Request for information: discussion on D&D attributes in the real world

[kabbes]

I remember somebody once linking on these forums to an excellent analysis of D&D attributes and how they relate to the real world. From memory, it basically pointed out that pretty much everybody in the real world would rate a 10 or 11 for everything, with the best in the world maybe getting up to 12 and the best of all time possibly a 13. And so on for levels.

The analysis was then applied to D&D rules for lifting, jumping, crafting items and so on.

Does this ring any bells with anybody?

[Vizzerdrix]

nope, but it sounds like hogwash to me.

[PersonMan]

I remember somebody once linking on these forums to an excellent analysis of D&D attributes and how they relate to the real world. From memory, it basically pointed out that pretty much everybody in the real world would rate a 10 or 11 for everything, with the best in the world maybe getting up to 12 and the best of all time possibly a 13.

I doubt that. Why? Because a 12 is high average. So, unless everyone actually has a 9 or 10 in Int, there are going to be people with high scores. They're gonna be rare, yes, but...Then again, will they be?

A person with, say, ~14 Cha might be a public speaker for some group, a 14 Dex could be a dancer, or even just be graceful in whatever they do. While a 14 Str could just be, say, a good football player.

The thing with the average is that most people are it. Most of us are between 11 and 9. But some of us are higher. I don't think that there are more than a few people on this planet with an Int score of 16 or higher, but that doesn't mean that 14s are super-rare, either.

[Riffington]

You'd want 18 (or maybe 20) to represent the Earth maxima.
Note that not all feats of strength (for example) are really represented by the Str attribute. Weightlifting is a skill, that adds to maximum lift for your strength. Mike Tyson in his prime, for example, was as strong as people get (usable hitting strength), but others are far better at lifting heavy weights than he was. So looking up your bench press and the D&D strength table would be inaccurate if you're a lifter.

[BobVosh]

Age rules utterly break this. You can easily get +3 wisdom for age. So the wisest young man can never be wiser than EVERY other old dude, according to that analysis.

*edit* Also means noone is above level 7 at the max.

[Zeta Kai]

Does this ring any bells with anybody?

You're looking for this article (http://www.thealexandrian.net/archive/archive2007-03.html), which I largely agree with. The Alexandrian rocks.

[Kurald Galain]

I remember somebody once linking on these forums to an excellent analysis of D&D attributes and how they relate to the real world. From memory, it basically pointed out that pretty much everybody in the real world would rate a 10 or 11 for everything, with the best in the world maybe getting up to 12 and the best of all time possibly a 13. And so on for levels.
That's a far cry from "excellent", and in fact more than a little silly. It means the "best of all time" in, say, strength would lose in arm wrestling against Joe Average, 30% of the time.

Honestly, every comparison of D&D stats to real world abilities I've seen breaks down immediately under mere casual scrutiny.


You're looking for this article (http://www.thealexandrian.net/archive/archive2007-03.html)
That being an excellent example of someone who completely failed to do the research. Regardless of whether the Alex "rocks" or not, this particular article is laughably bad.

[kabbes]

You're looking for this article (http://www.thealexandrian.net/archive/archive2007-03.html), which I largely agree with. The Alexandrian rocks.

That's exactly the one. Thanks muchly.

[Tengu_temp]

If we assume that everyone rolls 3d6 for all stats, then each person has roughly 2,74% chance of having 18 in at least one ability score. This means that one out of 37* people has at least one 18. Peak of human potential? I don't think so.

* - Actually it's more like 36.5, but let's round it up.

[Milskidasith]

I don't understand why people think that it's so hard to get above 16 intelligence. 18 in any stat is one in 216 in the D&D world, just using 3d6. Of course, it's probably because of the assumption int = 10 * IQ, but that's laughably bad as it is.

[Chrono22]

I don't think DnD attributes are good for representing any real attributes for a couple of reasons.
1. They are overly general.
2. They scale in an unusual fashion (18 = the best? not a ten?)

Since I wanted realistic attributes for my system, and a simple and intuitive method for tracking scores, I ended up splitting each of the 6 DnD ability scores in half (for a total of 12 attributes). I also changed the range to 1-10, with 1 being pathetic and 10 being extraordinary. I completely removed "bonuses", and apply attributes point for point to appropriate skills.

Real world comparisons and approximations are alot easier now.

[Milskidasith]

12 attributes just makes it more complicated.

[Kurald Galain]

12 attributes just makes it more complicated.

...the real world is complicated...

[Milskidasith]

True, but I don't see how adding more attributes would make it easier to describe people; a stat based system is inherently bad for describing real world people. Plus, since he apparently uses the stats for games, it would make the games needlessly complex.

[Tengu_temp]

I'd say the system that you can most easily apply to real world is WoD/Exalted, but that's because its attributes are only on a 1-5 scale (very few real life humans have 0 in some attribute, and even less have more than 5), with 2 being the norm, and each dot has a very wide range. Generals are easier to apply than specifics.

[Milskidasith]

I'd say the system that you can most easily apply to real world is WoD/Exalted, but that's because its attributes are on a 1-5 scale (very few real life humans have 0 in some attribute, and even less have more than 5), with 2 being the norm, and each dot has a very wide range.

That's a problem on the opposite end of the scale to D&D; D&D tries to be too precise, which works for a game but not for the real world, and exalted is too vague to be any use. If you had a three in a stat, you could be anywhere from being, say, pretty smart or you could be a certified genius, or you could just be really good at social skills, or anything else that falls under the umbrella.

[Chrono22]

True, but I don't see how adding more attributes would make it easier to describe people; a stat based system is inherently bad for describing real world people. Plus, since he apparently uses the stats for games, it would make the games needlessly complex.
Um, how so? You'd be surprised how freeing it is to turn flavorful descriptions into mechanics that are consistent with those descriptions. Rules that are consistent with reality as we perceive it mean that you can turn "normal roleplaying language" into mechanically viable action.
Consider this: Strength in DnD is a factor of your physical might and your body mass- a character that is physically mighty will also be more stable because the two- mass and might- have been tied together. But the truth of the matter is, they are mutually exclusive attributes. A person that is mighty need not be exceptionally stable. Nor is the reverse true- an obese man might be incredibly weak.
Having more attributes does add a small layer of complexity to the game- but it's the type of complexity that players appreciate... their character seems more defined, more real, since his scores more realistically reflect their concept.
I'd say halo 3's graphics are more complex than halo CE's cellshading and bump mapping- is that a bad thing?

[Tengu_temp]

That's a problem on the opposite end of the scale to D&D; D&D tries to be too precise, which works for a game but not for the real world, and exalted is too vague to be any use. If you had a three in a stat, you could be anywhere from being, say, pretty smart or you could be a certified genius, or you could just be really good at social skills, or anything else that falls under the umbrella.

Of course. But I'd say that the only way to make a non-vague system apply well and with some level of accuracy to real life is to create an overcomplicated monstrosity that will make FATAL's mechanics look like Risus in comparison.

[kamikasei]

If we assume that everyone rolls 3d6 for all stats

I was under the impression that only "heroes" do so (even if the default statgen method has shifted to 4d6b3), not the population at large.

This is a clear case where the character creation rules simply don't help in generating actual demographics for a world that makes sense; that's not what they're for.

[Tyndmyr]

Straight 3d6 rolls for stats are available as the weakest alternative means of rolling characters. Using those standards for ordinary people, in conjunction with the commoner class is definitely fair.

After all, the 4d6drop lowest works out to an average of about 12 pts.

[kamikasei]

Using those standards for ordinary people, in conjunction with the commoner class is definitely fair.

What's fair got to do with it? Sure, it's a good way to balance antagonists and other NPCs as weaker than, but comparable to, the heroes of the story. But there's nothing to indicate that it's how the inhabitants of a D&D setting come by their stats, that they follow a 3d6 bell curve, that one in every 216 people has an 18 in a given stat, etc. They're the character creation rules. They're not the "create an entire world using only an RNG and these rules" manual.


After all, the 4d6drop lowest works out to an average of about 12 pts.

I've never run the numbers myself (I must try at some point, since figuring out how will probably teach me something worthwhile about probability, which I know far too little about), but I thought 4d6b3 was supposed to work out to the 25-point Elite Array on average?

[Tyndmyr]

Given that commoner is among the weakest classes possible(quite possibly *the* weakest), and certainly the weakest class you could apply to humanity at large, it seems a reasonable choice. Likewise, the 3d6 roll is the weakest form of character creation available.

Thus, this is the floor for character creation in D&D. Is it realistic? Probably not...but if you're trying to recreate the world in D&D terms, this is the floor from which you start.

Edit: I'll link to fun stat stuff later. Basically, 4d6k3 is one of the better ways of getting stats. The most powerful given variant is that, plus rerolling 1s, if I remember correctly. All stats are 6 at minimum, and that's *extremely* unlikely.

[Random832]

Given that commoner is among the weakest classes possible(quite possibly *the* weakest), and certainly the weakest class you could apply to humanity at large, it seems a reasonable choice. Likewise, the 3d6 roll is the weakest form of character creation available.

Technically, point buy below 20 or so is weaker than 3d6.

The standard array is pretty weak too.

[kamikasei]

I'm afraid I don't see what the exercise is supposed to achieve. Are we trying to find out something about the attributes by seeing how they compare to real-world distributions of ability? That would require having rules for generating ability distributions for the population at large. We have no such rules. We do not need to pick the worst of a bad set of options, we can simply say "you can't make any useful statements (supported by or derived from the rules, at least) about these details on a large scale in a D&D setting, because the rules simply don't cover them".


Edit: I'll link to fun stat stuff later. Basically, 4d6k3 is one of the better ways of getting stats. The most powerful given variant is that, plus rerolling 1s, if I remember correctly. All stats are 6 at minimum, and that's *extremely* unlikely.

...Yes? I didn't say I don't know what 4d6b3 is. (What I don't know is how one goes from the probability of each value for a single ability, to the "average" set of six abilities.)

[Eldariel]

Technically, point buy below 20 or so is weaker than 3d6.

The standard array is pretty weak too.

Huh? 20pb is very high compared to 3d6 (which averages standard array). Standard array is 13pb.

[Matthew]

Attributes for "common" NPCs used to be generated on a basis of 3d6, treating all "6s" as "5s" and all "1s" as "2s", the idea being that the average spread was 6-15. The open ended attributes of D20/3e make "average" a bit difficult to figure, and probably a bit pointless.

Also, I agree with Kurald Galain as regards the Alexandrian article.

[kamikasei]

Also, I agree with Kurald Galain as regards the Alexandrian article.

Could you specify why? (Or, if you've already done so elsewhere, provide a link?)


Huh? 20pb is very high compared to 3d6 (which averages standard array). Standard array is 13pb.

Hmmm. How is that average being determined? (By standard array you mean either 10, 10, 10, 11, 11, 11, or 13, 12, 11, 10, 9, 8, right? Although both of those are 15pb, not 13...) Is it properly accounting for the fact that PB costs increase as the stat does?

[Eldariel]

Hmmm. How is that average being determined? (By standard array you mean either 10, 10, 10, 11, 11, 11, or 13, 12, 11, 10, 9, 8, right? Although both of those are 15pb, not 13...) Is it properly accounting for the fact that PB costs increase as the stat does?

Oh yeah, 6 stats, duh. So yeah, 15pb. And standard array = 10/10/10/11/11/11, nonelite array = 8/9/10/11/12/13.

[lesser_minion]

I agree with what the Alexandrian had to say on this really.

KG, could you elaborate on the problems you had with the article?

As far as I can tell, his conclusion is essentially valid - over the first few levels of the game, the capabilities of characters map reasonably closely to what a real-world observer might expect.

Admittedly, he only demonstrates this point for a few things, but his numbers seem pretty sound.



That's a far cry from "excellent", and in fact more than a little silly. It means the "best of all time" in, say, strength would lose in arm wrestling against Joe Average, 30% of the time.


By RAW, if something is "a straight test of ability with no luck involved", the character with the higher ability score wins, and ties are resolved by coin toss.

An Arm Wrestle is also specifically called out as being one of these.

[Random832]

Huh? 20pb is very high compared to 3d6 (which averages standard array). Standard array is 13pb.

You can't do the math that way - PB is non-linear, so "averages to standard array" does not mean "average PB value is standard array PB value". You're right that 20 is a high estimate though.

One other issue is that PB can't give you a value less than eight... If you reroll all stats less than eight, 3d6 gives 21.5; if you accept them at values of -1 through -5, it gives about 16.2, so in that sense I was wrong about my guess of 20.

Of course averages aren't the whole picture - it's not possible to have even a single 18 with 15pb, whereas almost one in every 36 people will have at least one 18 with 3d6.

216^6/(216^6-215^6) ~= 36.419

[Zeta Kai]

That being an excellent example of someone who completely failed to do the research. Regardless of whether the Alex "rocks" or not, this particular article is laughably bad.

That's a very strong statement, one that bears elaboration. It seems that the article demonstrates its various points with adequate math & numerous examples. IMO, the most disagreeable thing in the article is the assertion that Aragorn is a Rgr1/Ftr1/Pal3.

Can you please explain why you feel that the article is poorly researched & "laughably bad"?

[Foryn Gilnith]

On the topic of whether or not Alex "rocks": That article is his most notable work. He did a piece later attacking 4e, but that's one among a sea of others. Most of his other stuff is random theatrical commentary. His proposal for a Star Wars Sequel Trilogy was interesting, but his reputation pretty much rides on this.

Also, on the arm wrestling comment, IIRC arm wrestling was given as a specific example of higher strength just winning. You don't make a height check to see who's taller, and you don't make a Strength check in this scenario to see who happens to be stronger.

But yeah, details plox.

[Riffington]

Age rules utterly break this. You can easily get +3 wisdom for age. So the wisest young man can never be wiser than EVERY other old dude, according to that analysis.

*edit* Also means noone is above level 7 at the max.

That's all ok.
In the real world, you don't get +wisdom for age. It falls with age, albeit more slowly than Dex. And level 7 certainly seems higher than the Earth max.

[Tyndmyr]

His basic premise that it's easier to map reality to a smaller spread of character leves I agree with. We just don't see level 20 characters in real life, who can slaughter essentially infinite numbers of lesser beings. I can buy that nearly everyone ranges from level 1-10.

I have more difficulty with the stat ranges. I think there's much more variation in real life than he accounts for, and focusing on strength has his example stat may be why.

[Matthew]

Could you specify why? (Or, if you've already done so elsewhere, provide a link?)

There are some pretty hefty threads on the subject around here, probably a search for "Alexandrian" will turn them up (might have to use Google by the looks of it, searches here only go back six months). If I recall correctly, the problem was that on the surface what he says is plausible, but it does not hold up to close examination. The author had an account here once and was engaged in the debate, but he did not manage to convincingly defend his article. He did manage to launch a personal attack on a forum member from his website, though, as I recall...

To be fair, it is not that the article is completely wrong, it just stretches the point too far.

[Gan The Grey]

I don't think attribute generation needs to take into account the full range of humanly possible attributes. While there are people that have been born with a 3 intelligence because of genetic defects, I would view that less as a randomly determined attribute and more of a flaw. Also, why would 18 need to be achievable in a PB system? I can't think of a single person who was born as an Olympic body builder or super genius. They all got that way by training after the fact. Maybe they had a bit of an advantage over the general populace, but they never would have gotten where they are without serious training and discipline.

[Tyndmyr]

Well, you're not born with 18 in a stat. That represents your ability at more or less the prime of your life. Presumably you did something to get to that point.

After all, adventurers are not infants at first level. Hopefully.

[Zeta Kai]

I can't think of a single person who was born as an Olympic body builder or super genius. They all got that way by training after the fact.

My understanding is that genius intellect is inherent, & not achievable via training, no matter what those Baby Einstein videos try & sell us.

[Gan The Grey]

My understanding is that genius intellect is inherent, & not achievable via training, no matter what those Baby Einstein videos try & sell us.

No matter how smart you are, if you don't work out your brain, you won't get any smarter, and you will probably lose intelligence. The brain is a metaphorical muscle much in the same way that the bicep is. Myelin(Sp?) sheathes and all that.

The way I see it, people who are geniuses are simply born with a higher than average intellect (say 13 or 14), allowing them to more quickly understand the simpler functions of the universe during a time when they learn more easily (i.e. lower level means less experience to level up). This allows them to tackle more difficult concepts during this heightened stage of learning and prepare them for even more difficult concepts as they age.

Since most people choose to celebrate and hone their strengths as opposed to correcting their weaknesses, it only makes sense that they would continue to support their intellectual paths by increasing their Intelligence score through study and thought exercises(i.e. training their Int stat over time in D&D world), place additional attribute points gained by leveling in their intelligence (if we assume max level 10 expert, then an additional +2 to intelligence), and skill choices that support their interests (an engineering whiz may not understand the same things as a biochemistry professor, but they can have the same Intelligence score, they just seem smarter because of more ranks in Knowledge: Architecture & Engineering).

Quicker or slower learning can be also be attributed to feats, traits, flaws, extraordinary abilities/qualities and, who knows, supernatural abilities/qualities.