H78
PhD Student
Posts: 1458
Loc: Los Angeles, CA
Reg: 01-06-11
|
02-14-12 04:43 PM - Post#120663
In response to mrjames
Since we brought Adjusted +/- up, here's a great site that has a graphical representation of defensive and offensive adjusted +/-.
http://public.tableausoftware.com/views/2012NCAAAd...
You just have to change the Average Team Margin slider to include all values and filter to just the Ivy League. The funny thing is that adjusted +/- actually looks pretty good.
Thanks, mrjames. I'm having fun playing with it right now. Am I reading it right that Zack Rosen, Tyler Bernardini, Laurent Rivard and Ian Hummer look the strongest?
|
H78
PhD Student
Posts: 1458
Loc: Los Angeles, CA
Reg: 01-06-11
|
02-14-12 04:45 PM - Post#120664
In response to mrjames
P.S. Speakin of Plus/Minus, here's an article that just popped up in this week's issue of ESPN The Magazine, calling Plus/Minus a "controversial stat": Link
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-14-12 05:05 PM - Post#120667
In response to H78
Know your axes. Up-down is offense, left-right is defense. I think those guys are the strongest offensive +/- guys. The strongest defensive guys are Curry, Miller, Wright, Casey, Willhite and some others that make a ton of sense, while the weakest ones make a ton of sense as well.
But "sense" increases with playing time (and thus sample size). You'll see the results get more curious (and probably unreliable) as you lower the required minutes.
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
02-14-12 05:11 PM - Post#120668
In response to 1LotteryPick1969
The off-ball screen is still part of the basic Carril offense. They were running the "chin" play without touching the chin.
An off-ball screen and roll is not a pick and roll, and in this case Connolly did not roll when he set the pick - he kept Wright at the high post so Bray could shoot his layup/postup over a guard.
The time that Connolly did roll, it was instead of setting the pick because Wright jumped the screen. Davis was at the top of the key and started moving toward Connolly for a high pick. Wright moved from behind Connolly to hedge on Davis. Davis stopped his cut and Connolly rolled down the lane, caught the pass, made the layup and was fouled by Wright coming too late to the play. Definitely no travel on that play.
|
1LotteryPick1969
Postdoc
Posts: 2275
Age: 73
Loc: Sandy, Utah
Reg: 11-21-04
|
02-14-12 05:25 PM - Post#120670
In response to Brian Martin
But Umbrellaman is correct...if you look awkward, the travel will be called more often. I often see the very athletic spot-up shooters do the little hop step (very subtle) after receiveing the pass, but it is not called. Also change the pivot foot after receiving a pass and landing 1-2, then moving foot 1. When Wright split the double team late in the game, I was sure he had hopped through it, but on review, I think he was just very quick on his feet. The players have such elegant footwork now, it must be all the more difficult to be a ref.
|
TheLine
Professor
Posts: 5597
Age: 60
Reg: 07-07-09
|
02-14-12 05:31 PM - Post#120671
In response to Jeff2sf
Just answer me this, since you replied "THANKS ALL". Are what Brian and H78 posted stats or observations? They are stats. Mike did an observation.
While Brian and H78's stats were useful they didn't really explain what happened in the way that Mike's observation did. Any yes, I was already aware of the basic stats but they were so out of line with how Harvard has been playing D this year that I really did need further explanation.
After Mike's comment I looked at +/-, which is admittedly the last stat I look at. Lo and behold Miller - the Freshman who played the most minutes - is at +4, which means Harvard was -12 while he wasn't on the court. You can't blame Miller for that. You can't blame Travis too much either since Harvard was -11 when he wasn't in the game. I'm willing to believe Miller and Travis had trouble with the Princeton offense since everyone seems to agree on that point and there is no "got burned on D" stat in the box score.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-14-12 05:51 PM - Post#120676
In response to TheLine
The point about +/- is that there is a ton of noise in that stat. But over 1000 minutes or so, you can gain a bit of confidence that the noise is cancelling itself out and that there is something to the residual effect.
It's why game +/- or even a series of games' +/- is a really bad stat to look at, but season +/- can really tell a decent story. You also have to adjust the +/- for opponent. That leaves you with the chart from the link I posted above, which really makes a fair deal of sense given what I've seen and know about the league. But the cooler point that it illustrates is that at lower levels of minutes, the results sometimes run violently counter to what I'd expect, which suggests that noise is dominating the results.
|
H78
PhD Student
Posts: 1458
Loc: Los Angeles, CA
Reg: 01-06-11
|
Question on Harvard v. Princeton 02-14-12 06:03 PM - Post#120680
In response to mrjames
OK, so the top ASPM players are higher, and to the right. If I take minimums of 1.45 APSM, 10 games, about 10 minutes per game and don't look at winning margins, I get 16 "top" players:
H: Curry, Wright, Casey, Rivard, M-M, McNally
Y: Wilhite, Mangano
Pr: Hummer, Bray
Pe: Bernardini, Rosen, Belcore
Col: Cisco, Barbour
Cor: Miller
Right?
P.S. Sorry if this post is "out of order". I'm just a tyro at this stuff.
|
Silver Maple
Postdoc
Posts: 3777
Loc: Westfield, New Jersey
Reg: 11-23-04
|
02-14-12 06:06 PM - Post#120681
In response to mrjames
It seems to me that +/-, a measure that encompasses an unusually large amount of disparate data (thus the noise), can be of value in two ways: for evaluating a player's overall performance over a lot of minutes and games (such as an entire season), or for evaluating an individual player's impact on one specific game or portion of a game. The single game analysis is not at all useful in assessing the player's overall value to the team over any significant period of time, but it will tell you something about how well he performed on a specific night.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
Re: Question on Harvard v. Princeton 02-14-12 06:23 PM - Post#120683
In response to H78
I cut things at 300 minutes played. I get...
H: Curry, Wright, Casey, McNally
Y: Willhite, Mangano
Pr: Hummer, Bray
Pe: Belcore
Col: Cisco
Cor: Miller
That seems to make sense to me. I don't know much about Bray as a defender, but I feel good about most of those as being top defenders in our league.
Some of the Harvard guys that just miss the minutes cut (Saunders and Moundou-Missi) both are pretty solid defenders, and they rate highly as well. The worst Harvard defenders (Rivard, Miller and Webster) are guys that I often think of as really crappy defenders. So it seems directionally correct.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
Re: Question on Harvard v. Princeton 02-14-12 06:23 PM - Post#120684
In response to H78
I cut things at 300 minutes played. I get...
H: Curry, Wright, Casey, McNally
Y: Willhite, Mangano
Pr: Hummer, Bray
Pe: Belcore
Col: Cisco
Cor: Miller
That seems to make sense to me. I don't know much about Bray as a defender, but I feel good about most of those as being top defenders in our league.
Some of the Harvard guys that just miss the minutes cut (Saunders and Moundou-Missi) both are pretty solid defenders, and they rate highly as well. The worst Harvard defenders (Rivard, Miller and Webster) are guys that I often think of as really crappy defenders. So it seems directionally correct.
|
|