stat adjustments, positive or negative. Simply put, allow them to have the other racial features, just not the stats, if they so choose. That gets around your problem without having to worry about adjusting point buy at all, yet allowing for flexibility of stat choice if they wish to play a concept that the stat changes naturally work against. A DM I played with before used this, and it seemed reasonable to me.
Might make sense to be a bit selective with that. Shifter in particular would be significantly more powerful without its net -2 to stats.

[Vhaidara]

Might make sense to be a bit selective with that. Shifter in particular would be significantly more powerful without its net -2 to stats.

And don't forget warforged. Con bonus is nice, but the other abilities restrict their ability to be casters who aren't wizards.

[Troacctid]

Or hey, how about them Elans?

[Chronos]

OK, so let them remove penalties at the cost of an equal amount of bonuses. So you could play a warforged with only -2 Wis or -2 Cha (your choice), but you couldn't get rid of the elan's -2 Cha because there's nothing to give up for it.

[Greenish]

Shifter in particular would be significantly more powerful without its net -2 to stats.Shifter with net stat bonuses of ±0 would still be worse than human (or other powerful races), unless you were doing something funky with the Shapechanger subtype.

[eggynack]

Shifter with net stat bonuses of ±0 would still be worse than human (or other powerful races), unless you were doing something funky with the Shapechanger subtype.
The primary gain from shifter is all of the stuff unique to the race, like shifter druid substitution levels and moonspeaker, and those things have the potential to be more powerful than human. The loss of crappy stats would make them a decent bit more powerful, though not massively so.