College Threes are Shorter than NBA Threes
Oct 24, 2017 13:01:49 GMT -6
skrouse and BKay Jewelers like this
Post by eric on Oct 24, 2017 13:01:49 GMT -6
So what?
So when we look at players' college stats to try and project their NBA stats, we're dealing with multiple layers of uncertainty.
1. College stats are a small sample: most players will have an error bar of at least ±.04, and even Steph Curry who played four years and took looooots of threes (414 of 1004) ends up with ±.03. Since he shot 41% that didn't really matter because 38% is still really good, but for a guy like Gary Neal who was 36% ± 3% the low end of that range starts to look pretty worrisome.
2. College teams have poor fundamentals, coaching, teammates, refereeing, scouting, I could go on... the upshot of this is that we plausibly have effects on both sides of the ball when we put a player in a real NBA offense going against real NBA defenses in real NBA games.
3. College basketball has a cute little three point line at 20 feet 9 inches, which is shorter than the NBA line all the way across. Can we measure this, at least?
You betcha!
.
I took all the players who attempted at least 100 threes in their college career and rookie year since the 2007 season, when the one and done rule was instituted and two years before the NCAA moved their line back from the even more adorable 19 feet 9 inches. This ended up being a cool 111 players. Take their college 3P%, subtract rookie 3P%, bod the bing bod the boom and you get:
On average players go -.026 from three. 32% stayed the same or went up, the remaining 68% dropped by at least 1 percentage point.
So when we look at players' college stats to try and project their NBA stats, we're dealing with multiple layers of uncertainty.
1. College stats are a small sample: most players will have an error bar of at least ±.04, and even Steph Curry who played four years and took looooots of threes (414 of 1004) ends up with ±.03. Since he shot 41% that didn't really matter because 38% is still really good, but for a guy like Gary Neal who was 36% ± 3% the low end of that range starts to look pretty worrisome.
2. College teams have poor fundamentals, coaching, teammates, refereeing, scouting, I could go on... the upshot of this is that we plausibly have effects on both sides of the ball when we put a player in a real NBA offense going against real NBA defenses in real NBA games.
3. College basketball has a cute little three point line at 20 feet 9 inches, which is shorter than the NBA line all the way across. Can we measure this, at least?
You betcha!
.
I took all the players who attempted at least 100 threes in their college career and rookie year since the 2007 season, when the one and done rule was instituted and two years before the NCAA moved their line back from the even more adorable 19 feet 9 inches. This ended up being a cool 111 players. Take their college 3P%, subtract rookie 3P%, bod the bing bod the boom and you get:
On average players go -.026 from three. 32% stayed the same or went up, the remaining 68% dropped by at least 1 percentage point.
![[Account Deleted] Avatar](http://storage.proboards.com/5589790/avatar/ZhaFVHTxVPwmyuOiOKEd.png)