Per 40/Efficiency Stats

#26      

whovous

Washington, DC
At various points in this young season, I've thought that it was time for Alstork, Frazier, Kipper and Williams to move to the back of the line for playing time. So far, I've been wrong every time. I guess maybe the coach gets this better than I do.
 
#27      
At various points in this young season, I've thought that it was time for Alstork, Frazier, Kipper and Williams to move to the back of the line for playing time. So far, I've been wrong every time. I guess maybe the coach gets this better than I do.

Probably just lucky.
 
#28      
RE: Alstork

Did you even watch last night's game? He's one of the best, if not the best defender on the team and knocked down clutch 3s and free throws near the end of the game. Everyone is still getting a feel for this offense and how they contribute.

Alstork albeit not contributing a lot on offense as of yet, has proven indispensable on defense...thus he will stay in the rotation as a starter I think.
 
#29      
When will Aaron Jordan and Kipper Nichols be in the starting rotation????? They have been major contributors the whole season and the team does better when they are on the floor. Its time for Alstork to take a back seat. He is mid major talent and is playing like it.

Playing starter minutes>starting.

Also Alstork locked down the quickest pg in the league last night for long stretches and backed it up by hitting huge shots on O. His mistakes have been made from being too unselfish. He was one of our top 2 players last night, on a team full of players originally recruited for high major basketball. Stop. :hand:
 
#30      
This “who should start” argument is incredibly pointless.
Maybe in a Groce system it matters, but Underwood goes 9-10 deep.
Everyone gets a shot, and the players performing the best will edge others in minutes.
 
#31      
Playing starter minutes>starting.

Also Alstork locked down the quickest pg in the league last night for long stretches and backed it up by hitting huge shots on O. His mistakes have been made from being too unselfish. He was one of our top 2 players last night, on a team full of players originally recruited for high major basketball. Stop. :hand:

+1000. Alstork is the only reason Cowan didn't destroy us in the 2nd half too.
 
#33      
Mark and Trent need to curtail their selection and their conversion rates will go up. They take some deep, unnecessary 3s.

Deep is an understatement lol. They sure do love to launch it.

BU has said it before, Mark Smith's lack of rebounds annoys BU more than anything else. With his athleticism and strength, he could easily be averaging 5 or 6 per game. But a big portion of defense and rebounding is mental, knowing where to be and when to be there. Smith is behind in the mental aspect of the game, as expected. Damonte, on the other hand, is such a smart player. Which is a big reason why he is a great defender and rebounder.
 
Last edited:
#34      
Love the stats Townie. I was wondering if maybe you could filter out the cupcakes, once we play a few more "real" games. For example, after the Mizzou game could you just provide the stats for WF, MD, NW, UNLV, and Mizzou?
 
#35      

TownieMatt

CU Expat
Chicago
Love the stats Townie. I was wondering if maybe you could filter out the cupcakes, once we play a few more "real" games. For example, after the Mizzou game could you just provide the stats for WF, MD, NW, UNLV, and Mizzou?

It's a good thought, and others have asked for this, but the issue is that I don't pull stats from individual games. These are aggregate stats from Basketball Reference that I pull into a spreadsheet, so you can't simply filter by opponent. I'll look into this more, but my gut says doing it would be really labor intensive.

For now, treat these like KenPom. The further we get into the season, the more accurate these will become.
 
#36      

TownieMatt

CU Expat
Chicago
RE: Alstork and defense

Advanced analytics still have a really hard time quantifying the effect that players have on the defensive end. Defensive win shares is one of the better defensive stats available for CBB, but even that only takes into account minutes played, defensive rebounds, steals, and blocks. So much of what a good defender does isn't included in that calculation.

Take Alstork's great defensive performance against Cowan. How many times did that include a steal, block, or rebound? Hardly ever. Instead, his defense forced Cowan to give up the ball or take bad shots, which doesn't end up in the box-score.

IMO the best defensive stat is Plus-Minus, which is the scoring differential while a player is in the game (Real Plus-Minus is actually even better). This obviously takes offense into account as well, but points allowed while you're on the court is ultimately what wins or loses games.

The problem with Plus-Minus is that is requires pretty advanced score-keeping, since you need to keep track who was on the floor for every point scored. To my knowledge, that data isn't publicly available. Programs pay statistical services for analysis like that (Stats, Inc. for example), so I'm almost positive our staff has access to those stats.
 
#37      
RE: Alstork and defense

Advanced analytics still have a really hard time quantifying the effect that players have on the defensive end. Defensive win shares is one of the better defensive stats available for CBB, but even that only takes into account minutes played, defensive rebounds, steals, and blocks. So much of what a good defender does isn't included in that calculation.

Take Alstork's great defensive performance against Cowan. How many times did that include a steal, block, or rebound? Hardly ever. Instead, his defense forced Cowan to give up the ball or take bad shots, which doesn't end up in the box-score.

IMO the best defensive stat is Plus-Minus, which is the scoring differential while a player is in the game (Real Plus-Minus is actually even better). This obviously takes offense into account as well, but points allowed while you're on the court is ultimately what wins or loses games.

The problem with Plus-Minus is that is requires pretty advanced score-keeping, since you need to keep track who was on the floor for every point scored. To my knowledge, that data isn't publicly available. Programs pay statistical services for analysis like that, so I'm almost positive our staff has access to those stats.

Say you have a coach or manager keep track of what you would consider “forced bad shots” or “forced (tough) passes + steals/blocks/rebounds/deflections.

How could you combine those into one defensive metric? Asking for a friend.
 
#38      

TownieMatt

CU Expat
Chicago
Say you have a coach or manager keep track of what you would consider “forced bad shots” or “forced (tough) passes + steals/blocks/rebounds/deflections.

How could you combine those into one defensive metric? Asking for a friend.

The problem there is that you're introducing subjectivity into the metric. What is a bad shot? What is a bad pass? Reasonable people will disagree.

IMO the way to do it is to create a variant of Plus-Minus that only takes into account how many points your team allows while you're on the floor (maybe this exists somewhere?). Obviously not every point your team allows while you're on the court is your fault, but over the course of the season you'd have a really good idea who your best defenders are.
 
#39      
The problem there is that you're introducing subjectivity into the metric. What is a bad shot? What is a bad pass? Reasonable people will disagree.

I get that but if you’re only doing it for your team I think subjectivity matters less. I get that analytics is about objective numbers painting a picture etc but I don’t want it generalized for comparison, just for my team.

Do you think something like

(forced bad pass+Forced bad shot+steals+deflections+Defensive rebounds+Forced TOs+charges taken)/minutes played

Could give you a good indication? Or change (minutes played) to (defensive possessions). I can literally everything in bold from a hudl breakdown after a game and if I had a simple equation to evaluate from i think it would be really interesting.
 
#40      

TownieMatt

CU Expat
Chicago
I get that but if you’re only doing it for your team I think subjectivity matters less. I get that analytics is about objective numbers painting a picture etc but I don’t want it generalized for comparison, just for my team.

Do you think something like

(forced bad pass+Forced bad shot+steals+deflections+Defensive rebounds+Forced TOs+charges taken)/minutes played

Could give you a good indication? Or change (minutes played) to (defensive possessions). I can literally everything in bold from a hudl breakdown after a game and if I had a simple equation to evaluate from i think it would be really interesting.

Deflections, charges taken, forced TOs would definitely be a good additions. I'd also include blocks. You'd want to make sure you're not weighting things toward guards or forwards too much.

This is just my opinion (I'm no Bill James), but I wouldn't include subjective measures in a metric. IMO, objective measurements + subjective insights from watching games = good analysis.
 
#41      
Deflections, charges taken, forced TOs would definitely be a good additions. I'd also include blocks. You'd want to make sure you're not weighting things toward guards or forwards too much.

This is just my opinion (I'm no Bill James), but I wouldn't include subjective measures in a metric. IMO, objective measurements + subjective insights from watching games = good analysis.

Time to add blocks and do some analytics for fun
 
#42      
I might suggest dividing by something other than minutes. If you are ok with custom subjective stats and are looking at film/keeping stats to the point where you have (or can assemble) a stat for defensive mistakes (times beaten off the dribble/points scored against/dumb fouls/other things that you as a coach abhor in a defensive possesion), comparing the defensive positives with the defensive negatives will give you a stat along the lines of assist to turnover.

I *think* dividing good stats by the minute will skew towards rewarding players for risky behavior. Someone who is always going for the steal by overplaying or hacking would be rewarded even though they may also be giving up easy buckets and free throws.
 
#43      

Epsilon

M tipping over
Pdx
Seems to me that it would also be useful to see how much a player affects the person who they are guarding, in terms of that person's performance vs their averages. This might also help take into account the fact that certain individuals are always tasked with guarding the best offensive players.
 
#44      

TownieMatt

CU Expat
Chicago
I didn't get to see the Maryland game, but just going by Cowan's stats it seems like maybe he had a great game even though he was "locked down" for long stretches by our best defender. I would LOVE to see the +/- chart for the whole season and one that just includes the more difficult games. When I look at our stats for the season, Alstork is just a middle of the pack player with Kipper and Aaron playing less minutes and contributing more points, better field goal percentage and more rebounds for Kipper and Aaron only 4 behind Alstork. All I am saying is that Kipper and Aaron should be playing more than Alstork. If Alstrok is so indispensable, then let him play the point or something more often. His assist to turnover ratio is good and he may be worth more there anyway.

I am just tired of watching these mid-major guys come in to Illinois when they finish at a weaker school and fade when the schedule gets tough. It feels like a few years in a row now. The only difference is that this year we have a mid-major guy who hasn't even done all that well against the non-conference teams either.

Watch the Maryland game. Statistically you're right, both AJ and Kipper have had more productive starts to the season offensively. But there was only one player on our team that was capable of slowing down Cowan, and that was Alstork.
 
#45      

TownieMatt

CU Expat
Chicago
Is there a place to go to find a +/- chart for each game? I am wasting too much time trying to find one.

I mentioned it above, but I don't believe anyone makes Plus-Minus available to the public for CBB. Would love to use it in these stats, so if anyone finds them, please let me know.
 
#46      
When will Aaron Jordan and Kipper Nichols be in the starting rotation????? They have been major contributors the whole season and the team does better when they are on the floor. Its time for Alstork to take a back seat. He is mid major talent and is playing like it.

Disagree. It seems like you are only counting his point production. He's our best defensive player and a really good passer and solid rebounder. As long as everyone is getting plenty of minutes, I can't see why it matters if one player is starting and another coming off of the bench. Bench production is an important factor and both guys are bringing it.

Alstork played really well against Maryland. Cohen had 16 or so in the first half and was basically unstoppable. Alstork took the lead on him in the 2nd half and he scored 2 points.
 
Last edited:
#47      
Instead of using time, use points per possession during players time. Still difficult to easily calculate but accounts for total defense. Not comparable to other teams but a good metric vs teammates.
 
#48      

blmillini

Bloomington, IL
The other measure that is missing is defensive denial. You could have a defender that does an extremely good job of making sure his player never gets the ball or keeps his player from ever driving or getting off a shot. That would never show up anywhere but is probably the single most impressive thing a defender could do but really hard to measure and probably subjective again.

In my mind it is very much similar to range factor for infielders in baseball and making sure it is determined not just by how fast a player can move but also how good he is at reading the play, understanding the impact of pitches and other factors. Even guys that understand the game sometimes don't understand those things.
 
#49      
Deflections, charges taken, forced TOs would definitely be a good additions. I'd also include blocks. You'd want to make sure you're not weighting things toward guards or forwards too much.

This is just my opinion (I'm no Bill James), but I wouldn't include subjective measures in a metric. IMO, objective measurements + subjective insights from watching games = good analysis.

Would I be correct in assuming, that offensive and defensive possession numbers would mirror each other +/- 1 because it depends on when the sub is made?

Edit: also I’m talking by lineup hence the sub part.
 
Last edited:
#50      

TEYPAY

Springfield
Blocks are only good if it is a change of possession -- if you block the shot and he gets it right back and scores a and+1 -- did the block help?