Probably boring - how I would apply analytics to coach Purdue Basketball

Apr 5, 2022
136
231
43
I was bored today and found an interesting trend - don't read unless you're a stats nerd, as I'll probably waste your time. The last time I got bored and looked into Purdue basketball statistics, I tried to prove that Painter hasn't underperformed in the tournament - but, for the sake of argument I'll say this is my explanation as to why Purdue underperforms: play style decisions.

Purdue had Zach Edey and Jaden Ivey last season. Needless to say, it got fouled a lot (15th highest free throw rate in the country in 2022), and this is a typical trend with its year-in-year out low post offense. On the other hand, Painter always plays a low pressure defensive style and doesn't try to force turnovers or foul much (10th lowest rate in country in 2022), as he needs to slow down the game to allow his bigs to catch a breath, which explains their slow tempo. Basically, Purdue played the far extreme on both ends of the court and was the biggest outlier in the country.

What is interesting is that it appears that teams are disproportionately rewarded with defensive success for fouling (see chart linked); there is a direct correlation where fouling more led to better defense in 2022. This correlation is strong enough that if Purdue were to have, instead, inverted its foul rates to foul a lot and rarely get fouled, the data shows an expected +5.02 points per game in scoring margin (data analyzed the top 150 teams in 2022).


How did this affect Purdue in the tournament? Purdue played 3 teams that foul a ton (Yale = 233rd most, Texas = 336th, St. Peter's = 347th). Purdue ran into teams that played high pressure defense that knew they were relatively undersized but had a quickness advantage. As a result, they had game plans that wanted to force turnovers and force a lot of whistles. Purdue was +22 on free throws against Yale (even TOVs) and +34 (-5 TOVs) against Texas during the first weekend. This got a lot of media attention. Crazily enough, this attention coupled with the fact that Courtney Green refereed the next game (Purdue is 1-15 against the spread in games he referees since 2020) led to Purdue being -6 against St. Peter's (-7 TOVs).

I think it's fair to say that Edey and Haas draw less fouls than they might deserve - there is a rule in college basketball more or less dedicated to Haas for this reason. However, my biggest takeaway is that Purdue plays a very unusual style where it gets fouled more than almost anyone and fouls less than almost anyone, and that this strategy may cost it. If it decided to press and get in ball handlers' faces, I believe that Purdue would get called for more fouls, which would, in turn, allow them to draw more fouls, as well, but with the benefit of forcing more turnovers.

My conclusion is that this odd play style of having a huge big man that is fed on every play is very unusual and forces referees to call a game that is outside of their comfort zone, especially in the tournament where you have two random teams from different conferences with different play styles. As a result, Painter introduces his team to more game-to-game variance in outcomes than almost any other team. Courtney Green seems to favor calling fouls on guards over bigs. This is fine - I'm not going to say he is right or wrong in doing so, but I would say that this style of his doesn't lend itself to helping Purdue win games. On the other hand, I'm sure there are officials who greatly benefit Purdue, and this plays into my theory of greater game-to-game variance with such a polarizing team strategy. Purdue may have benefitted big time against Texas, but it may have caught the opposite luck against St. Peter's. You could probably play that game 100 more times and never see the team that is in the 97th percentile of FT shooting rate end up -6 against the team that fouls more than 97% of teams in the country - statistically, Purdue was expected to be +14, and 20 is a huge net difference - that's going to the line 10 extra times!

If I were the coach of Purdue Basketball, I would consider this data a detractor to the idea that the 7 foot big man system is ideal for tournament success. I find it hard to believe that this strategy actually costs Purdue 5 points per game, but I think 2 points per game (averaged) is reasonable. 2 extra points would have saved Purdue from losing on 4 buzzer beaters last year. This study was concluded upon a small data set, but in 2019, Purdue was 264th in drawing fouls and 144th in committing them. Coincidence? Who knows. I didn't feel like spending more than a half hour on collecting data for other years to look into it, so maybe my theory just so happened to align with exactly what seemed to happen in 2022 lol.
 

FirstDownB

All-American
Oct 12, 2015
9,754
13,857
113
tl;dr
Painter builds teams to win a lot of games in the regular season, chase the Big Ten title, and keep his job
 
Feb 24, 2020
7
5
3
Appreciate your analysis. Also plays in that a certain percent of fouls never get called so heavy fouling teams really junk up games and give them an advantage over smooth rhythm centric teams.
 

Boiler Buck

All-American
Mar 11, 2010
11,790
10,223
113
Purdue had Zach Edey and Jaden Ivey last season. Needless to say, it got fouled a lot (15th highest free throw rate in the country in 2022), and this is a typical trend with its year-in-year out low post offense. On the other hand, Painter always plays a low pressure defensive style and doesn't try to force turnovers or foul much (10th lowest rate in country in 2022), as he needs to slow down the game to allow his bigs to catch a breath, which explains their slow tempo. Basically, Purdue played the far extreme on both ends of the court and was the biggest outlier in the country.

Not sure I would consider Purdue "slow tempo". But willing to listen to good data....Do you have data to support that in terms of possessions per game as compared to rest of B10?
 

Hausta01

All-American
Oct 2, 2017
6,192
7,691
113
Not sure I would consider Purdue "slow tempo". But willing to listen to good data....Do you have data to support that in terms of possessions per game as compared to rest of B10?

Purdue was 237th last year in tempo, per KenPom and 278th the year prior. We generally do play a slower tempo in terms of total possessions per game.

Some of that I am sure is impacted by the B1G overall who has a number of teams who love to slow the game down.
 
  • Like
Reactions: Boiler Buck

northside100

All-American
Gold Member
Aug 11, 2001
5,017
4,128
113
I was bored today and found an interesting trend - don't read unless you're a stats nerd, as I'll probably waste your time. The last time I got bored and looked into Purdue basketball statistics, I tried to prove that Painter hasn't underperformed in the tournament - but, for the sake of argument I'll say this is my explanation as to why Purdue underperforms: play style decisions.

Purdue had Zach Edey and Jaden Ivey last season. Needless to say, it got fouled a lot (15th highest free throw rate in the country in 2022), and this is a typical trend with its year-in-year out low post offense. On the other hand, Painter always plays a low pressure defensive style and doesn't try to force turnovers or foul much (10th lowest rate in country in 2022), as he needs to slow down the game to allow his bigs to catch a breath, which explains their slow tempo. Basically, Purdue played the far extreme on both ends of the court and was the biggest outlier in the country.

What is interesting is that it appears that teams are disproportionately rewarded with defensive success for fouling (see chart linked); there is a direct correlation where fouling more led to better defense in 2022. This correlation is strong enough that if Purdue were to have, instead, inverted its foul rates to foul a lot and rarely get fouled, the data shows an expected +5.02 points per game in scoring margin (data analyzed the top 150 teams in 2022).



How did this affect Purdue in the tournament? Purdue played 3 teams that foul a ton (Yale = 233rd most, Texas = 336th, St. Peter's = 347th). Purdue ran into teams that played high pressure defense that knew they were relatively undersized but had a quickness advantage. As a result, they had game plans that wanted to force turnovers and force a lot of whistles. Purdue was +22 on free throws against Yale (even TOVs) and +34 (-5 TOVs) against Texas during the first weekend. This got a lot of media attention. Crazily enough, this attention coupled with the fact that Courtney Green refereed the next game (Purdue is 1-15 against the spread in games he referees since 2020) led to Purdue being -6 against St. Peter's (-7 TOVs).

I think it's fair to say that Edey and Haas draw less fouls than they might deserve - there is a rule in college basketball more or less dedicated to Haas for this reason. However, my biggest takeaway is that Purdue plays a very unusual style where it gets fouled more than almost anyone and fouls less than almost anyone, and that this strategy may cost it. If it decided to press and get in ball handlers' faces, I believe that Purdue would get called for more fouls, which would, in turn, allow them to draw more fouls, as well, but with the benefit of forcing more turnovers.

My conclusion is that this odd play style of having a huge big man that is fed on every play is very unusual and forces referees to call a game that is outside of their comfort zone, especially in the tournament where you have two random teams from different conferences with different play styles. As a result, Painter introduces his team to more game-to-game variance in outcomes than almost any other team. Courtney Green seems to favor calling fouls on guards over bigs. This is fine - I'm not going to say he is right or wrong in doing so, but I would say that this style of his doesn't lend itself to helping Purdue win games. On the other hand, I'm sure there are officials who greatly benefit Purdue, and this plays into my theory of greater game-to-game variance with such a polarizing team strategy. Purdue may have benefitted big time against Texas, but it may have caught the opposite luck against St. Peter's. You could probably play that game 100 more times and never see the team that is in the 97th percentile of FT shooting rate end up -6 against the team that fouls more than 97% of teams in the country - statistically, Purdue was expected to be +14, and 20 is a huge net difference - that's going to the line 10 extra times!

If I were the coach of Purdue Basketball, I would consider this data a detractor to the idea that the 7 foot big man system is ideal for tournament success. I find it hard to believe that this strategy actually costs Purdue 5 points per game, but I think 2 points per game (averaged) is reasonable. 2 extra points would have saved Purdue from losing on 4 buzzer beaters last year. This study was concluded upon a small data set, but in 2019, Purdue was 264th in drawing fouls and 144th in committing them. Coincidence? Who knows. I didn't feel like spending more than a half hour on collecting data for other years to look into it, so maybe my theory just so happened to align with exactly what seemed to happen in 2022 lol.
This is an excellent post and deserves more attention than it's received. Love the use of data versus the typical mindless pontification we tend to see.

Your observation that Purdue, because of it's style of play, is subject to greater variation in how their games are called is very astute.

I'd call out that whether Purdue actually wants to move away from their current style of play depends on their next best alternative, i.e. they've been able to attract elite post big men, can they attract elite players at other positions or other styles of play?
 

depthcharge623

Redshirt Freshman
Dec 31, 2013
1,317
1,563
113
This is an excellent post and deserves more attention than it's received. Love the use of data versus the typical mindless pontification we tend to see.

Your observation that Purdue, because of it's style of play, is subject to greater variation in how their games are called is very astute.

I'd call out that whether Purdue actually wants to move away from their current style of play depends on their next best alternative, i.e. they've been able to attract elite post big men, can they attract elite players at other positions or other styles of play?
That's the right question. As I pointed out at the end of last year, you can't just assume that we would play equally as well with another style so it isn't apples to apples. I think that the trend of Carsen and Ivey is helping us move in the right direction with recruiting, and there may be a tipping point where we can start to get the right recruits to play other styles. But we aren't there yet. So in the near term it seems like our options are to play at an extremely high level through a traditional big, or play at an average level with another style.