Plus/Minus for the Purdue snoozer

[PurpleWhiteBoy]

Player	Minutes	NU pts	PU pts	Team +/-	Player Adj	Net +/-
Pete Nance	25	35	50	-15	+3.4	-11.6
Ryan Young	18	28	24	+4	+2.8	+6.8
Boo Buie	31	42	51	-9	+0.0	-9.0
Chase Audige	17	23	29	-6	-2.5	-8.5
Robbie Beran	26	39	45	-6	+2.9	-3.1
Ryan Greer	22	36	34	+2	+1.3	+3.3
Ty Berry	17	34	30	+4	-6.7	-2.7
Julian Roper	27	35	46	-11	+3.5	-7.5
Elyjah Williams	6	11	13	-2	-2.5	-4.5
Casey Simmons	6	7	13	-6	-2.1	-8.1

Pete Nance was unable to guard the post and we allowed 2 pts per minute when he was in the game. He was an adjusted -11.6 because he had an "ok" game offensively.
Ryan Young and Ryan Greer were the two bright spots, both playing winning games.
Ty Berry shot the ball poorly and held his teammates back.
Chase Audige and Boo Buie both had subpar outings.
Casey Simmons indicated again that he should not be playing right now.
Robbie Beran and Julian Roper both did some positive things but were on the floor a lot with Nance, who couldn't guard Edey (as everyone knew going into the game).

The starters (Nance, Beran, Roper, Audige, Buie) got beat up 14-5 in 6:24 of game time. They scored 5 points in 6:24.

Buie and Greer (regularly mentioned as having success) went 20-18 in about 14 minutes.
Audige and Berry (another good pairing) went 13-6 in a limited 5:36.
When either of those pairings was on the court, we won 33-24. Otherwise, we lost 46-31.
I'd suggest we try to use those pairings as much as possible.
Young, Williams, Buie and Greer (regularly successful) did not play together at all.
Young, Williams, Audige, Greer and Berry entered the game after the pathetic opening 4:30, with NU trailing 4-2. That group went on an 8-4 run over the next 2:30. Collins stopped that momentum by subbing in Simmons and Nance for Audige and Young. Purdue immediately went on a 14-5 run over the next 5:40. Young returned to the game and NU went on a 7-4 run.

The frustration and underperformance continue.

 

[Purple Pile Driver]

Player	Minutes	NU pts	PU pts	Team +/-	Player Adj	Net +/-
Pete Nance	25	35	50	-15	+3.4	-11.6
Ryan Young	18	28	24	+4	+2.8	+6.8
Boo Buie	31	42	51	-9	+0.0	-9.0
Chase Audige	17	23	29	-6	-2.5	-8.5
Robbie Beran	26	39	45	-6	+2.9	-3.1
Ryan Greer	22	36	34	+2	+1.3	+3.3
Ty Berry	17	34	30	+4	-6.7	-2.7
Julian Roper	27	35	46	-11	+3.5	-7.5
Elyjah Williams	6	11	13	-2	-2.5	-4.5
Casey Simmons	6	7	13	-6	-2.1	-8.1

Pete Nance was unable to guard the post and we allowed 2 pts per minute when he was in the game. He was an adjusted -11.6 because he had an "ok" game offensively.
Ryan Young and Ryan Greer were the two bright spots, both playing winning games.
Ty Berry shot the ball poorly and held his teammates back.
Chase Audige and Boo Buie both had subpar outings.
Casey Simmons indicated again that he should not be playing right now.
Robbie Beran and Julian Roper both did some positive things but were on the floor a lot with Nance, who couldn't guard Edey (as everyone knew going into the game).

The starters (Nance, Beran, Roper, Audige, Buie) got beat up 14-5 in 6:24 of game time. They scored 5 points in 6:24.

Buie and Greer (regularly mentioned as having success) went 20-18 in about 14 minutes.
Audige and Berry (another good pairing) went 13-6 in a limited 5:36.
When either of those pairings was on the court, we won 33-24. Otherwise, we lost 46-31.
I'd suggest we try to use those pairings as much as possible.
Young, Williams, Buie and Greer (regularly successful) did not play together at all.
Young, Williams, Audige, Greer and Berry entered the game after the pathetic opening 4:30, with NU trailing 4-2. That group went on an 8-4 run over the next 2:30. Collins stopped that momentum by subbing in Simmons and Nance for Audige and Young. Purdue immediately went on a 14-5 run over the next 5:40. Young returned to the game and NU went on a 7-4 run.

The frustration and underperformance continue.

I wish you could assume the reigns of HC for the remainder of the season.

 

[Hungry Jack]

Player	Minutes	NU pts	PU pts	Team +/-	Player Adj	Net +/-
Pete Nance	25	35	50	-15	+3.4	-11.6
Ryan Young	18	28	24	+4	+2.8	+6.8
Boo Buie	31	42	51	-9	+0.0	-9.0
Chase Audige	17	23	29	-6	-2.5	-8.5
Robbie Beran	26	39	45	-6	+2.9	-3.1
Ryan Greer	22	36	34	+2	+1.3	+3.3
Ty Berry	17	34	30	+4	-6.7	-2.7
Julian Roper	27	35	46	-11	+3.5	-7.5
Elyjah Williams	6	11	13	-2	-2.5	-4.5
Casey Simmons	6	7	13	-6	-2.1	-8.1

Pete Nance was unable to guard the post and we allowed 2 pts per minute when he was in the game. He was an adjusted -11.6 because he had an "ok" game offensively.
Ryan Young and Ryan Greer were the two bright spots, both playing winning games.
Ty Berry shot the ball poorly and held his teammates back.
Chase Audige and Boo Buie both had subpar outings.
Casey Simmons indicated again that he should not be playing right now.
Robbie Beran and Julian Roper both did some positive things but were on the floor a lot with Nance, who couldn't guard Edey (as everyone knew going into the game).

The starters (Nance, Beran, Roper, Audige, Buie) got beat up 14-5 in 6:24 of game time. They scored 5 points in 6:24.

Buie and Greer (regularly mentioned as having success) went 20-18 in about 14 minutes.
Audige and Berry (another good pairing) went 13-6 in a limited 5:36.
When either of those pairings was on the court, we won 33-24. Otherwise, we lost 46-31.
I'd suggest we try to use those pairings as much as possible.
Young, Williams, Buie and Greer (regularly successful) did not play together at all.
Young, Williams, Audige, Greer and Berry entered the game after the pathetic opening 4:30, with NU trailing 4-2. That group went on an 8-4 run over the next 2:30. Collins stopped that momentum by subbing in Simmons and Nance for Audige and Young. Purdue immediately went on a 14-5 run over the next 5:40. Young returned to the game and NU went on a 7-4 run.

The frustration and underperformance continue.

Thank you for these analyses, PWB. They really are interesting.

Long live the Nicholbomb!

 

[SDakaGordie]

Thank you for these analyses, PWB. They really are interesting.

Long live the Nicholbomb!

In the end, all of this borders on the absurd. You could go through every other game in the NCAA and play a similar blame game. It does not mean it’s useful.

 

[its_the_sauce]

The biggest minus was Blue Stone being out of Malört..

 

[phatcat_rivals223240]

I wish you could assume the reigns of HC for the remainder of the season.

Heck, he could do it from his house.

 

[Titanium999]

I wish you could assume the reigns of HC for the remainder of the season.

Me too! I said the same thing before the game!

 

[Titanium999]

Heck, he could do it from his house.

We have too many ball handling turnovers

 

[SmellyCat]

How is "Player Adj" calculated?

 

[NUCat320]

@PurpleWhiteBoy , thanks for doing this.
First game I’ve seen live in a while, and Snoozer is an accurate description.

A few questions:
- do you track # of lineups used in a game?
- do you track # of substitutions in a game?
- have those trended in any way this season?
- do you track opponents on +/- or substitutions or anything?
- if so, how do NU’s numbers compare to their opponents?

My guess is that NU substitutes more and uses more lineups. I’m interested in the gap.

 

[willycat]

We have too many ball handling turnovers

Right and missed shots in the paint.

 

[Hungry Jack]

How is "Player Adj" calculated?

(Player SOS / [Turk Coefficient])^.5

 

[gipnov23]

The starters (Nance, Beran, Roper, Audige, Buie) got beat up 14-5 in 6:24 of game time. They scored 5 points in 6:24.

This is the biggest indictment on Collins. He wastes minutes with a terrible starting lineup and it puts us behind the eight ball. Take these minutes out and it's an even game. It's not his fault we can't hit three pointers but it is his fault for leaving inefficient shooters in the game way too long

 

[SmellyCat]

(Player SOS / [Turk Coefficient])^.5

Oh man, how do we calculate the Turk Coefficient? Is it adjusted depending on how flat the earth is?

 

[GatoLouco]

This is the biggest indictment on Collins. He wastes minutes with a terrible starting lineup and it puts us behind the eight ball. Take these minutes out and it's an even game. It's not his fault we can't hit three pointers but it is his fault for leaving inefficient shooters in the game way too long

Why change anything when what we have been doing has been working so well? IN, IL and now PU, we came out of the gates with some superb offense.

 

[PurpleWhiteBoy]

How is "Player Adj" calculated?

Smelly:

Obviously, Raw +/- is just the score of the games when each guy is on the floor (ignoring some garbage time and some Free throw contests).

The Adjustment is something I did in response to the reasonable criticism that if one guy is playing great and his 4 teammates are playing terribly they'd all have the same +/-. My original thought was that over the course of the season the effect of "my teammates are playing well or poorly" would all even out. But I added the Adjustment to augment the analysis on a game by game basis.

We start with "performance points." I award 1 or 2 performance points for a made shot, deduct 1 performance point for a missed shot. Add a point for a steal, subtract a point for a turnover. Add a point for an offensive rebound and 0.5 points for a defensive rebound. Add a half point for an assist. Add 0.75 points for a block. Add 0.4 points for a made free throw. Subtract 0.6 for a missed free throw. If you miss the front end of a one-and-one you get a 1.2 point deduction.

So I do that for every lineup during the course of the game. For example, with 16:05 left in the game yesterday, Collins put Young in for Beran and we had Nance, Young, Berry, Buie and Roper on the court.
They played about 1:31. During that segment, Roper had 2 steals and a block (+2.75). Berry missed two shots (-2). Nance missed a shot (-1). Young got an offensive rebound (+1). Buie didn't do anything specific (0). However, we didn't score and Purdue scored 2 points. All 5 of those guys got a -2 for their +/-. The performance adjustments total up to +0.75, meaning we should have won the segment by 0.75 points, on average.

To get the adjustment for the impact of each player, you have to make sure that the total of the adjustments nets out to zero for every segment. This is basically allocating credit or blame for the success or failure of the lineup. So if Young has a +1 performance, that means his 4 teammates have to take a -0.25 for their own adjustment. In the example above, this results in...

Roper +3.25, Young +1.0625, Buie -0.1875, Nance -1.4375, Berry -2.6875.

You add up all the segments for each guy and get his total adjustment.

I think its a good method to evaluate individual performance in the context of the lineups we deployed.
Raw +/- is really about the success of the 5 man lineup and captures a ton of intangibles that the box score cannot.

If thats confusing, think of it this way... If Young scores a basket and we stop Purdue from scoring, then Young grabs the rebound, we win the segment 2-0. Young and his linemates are all raw +2. But Young did most of the work. So he has 1.5 performance points. The total adjustment has to be zero, so the 4 linemates are all adjusted down -1.5/4 = -0.375. Each player still has a raw +/- of +2, but Young has a Net +/- of +3.5, everybody else in the lineup has a Net +/- of +1.625.

 

[SmellyCat]

Thanks for the explanation. I get why you do it. I wonder if there is anything behind the numbers you're assigning or if it's just a gut thing on your part? Like blocks are worth 0.75 and steals worth 1. I understand why one is worth less than the other, since a block doesn't ALWAYS lead to a turnover, but where did you arrive at 0.75?

 

[PurpleWhiteBoy]

Thanks for the explanation. I get why you do it. I wonder if there is anything behind the numbers you're assigning or if it's just a gut thing on your part? Like blocks are worth 0.75 and steals worth 1. I understand why one is worth less than the other, since a block doesn't ALWAYS lead to a turnover, but where did you arrive at 0.75?

A block doesn't guarantee we get the ball.
Its really based on possessions - the point assignments I mean.
A block negates a shot, so it is better than a defensive rebound.
But it isn't as good as a steal.

 

[Purple Pile Driver]

Smelly:

Obviously, Raw +/- is just the score of the games when each guy is on the floor (ignoring some garbage time and some Free throw contests).

The Adjustment is something I did in response to the reasonable criticism that if one guy is playing great and his 4 teammates are playing terribly they'd all have the same +/-. My original thought was that over the course of the season the effect of "my teammates are playing well or poorly" would all even out. But I added the Adjustment to augment the analysis on a game by game basis.

We start with "performance points." I award 1 or 2 performance points for a made shot, deduct 1 performance point for a missed shot. Add a point for a steal, subtract a point for a turnover. Add a point for an offensive rebound and 0.5 points for a defensive rebound. Add a half point for an assist. Add 0.75 points for a block. Add 0.4 points for a made free throw. Subtract 0.6 for a missed free throw. If you miss the front end of a one-and-one you get a 1.2 point deduction.

So I do that for every lineup during the course of the game. For example, with 16:05 left in the game yesterday, Collins put Young in for Beran and we had Nance, Young, Berry, Buie and Roper on the court.
They played about 1:31. During that segment, Roper had 2 steals and a block (+2.75). Berry missed two shots (-2). Nance missed a shot (-1). Young got an offensive rebound (+1). Buie didn't do anything specific (0). However, we didn't score and Purdue scored 2 points. All 5 of those guys got a -2 for their +/-. The performance adjustments total up to +0.75, meaning we should have won the segment by 0.75 points, on average.

To get the adjustment for the impact of each player, you have to make sure that the total of the adjustments nets out to zero for every segment. This is basically allocating credit or blame for the success or failure of the lineup. So if Young has a +1 performance, that means his 4 teammates have to take a -0.25 for their own adjustment. In the example above, this results in...

Roper +3.25, Young +1.0625, Buie -0.1875, Nance -1.4375, Berry -2.6875.

You add up all the segments for each guy and get his total adjustment.

I think its a good method to evaluate individual performance in the context of the lineups we deployed.
Raw +/- is really about the success of the 5 man lineup and captures a ton of intangibles that the box score cannot.

If thats confusing, think of it this way... If Young scores a basket and we stop Purdue from scoring, then Young grabs the rebound, we win the segment 2-0. Young and his linemates are all raw +2. But Young did most of the work. So he has 1.5 performance points. The total adjustment has to be zero, so the 4 linemates are all adjusted down -1.5/4 = -0.375. Each player still has a raw +/- of +2, but Young has a Net +/- of +3.5, everybody else in the lineup has a Net +/- of +1.625.

Do you play fantasy sports? Draft kings has a daily game where you select players to generate a team. All within a salary cap. You can play for peanuts, might want to give it a view.

 

[peatymeanis]

In the end, all of this borders on the absurd. You could go through every other game in the NCAA and play a similar blame game. It does not mean it’s useful.

Wait a minute, I thought you loved “looking deeper”

 

[rwhitney014]

I would like for us just to hit some more shots, particularly the wide open ones.

 

[PurpleWhiteBoy]

@PurpleWhiteBoy , thanks for doing this.
First game I’ve seen live in a while, and Snoozer is an accurate description.

A few questions:
- do you track # of lineups used in a game?
- do you track # of substitutions in a game?
- have those trended in any way this season?
- do you track opponents on +/- or substitutions or anything?
- if so, how do NU’s numbers compare to their opponents?

My guess is that NU substitutes more and uses more lineups. I’m interested in the gap.

Hi NUCat320.
We fell behind early with a lineup that didn't make sense. We never really looked inspired or capable of winning.

Regarding your questions..
Yes I have every lineup and playing time and result for every Power 5 game. For last year too.

Yes I have every substitution, other than the end of game garbage in some games.

I can't say there's a trend to the number of substitutions. I have looked at it and it just seems random other than "Robbie, Pete, Chase and Boo must start both halves" and "I'm going to sub at the first tv timeout." Collins almost always plays the last 5-8 minutes of the game differently than the rest of the game. Lineups that succeed in the first half don't seem to get preference for playing time in the rest of the game. Also Collins is quick to sub guys out if they commit a foul in the first half. I would say he seems very cautious in that regard.

I don't track the opponent as part of the process. Sometimes I look at who played against whom on an ad hoc basis. It is way too much work to do the entire league.

NU substitutes a lot. I have not studied what other teams do, but if you think about how regularly we have guys sitting at the scorer's table when our opponents don't, I don't think it requires a study.

 

[SDakaGordie]

Wait a minute, I thought you loved “looking deeper”

It makes me feel good that you care enough to listen.