NCAAWB team free throw percentages -- rising tide lifting all boats?

[connie]

My question of late: is UCONNWB the leading edge of increased excellence across NCAAWB?

One very small step towards a statistical analysis of that question is to determine statistically how much better UConn is compared to other teams, and whether the degree of difference is shrinking. By way of exploring, I decided to begin by checking whether the standard deviation between/among the best and worst team free-throw percentages. (I may catch hell for saying this way . . . but crudely put, the standard deviation measures variation within a group.) I choose free throws because they are objectively easy to measure. Same distance to the hoop for everyone. And there has been no revolution in technique as might otherwise skew the data. My hypothesis is that as the competition in NCAAWB sharpens--as the sport develops, players get better, training and coaching improves, the pool of candidates expands, etc.--all players will on average become better free throw shooters. Moreover, presumably the gulf between the worst and the best will narrow. If so, then the big question is: where does UConnWB fit in the big picture? And can we then demonstrate that UConn is perhaps good for women's basketball? (Answer: no duh!)

So . . . I went to the NCAA Women's Basketball Archived Statistics and looked up data for team free throw percentages for all Div. I teams from the 2001-02 season to 2015-16 season. (The archive only goes back to 2001-02.) I realize that this is a very small sample. Here's what I found. Free throw shooting may be getting slightly better overall. Okay. But more interesting, the difference between the worst and the best appears to be shrinking. Below is a primitive graph of the results, with the X-axis representing the year (1 = 2001-02), and the Y-axis showing the standard deviation among all 335-350 teams (depending on the year and the data).

Hard to say. But it certainly could be argued that at least as far back as 2003-04, the standard deviation for team free throw percentage is dropping. The earlier years (2001-02, 02-03) could be anomalies. Can't be sure since I don't have the data for the previous years. But what is arguably the case is that more teams are becoming better free throw shooters. This suggests improvement in play in at least on area. The inference is that play in other areas is also improving.

More important, because the standard deviation is dropping this information arguably supports the thesis: NCAAWB shows increased parity along with increased performance.

 

[connie]

Forgot to mention: the standard deviation has dropped most precipitously during the period of UConn's dominance (2012-13 to 2015-16). Bad for basketball, right?

 

[tomcat]

First off, this is fascinating. Second, for someone to explain statistics, which I understood but with great difficulty back more than 40 years ago, simply and easily, suggests to me that our OP (welcome to the BY, by the way) has some very useful skills. Keep going. You've analyzed free throws. What's next?

 

[CTtransplant]

My question of late: is UCONNWB the leading edge of increased excellence across NCAAWB?

One very small step towards a statistical analysis of that question is to determine statistically how much better UConn is compared to other teams, and whether the degree of difference is shrinking. By way of exploring, I decided to begin by checking whether the standard deviation between/among the best and worst team free-throw percentages. (I may catch hell for saying this way . . . but crudely put, the standard deviation measures variation within a group.) I choose free throws because they are objectively easy to measure. Same distance to the hoop for everyone. And there has been no revolution in technique as might otherwise skew the data. My hypothesis is that as the competition in NCAAWB sharpens--as the sport develops, players get better, training and coaching improves, the pool of candidates expands, etc.--all players will on average become better free throw shooters. Moreover, presumably the gulf between the worst and the best will narrow. If so, then the big question is: where does UConnWB fit in the big picture? And can we then demonstrate that UConn is perhaps good for women's basketball? (Answer: no duh!)

So . . . I went to the NCAA Women's Basketball Archived Statistics and looked up data for team free throw percentages for all Div. I teams from the 2001-02 season to 2015-16 season. (The archive only goes back to 2001-02.) I realize that this is a very small sample. Here's what I found. Free throw shooting may be getting slightly better overall. Okay. But more interesting, the difference between the worst and the best appears to be shrinking. Below is a primitive graph of the results, with the X-axis representing the year (1 = 2001-02), and the Y-axis showing the standard deviation among all 335-350 teams (depending on the year and the data).

View attachment 19901

Hard to say. But it certainly could be argued that at least as far back as 2003-04, the standard deviation for team free throw percentage is dropping. The earlier years (2001-02, 02-03) could be anomalies. Can't be sure since I don't have the data for the previous years. But what is arguably the case is that more teams are becoming better free throw shooters. This suggests improvement in play in at least on area. The inference is that play in other areas is also improving.

More important, because the standard deviation is dropping this information arguably supports the thesis: NCAAWB shows increased parity along with increased performance.

The small problem with the data, as you present it, is that you are only proving that the difference between the best and the worst is narrowing. You appear to assume that the best have remained the same (or gotten better). It is possible that the narrowing has occurred because the best FT shooting teams are not as good as they used to be.

Did you check to see if the FT shooting percentage of the top 25 or 50 teams is the same over the period? It would change your conclusion if it were not so.

That said, I really like the way you think, and I appreciate the effort you made to be analytically correct!

 

[connie]

You've analyzed free throws. What's next?

How about turnovers? My thinking . . . turnovers are a reflection of relative abilities. The defense tries to take the ball away; the offense tries to prevent that from happening. As NCAAWB has matured is there a decline in the difference between teams with the least turnovers and the most turnovers, teams with the least steal and the most steals?

Let's find out!

 

[connie]

The small problem with the data, as you present it, is that you are only proving that the difference between the best and the worst is narrowing. You appear to assume that the best have remained the same (or gotten better). It is possible that the narrowing has occurred because the best FT shooting teams are not as good as they used to be.

Did you check to see if the FT shooting percentage of the top 25 or 50 teams is the same over the period? It would change your conclusion if it were not so.

That said, I really like the way you think, and I appreciate the effort you made to be analytically correct!

You are correct! It could be that the narrowing difference is because all the teams are shooting more poorly, with the best teams shooting more poorly than the worst teams. In that event, parity increases while overall performance gets worse! I'll see what we can learn.

Good eye!

 

[connie]

I went back and gathered the team free-throw percentages for all Division I teams for each year, 2001-02 to 2015-16. I determined the average for all teams. Unlike the previous calculation, I left out the few teams every year or so that were reclassified mid-season. (I also caught a few data collection errors.) In light of the roughly 340+ data points (teams) each year, I do not regard the excluded data as significant. But I leave it to others to check me on that.

Here is the graph, showing that overall team free throw shooting generally improved over a 14-year time-frame:

NCAAWB Team free throw percentage 2001-02 to 2015-16
Average for all teams


I also determined the standard deviation for each year. If the average goes up each year and the deviation drops each year, then that tells you that not only is free throw shooting improving overall, but also that all teams are generally improving: i.e., increasing performance and increasing parity.

And guess what? That is exactly what we find. Here is the graph, showing declining standard deviation (the difference between best and worst is shrinking).

NCAAWB Team free throw 2001-02 to 2015-16 -- Standard Deviation

Conclusion: free throw performance is improving, and there is a decline in the difference between the best and the worst.

I'd love to prove that the rising tide of UConn excellence is lifting all boats. But that's for another day.

 

[j66kicker]

I'd love to prove that the rising tide of UConn excellence is lifting all boats. But that's for another day.

Looking forward to it! That will take some serious cause/effect math. But could help shut up the ,"UConn is bad for the game" crowd.

 

[CTyankee]

Free throws??? Statistics??? Tell that to Geno after his PG of the future missed a bunch of big ones at the end of the Tulane game tonight... He'll have her shooting free throws in practice until her arm falls off for the next three days...

 

[CTtransplant]

Nicely done Connie. Big Data come to the Boneyard! Pretty clear that shooting proficiency in women's basketball is improving, and that parity has increased over the past 10 years.

I look forward to your next foray into the data.

 

[NVUConnFan]

You are a statistical maven. I grok means and standard deviations. But to analyze the statistical data takes a special ability. Keep going.