D&D Scores Translated into Real Life Numbers

[Aurini]

So this morning I finally got around to doing something I've been meaning to do for some time: come up with the specific percentiles of Strength and IQ, and equate them to your character scores in Dungeons & Dragons.

Here's the detailed explanation with links to the original data (http://www.staresattheworld.com/2018/03/realistic-dd-comes-strength-iq/) - but I summed up the information in the following charts:

https://i1.wp.com/www.staresattheworld.com/wp-content/uploads/2018/03/Bench-Press-IQ-by-Ability-Score.png?w=481
https://i0.wp.com/www.staresattheworld.com/wp-content/uploads/2018/03/DD-Strength-Max-Lift.jpg?w=480

Note: the green line is real world scores, the brown line is what D&D lists.

My bench is 185 lbs, which puts me at a Strength of 14. As for IQ - I haven't been tested since Elementary School, so I consider that result to be questionable.

Here's the table I used to make these charts:

https://i2.wp.com/www.staresattheworld.com/wp-content/uploads/2018/03/DD.png?resize=1024%2C112

So how are y'all shaping up?

[Frozen_Feet]

Actually, physical stamina correlates strongly with so many things associated with Constitution that the results for a 12 minute running test make a fair approximate measure for Con score. I did a fairly extensive mapping of this, but the data was lost when the computer I used went belly up.

Anyways, your tables would peg me at 13 for Intelligence, 14 for strength based on bench press and 13 for strength based on deadlift.

EDIT: oh yeah, I forgot: I don't know if the values you used for D&D really originated from Gygax and Arneson, but they did use then-current Olympic records for calibrating Strength for AD&D. Likewise, Intelligence was explicitly called out to be analogous measure to IQ, even if the actual mapping between IQ and Int was not done and was later filled in by various erroneous rules of thumb (such as Int × 10 = IQ).

[Aurini]

If you could point me in the right direction, I'd love to see that data, Frozen_Feet. :)

[Frozen_Feet]

12 minute running test AKA Cooper's test has been and is used by militaries and school systems worldwide, so check of your government has made any of the statistics publicly available. The typical tables used for test interpetation can be found even on Wikipedia.

[AMFV]

A big part of the problem is that "Strength" in D&D models several very different things. It would model lifting ability, yes. But it also would model punching power and accuracy. As well as innate ability in several skills that are not inherently related. As such it doesn't really parallel to any real world ability. A guy could be able to deadlift a lot, but not punch fast or accurately. A dude could be good at bench press but not inherently good at climbing. Because it's an abstraction it doesn't really model well in the real world.

[Brother Oni]

Further AMFV's comments, there's differences even in activities. My bench press weight puts me about STR 12, but my clean and press puts me about STR 15 as I train for fast twitch power.

Even the Cooper's test measures only one thing; cardiovascular fitness. Personally, I've never been able to get below an 11 minute mile as I'm not built for running, but years of martial arts training means I can take heavy hits.
Objectively measuring pathogen and toxin resistance which Constitution also models is also somewhat dubious; does the genetic alcohol intolerance trait of SE Asians mean they have a racial CON penalty for example?

[Lvl 2 Expert]

I have never tried deadlifting, but I have won a free beer ones by lifting 200 kg (440 pounds) using a similar motion. So that would put me around 16. Let's say 15 because I probably started a bit higher and the weight was supported against wobbling better. My "high score" on an IQ test is 150, which would be an 18. But I feel like that's mostly because those test are all about quick insight questions which I just happen to be extraordinarily good at. (Also, I'd probably score closer to 130 on an average well made test. But the 150 is just nice to brag about.) If we're talking practical intelligence, being good at thinking about real world problems, I'm probably more like top 20% at best, so a 13 or 14 or so, "practical IQ" 115-120.

I also somehow get the feeling that in some to most D&D settings extreme stats are actually supposed to be more special than the 3d6 distribution implies, as nice as the lifting weight curve does match up. The elite array for instance gives a character a single 15. Now I'm a big strong guy, but I'm not going to be stronger or more use for my strength than say an elite marine. Nor am I going to out think a professor. Even just a single 16 makes you very special in that world. But maybe that's just my take on the stats.

[Florian]

I also somehow get the feeling that in some to most D&D settings extreme stats are actually supposed to be more special than the 3d6 distribution implies, as nice as the lifting weight curve does match up. The elite array for instance gives a character a single 15. Now I'm a big strong guy, but I'm not going to be stronger or more use for my strength than say an elite marine. Nor am I going to out think a professor. Even just a single 16 makes you very special in that world. But maybe that's just my take on the stats.

Personally, I think it´s pretty stupid to try and translate D20 in-game stats to map with RL capabilities. They don't match up, ever.

[Frozen_Feet]

Even the Cooper's test measures only one thing; cardiovascular fitness.

Technically true, in practice borders on irrelevant criticism, because cardio-vascular fitness correlates well with plenty of other aspects of good health, from pain and shock resistance to strength of the immune system.

Same goes for press and grip strength tests as pertains to striking power, explosive strength etc. That is why these tests have become standards in health and fitness check-ups.

Remember that even in D&D, stats are not be-all-end-all of any action. For example, high strength correlates with damage, to-hit etc., but it's actual causal impact is minor when you view it in the context of the whole system.

In D&D terms, when you're thinking of a strong person who can't hit things, chances are you're thinking of a high STR score non-combatant and contrasting it with an actual Fighter. To use d20 rules to illustrate the point, a STR 18 1st level Commoner in a boxing match is actually attacking at just +0 due to non-proficiency and eating an attack of opportunity for each attack he makes. His attacks are also non-lethal. While a STR 10 Fighter with Imp. Unarmed Strike is attacking at +1, effectively gets two blows for each of the commoner's, and does lethal damage.

[Brother Oni]

Technically true, in practice borders on irrelevant criticism, because cardio-vascular fitness correlates well with plenty of other aspects of good health, from pain and shock resistance to strength of the immune system.

While I agree on some aspects like immune system and general health, I disagree on others like pain and shock resistance - I know a lady from work who can run a sub 8 minute mile, so theoretically by the Cooper's Test yardstick, despite being more than 4 stone lighter than me, she should be able to take a hit better than me.


Same goes for press and grip strength tests as pertains to striking power, explosive strength etc. That is why these tests have become standards in health and fitness check-ups.

I suspect some of those tests have become standard because they're so easy to do and give vaguely useful information, much like BMI is a measure of general health. For example, the pectorals are one of the major muscle groups exercised during a standard bench press, but you don't use those muscles during punching.

I remember reading about a retired Royal Marine officer complaining about the general fitness of new recruits in that they turn up gym fit (ie vanity muscles), but unused to training in hard shoes and unable to run with a heavy load.

That said, I concede the rest of the argument as my D&D knowledge is very rusty.

[FreddyNoNose]

So this morning I finally got around to doing something I've been meaning to do for some time: come up with the specific percentiles of Strength and IQ, and equate them to your character scores in Dungeons & Dragons.

Here's the detailed explanation with links to the original data (http://www.staresattheworld.com/2018/03/realistic-dd-comes-strength-iq/) - but I summed up the information in the following charts:

https://i1.wp.com/www.staresattheworld.com/wp-content/uploads/2018/03/Bench-Press-IQ-by-Ability-Score.png?w=481
https://i0.wp.com/www.staresattheworld.com/wp-content/uploads/2018/03/DD-Strength-Max-Lift.jpg?w=480

Note: the green line is real world scores, the brown line is what D&D lists.

My bench is 185 lbs, which puts me at a Strength of 14. As for IQ - I haven't been tested since Elementary School, so I consider that result to be questionable.

Here's the table I used to make these charts:

https://i2.wp.com/www.staresattheworld.com/wp-content/uploads/2018/03/DD.png?resize=1024%2C112

So how are y'all shaping up?

I saw that over the weekend. What a pretentious load of claptrap.

[Frozen_Feet]

While I agree on some aspects like immune system and general health, I disagree on others like pain and shock resistance - I know a lady from work who can run a sub 8 minute mile, so theoretically by the Cooper's Test yardstick, despite being more than 4 stone lighter than me, she should be able to take a hit better than me.

Your conclusion (that she'd be able to take a hit better than you) would only hold true in D&D if you'd already corrected for all other factors, such as class, level, feats etc.. To give an extreme example: A barbarian has d12 hit die and a commoner has d4. A commoner with Con 18 has potential maximum hitpoints of 8, but so does a Barbarian with Con 3 (respective stat modifiers +4 and -4).

As the formula also contains a variable, the commoner could end up with just 5 hitpoints - less than the average Fighter. (Fighter has d10 hit die so with +0 modifier the average is 5.5.)

The point of the illustration here is to show that the causal link between Constitution and durability only becomes apparent when studying equals, or great masses of people.

Same goes for real-life correlates. That lady from your workplace may not be able to take a hit as well as you, but compared to other women of same size and general lifestyle, she most likely can.


I suspect some of those tests have become standard because they're so easy to do and give vaguely useful information, much like BMI is a measure of general health..

You can stop suspecting, I can confirm that is the reason they're used. Measures such as BMI, IQ, 12 minute running test, grip strength test etc. are used because they are useful statistical correlates: when dealing with large sample sizes (such as entire age cohorts in case of the military), it becomes apparent how they relate to physical fitness. Their predictive strength for individuals is weak, but for large groups it becomes massive.

What I'm after here is that this is how you should view these tests when trying to reverse-engineer ability scores. We're not trying to find a be-all-end-all measure which would explain the case of every individual, because the explaining power of ability scores even for the individual is limited. We're looking for measures which show general trends across populations.


For example, the pectorals are one of the major muscle groups exercised during a standard bench press, but you don't use those muscles during punching

I remember reading about a retired Royal Marine officer complaining about the general fitness of new recruits in that they turn up gym fit (ie vanity muscles), but unused to training in hard shoes and unable to run with a heavy load.

That said, I concede the rest of the argument as my D&D knowledge is very rusty.

Again, you are technically correct, but the criticism borders on irrelevant. As noted, in D&D other factors influence punching as well, so you can get the exact same situation there: a person with high Str but poor ability to fight unarmed. The earlier example of a commoner in a boxing match applies here.

It's even more obvious with the example your officer gave, because in D&D (again using d20 rules as a discussion point), running is covered by Constitution and Feats rather than Strength. So it would be wrong to expect a high Str person to be able to run well, because those things don't correlate well in the system.

I'm using rules minutiae as an example, but the more general and more interesting point is again about the correct use of statistical correlates.

[Anonymouswizard]

Really, most people should be in the 8-13 range, your most remarkable stat is probably a +1. The strongest or must intelligent person you know is probably a +2, maybe a +3 (I know a handful of the latter, I think, but that's because many people would peg me as +2 and I'm the dumb one of the family). In D&D somebody with a 16 in Strength is probably the strongest person in their town. But that's eclipsed by potential training, or class levels.

[AMFV]

Your conclusion (that she'd be able to take a hit better than you) would only hold true in D&D if you'd already corrected for all other factors, such as class, level, feats etc.. To give an extreme example: A barbarian has d12 hit die and a commoner has d4. A commoner with Con 18 has potential maximum hitpoints of 8, but so does a Barbarian with Con 3 (respective stat modifiers +4 and -4).

Yes, but constitution also includes several things that are not related to running endurance and speed at all. Like ability to resist poisons. If you are faster at running you won't be inherently better at resisting poisons. If you are faster at running you won't be inherently better at focusing on a task (Concentration).

Although to be fair, Con is probably the skill that maps most readily from a real life perspective, any of the other physical skills it starts getting a lot dicier. Strength for example not only maps your ability to punch, but also your vertical, those are often unrelated. Also your ability to climb.



The point of the illustration here is to show that the causal link between Constitution and durability only becomes apparent when studying equals, or great masses of people.

But we aren't dealing with great masses of people or equals. Now to be fair this is a rules abstraction and as such it's very reasonable. But when you start moving into trying to use it for more real world modelling that isn't playing a game it breaks down in short order.



Same goes for real-life correlates. That lady from your workplace may not be able to take a hit as well as you, but compared to other women of same size and general lifestyle, she most likely can.


Actually, probably not. Cardiovascular endurance has very little to do with any of the variables involved in taking a punch, outside of pain tolerance. I mean you're looking at hardness of bones, glass jaw, all kinds of factors that a run does not effectively replicate.



You can stop suspecting, I can confirm that is the reason they're used. Measures such as BMI, IQ, 12 minute running test, grip strength test etc. are used because they are useful statistical correlates: when dealing with large sample sizes (such as entire age cohorts in case of the military), it becomes apparent how they relate to physical fitness. Their predictive strength for individuals is weak, but for large groups it becomes massive.


Yes, which is good for coming up with the large scale numbers you see in D&D, but not so useful in terms of figuring out what an individual would score in terms of their strength.



What I'm after here is that this is how you should view these tests when trying to reverse-engineer ability scores. We're not trying to find a be-all-end-all measure which would explain the case of every individual, because the explaining power of ability scores even for the individual is limited. We're looking for measures which show general trends across populations.


But that is NOT what the article is doing, it's measuring absolute individual performance. And it's trying to evaluate how close the abstraction is to the real. For the latter you're correct, but not so much for the former. So it's not useful in terms of figuring out your own strength, but it is useful in saying "D&D strength is pretty close to the trendlines"



Again, you are technically correct, but the criticism borders on irrelevant. As noted, in D&D other factors influence punching as well, so you can get the exact same situation there: a person with high Str but poor ability to fight unarmed. The earlier example of a commoner in a boxing match applies here.

But those other factors aren't innate ability, and many of the factors in boxing are. There's a reason why people who are deeply into sports get more and more interested in innate ability things than people who are not. Because it starts to matter more and more. Now for a game figuring out arm reach and jab speed and a thousand other variables is probably not useful, but in terms of figuring out how "real" a model is, it certainly is.



Really, most people should be in the 8-13 range, your most remarkable stat is probably a +1. The strongest or must intelligent person you know is probably a +2, maybe a +3 (I know a handful of the latter, I think, but that's because many people would peg me as +2 and I'm the dumb one of the family). In D&D somebody with a 16 in Strength is probably the strongest person in their town. But that's eclipsed by potential training, or class levels.

That's a flawed line of thinking. You are assuming that all people are average, not that most people are average. What if somebody here is the strongest person in their town? I mean what if I know the strongest dude in town, or the strongest dude in my state (I do know the strongest dude in my state), are you telling me that a guy who can overhead more than 300 lbs comfortably and can deadlift over 900 lbs is strength 16? Cause that doesn't track.

[Anonymouswizard]

That's a flawed line of thinking. You are assuming that all people are average, not that most people are average. What if somebody here is the strongest person in their town? I mean what if I know the strongest dude in town, or the strongest dude in my state (I do know the strongest dude in my state), are you telling me that a guy who can overhead more than 300 lbs comfortably and can deadlift over 900 lbs is strength 16? Cause that doesn't track.

No, I'm saying that a 16 is rare enough that if you have one you're probably the one best at X in your town (cities are more likely to have a 17 or 18 as their strongest). I'm not saying 'the strongest in the town is Strength 16', I'm saying that if you have a 16 and you live in a community of thousands odds are you're the strongest.

Let me put it this way, if we assume that one person in a million has one or more 18s (just for the nice numbers) then the United Kingdom has approximately sixty people with at least one 18. The world has seven million. The village I grew up in probably doesn't have one, although due to the way statistics work it could have several. 16s are more common, but a town of thousands might still only have one. However it could still believably have people with higher stats, it's just not likely.

P.S. use real units. Translating pounds into Newtons and Kilos is a headache.

[Corsair14]

8-12 is going to be the average person's stats. 13-15 is going to be your above average person in whatever stat and unlikely for that person to have more than one stat above that. 16-17 is going to be a stat for your exceptional people and relatively rare. 18+ is genius level, likely no more than 1 in 10,000 people will have this stat.

I disagree with the OP's chart though. Bench Pressing 200 pounds is strength 16? I'm much higher than that and if I was going to stat myself out I wouldn't put myself higher than a 15 tops. With some of the people I know they would be in the mid 20s according to the chart. That said, one exercise shouldn't be the end all on what your stat is. Also it is also the easiest stat to increase through work followed by Con then dex then charisma. The last two would be very difficult to increase. How do you increase your ability to interpret logic or increase common sense?

[AMFV]

No, I'm saying that a 16 is rare enough that if you have one you're probably the one best at X in your town (cities are more likely to have a 17 or 18 as their strongest). I'm not saying 'the strongest in the town is Strength 16', I'm saying that if you have a 16 and you live in a community of thousands odds are you're the strongest.

Let me put it this way, if we assume that one person in a million has one or more 18s (just for the nice numbers) then the United Kingdom has approximately sixty people with at least one 18. The world has seven million. The village I grew up in probably doesn't have one, although due to the way statistics work it could have several. 16s are more common, but a town of thousands might still only have one. However it could still believably have people with higher stats, it's just not likely.

P.S. use real units. Translating pounds into Newtons and Kilos is a headache.

The problem is you're ignoring the probability distribution of the dice. It's not "an 18 is the highest human stat ever and there has never been higher", it's "18 Strength would put you in the .07th percentile", which is very high but I doubt is the limit for human strength achievement. Particularly since you can find people who have strength achievements that are much smaller group wise than that. So an 18 strength would probably be just under one in 10,000, not even approaching one in a million.

So the strongest person in a town of 20,000 would probably have a strength 18, unless we are completely ignoring how the numbers are distributed.

[SaintRidley]

This is silly. D&D attribute numbers don't and can't fit to anything in real life because they're hideously imprecise abstractions. Also taking IQ terribly seriously. That's another ludicrous aspect of this whole thing.

Mark Henry's got a 585 lb bench press and a 903 lb deadlift. Since your scores seem based on the bench press and go up roughly 10 lbs per point, it looks like Mark Henry's strength score by your logic is around a 53.

[AMFV]

This is silly. D&D attribute numbers don't and can't fit to anything in real life because they're hideously imprecise abstractions. Also taking IQ terribly seriously. That's another ludicrous aspect of this whole thing.

Mark Henry's got a 585 lb bench press and a 903 lb deadlift. Since your scores seem based on the bench press and go up roughly 10 lbs per point, it looks like Mark Henry's strength score by your logic is around a 53.

To be fair though, Gygax would have used the deadlift, since there is no rules for benching in D&D. You can "lift off the ground double your maximum load" so that would mean Mark Henry would have a strength of approximately 22, which is appropriate for somebody that's made that their life goal, and possibly also would have the benefits of a rage like state. Using deadlift and OHP works much better for the abstraction, now it is a ridiculous abstraction but dumping the bench press makes it a lot less ridiculous.

[DavidSh]

What stat distribution should we assume for the general populace? 3D6 independently rolled for each stat? Then you expect 1/216 of the people to have 18 strength.

Maybe 3D6 has a higher average, or more variability, than the population in general has.

[Anonymouswizard]

The problem is you're ignoring the probability distribution of the dice. It's not "an 18 is the highest human stat ever and there has never been higher", it's "18 Strength would put you in the .07th percentile", which is very high but I doubt is the limit for human strength achievement. Particularly since you can find people who have strength achievements that are much smaller group wise than that. So an 18 strength would probably be just under one in 10,000, not even approaching one in a million.

So the strongest person in a town of 20,000 would probably have a strength 18, unless we are completely ignoring how the numbers are distributed.

Ah, we have different ideas on how stats are generated. I assume PCs roll because they're supposed to be interesting, and unusually high or low stats are interesting. If you assume that you roll for everybody then yes, exceptional stats become more common. I assume that most NPCs will have an array of 8-13, and a total +3 to +8 bonus in skills they care about. Then again I really don't care about such things as assigning stats to real world values because eventually we're into at least the mid twenties just because somebody wants to give different values to Einstein, Tesla, Hawking, and every other famous scientist.

Plus I just like the idea that in a fantasy game PCs can be stronger than the strongest human in the teak world. Therefore I like 18=real world peak because it makes that easier. Especially because for everything bat Strength and maybe constitution there's no decent real world values you can user to measure them (depending on what you user to measure it my Dexterity varies between 6 and 10, my Into between about 8 and 16, my charisma between about 10 and 14, and my Wisdom between 3 and 9. My Strength had a value of 'meh, goodish I guess but nowhere near professional weightlifters', or probably 10-12, and my Constitution 'alright if I don't have to run', so between about 6 and 10 I guess).

I'll also like to point out that this is why I preferred unleveled systems, those who really care about being strong can just invest their points in Strength.

Anyway, this is why I'm beginning to prefer systems with smaller distributions. 1-5 and 1-6 being the must common, which tend to differentiate being good more than being bad (kind of like modern D&D and it's tendency for characters without anything below an 8). I even own a system that limits Attributes to the 2-6 range, because below a two and you can't succeed at any tasks. I also rarely care how much you can lift, as long as your Strength is 'very strong' you can probably lift 'very heavy' things. I record encumbrance in significant items already.

[hamishspence]

To be fair though, Gygax would have used the deadlift, since there is no rules for benching in D&D. You can "lift off the ground double your maximum load" so that would mean Mark Henry would have a strength of approximately 22, which is appropriate for somebody that's made that their life goal, and possibly also would have the benefits of a rage like state. Using deadlift and OHP works much better for the abstraction, now it is a ridiculous abstraction but dumping the bench press makes it a lot less ridiculous.

It's worth keeping in mind that you can lift "your maximum load" over your head, and also move around with it.

Whereas you can lift "double your maximum load" off the ground, but generally speaking, not over your head. And you cannot move quickly - only 5 ft per round:

http://www.d20srd.org/srd/carryingCapacity.htm

In the context of 3.5, the "prodigy of (stat)" trait from DMG2 (+2 to Stat, and a +4 untyped bonus to all checks based on that stat), is a good way to beef up "low level" characters - makes sure your ultra-high strength expert still has very few "hit points", "skill points", etc.

I could see a World's Strongest Man contender being a prodigy of Strength, or a teenager with a high-end degree, as a prodigy of Intelligence.

[danzibr]

I think the real problem with using 3-18 is that it caps at 18. Yeah, 3d6 is approximately normal. Yeah, IQ is normal, and I’d guess there are ways to quantify strength, dexterity, and constitution which would also be approximately normal. But you will have extreme outliers for IQ and whatnot, several standard deviations above the mean, which would be well above an 18.

[Frozen_Feet]

What stat distribution should we assume for the general populace? 3D6 independently rolled for each stat? Then you expect 1/216 of the people to have 18 strength.

The original post is assuming a normal distribution and looking for measures which map to it reasonably well. The 3d6 roll approximates a normal distribution for stat range 3 to 18.

Still, even taking 3d6 at face value (which is a workable starting point), you shouldn't expect a fraction of 1/216. Why? Because that 1/216 only applies to young adults who have not suffered any sort of lasting disease, poisoning, injury or malnutrition. This applies to greater or lesser extent to all ability scores.

---

@Danzibr: 18 is not the actual cap in d20 versions of D&D, it wasn't the cap for all stats even in AD&D. That's why assuming 18 as some super-rare manifestation of human peak does not make sense - that assumption is not actually shared by the rules.

[2D8HP]

......even taking 3d6 at face value (which is a workable starting point), you shouldn't expect a fraction of 1/216. Why? Because that 1/216 only applies to young adults who have not suffered any sort of lasting disease, poisoning, injury or malnutrition. This applies to greater or lesser extent to all ability scores...


Boy howdy.

I'd be suprised if I had any "stat" above an 8 these days.

Dagnabbit.

[mtatosky]

What about Charisma??

[danzibr]

@Danzibr: 18 is not the actual cap in d20 versions of D&D, it wasn't the cap for all stats even in AD&D. That's why assuming 18 as some super-rare manifestation of human peak does not make sense - that assumption is not actually shared by the rules.
Roger!

Wait, how can a human start with an ability score over 18 in 3.5? Core anyway.

[Lvl 2 Expert]

What about Charisma??

I know a person we can use as a benchmark for whatever the highest reasonably achievable score should be.

That man is almost universally likable, the best instructor you've ever had and almost scary good at social situations. We spend a whole evening playing "Avalon", one of those games where you have to figure out who the traitors are. Whoevers side this guy was on won, literally all the time.

Me, I can roll the skills I use the most without a penalty now after enough practice. And looking at how this guy does things and trying to consciously emulate them.

[Frozen_Feet]

For Intelligence, Wisdom and Charisma, you get bonuses based on age, so if a character is allowed to start at venerable, they might have a score as high as 21.

I don't remember any similar permanent boosts in core only for physical stats, but a Barbarian's rage can temporarily push both Strength and Constitution as high as 22.

Once you move outside core, there are things like the Paragon template, and once you move beyond starting characters, there's 4th level ability increase etc. These are important, because d20 D&D doesn't model truly exceptional people just by their ability scores, it also adds class, experience and templates in the mix.

Anyways, the overall point is not to try overfit all of humanity in the 3d6 range, because the game doesn't really do that either.

EDIT: oh, and just to give contrast for earlier editions: in 1st Edition AD&D, Wisdom increased with age, and was one of the few stats allowed to rise to 19. For Strength, only Non-Fighters were restricted to 18. Fighters also rolled d100 to get a percentage score. The highest strength for human male fighter was 18/00, and the lifting values for this score were calibrated to be in line with then-current Olympic records. (These records have since been eclipsed.)

[Genon]

So this morning I finally got around to doing something I've been meaning to do for some time: come up with the specific percentiles of Strength and IQ, and equate them to your character scores in Dungeons & Dragons.
- but I summed up the information in the following charts:


Note: the green line is real world scores, the brown line is what D&D lists.

My bench is 185 lbs, which puts me at a Strength of 14. As for IQ - I haven't been tested since Elementary School, so I consider that result to be questionable.

Here's the table I used to make these charts:

So how are y'all shaping up?

I personally think that this scale underestimates how strong humans can actually be, with proper training. The world record benchpress is 1,102 pounds, according to Chron, but I sadly can't post the link because this is my first post. My point is, the Strength scale's maximum benchpress stat (not even looking at IQ scores and deadlifts) caps out vastly below the maximum known strength a human can achieve unassisted. Keep in mind that in the real world we don't have fancy strength-boosting gear or potions either, as powerlifting competitions naturally ban any sort of performance enhancers (and while there was a documentary I saw in which an out-of-shape middle-aged guy with a homemade exoskeleton entered a weightlifting competition to demonstrate it to the public, that was explicitly a special case for obvious reasons), so the idea that human strength caps out at only around 203 pounds benchpressed is rather disingenuous.

In short, I think this scale needs revision.

[Frozen_Feet]

The world records for various lifts indeed go further than the table, but they're also made by people who are likely more than 3 standard deviations from the expected value. So it won't necessarily make sense to cram them in the 3d6 range to begin with. (See above about 18 not actually being the cap.)

[AMFV]

I personally think that this scale underestimates how strong humans can actually be, with proper training. The world record benchpress is 1,102 pounds, according to Chron, but I sadly can't post the link because this is my first post. My point is, the Strength scale's maximum benchpress stat (not even looking at IQ scores and deadlifts) caps out vastly below the maximum known strength a human can achieve unassisted. Keep in mind that in the real world we don't have fancy strength-boosting gear or potions either, as powerlifting competitions naturally ban any sort of performance enhancers (and while there was a documentary I saw in which an out-of-shape middle-aged guy with a homemade exoskeleton entered a weightlifting competition to demonstrate it to the public, that was explicitly a special case for obvious reasons), so the idea that human strength caps out at only around 203 pounds benchpressed is rather disingenuous.

In short, I think this scale needs revision.

That isn't unassisted. That's with a bench shirt. It adds A LOT. Like the world raw record is currently 738 lbs, so not quite so much over the maximal record. Also worth noting that benching includes a lot of technical options a person can do to make their weights heavier without actually getting stronger, like by working on arch and technique. The bench press is a really technical lift and as such is a bad choice for this sort of thing because it's more technical than raw force. The deadlift or push press records would be much better to use for this.

It's also worth noting that what "move around with the weight overhead" means is significant. I think you could probably find strongmen who could walk around with their heaviest OHP overhead. And as far as the deadlift and 5' step goes, we see that sort of thing in frame carries and what-not.

[Lvl 2 Expert]

(See above about 18 not actually being the cap.)

Either that, or 3d6 should not be treated as the distribution for random characters, it's just a method for generating heroes worthy of legend (as is 4d6b1). The distribution from the opening post after all does not take skill and such into account. Ones you start doing that, a level 5 character with an 18 in their best stat and the right build can challenge an Olympian (http://thealexandrian.net/wordpress/587/roleplaying-games/dd-calibrating-your-expectations-2). (If this does not work for strenght feats it's probably because your dm does not use athletics checks for them.)

I prefer that version of the idea of d&d myself. A 16 is really good, a cut above "regular" elite performances, as per the elite array. As a bonus this makes racial bonuses more distinctive. Orcs are not just kind of strong, they are really strong. I'll take the small differences in for instance lifting weight for situations where skills don't apply as an inaccuracy in that.

[Tvtyrant]

I think the real problem with using 3-18 is that it caps at 18. Yeah, 3d6 is approximately normal. Yeah, IQ is normal, and I’d guess there are ways to quantify strength, dexterity, and constitution which would also be approximately normal. But you will have extreme outliers for IQ and whatnot, several standard deviations above the mean, which would be well above an 18.

I always likes 5d4 myself. Higher maximum and tighter curve.

[Jay R]

Back in the 1970s, a very early issue of The Dragon had an article describing how to find your own real-world stats. Much of it was nonsense, and the only part I remember was the definition for Wisdom.

Your Wisdom score was defined as 20 minus the number of hours per week you spend on D&D.

At the time, that gave me a Wisdom of about -10.

[hamishspence]

Games like D20 Modern and D20 Call of Cthulhu might be a good starting point for translating real-life terms like "puny" or "graceful" into D&D stats.

Int 12: "Sharp witted"
Int 14: "clever person"
Int 18:"Genius"

Wis 6-7: "foolhardy"
Wis 18: "guru"

Cha 8-9:"shy"
Cha 12: "Attractive person"
Cha 14: "Take-charge type"
Cha 18: "Natural-born leader"

6-7 Con: "puny"
4-5 Con: frail"
Con 12: "healthy person"

Str 16-17: "weightlifter"
Str 18-19 is "olympic athlete"

Dex 12 "graceful person"
Dex 17-18 "circus acrobat"

[AMFV]

Games like D20 Modern and D20 Call of Cthulhu might be a good starting point for translating real-life terms like "puny" or "graceful" into D&D stats.
<snip>
Str 16-17: "weightlifter"
Str 18-19 is "olympic athlete"
<snip>


The problem is that strength 19 is NOT an olympic caliber strength score. At least not in d20 or 3.5. The problem is that by tying it back to concrete numbers of weights lifted they kind of worked themselves into a corner. I think it might be better to just have the strength scores be an actual abstraction. I mean it's not like people are saying "Hey that weighs 502 lbs and your strength score only allows you to lift 501." So I mean it'd be easier to have an abstraction. Like for the other stats which are all a good deal more abstract.

So you could have something like this:

Strength 10 : Average, should not have problems completing average human tasks.

Strength 12-14 : Above average strength, should be able to lift fairly heavy objects, roughly as strong as a laborer or field worker.

Strength 15-17 : Significantly above average strength. Able to lift very heavy objects. Roughly as strong as a heavy laborer or heavy construction worker.

Strength 18 : As strong as most humans can get unassisted, able to lift heavy objects easily and move them in awkward ways.

That works better because it's largely abstract, which means that the whole system doesn't collapse. Also because strength in D20 largely deals with weight moved, your strength score should be tied back into your actual weight, whereas they don't. So it's probably easier to abstract it.

[hamishspence]

The problem is that strength 19 is NOT an olympic caliber strength score. At least not in d20 or 3.5.

It's a little low for an olympic weightlifter, true. Olympic-standard in other Events might work though. Keep in mind that 18's only the highest starting stat - the Human Paragon class from Unearthed Arcana

http://www.d20srd.org/srd/variant/races/racialParagonClasses.htm#humanParagon

and the Prodigy (DMG2, +2 to one stat) traits, allow a low level character to have a stat in the low 20s - in line with The Alexandrian's argument that level 5 should be the upper limit.



For "non-max Strength Olympian events" like Swimming (nobody would expect the world's greatest swimmer to also be a great weightlifter) = Prodigy is a good way to beef up their best Skill stat = in addition to the stat bonus, it grants a +4 untyped bonus to all skills the stat governs.

So you could have a Str 13 to Str 15 olympic swimmer, with Prodigy of Str, Skill Focus Swim, maxed Swim skill, etc, and they would have a decently high Swim score.


Using DMG2, the highest a "1 Hit Dice character" can get in a stat is 20, thanks to Prodigy. With Human Paragon and levelling, you can reach world record levels, nice and early.

[AMFV]

It's a little low for an olympic weightlifter, true. Olympic-standard in other Events might work though. Keep in mind that 18's only the highest starting stat - the Human Paragon class from Unearthed Arcana

http://www.d20srd.org/srd/variant/races/racialParagonClasses.htm#humanParagon

and the Prodigy (DMG2, +2 to one stat) traits, allow a low level character to have a stat in the low 20s - in line with The Alexandrian's argument that level 5 should be the upper limit.



For "non-max Strength Olympian events" like Swimming (nobody would expect the world's greatest swimmer to also be a great weightlifter) = Prodigy is a good way to beef up their best Skill stat = in addition to the stat bonus, it grants a +4 untyped bonus to all skills the stat governs.

So you could have a Str 13 to Str 15 olympic swimmer, with Prodigy of Str, Skill Focus Swim, maxed Swim skill, etc, and they would have a decently high Swim score.

You'd still have the problem of the model winding up failing as a modelling tool. D&D is not a good sports simulator. It's an acceptable combat simulator. I mean if you wanted something that would really reflect real life athletics D&D is not your best bet, simply because there are too many things that increase when other things do. Like a world champion swimmer shouldn't necessarily be an awesome weightlifter and they certainly shouldn't be able to go many rounds with a boxer as the HP required to be level five would allow them to do.

I think the best bet is to treat the strength scores more abstractly rather than as examples of specific benchmarks, because assigning those really specific benchmarks is what's causing it to break down. I mean you'd even have cases where you'd have a guy who could deadlift or overhead too much respectively. Also the difference between an 800 lb deadlift and a 900 lb deadlift is much more significant than the different between a 300 and a 400. That's not well modeled in the system.

[hamishspence]

Like a world champion swimmer shouldn't necessarily be an awesome weightlifter and they certainly shouldn't be able to go many rounds with a boxer as the HP required to be level five would allow them to do.

Expert (or even Commoner?) hit dice for the first one, and "minimum number of hit points rolled" can partially account for that.

While it'll never be perfect - I think 3.5 can produce very rough approximations of extremely talented and gifted real-world people.

[AMFV]

Expert (or even Commoner?) hit dice for the first one, and "minimum number of hit points rolled" can partially account for that.


Partially, but not significantly enough though. Again, D&D is not designed to model athletes, it's designed to model adventurers who are basically soldiers. So assuming that their strength and constitution will directly tie back into those thins is reasonable. But not for athletes.

Look if you're having to massage the model that much to produce even mildly believable results, it's a bad model. And that's to be expected because you're trying to use it for something it is not designed for.



While it'll never be perfect - I think 3.5 can produce very rough approximations of extremely talented and gifted real-world people.

Well it depends on the purpose that you are approximating them for. If you're trying to figure out if Micheal Phelps would win in a fight with a bear, it's a horrible approximation. If you're trying to figure out which athletes would win in any kind of athletic competition, it's again a horrible approximation.

It's really only good at measuring characters who are roughly similar and how they perform in combat, which is what it's supposed to do. It's when people start trying to use the system for things that it is definitely not designed for that you start running into problems.

[hamishspence]

Partially, but not significantly enough though. Again, D&D is not designed to model athletes, it's designed to model adventurers who are basically soldiers. So assuming that their strength and constitution will directly tie back into those thins is reasonable. But not for athletes.


3.0-3.5 D&D is part of the wider D20 system though.

The OPs question of what the stats correlate to is still worth considering - even if "a few provisos, a couple of quid pro pros" :smallsmile: is necessary to answer.


For Intelligence, Wisdom and Charisma, you get bonuses based on age, so if a character is allowed to start at venerable, they might have a score as high as 21.

I don't remember any similar permanent boosts in core only for physical stats, but a Barbarian's rage can temporarily push both Strength and Constitution as high as 22.

Once you move outside core, there are things like the Paragon template, and once you move beyond starting characters, there's 4th level ability increase etc. These are important, because d20 D&D doesn't model truly exceptional people just by their ability scores, it also adds class, experience and templates in the mix.

Anyways, the overall point is not to try overfit all of humanity in the 3d6 range, because the game doesn't really do that either.

Yup. Hence my useage of Paragon, Prodigy, and 5 or so levels - to get 23s as top end "non-superhuman" stats for people who are both very gifted and very experienced.

[tensai_oni]

Ugh, this thread again. Let's whip out my checklist of erroneous statements that always show up in this kind of threads, and see which ones were said or at least implied so far:

[X] DnD scores can (easily and directly) translate into real life numbers
[X] INT is IQ (while we're at it, taking IQ seriously? In the 21st century?)
[X] Most RL people would have all stats in the 8-11 range
[X] 18 is the highest possible score a RL human being can have
[X] Most RL people would be level 1
[ ] Most RL people would be Commoners, class-wise

Uh oh, looks like we missed one. I guess the thread isn't over yet, folks.

[Knaight]

A dude could be good at bench press but not inherently good at climbing.

This is a particularly big issue with strength in D&D. There are a fair few things that it covers where strength makes you better at something but you're working against your weight, with climbing and jumping being two obvious examples. This gets particularly screwy given the D&D strength scale. An elephant is much, much stronger than a squirrel. That doesn't mean it jumps or climbs better.

Sticking to roughly same shaped creatures, you run into similar issues. It's bad enough with people (where a big 350 pound person in good shape is almost certainly much stronger than a small 120 pound person in good shape, and there's roughly no reason to assume that they're the better climber), bringing in halflings and giants just makes it ridiculous.

[Lvl 2 Expert]

[ ] Most RL people would be Commoners, class-wise

Okay, I'll complete the list for you. Most people from a RL a medieval setting would be commoners, class wise. An educated professional in anything other than combat would usually be an expert, which brings with it a free choice of class skills. You could build a pretty good swimmer with 5 levels of expert. A boxer or a rugby player probably even qualifies for a combat based class. This in turn helps with discriminating between different uses of the same base stat. An Olympic swimming champion at lvl 5 could still be beaten up by a lvl 3 sort of decent American Football pro, because that guy has D12 hit dice (12+6.5*2=25>6+3.5*4=20), just as much of a base attack bonus (+3) and a feet for tripping plus another for grappling. The Olympic swimmer meanwhile has skill points to spare to be better at climbing and various athletics than the American Football player is, on top of creaming the guy in a swimming contest with skill focus and such.

The most people are commoners can also help with the confusion of lifting strength. D&D assumes you'll be using skills or combat mechanics, not raw stats unless you can't help it. If you try to model weight lifting records on base stats alone you need ridiculous numbers. But given how hard it is to gain base stat points during play, this is not how the game tries to model it. If the game was at any point serious about weight lifting as a profession needing to be modeled well, it would made made a skill for it, including a skill focus feat and possibly a synergy with some other skill. If most normal people do fit the curve it's because most people are low level commoners and experts or they simply focus on other fields, just like a wizard in the game does not lift weights very often. In fact, you can probably shrink the curve a little, since people with high base stats are also more likely to have one to four levels of training in related field. So suddenly "x% of the people need to have an 18 because they lift with a bonus of +4" becomes "O, maybe that's just a 12 with 3 skill ranks, never mind". And yes, that's perfectly within the reach of a lvl 1 or 2 commoner or expert using standard array or commoner array (except commoner array people would need 4 ranks because they don't get a 12, just 10's and 11's, at which point they're pretty serious about lifting, as it's one of their best skills.)

The fact that there isn't a "goes to the fitness center a lot" skill and the rulebooks specifically tell you to use raw strength in these cases just tells us the writers think the topic is boring and we should go do something better with our time, like fighting some orcs (the answer is always fighting some orcs). It does not mean that a thing real life people train in to become good at it should be modeled as if it's not something you can train in. Training is what the skill system is for.