Hokepoints is a Basketball State Comment Count

Seth

Regions

My regional breakdown, still.

After I did that regional study of football talent production by state, Michael Elkon (Braves & Birds, SB Nation, regular HTTV contributor) asked if I'd do the same with hoops recruiting. I responded that I'd love to, but we just had our first child and I need some time to stare at her. This is also my response for why I didn't have any content last week. In fact it is my excuse for everything; to those who don't have kids I can say "you don't understand" and they have to shut up because this is the ultimate trump card. Those who are already parents keep quiet because they're in on it. Having kids is AWESOME!

Anyway it's back to work, and because it's me that means charts. So back to charts.

This is NOT exactly accurate

Data are from the Rivals (most easily accessible) databases since 2003. Putting lists of football and basketball recruits against each other is not a one-for-one comparison. Basketball has more teams, fewer recruits per team, way more international players, and players who went directly to the NBA or committed to Kentucky or some other stupid one before they're done with the pretense.

Top basketball players are also far more likely to go to prep schools, and these are often nowhere near their hometowns. The Rivals database lists actual hometowns for many prep players, but not international ones, so, e.g., Canadian from Canada Nik Stauskas registers as a Massachusetts recruit despite being from Canada. Where a hometown was noted I used that. Some states will appear disproportionately large because their prep programs draw kids from around the region, but that is also an advantage to the schools near the prep programs.

Talent Supply By Region

As with football, the Southeast appears to produce a disproportionate amount of talent compared to its population, but to nowhere near the extreme as it is with football. Observe:

Region % U.S. pop
(2010)
% of Top ~400
Hoops Recruits
% of Top ~400
FB Recruits
Atlantic 22% 20% (-2) 15% (-7)
Midwest 18% 18% ( - ) 14% (-4)
Northeast 5% 6% (+1) 1% (-4)
Pacific 19% 14% (-5) 14% (-5)
Plains 17% 17% ( - ) 18% (+1)
Southeast 19% 25% (+6) 38% (+19)

The Atlantic, Midwest, and Northeast are considerably better represented, suggesting a marginally higher basketball orientation than the national average. My guess is this has a lot to do with the fact that it doesn't snow in gyms.

The list of top states in proportionally producing more basketball talent was heavily influenced by the prep school effect: New Hampshire (more than 3x their share of hoops talent) was done by three schools: Tilton, New Hampton, and the Brewster Academy. Most of Nevada was Findlay Prep, and Bishop Gorman sent most of the rest. Leaving those aside, the big basketball states (proportional to their population) were Kansas (209%), D.C. (202%), Mississippi (185%), Georgia (183%), Iowa (172%), Virginia (166%), North Carolina (154%), and Indiana (150%).

State Hoops Football % Pop
Illinois 28% 21% 23%
Indiana 18% 10% 12%
Iowa 10% 2% 6%
Michigan 14% 14% 18%
Minnesota 6% 5% 10%
Ohio 17% 45% 21%
Wisconsin 8% 4% 10%

There's a reverse prep effect at the bottom: Vermont and Rhode Island were drained by New Hampshire it appears, and Delaware seems to have sent their kids to Virginia or D.C. The remainder to produce less than half as much talent as you would expect from their populations: Alaska (17%), Montana (25%), Colorado (34%), Nebraska (40%), New York (41%), South Dakota (45%), and New Mexico (47%).

Michigan (3% of the U.S. population, 2.4% of the top basketball talent) was about in the middle, about even with Wisconsin, Oklahoma, Missouri, Ohio, and Arizona. Straight-up Michigan is the 14th biggest producer of basketball talent, and the 12th biggest producer of football talent. I thought the more interesting stat was within the Midwest (that above table), where Ohio produces nearly half of the top football prospects the basketball talent is shared.

[jump for where they go]

Regional Retention

Basketball seemed to have way more kids leaving their home region than football:

Chose School in Home Region
Region Basketball Football
Atlantic 49% 52%
Midwest 65% 75%
Northeast 18% 35%
Pacific 58% 75%
Plains 54% 72%
Southeast 58% 79%
ALL 54% 72%

This makes sense considering there are so few spots, so many teams, and most of all coaches are looking for a lot more specific guys for their systems. Those numbers are still pretty jarring. The upside is you can see here where the Big Ten is getting an advantage: the highest regional retention in the country. It doesn't explain why the SEC is terrible. A breakdown of the Midwest vs the Southeast:

Southeast Talent   Midwest Talent
To SEC 32%   To Big Ten 36%
Other major 39%   Other major 35%
Mid-major 14%   Mid-major 10%
Minor (FCS) 15%   Minor (FCS) 19%

The retention story doesn't seem to be it either. Could it be the Big Ten just recruits nationally better than the others? Final table: destination of recruits who leave their home region (football FCS schools left out to get a better comparison):

Conference Basketball Football
SEC 15.2% 19.3%
Big 12 14.3% 9.2%
ACC 12.9% 20.8%
Pac-12 12.2% 12.3%
Big East 11.4% -
American 10.7% 2.0%
Big Ten 8.8% 22.6%
C-USA 6.8% 1.3%
Mountain West 3.0% 1.2%
Sun Belt 1.8% 0.0%
Independent 1.6% 9.2%
MAC 1.4% 0.8%
FCS/minor 14.8% 1.3%

Nope. The Big Ten since 2006 has dragged more talent to its arms than any other…in football. In basketball they're not doing anything different; it's the SEC in fact that's drawing kids from elsewhere. Culprits: the Big XII's big number is from West Virginia and Kansas, the #1 and #2 out-of-region destinations in the country. Kentucky, Missouri, UConn, Cincy, and Marquette were some other big national recruiters.

What's keeping the Big Ten in the pack is Michigan State, Wisconsin, Purdue, three programs that have historically recruited a lot of 3-, 4-, and 5-star players, 95% of whom were from the Midwest footprint. No other school with more than 15 recruits over the period was 90% regional. You don't get to SEC schools until you dip that to 70% (LSU and the Mississippi schools), where you'll also find the rest of the Big Ten except Minnesota.

Conclusion: It appears the Big Ten is doing something else. Maybe better coaching, longer development periods, ??? It doesn't seem to be a bigger talent base though.

Comments

jmblue

March 11th, 2014 at 1:28 PM ^

I would not have guessed that Iowa would produce more basketball talent than Minnesota or Wisconsin, given that it's easily the least-populous state in the Big Ten region.  Interesting.

Illinois is quite the hotbed.  Makes you wonder why UI isn't more consistent.

Shop Smart Sho…

March 11th, 2014 at 1:32 PM ^

I'm guessing it isn't Illinois as a state that is the hotbed.  It is Chicago and the suburbs that produces the majority of the talent, would be my guess.  Compare that to Indiana, where the talent isn't as centralized in the largest city.  While the hoops high schools in Indianpolis are really good, you are still going to be able to find Mid-Major to low Major talent all over the state.

JeepinBen

March 11th, 2014 at 2:26 PM ^

Don't forget when it comes to population the city/suburbs of Chicago has around 10 million people. The city itself has produced a whole lot of elite basketball talent (Isaiah Thomas, D Rose, Garnett, Jabari Parker, etc etc etc) in addition to the city having lots of smaller, historically (if not recently) good basketball schools (DePaul, etc) plus the impact of Michael Jordan, when everyone really did Wanna Be Like Mike.

I guess what I'm saying is there are a whole lot of influences that make Illinois and specifically Chicago a good incubator for hoops talent.

 

As to why that hasn't translated into wins in Champaign... I'm not sure.

funkywolve

March 11th, 2014 at 1:30 PM ^

Congrats on the baby girl!!

It's be interesting to know how many kids leave their conference footprint.  For a league like the ACC, a kid might be leaving their region but not necessarily their conference footprint.  It might be tougher with a lot of the smaller leagues and the Big East in basketball but it'd be interesting with regards to football BCS conferences.

I know with expansion recently it would make the last few years tough, but the ACC has had their main footprint (Miami to Boston) for probably close to a decade.

sadeto

March 11th, 2014 at 1:52 PM ^

First of all congratulations on the birth of your daughter. The journey begins for you and your spouse.

Since about 57% of NCAA Division I basketball players are African American, and just under 50% of DI football players are - both hugely disproportionate to their population percentages -  you need to adjust your regional population estimates to account for this. It isn't completely accurate to do so, that is to go backward from the DI or even top 400 ethnic/racial distribution and adjust the regional population figures accordingly, but it's more accurate than simply using the total population estimate as you do. I don't have time to do it, but I would think you would create a weight variable that adjusts the relative proportions across regions to reflect the differential proportions of the African American population. It would explain/eliminate some of the regional variations - e.g. the over-representation of the Southeast, and the underrepresentation of the Pacific, but not all. 

Seth

March 11th, 2014 at 3:04 PM ^

I lost the link to a 2004 piece by ESPN 2's LZ Granderson that debunked the theory that race was the direct link to athleticism. It seemed there was just as strong of a correlation with poverty to athletic success than there was to race, which is so strongly correlated to poverty in this country (our greatest shame) that one gets mistaken for the other. And then the correlation with education standards was even stronger, suggesting that the culprit for the disparity in racial representation among high-level athletes was due to fewer educational opportunities, which correlated with poverty, which correlated with race.

There are plausible explanations for all of them, especially since the sample of high-performing athletes is so specialized that multiple factors (genetic ability, willpower, level of commitment to the sport) must be in play in order to produce one.

Since those variables are regional characteristics, I think it's important that they stay. We're measuring what produces athletes, even if it's just shitty, underfunded, segregated schools in the asshole of America.

turd ferguson

March 11th, 2014 at 3:38 PM ^

Makes sense.  Really, it's just about which question you're trying to answer.  

As you've set it up, this suggests that if you were a basketball coach who had to blindly offer recruits knowing only the state in which they live, you should probably start with Kansas and the prep school states (and stay the hell away from Alaska).  

Another interesting question is whether some states seem better than others at developing the kids who show up in their schools (as basketball players), given who shows up in those schools.  That's a much messier question that would require controlling for a lot of variables, almost all of which aren't measured exactly as we'd like. 

An analogy:  Thad Matta obviously sees a higher percentage of his players make it to the NBA than Tim Miles does.  If we wanted to know who's better at developing talent, we'd want to control for variables associated with how talented the kids are, how much they want to go to the NBA, etc.

Regardless, cool stuff and impressive that you pulled this off between diaper changes.

PM

March 11th, 2014 at 4:55 PM ^

but consider this.  While a significantly higher percentage of african americans live in poverty compared to caucasians, I can (almost) assure you that there are more caucasians living in poverty than african americans... simply because there are so many more caucasians in the U.S.  Now, if you adjust for living in an urban setting that might change... of course there's always Larry Bird to represent the rural white boy contingent.  

Personally, I would argue that basketball talent comes from growing up playing basketball (combined with some physical talent of course).  Go jump into pickup games in Indiana or Utah where middle aged white guys will take you to school if you don't have any game (I don't, btw, but have been exposed to enough ;-)  

 

 

JamieH

March 11th, 2014 at 5:33 PM ^

Wnile I agree with the poverty-athletic success angle to a certain degree, poverty does not make someone jump higher or run faster.  While there are certainly white athletes that jump high and run fast, and certainly black athletes that do neither, the raw number of black athletes who excel at these two things most certainly has to outweigh the overall black percentage of the population in the United States.  Otherwise we would see more elite white sprinters in the Olympics etc.  You can't tell me that white guys just don't WANT to run in the Olympics.  Even professional white athletes who make it in the NBA or NFL don't tend to be known for things like explosive jumping ability.  That can't just all be an accident or confirmation bias, can it?

PM

March 11th, 2014 at 10:08 PM ^

with you. Reading my post over it looks like I started out on track but went astray. To clarify, I don't think the poverty angle really works when you consider there are more caucasians (let's add hispanic and asians to make it even more lopsided) than african americans living in poverty in the U.S.  

Seth

March 11th, 2014 at 11:18 PM ^

I think the point is culture. Where kids don't see a clear way to career success through education they seek other avenues available to them. Athletic ability is just one ingredient: honing the quick twitch athleticism and skills to make that useful on a team takes years and ridiculous commitment. Parents can identify that and take on the role of obsessive coach, but poverty is its own motivation, and the positive reinforcement in cultures that highly value athletes is a very typical route.



Genetically humans are one of the most homogenous species on the planet. Our bodies are quite fungible, and our minds adaptable, especially when we're young. I won't deny that certain hereditary traits from West African populations have proliferated among athletes in America, but that is SO overstated: Swedish, Dutch, and Celtic populations have strong genetic alleles that made their way into sports too. Generalizing gets you into trouble: Europeans have more Neanderthal genes in them, and the Neanderthals stocky, tough Neanderthal body plan would make for kick-ass power backs, fullbacks and defensive linemen, so therefore...? In general, with humans, whenever you find a perceived difference, look for culture first and genetics last.

snarling wolverine

March 11th, 2014 at 6:35 PM ^

I lost the link to a 2004 piece by ESPN 2's LZ Granderson that debunked the theory that race was the direct link to athleticism

I'd be really curious to read this piece, because there is pretty clear evidence that people of West African ancestry excel at sprinting and jumping (and in sports that make extensive use of these traits), not just in the United States but pretty much everywhere they live, including the Caribbean and Europe.  Black players are massively overrepresented (compared to their share of the population) on the national soccer teams of France, England and the Netherlands, for instance, as well as on the national basketball team of Canada. 

 

sadeto

March 11th, 2014 at 6:45 PM ^

No matter whether so-called "athleticism" is directly related to ancestry or poverty, it doesn't matter at all for what I was saying about the need to adjust the population totals to account for the disproportionate representation of African American athletes among D1 basketball and football programs. You still have to do it. For example, the Southeast appears to be wildly over-represented in football, and still well-overrepresented in basketball. But if you account for the fact that the African American population is a higher percentage of the total population in the Southeast, by a large factor, and that such athletes are a higher percentage of D1 athletes, then the argument is far less compelling. I suspect the Southeast will still be over-represented somewhat, due to weather and resulting access to outdoor activities, but not nearly as much so as the table indicates. 

Now, regarding that article by ESPN's LZ Granderson, a journalist, I would like to know how his undergraduate degree at WMU and his career as a sportswriter qualify him in any way to make such an argument, or to "debunk" the link between ethnicity and athleticism (a loaded word). That's preposterous. Tell me that a credentialed scientist did so in a peer-reviewed journal and I might pay attention. 

Colin M

March 11th, 2014 at 7:51 PM ^

I haven't read the article by Granderson, but I think Seth's point has to do with Omitted Variable Bias. If you add a race variable and attempt to measure the effect of race without also controlling for variables that are correlated with both race and the output variable, your estimates will most likely be biased. If you don't include a measure of poverty, then you won't be able to parse out what portion of the effect is due to race and what portion is due to poverty. Since both are also correlated with region, I think Seth made a prudent decision to keep his model simple and basically say, "it's hard to tell what's happening." 

Excelsior

March 11th, 2014 at 2:02 PM ^

So how does someone like McGary or ALbrecht or Stauskas fit into this though? They did not go to high school in the place where they are "from" so is McGary considered to be "from" New Hampshire or Indiana?

 

I see you mentioned it above for stauskas but is that true for everyone? And how do you know whether someone transferred to a basketball factory instead of going to their natural high school?

turd ferguson

March 11th, 2014 at 2:15 PM ^

I came on to say three things.

1.  Congrats!

2.  I really like this analysis (and I'm kind of a picky asshole with these things).

3.  It'd be interesting to see a separate analysis (column, really) with Black populations for the reasons that sadeto mentions above.  That'd just be # of basketball players coming out of the state divided by # of African-Americans in the state.  That would help us tease apart whether this is really about demographics or really about some states being basketball states.

Really, though, just congratulations.

turd ferguson

March 11th, 2014 at 4:21 PM ^

 

It's my fault for picking a name referencing something that fewer people know than I realized.  An SNL skit from about a decade ago (Celebrity Jeopardy).  It's actually just a coincidence that there's a mustache on my avatar and on the Burt Reynolds character.

notYOURmom

March 11th, 2014 at 2:48 PM ^

Did not one but two posters just suggest that the figures are distorted by the presence of actual black players who should be removed so we could calculate whether a state is a "real" basketball state?

Words escape me.

 

 

turd ferguson

March 11th, 2014 at 3:17 PM ^

What?

I'm one of those two posters, and can you point out exactly where I suggested that he remove the black players from his analysis (or where the other poster, sadeto, suggested that)?  My suggestion was to change leave the numerator the same (# of players, regardless of race) and run a separate set of analyses with the states' black populations only in the denominators.  This is a really simple (and less than ideal) way of controlling for an obviously important variable.

I'm guessing that this conversation won't go anywhere with you, but at least others will see how off-base you are.

notYOURmom

March 11th, 2014 at 5:09 PM ^

You said:

 

That would help us tease apart whether this is really about demographics or really about some states being basketball state

So my difficulty is the implication that some states do well because of "demographics" (.e.g race) and others because they are "basketball states."

That is, "controlling for" the presence of black players in order to determine the remaining effect (that is, removing the effects of demographics from the rest of the analysis).

Also I think perhaps you have the numerator and denominator mixed up.  

You are proposing:

# of players regardless of race

divided by

"Total Black population"

Could that really be what you meant?

turd ferguson

March 11th, 2014 at 5:29 PM ^

No, I didn't confuse the numerator and denominator.  The numerator is the one on top.

I explained my point a little further in a response to Seth above.  Depending on which question you're trying to answer, you might need to control for confounding variables.  This is common practice in many, many disciplines and doesn't reflect some kind of racist tendency, despite what you suggest in your uncharitable interpretation above.

I'll reiterate a point I made earlier.  Let's say that you want to know whether Thad Matta or Tim Miles is a better developer of talent.  You could look to the % of their players who make it to the NBA and say, "Well, all kinds of guys can become NBA players, so this indicates that Matta is a better developer of talent."  I think that's misleading.  I would want to control for some variables related to who they were when they arrived (probably starting with recruiting rankings).  In this case, you could just look at the % of a state that turns into top basketball prospects, which is essentially what this analysis does, and that's honestly interesting.  If you wanted to know which state develops talent best (e.g., through its basketball programs), you'd probably want to control for variables related to who lives in that state and how likely they are - before basketball coaches get their hands on them - to become top prospects.  For example, kids in Minnesota probably aren't thinking about basketball nearly as much as they're thinking about hockey.  If this were about testing how well basketball players develop, then it would be misleading to just look at what % of the state becomes top basketball prospects.  Race is a pretty lousy proxy for the kinds of things that make people more likely to become star basketball players, but it's a variable that's easy to get and it would be a start (if this were the goal).

Long story short:  never mind.

JamieH

March 11th, 2014 at 6:22 PM ^

It's a data point.  Ignoring it because you are afraid of talking about race is dumb. 

 

Look at college basketball.  Do you see a lot of Asian or Hispanic basketball players?  I don't.  So it would be entirely logical to assume that areas of the country that have heavy Asian and Hispanic populations will probably be underrepresented (vs. their overall population size) in terms of the number of college basketball players they produce.

 

Nowhere in that statement am I making any judgments on the relative abilities of HIspanic or Asian basketball players.  It is merely an observation that college basketball does not currently, not has it historically, has a large number of Hispanic or Asian participants, for whatever reason.  Therefore it is not logical to assume it suddenly will in the immediate future.

 

The whole point of this endeavor is to find areas of the country that produce more/less basketball players than "expected" right?  Well in order to really do that, you have to properly define the pool of possible players to begin with, and just using raw population numbers doesn't really cut it.  Just culling by race doesn't cut it either, but it is an easily available statistic that could be used to make the numbers somewhat more accurate, by using the approximate racial makeup of college basketball.  Obviously that isn't a constant value and will change over time, but it could still be used to make this a more interesting evaluation, IMO.