How Realistic is D&D When it Comes to Strength and IQ?

One of the things that’s always struck me about Dungeons & Dragons is its realism.  This is especially true today, when most RPG mechanics are built up around Skinner Box techniques rather than meaningful progress, and it speaks to the effort that Gary Gygax and Dave Arneson put in to their game system – especially when you take into account that this was before the Internet allowed you to find weird, obscure information in a matter of clicks.

Can you imagine sitting down in 1974 and trying to figure out “How much would a chicken cost in 1350 A.D.?” The answer they came up with is relatively accurate (2 copper pieces), unlike modern games such as Fable where a loaf of bread goes for one 1 gold piece ($1320 USD at the time of this writing).

The D&D monetary system might be worth analyzing in detail some other time.  For now, let’s focus on the base mechanics.

Dungeons & Dragons breaks down the human animal to six variables, three physical and three mental: Strength, Dexterity, Constitution, Intelligence, Wisdom, and Charisma.  Each of these Base Stats is then given a score between 3 and 18, which is determined by adding up the sum of three six-sided die.  When you plot out these rolls, you wind up with a normal distribution that appears as follows:

Note: the probability of receiving a certain score is based upon the possible ways to roll it. Scores of 3 and 18 only have one possible roll – 1/1/1 and 6/6/6 respectively. A score of 5 has 6 possible rolls – 2/2/1, 2/1/2, 1/2/2, 3/1/1, 1/1/3, & 1/3/1. Et cetera.

Ability Scores of 10 & 11 are considered average, with neither bonuses nor penalties applied to them.  18 is considered superhuman, while 3 is the lowest of the low.  The scores continue upwards indefinitely – beings such as ogres and or rhinoceroses are stronger than the strongest man, and would have Strength scores to match – and while in theory it continues indefinitely leftward into the negative numbers, for practical game play purposes a score of 0 is generally considered ‘Non-corporeal/brain-dead’.  A tree, for instance, might have a fairly high constitution score, but it’s Str, Dex, Int, and Chr would be set at 0 (the fact that trees communicate with pheromones, organizing to ward off parasites, suggests that there might be some sort of Wisdom score there, too – but that’s only if you want to be really autistic about the whole thing).

But what do these scores mean?  Obviously they fall into the Gaussian/Bell Curve distribution, so if we can determine the averages amongst humans in the real world, we can apply these averages to the Ability Scores and translate them into real world numbers.

Out of the six Abilities, only two of them are easily translatable into measurable traits.  Dexterity (your reflexes, balance, hand-eye coordination) and Constitution (overall health, resistance to poison, endurance) are ‘catch all’ traits which simplify a number of diverse attributes into a single score.  In the real world, these traits can vary: for instance, the long-distance runner who dies from a heart attack.  His running would suggest a high Con score, while his heart attack would suggest it was low.  Similarly, Wisdom (your ‘in tuneness’ with nature – how observant you are, how much common sense you display) and Charisma (your self-possession and force of personality) are obvious traits that we notice, but I’m unaware of any testing that can definitively say “This man is more charismatic than X% of the population.”  This leaves us with Strength and Intelligence – two abilities for which we do have accurate measurements.

IQ is extremely well-defined, given that it was premised upon the Bell Curve in the first place.  The Percentage and Cumulative rows in the above chart refer to the number of Ability Scores rolled in the general population: what each score’s percentile score is.  The IQ listed below it is a direct translation of percentile into IQ. (Confession time: I forget the math behind this, so I just used this handy chart).

For Strength things get a little bit more difficult.  While there are plenty of scales out there that list ‘Average, healthy, strong’ or whatnot, they’re mostly focused on fitness plans and marketing.  I’m not interested in what Health Canada thinks ‘healthy’ is – I want to know what the population averages are!

Eventually I found this chart, referenced by this post on the FlashFlash Revolution forum:

If you’re a gym goer, perhaps these numbers seem inflated to you: I’ll simply quote the original poster Arch0wl who makes some educated guesses about how many people work out:

This data gives absolute numbers for bench press percentiles, — i.e. the average man under 30 has a bench press of 146lb.

Note: they are not saying “average gymgoer” but average man. Only fifteen percent of Americans have gym memberships but this does not say how many actually go. This says that “50% of all new health club members quit within the first six months of signing up and 90% of those who join health and fitness clubs will stop going regularly within the first 90 days.” Meanwhile this data states that 67% of people who have gym memberships never use them.

So, we can reasonably conclude that of the 15% with gym memberships, maybe 5% of the population has a gym membership and actually uses it.

Now here’s where some guesswork comes in, but think about what counts as a gym: planet fitness counts as a gym. Bally Fitness counts as a gym. Lifetime fitness counts as a gym. That one gym ‘curves’ counts as a gym. YMCA counts as a gym. For every gym like Gold’s or Metroflex, there is an equal and opposite Planet Fitness full of people who bench 100lb or less on a smith machine.

I go to Gold’s gym. I can even give you a google link to see inside my gym. If you click a bit further you can see the weightroom. Notice that no one is there — most people are doing cardio and machines. I know that at about 6pm they have these ridiculous aerobics classes where something like 40 women follow an instructor who helps them spend an hour burning maybe 200 or 300 calories to cheesy music. They do this several times a day.

So of *my* gym, where benching 315lb isn’t an uncommon thing, how many people actually lift? I’d say 1/3.

But let’s be generous and say 1/2 of gymgoers lift period, because I know that at Planet Fitness a lot of people lift with really light dumbbells. Bad lifting is still lifting. So of the people with gym memberships who actually go, half of them lift. That means only 2.5% of america actually lifts. Let’s round this to 3% for talking point purposes.

Keep this in perspective when you think the average man benching something like 145lb is way too low to be accurate. By going to a gym and lifting weights, you are the 3%. You’re part of at most 3% of the population who has the motivation to go and do that. The top 5% of that lifting charts starts at around 205lb, and only 3% of America actually lifts. So even though we know this isn’t true, the 3% of America who actually lifts could still be in the top 3% of bench press stats and that would still leave a mystery 2% to fill out considering the 95th percentile starts at 203lb.

If you’re curious, 99th percentile is also obtainable from this data since they list the SD; it’s a little over 235lb.

Now keep in mind that the above table only goes as high/low as the 95th/5th percentile: beyond these thresholds scores are hard to process (you’ve entered into the Stephen Hawking/Olympic Champion category, where numbers are small, and achievements are all over the place).  And yet, when we look at the D&D scores, we find that only 5% of the population has a score above 15.  In other words, we only have bench press data for those whose Ability Score is between 5 and 15.  Because IQ and bench press just so happen to have similar numbers, we can plot both of these out on the same chart (we’ll use the numbers for the 18-30 category):

If you’re curious, the author can currently do two reps at 185; this gives me a Strength score of 14.

Now D&D doesn’t include a scale for bench press, but it does have a scale for max lift.  Over on the now-defunct blog SimAntics, the author compiled the numbers on dead lift (it involves some rough guess work and some dead links, but we’ll take what we can get):

Plugging the rough data into our chart (going off of the curve to avoid the apparent artifact between 0.50 and 0.80) we get the following:

So how does this compare to D&D?  We’re provided with the following chart which defines what sort of loads a character can carry based off of their Strength score:

Given that Max Lift is defined as twice a character’s heavy load, we can then plot out the following:

In other words – the real world numbers we find for deadlift are almost identical to the numbers that Gygax and Arneson wrote down!  Luck?  Intuition?  Or good research and hard work?

A final topic worth considering is the difference between male and female scores when it comes to Strength.  In D&D (as well as most RPGs) sex is routinely ignored.  But the real world tells a different story.  When we plot out the numbers we got from the Hockey, Physical Fitness table we see the following:

The top 5th percentile of women are as strong as the bottom 10th percentile of men.  Furthermore, there is little variance across this chart; female bench press goes from 65 to 105 (40 lbs), compared to men’s going from 89 to 203 (114 lbs).

If I were running a D&D campaign, I’d institute the following house rule: female characters receive a -4 to Str, and a +2 to Dex… and even that’s far more optimistic than what these numbers suggest.

Fair?  Hardly.  But that’s what I always liked about D&D.  It’s realistic foundation – the fundamental unfairness of it – helped in making the fantastical elements easier to accept.

Now if only the Hit Points didn’t accrue like rust on an Maritime automobile…

ͼ-Ѻ-ͽ