Nested comments on Substack become unreadable after a certain number of levels, so I’m using this post to continue a conversation.
Jeff: We know that athletic ability is heritable in individuals, but you’d look pretty dumb if you said it’s plausible that Jamaicans are naturally better sprinters, Kenyans better distance runners, and African-Americans better basketball players.
Now if you said Kenyans are better distance runners and Norwegians better cross country skiiers, one could come up with a dumb but superficially plausible (because they look different, so they must be really different deep down) just-so story about different selective pressures in Norway vs Kenya.
And then I’d tell you it’s a dumb just-so story, and you could tell me that “explaining via heredity is plausible, but not proven. More research is needed to test the hypothesis.”
TGGP: I actually do think it’s likely that people of East African descent are naturally better long distance runners, while people of West African descent are naturally better sprinters. https://westhunt.wordpress.com/2018/04/08/live-not-by-lies/
That link doesn’t discuss it, but East Africans tend to be taller & thinner than West Africans, resulting in advantages in different sports.
Jeff: West Africans in general are not notably great sprinters. It’s specifically a few Caribbean islands that excel at it, and the best of them, Usain Bolt, is tall and thin. It’s not because Caribbeans have genes to be good sprinters. It’s because in Jamaica sprinting is pretty much the national sport and has been for decades.
Countries excel at whatever sport kids there are disproportionately into. Brazil, soccer. Japan, baseball. America, basketball. Norway, winter sports. And so on. With team sports, it’s obvious that it’s not genetic. With individual sports, if you’re prone to over-theorizing stuff, it’s easy to convince yourself that it’s genetic. But it’s really just how many kids went into the funnel and whether the talented among them got support.
TGGP: The West African advantage in sprinting isn’t just Jamaicans/Caribbeans. Look at this list of the top 25 men in the 100m:
https://en.wikipedia.org/wiki/100_metres#All-time_top_25_men
There are 5 Jamaicans and 3 from Trinidad/Tobago. There are 10 Americans and 2 Canadians (recall that the Middle Passage mostly brought West Africans to the New World). The one remaining country with multiple people is Nigeria with 2. Aside from them, Namibia also has one entry, as does France and Portugal (again, of West African descent). Talking about the culture of the Caribbean can’t explain why Nigeria has more than the entire continent of Asia. Running fast for short distances is something basically all kids do all over the world, and no specific culture had to invent. The fact that continental West Africa has fewer entries per capita than the West African diaspora does indicate that some non-genetic factors probably play a role, but you can’t explain the odd coincidence of every entry being West African descended even though they live in multiple different regions/cultures.
Jeff: There are 3 million Jamaicans vs 44 million African-Americans, but Jamaica produces as many top 10 sprinters as the US and half as many top 25 sprinters as the US. Do you think all the fast slaves got sent to Jamaica or something?
TGGP: As I said, the per capita numbers indicate to me that there are some non-genetic factors. But the complete absence of any people not of West African ancestry (which is to say, the overwhelming majority of the Earth’s population) speaks much louder than 3 million vs 44 million having equal numbers in the top 10.
Jeff: It’s totally possible that West African genes help, but you have to add a bunch of epicycles to the theory to explain why West Africans make great sprinters in Jamaica, good sprinters in America, and are nothing special anywhere else. And if you just drop the West African bit and keep the epicycles, you have a more parsimonious theory with just as much explanatory power.
 |
TGGP: I disagree, I think a theory about West African ancestry explains much more than your Jamaican/Carribbean specific theory. A simple version would be that there’s some threshold to be at all competitive at the top levels of the 100m, and only people with West African ancestry can reach that level. Right away you’ve explained the astonishing factoid that the vast majority of the Earth’s population which is NOT of West African descent is absent. Then afterward we can say that there are smaller factors which have effects on the margin. We place less weight on those factors because they explain less than the theory about West Africans. In your Jamaican/Carribbean theory, you ONLY explain those 8 out of 25. It’s an astonishing coincidence that the 17 non-Carribbeans in that group are also of West African descent. Even the two from Europe! Combine with that, we have what Greg Cochran noted: there are physical differences we know about which are quite relevant for sprinting! My bet is that the next time a non-Carribbean gets added to that list, they will also be of West African descent. How much would you wager against me? |
Jeff: I wouldn’t wager much because the next guy will be African-American. But that’s mostly a function of US high school track demographics (plus America’s ridiculously good Olympic talent pipeline), not West African genes. If China wanted to churn out world-class sprinters like they do gymnasts and divers, it’d be a Chinese guy instead.
The nesting became too much after that last comment from Jeff, so I decided to move the discussion here.
TGGP: WHY are “US high school track demographics” such that people of West African descent dominate them? They’re a minority of the U.S population! Don’t you find it odd that both they and the Caribbean sprinters have that shared ancestry? Would you be willing to bet against me of whether the next person on the list who is NEITHER Caribbean nor American will also be of West African descent?
February 18, 2021 at 9:37 pm
One more comment I have in response to his last one: do you really think it’s arbitrary that the Chinese are focusing on gymnastics/diving rather than sprinting, while Jamaicans are doing the reverse? Do you think it’s arbitrary that Samoans compete a lot in sumo wrestling and American football rather than any of the aforementioned sports? Are there any East Asian countries which (arbitrarily) decided to focus on sprinting? Any countries predominately of West African descent which (arbitrarily) decided to focus on one of those other areas?
February 19, 2021 at 1:21 am
TL;DR: America is good at everything, and black Americans do track a lot. Jamaica is really, really good at sprinting. The rest mostly seems like noise to me.
I don’t think it’s *random* which sports people wind up in. Tall people disproportionately like basketball, regardless of background. You won’t be a good linebacker if you’re average size. But to get a lot of people into the funnel for a sport, it has to have neighborhood-level appeal — kids and parents have to think it’s cool, there have to be facilities for it, and so on. A whole culture has to get organized around playing the sport.
Then, to turn the kids into Olympians, you need all sorts of national-level infrastructure (teams, coaches, consultants, competitions, sponsorships, governing bodies, etc). Most countries can’t sustain the cost of that infrastructure for more than a handful of sports.
America actually has this infrastructure for almost every sport (including track & field), so here you notice more when it’s missing than when it’s present — soccer, notably, doesn’t have much of a pipeline from gifted middle school player to international player. For everything else, we have a high school -> college -> professional pipeline that’s well-funded and crazy effective. But in most countries, in most sports, the opposite is true. A gifted middle schooler has no opportunity to develop their skills.
Sprinting has had the “cool” factor going for it for decades in Jamaica, and has also built up the infrastructure to recruit and train their talent. It’s so much of a part of the culture that elite sprinters *outside* of Jamaica are often Jamaican emigres or their kids, even though Jamaica has a population similar to Queens, NYC.
Once you exclude Americans and Jamaicans (counting Jamaican emigres and their kids as Jamaican), there aren’t that many world class sprinters out there. Nigerians and Nigerian diaspora are the biggest group, but there’s a smattering of white and Asian and non-West Africans too.
Ok, but all the Americans *are* black — what’s with that? Well, black athletes are wildly overrepresented at the top tier of American sports, but it’s not uniform across sports. The NFL is 70% black and the NBA is 75% black, while the MLB is 7.7% black and NHL is only 7% non-white (not sure what share of those are black). Tennis mostly white, but the best three female tennis players are at least part black (Serena, Venus, Osaka), and golf similarly is mostly white but the most famous golfer is part black (Tiger).
Well, yeah, but those are at least partly not black. Not 100%. Fair enough. But what’s actually anomalous is that America should be able to produce good white sprinters — France and Russia and the UK have managed. My guess, from what I remember of high school sports, is the talented white sprinters wind up as football wide receivers and soccer strikers, play that in college, too, and then never pursue professional track.
February 19, 2021 at 1:39 am
Another thought… most African-Americans are of partly European heritage, certainly more so than Nigerians (and possibly Jamaicans?) but diluting the allegedly performance-enhancing West African genes doesn’t seem to put them at much of a disadvantage.
February 19, 2021 at 10:09 am
Yes, I’m pretty sure Jamaicans are less European than African-Americans. In 1850 the black population outnumbered whites 20 to 1, and slavery ended there earlier than it did in the US. African-Americans are 87% African by ancestry.
You’re simultaneously arguing that the per capita representation of Jamaica vs the US shows that the important factor is Jamaica/Caribbean specific AND that the different demographics of the US (that 13% non-African ancestry) “doesn’t seem to put them at much of a disadvantage”. The non-African ancestry of the large majority of the US population DOES seem to put them at a disadvantage!
February 19, 2021 at 8:39 am
I think you make a number of good points, but when I look at the list of top sprinters, those cultural/country-specific factors look more marginal compared to the complete dominance of people of West African ancestry. Even if you exclude the 5 Jamaicans & 10 Americans, the remaining 10 are all West African in ancestry. That seems like way too much of a coincidence. You mention that France is able to produce good white sprinters, but the one Frenchman on that list is Ivorian. What are the odds!?
I don’t know if Tiger is still the best golfer, but he is well known in part because it’s unusual for a top-level golfer to be black. Golfing is not a sport that draws on sprinting ability. Basketball is. Even within the NFL, there is varying distribution by position. When was the last time a white player regularly started as cornerback? Quarterbacks can pass instead of running, and they are a lot more white. Catchers in baseball don’t have to run. How many black catchers have there been (I’m ignorant of baseball, so I really don’t know). Speaking of more things I don’t know, are whites overrepresented among wide receivers? Years ago “Caste Football” was arguing that teams were considering that a “black position” off-limits to whites, although Steve Sailer wasn’t convinced it was actual bias.
February 20, 2021 at 4:14 pm
FWIW I don’t think it’s totally crazy to say there’s a genetic component to people of West African descent doing so well in sprinting. It’s definitely _possible_. Tall people are better at basketball, so a tribe of freakishly tall people (like, 7 foot plus) would presumably produce way more NBA players than a tribe of normal height people.
I just think there’s some p-hacking going on, as in https://xkcd.com/882/
There are many genetic just-so stories one can dream up for why group X does better than group Y in some sport because every sport has some version of that dynamic.
For many of them, it’s obviously not genetic because the sport is confined to one geographic area (baseball in the US and Japan), or the dominant group changed recently (the Chinese took over gymnastics from the Russians).
For some, it’s obvious because the group that overperforms is not defined genetically. Brazilians are great at soccer, but Brazil is an ethnic melting pot. Germans are great at soccer, but some of those German nationals are of Turkish or African descent, and German isn’t a distinct genetic group anyway.
For others, it’s not obvious until you look at how they live. White Americans excel at swimming relative to African-Americans because white kids disproportionately grow up participating in swim teams.
For others, it’s peer effects. White kids run cross country more than African-Americans. African-Americans run track more (and play football during cross country season). Maybe at the elite level you could say that’s genetic, but at the level of JV high school sports, it’s just kids doing what their friends are doing.
For others, it’s parental influence. White parents disproportionately don’t let their kids play football, and boxing stopped being ok long before that.
For others, it’s the gifted amateur -> professional pipeline. That’s probably why Jamaica produces 10x as many world-class sprinters per capita as African-Americans (and at least 100x as many per capita as Nigeria).
Sometimes the sport changes. We used to think sprinters couldn’t be tall, so any group that tends to be shorter should disproportionately produce better sprinters, but then a Jamaican coach figured out how to make a tall guy run fast so now we have Usain Bolt. This is related to the gifted amateur -> professional pipeline in that the best coaches will be present wherever you already have lots of elite athletes.
Sometimes there’s discrimination. A lot of people used to think black guys couldn’t play quarterback, now Pat Mahomes is the best QB in football. And according to Moneyball, blond baseball players were more successful in baseball than dark-haired players… because they stood out more to scouts in a field full of dark-haired players.
If after sifting through tons of sports, it turns out that there are a handful where there’s somewhat well-defined genetic group that’s dominant, and there’s a semi-plausible genetic thing that differentiates them from others, and I can’t debunk the theory with trivia I already happen to know, I’m inclined to say it’s most likely a coincidence, or that an expert on the sport could debunk it.
To show it is genetic, and not just an ex post rationalization (i.e., “West African” is more than just a union of African American, Nigerian and Jamaican) I’d want to see some kind of interesting/surprising result in another sport that’s predicted by this. But I think those are hard to find because at most levels in most sports, any possible effect of genetics is swamped by all the confounders.
You mention cornerbacks in football and catchers in baseball.
Catchers: Pro baseball players are only 8% black, so there aren’t a lot of black guys at any position, but IIRC that number used to be higher when the amateur -> pro pipeline that originally supplied the Negro Leagues still existed. And back then black players excelled at every position. Home run hitters are usually slow and pudgy-for-a-pro-athlete (i.e., opposite of sprinters), but the home run leaderboard is mostly African-American and Dominican.
Cornerbacks: no idea what’s going on there. My level of football fandom is “I watch the Super Bowl if I remember to”, but the history of black players at quarterback suggests discrimination is (or was, until very recently) alive and well in football.
February 20, 2021 at 5:44 pm
You link to the comic, but you don’t seem to understand what p-hacking is. I never brought up statistical significance. The complete absence of people without West African ancestry from the top ranks of sprinters was so obvious I didn’t need to! I actually prefer Andrew Gelman’s phrase “garden of forking paths“, in which there are many possible analyses and the fraction which give rise to statistical significance is small enough that it would not be notable if we knew about all the other analyses that didn’t. But this isn’t an academic paper in which I needed to hit statistical significance to publish (or perish). This is something multiple people have written entire books about, not mere articles based on a single p-value.
It’s a known fact that there are differences in bone density, which affects bouyancy, between the races. You haven’t “looked into” the issue at all before announcing that the difference must be swim teams. There are known differences in how frequently different groups drown in swimming pools, so this is not merely a matter of sports trivia.
That logic doesn’t make any sense to me. Why wouldn’t genetics have an effect at the JV level? You mentioned that height is partly a matter of genetics, wouldn’t that also matter there? How can you assume peer effects rather than kids noticing what they’re relatively good at? WHY does one demographic do one sport more than another? It’s not like basketball was dominated by African-Americans from the beginning.
You might be right about those ratios, but there’s an even larger ratio of world-class Jamaican sprinters (or African-American sprinters) to world-class sprinters without West African ancestry.
I’m not an expert on football, but I think the consensus would be “one of the best” rather than “the best”. People have been writing for a long time about quarterback being whiter than other football positions, and the examples of multiple talented black quarterbacks hasn’t actually ended that.
Sailer pointed out that David Epstein wrote a book on the subject, Jon Entine had done so earlier. I pointed you toward Greg Cochran (who has actually published on genetics) talking about how WE ALREADY KNOW the different phenotypes that give an advantage in sprinting. You assume an expert could debunk it, but you haven’t pointed to any expert.
It’s not! There were 5 Jamaicans, 10 Americans & 2 Nigerians on that list of 25 which leaves 8 more people who all had West African ancestry!
What would you consider “interesting/surprising”? The basic idea is that there’s an advantage in sprinting, and so sports/positions which draw on that skill will see an effect.
February 20, 2021 at 9:08 pm
If you think black people don’t make good swimmers because of their buoyancy, and don’t make good quarterbacks because of their cognitive abilities, etc., I don’t really think this is a conversation worth continuing. I guess I could’ve also just inferred all of the above from your approvingly citing Steve Sailer :/
But I’ll go one more round, I guess. :shrug:
I can swim because I was taught to swim as a kid, not because being a white guy makes me a tad more buoyant. Buoyancy really doesn’t help much if you don’t know how to swim — if you’ve ever seen an adult who doesn’t know how to swim fall in a pool, it’s pretty funny, until it becomes very not funny. Go to your local YMCA and watch an adult swimming class if you don’t believe me.
JV sports are instructive because JV athletes by and large are terrible athletes which is why they’re stuck at the JV level. There’s not much selection bias for actually being good at a sport — high school JV is where you land when you’ve been selected *out* of the sport. It lets you see what kids like doing enough to keep at it even though they suck.
The XKCD cartoon is partly about p-hacking, but more generally it describes analysis of the “throw stuff at the wall and see what sticks” variety. Racial theories of athletic ability do that — “throw just-so stories at the wall and see which ones can’t be debunked by someone with Google and some spare time”. The number of them that turn out to be transparent bullshit are proof of this. For example…
…the black-people-can’t-be-QBs-because-they’re-dumb theory. QBs aren’t the smartest players on the field. The O Line has the smartest players (going by Wonderlic scores), and the O Line is mostly made up of guards and tackles, half of whom are black in the NFL. Centers are mostly white, and have lower Wonderlic scores than guards and tackles. QBs, also mostly white, score lower still.
> WE ALREADY KNOW the different phenotypes that give an advantage in sprinting
Until Usain Bolt came along, one of those phenotypes was supposedly being short (or at least not tall). Until 1968, 10 seconds was a hard speed limit. Now as many people break the barrier every year as broke it from 1968 to 1990. We don’t understand biology and biomechanics all that well. To the degree that we do understand it, it’s often not public because it’s a trade secret for the doctors and coaches who figure it out.
> David Epstein wrote a book on the subject
Sickle cell is clearly an adaptation to malaria, but beyond that these theories are just ex post rationalizations for lots of good sprinters having West African ancestry. One could as easily assume that West Africans have a slightly higher frequency of some gene for enjoying some aspect of sprint training.
> What would you consider “interesting/surprising”?
For example: if there’s a genetically distinct subset of West Africans who never become good sprinters. Better still if they become, like, really good shotputters instead.
February 21, 2021 at 12:11 am
I didn’t actually cite cognitive ability as a factor for quarterbacks. I just noted that it relies less on sprinting than other positions (because they can pass instead of run). And yes, I do think bouyancy is a factor for swimming! That should be obvious! You keep saying things are laughable when you are actually ignorant of the subject, and you haven’t cited any supposed “expert” like the ones you imagine might debunk such theories.
Of course knowing how to swim is relevant, but I don’t think you can dismiss bouyancy. Increase an organism’s density sufficiently and there’s no way it can swim no matter how much practice it has. The ocean of full of creatures whose bodies are adapted to swimming (in addition to ones that can live on the ocean floor), they don’t just rely on spending so much time swimming that they’re good at it.
JV are kids below a certain cutoff, but that doesn’t make them worse than the broader population of kids they’re drawn from.
Not partly, it’s EXACTLY about that.
That sounds like “hypothesis testing”. And since I’ve cited Andrew Gelman’s complaints about the term “p-hacking” I should note that he has a broader problem with the broader approach to hypothesis testing and thinks more research should be framed as exploratory rather than confirmatory. But the way exploratory research gets confirmed is with replication, and it’s the replication crisis which exposed the problem with p-hacking. You haven’t noted any failures to replicate anything.
Selection on a trait only requires that they be above average on that trait. I think there are multiple traits involved in being a good QB, and a balance of multiple selection pressures will result in tradeoffs.
It wasn’t the speed of light, it was just a record that hadn’t been broken yet.
People have been studying the biology of running a very long time. This is not one of those “scientists don’t know how bumblebees fly” situations.
That’s entirely possible and would be consistent with the evidence of West African dominance at the highest levels (dismissing it all as a coincidence is implausible in my view). I don’t know if there’s any research to back it up, but it’s at least an hypothesis to investigate (we could also test the sprint speeds of kids young enough not to have any training). But we actually do know for a fact there are phenotypic differences which you’re just dismissing as “We don’t understand biology”.
There are people with crippling mutations, but I don’t know if you’d consider that a “genetically distinct subset”.
I don’t see any reason to expect talent in shotput in particular given the theory we’re discussing testing.