Untitled Document
Brown Columbia Cornell Dartmouth Harvard Penn Princeton Yale



Username Post: usage rates        (Topic#20981)
Jeff2sf 
Postdoc
Posts: 3751

Reg: 11-22-04
01-15-18 12:13 PM - Post#243607    

someguy wrote:

"One thing I'm digging about this team is how we seem to find different ways to win. I complained a little about us letting Princeton take Betley away in the 2nd half, and I felt similarly about AJ in half 2 against Columbia. In both cases, we seemed to say ok we'll do something else. But we've reached the point where we have the options to do that."

I don't know if I'm digging this or not but it sure seems like it's planned with Steve Donahue saying something after the Princeton game about wanting most of his players around a 20% USG and how Analytics says it's good. Here's the thing:

Our two best players, Ryan and AJ are putting up 112 and 110 ORats on 21.7% and 20.7% Usage. They're taking 2.5% more of the possessions combined than they "should" if everyone was equally good and shared the ball equally. This democratic offense results in AJ getting less possessions than more inefficient players like Darnell and Max. It seems to me we could benefit from AJ and Ryan being more assertive. Everyone gives Darnell crap about the "ten shots or less" thing but it should apply equally to Max.

Anyway, this is just one team, so I decided to look at two other things: how the big boys do it and how the last successful-ish Penn team did it.

I looked at the top 5 in KenPom and with Villanova being the only exception, the other 4 all had a "star" player using up at least 25% of possessions and sometimes as much as 30. This went hand in hand with having one of their starters serve as a glue guy who would have a usage rate below 15% while being pretty damn efficient with the shots they did take. Villanova is more in our ballpark of distribution but at least had Jalen Brunson over a 24% usage (and side note, could you even imagine having a guy putting up a 147 ORAT on 24% usage and that many minutes? Just a bunch of wine and roses, peaches and cream, etc.)

Wait, some of you Penn nerds must be thinking... "glue guy", "successfulish"? that sounds like BELCORE!!! So yeah, I checked Penn's 2012 team. Zack and Tyler put up 113.5 ORats on 27 and 23 Usages. If that happens, then someone needs to use a whole lot less and sure enough, Rob Belcore was using only 14% of possessions. Importantly, unlike his awful sophomore and junior seasons, he was well over 100 ORAT so he was efficient when he did assert (you also had Steve Rennard and Mike Howlett doing God's work with their low Usg/High ORATS just on fewer minutes and you had Henry Brooks positively vomiting up possessions at an unheard of 25% USG/81 ORat combo, have mercy.)

Anyway, I don't totally know where I'm going with this, but I'd just say that given we're at a 150 KenPom, the jury's out on whether this democracy works. I'd rather have two shooters in the corners (what up Sam?) and let the spacing/gravity do it's thing but I concede the defense might not be as good. I think we're closer than people want to admit to asking whether Jarrod, who I thought was brilliant Saturday night with his rebounding and made 3, would make the offense go better than Max while not giving up too much on the D end.

 
SomeGuy 
Postdoc
Posts: 4583

Reg: 11-22-04
Re: usage rates
01-15-18 02:16 PM - Post#243618    
    In response to Jeff2sf

Isn't a related question how much the defense impacts the ORAT? Betley playing one on one with Stephens hits everything. They switch Bell to him, and now presumably Betley's ORAT goes down, and the other guys go up, because now Princeton's best defender is pulled to the corners and can't help out. Similarly, Faulds and Tape couldn't guard AJ and got torched inside and out. Switching Meisner to him and having him push AJ out of the paint and guard him on the perimeter reduces AJ's likely ORAT, but leaves Columbia vulnerable to Betley and Foreman getting to the basket.

it's more or less how Princeton won last year -- here are 5 guys you have to guard, feel free to focus on taking 1 or 2 away. We'll just beat you with whoever you choose not to focus on. Princeton struggled early this year when they couldn't get anything from anyone other than their big 3. The development of Much has made them a lot tougher.

 
palestra38 
Professor
Posts: 18292

Reg: 11-21-04
Re: usage rates
01-15-18 02:24 PM - Post#243622    
    In response to SomeGuy

I'm not sure what you mean about the development of Much---they still are getting only 20 minutes from him and not great numbers. They are playing an Iron Man 3 and then trying to find the remaining minutes elsewhere. If you can hold Cannaday, Bell and Stephens in check, you beat Princeton. I don't think either C was able to do that. That is the advantage of having Woods and Foreman this year--the defense on opposing guards is superior. We beat the crap out of Morgan and Smith had nowhere to go until given a free hand to try anything and he did just that. By that time, we had a pretty good lead in both games. That 10 of 10 FT shooting down the stretch against Columbia is something we need to keep doing.

 
Jeff2sf 
Postdoc
Posts: 3751

Reg: 11-22-04
Re: usage rates
01-15-18 03:41 PM - Post#243634    
    In response to SomeGuy

SomeGuy, maybe, but the fact that they're putting up ~100 ORats may mean they can't quite handle that load and that maybe a tradeoff that involves Ryan and AJ being less efficient (but still more efficient than the other 3) would be better. We're putting up 105.4 per possession (including the Brandywine game). Could we up that a point or two just by re-distributing possessions and not minutes?

We all said Princeton made an adjustment on Betley. It's not clear to me that we didn't adjust before taking in the adjustment. Maybe that's prudent, if you "know" something's going to happen, you move away from what worked but we moved away from it before we knew it had been stopped. Ryan had only 2 shots in the entire 2nd half. I'd have figured out more ways to let him shoot. A heat check for crying out loud.

 
SomeGuy 
Postdoc
Posts: 4583

Reg: 11-22-04
Re: usage rates
01-15-18 03:48 PM - Post#243636    
    In response to palestra38

What I mean by the development of Much is that he was averaging about 5 per game, including going scoreless three times, in the first 8 games. Princeton went 2-6.

Since then he has averaged 11 per game (in other words his average over the last 9 is higher than his high game over the first 8), and Princeton has gone 7-2. They've also played like a top 100 team over that stretch, after playing like a sub 200 team over the first 8.

That's what I mean by the development of Much.

 
SomeGuy 
Postdoc
Posts: 4583

Reg: 11-22-04
Re: usage rates
01-15-18 03:53 PM - Post#243637    
    In response to Jeff2sf

Yes, I was the one who made that point after the Princeton game. I would have continued to run Betley around and worked to get him open. I do think Princeton adjusted first, as we were moving Betley up until the 16 timeout. After that, they had him stand in the corner. While that is what I would have done, what I am digging is that we no longer HAVE to spring Betley to win. We have choices. We can argue about 'em, but we made a choice, and it was enough to win.

On the ORAT thing, you should look at college basketball reference like I do. In their ORAT system, Max and Antonio are both over 100. Problem solved.



 
palestra38 
Professor
Posts: 18292

Reg: 11-21-04
Re: usage rates
01-15-18 04:07 PM - Post#243639    
    In response to SomeGuy

This is where the use of mean averages is seriously misleading. Starting in Princeton's 3rd game, when he started playing more, Much has scored, 10,9,10 0(against Miami), 9, 9, 12, 19 (against USC in an outlier 103 point explosion), 10, 12, 5, 11, 13 and 8.

That doesn't show signficant offensive improvement, other than the effect of the 0 and the 19. It does show some consistency, which is admirable for a young player, but your conclusion of his improvement isn't supported by the numbers as a whole.

 
SomeGuy 
Postdoc
Posts: 4583

Reg: 11-22-04
Re: usage rates
01-15-18 04:11 PM - Post#243640    
    In response to palestra38

Also, in regard to holding Princeton's big three in check, the numbers suggest the opposite. Whether they win or lose relates pretty directly to how much everybody else scores.

When the big three scores 50 combined, Princeton has lost every game but USC. On the other hand, when the rest of the team scores 30, Princeton is 7-1, with the only loss the squeaker to Middle Tennessee.

So again, that's what I mean by the development of Sebastian Much. Princeton has figured out how to get production outside the big three, and that has turned their season around.

 
SomeGuy 
Postdoc
Posts: 4583

Reg: 11-22-04
Re: usage rates
01-15-18 04:19 PM - Post#243641    
    In response to palestra38

Is this a good time to point out that Much has a higher ORAT, assist percentage, rebound percentage, and player efficiency rating than Antonio Woods?

 
SRP 
Postdoc
Posts: 3503

Reg: 02-04-06
01-15-18 04:28 PM - Post#243643    
    In response to SomeGuy

Much has been terrific as he has gotten used to D1 play. The D needs some work, especially if he gets dragged into the post, but he is a huge plus shooting, driving, and moving the basketball.

 
palestra38 
Professor
Posts: 18292

Reg: 11-21-04
Re: usage rates
01-15-18 04:42 PM - Post#243645    
    In response to SomeGuy

apples and oranges, but sure. Woods has some holes in his game, but the primary improvement of the team is against guys like Stephens, Bell and Cannady, Morgan and Smith....guys who were killing us in prior years Again, it's not so much about Much, but in the guys they lost to graduation---4 of them providing a 30 ppg (40 if you include Caruso) and 15 rebounds (20 if you include Caruso)

 
mrjames 
Professor
Posts: 5236

Loc: Montclair, NJ
Reg: 11-21-04
Re: usage rates
01-15-18 04:49 PM - Post#243649    
    In response to Jeff2sf

  • Quote:
I don't know if I'm digging this or not but it sure seems like it's planned with Steve Donahue saying something after the Princeton game about wanting most of his players around a 20% USG and how Analytics says it's good. Here's the thing:



Funny - this is something that Stu passed along upon hearing it and that I shot him back my thoughts on.

I didn't hear all the context here, so I don't know exactly what Steve meant. If he wants his players ON AVERAGE around 20% usage because it's a marker of balanced contributions over time, then he's definitely succeeding in making that happen. Whether or not that's "analytically" good depends on whether or not the Dean Oliver skill curves would indicate any player could take on more usage without sacrificing efficiency while another could drop usage and gain efficiency. That's a fancy way of saying, I think whether or not having everyone at 20% usage is a good thing depends on your personnel and probably isn't a universal truth. Steve could be saying that for his personnel, that is the case.

The odd part about that is that, with limited exceptions, on a game-by-game basis, some member of the starting lineup or a reserve getting a ton of minutes has hit the 30% usage mark or at least the 27-28% neighborhood. The Cornell game was a rarity where, other than Caleb Wood who only got 10 mins, no player went over 24% usage. The more common game for Penn involves having someone at 27-28% and often hitting 30+%.

Now, Penn certainly is doing it by committee with different players going big on usage in different games, which is why, for the year, everyone is bunched around 20. But on a game-by-game basis, Penn mostly plays like a team with a high usage player, it just happens to be a different one every night.

 
Chip Bayers 
Postdoc
Posts: 4682
Chip Bayers
Loc: New York
Reg: 11-21-04
Re: usage rates
01-15-18 05:04 PM - Post#243651    
    In response to mrjames

  • Quote:
Now, Penn certainly is doing it by committee with different players going big on usage in different games, which is why, for the year, everyone is bunched around 20. But on a game-by-game basis, Penn mostly plays like a team with a high usage player, it just happens to be a different one every night.




This points to successful scouting, game prep, and/or in-game adjustments by the coaching staff in identifying the mismatches or exploitable holes in an opponent’s defense, and then riding the guy who can take advantage on a given night.


Edited by Chip Bayers on 01-15-18 05:14 PM. Reason for edit: No reason given.

 
TheLine 
Postdoc
Posts: 3611

Age: 54
Reg: 07-07-09
01-15-18 06:17 PM - Post#243655    
    In response to Chip Bayers

Yes, agree, it seems purposeful.

Mike is there any easy way to quantify the negative contribution Penn's bad FT shooting is contributing to its ORat? Like what would the ORat be if Penn was an average FT shooting team?


 
SomeGuy 
Postdoc
Posts: 4583

Reg: 11-22-04
Re: usage rates
01-15-18 07:40 PM - Post#243659    
    In response to palestra38

Replacing those numbers is the point though. In the first 8 games, Princeton wasn't replacing that stuff. Over the last 9 they have been.

And for the record on Much, it wasn't until the 5th game that he played more than 11 minutes in a game (not the 3rd).

That's Princeton though. On Penn, I tend to agree. Whether it is simply that everything is working a little better, or whatever, I find that watching I am much more comfortable with the perimeter defense than I have been in a while. I thought Foreman was really really good defensively on Saturday. May be my giddiness about being 3-0 talking, given that his primary responsibility scored 27, but I thought he made Smith work hard all night. Favorite play was probably in the first half where he pressured full court and denied him the ball entirely.

 
Icon Legend Permissions Topic Options
Report Post

Quote Post

Quick Reply

Print Topic

Email Topic

457 Views





Copyright © 2004-2012 Basketball U. Terms of Use for our Site and Privacy Policy are applicable to you. All rights reserved.
Basketball U. and its subsidiaries are not affiliated in any way with any NCAA athletic conference or member institution.
FusionBB™ Version 2.1 | ©2003-2007 InteractivePHP, Inc.
Execution time: 0.023 seconds.   Total Queries: 8   Zlib Compression is on.
All times are (GMT -0500) Eastern. Current time is 09:37 PM
Top