mountainred
Masters Student
Posts: 513
Age: 57
Loc: Charleston, WV
Reg: 04-11-10
|
02-17-12 11:24 AM - Post#120929
Thought it might be a nice change of pace to talk about the actual team.
Cornell is clearly an underdog in both games, but the Big Red have a chance of stealing one or the other.
For my money, Penn is the second best team in the league and the game in Newman wasn't that close, but I still think Cornell matches up reasonably well with the Quakers. Looking back, in game one Cornell won three of the "four factors," but Penn shot the ball so well (and Cornell so poorly) it didn't matter. While Brooks is emerging, Penn hasn't gotten the play from their front line to exploit Cornell's biggest weakness. Penn in the Palestra is never fun, but if Cornell can keep the Quakers from shooting 10 for 20 from behind the arc, and the emergence of Gray as an all-Ivy guard is real (he's 12th in avg. eff. in league games and, no, I did not see that coming), this could go down to the wire.
Meanwhile, even though Cornell knocked off Princeton last time, I don't like the match-up with the Tigers. The last time, Cornell's pressure gave Princeton fits (13 turnovers) but the Tigers insistence on taking 28 threes despite a huge front line advantage was a big factor. I just don't see Princeton doing that again. And Davis isn't going 7 for 22 again. I hope he does, of course, but I don't expect it. One plus, Princeton does a very nice job of denying their opponents three point shots. That actually helped Cornell last time by forcing the Big Red to do more than chuck bombs (50 two point FGA v. 11 3's)
Cornell doesn't need to win either game to stay on pace for 7-7 and improvement from last season. Steal one, and we are looking at a possible winning league record. Not the ultimate goal, but progress.
|
scoop85
Freshman
Posts: 62
Loc: Goshen, NY
Reg: 02-16-12
|
02-17-12 02:40 PM - Post#120950
In response to mountainred
Solid analysis - My perception as well is that there's a better chance stealing one from Penn than from Princeton this weekend.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-17-12 03:19 PM - Post#120954
In response to scoop85
Big Red is 42% to take at least one this weekend (6% to take both).
The Penn read is the way I feel about the Quakers as well. Zack and Tyler can get hot and bury you. If they go off from three, you're not winning. The secondary options are not as scary, which is why I think people like the idea of being Penn more than a Princeton squad with a lot of weapons.
Gotta hit your shots against Princeton and play disciplined defense. That's at least more straightforward than the Penn game, which is basically reliant on the performance of the other team's scorers.
|
mountainred
Masters Student
Posts: 513
Age: 57
Loc: Charleston, WV
Reg: 04-11-10
|
02-17-12 04:26 PM - Post#120958
In response to mrjames
Big Red is 42% to take at least one this weekend (6% to take both).
The Penn read is the way I feel about the Quakers as well. Zack and Tyler can get hot and bury you. If they go off from three, you're not winning. The secondary options are not as scary, which is why I think people like the idea of being Penn more than a Princeton squad with a lot of weapons.
Gotta hit your shots against Princeton and play disciplined defense. That's at least more straightforward than the Penn game, which is basically reliant on the performance of the other team's scorers.
42% -- better odds than I thought. I'll take that in a heartbeat.
Zach and Tyler accounted for 36 of Penn's 64 points last time, so I think it is safe to say that tandem buried us. One thing Cornell has done well all season is defend the 3 -- they are #14 in the country -- but Penn won that battle the first time around.
|
mountainred
Masters Student
Posts: 513
Age: 57
Loc: Charleston, WV
Reg: 04-11-10
|
At Penn/Princeton
02-17-12 04:27 PM - Post#120959
In response to mrjames
Edited by mountainred on 02-17-12 04:28 PM. Reason for edit: Double post.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-17-12 06:06 PM - Post#120967
In response to mountainred
Just in time for that comment...
http://kenpom.com/blog/index.php/weblog/defense_ha...
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
At Penn/Princeton
02-17-12 06:54 PM - Post#120975
In response to mrjames
How does that prove that defense does not affect 3 point percentage or 3 point attempt rate?
Edit to add:
I don't know for sure about Cornell, but Princeton has a definite defensive strategy of preventing three point attempts and making teams make twos. Princeton has the lowest opponents 3-point rate (22.8) in Division I. Next are Wisconsin and Duke who also make a point of overplaying the arc. Opponents make 34% of threes against Princeton, but they don't get as many attempts unless they shoot 30 footers as Penn did.
On the other hand, opponents make only 27.7% of threes against Georgetown, 5th lowest in the country. JTIII plays a long lineup often with 6'8" players at the 2 & 3 positions. They contest threes very well. Last year, when Georgetown played small ball with three guards 6'3" and shorter most of the time, opponents made 34.4% of threes.
These are not random numbers.
Edited by Brian Martin on 02-17-12 07:23 PM. Reason for edit: No reason given.
|
mountainred
Masters Student
Posts: 513
Age: 57
Loc: Charleston, WV
Reg: 04-11-10
|
 At Penn/Princeton
02-17-12 08:39 PM - Post#120993
In response to mrjames
Do I at least get credit for an assist? 
So, opponent 3pt FG% is like opponent FT% (or at least I assume opp. FT% is "random"). A lot of the missed 3's against the Big Red appeared to be good looks to me, but I wanted to think the defense was doing something.
Brian, what Princeton does is different. They take away the shot entirely, force you to take shots inside the arc. The second chart shows a pretty significant correlation from the first half of a conference season to the second.
Edited by mountainred on 02-17-12 08:44 PM. Reason for edit: No reason given.
|
mountainred
Masters Student
Posts: 513
Age: 57
Loc: Charleston, WV
Reg: 04-11-10
|
Re: At Penn/Princeton
02-17-12 10:09 PM - Post#121049
In response to mountainred
.... if Cornell can keep the Quakers from shooting 10 for 20 from behind the arc, ...
Penn didn't shoot 10 for 20 -- it was 12 for 22. 
|
gokinsmen
Postdoc
Posts: 3666
Reg: 02-06-10
|
Re: At Penn/Princeton
02-17-12 10:26 PM - Post#121050
In response to mountainred
So, opponent 3pt FG% is like opponent FT% (or at least I assume opp. FT% is "random").
I don't think opp. FT% is random. As most people will tell you, when you shoot poorly from the line, it's usually due to fatigue. And you produce fatigue in the other team when you have a big, physical team that is creating lots of contact -- whether it's jostling inside for shots/rebounds or simply standing your ground when a smaller player attacks the rim.
I'll leave it to the stat experts to crunch the numbers before I declare this a fact, but it doesn't seem haphazard to me.
|
mountainred
Masters Student
Posts: 513
Age: 57
Loc: Charleston, WV
Reg: 04-11-10
|
Re: At Penn/Princeton
02-17-12 10:38 PM - Post#121052
In response to gokinsmen
So, opponent 3pt FG% is like opponent FT% (or at least I assume opp. FT% is "random").
I don't think opp. FT% is random. As most people will tell you, when you shoot poorly from the line, it's usually due to fatigue. And you produce fatigue in the other team when you have a big, physical team that is creating lots of contact -- whether it's jostling inside for shots/rebounds or simply standing your ground when a smaller player attacks the rim.
I'll leave it to the stat experts to crunch the numbers before I declare this a fact, but it doesn't seem haphazard to me.
Hmm, perhaps, though most teams I have followed have been rather consistent in their FT ability or lack thereof over the course of the season. FT defense seems to me to more of a testament of who you scheduled (which isn't truly random, I know).
But that's barely anecdotal and a long way from evidence.
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
02-18-12 12:20 AM - Post#121072
In response to mountainred
Opp ft% is not necessarily random. Generally, teams that have "good ft defense" are teams that put big men on the line more than they put guards on the line. Some teams will plan to hack big men who are bad foul shooters and try to avoid fouling 90% foul shooters.
As for 3 point %, it is a highly variable statistic. The best three defense last year gave up 28.5% and the worst gave up 41.6%, but in individual games, the 3-point shooting varies from 0% to 70%. A good 3 point defensive team will still give up a few 40% games but they will have more under 30% games.
The other important point is that defenses are not just trying to contest threes. They have to also guard against interior play, penetration, etc. The teams that defend the three best try to defend the paint without doubling inside with a perimeter defender, but sometimes your big men can't get that done and you have to pick your poison.
I will go back to the example of Georgetown. They were great at defending the three in 2007 and 2008 when they had Roy Hibbert. His defensive presence in the paint allowed his teammates to play tight on perimeter shooters knowing that Roy could clean up any dribble penetration. Then in 2009 through 2011 Georgetown played small ball with three guard lineups and couldn't defend the perimeter nearly as well. Now with a bunch of 6'8" players closing out on shooters they are great at defending the three again. The numbers don't lie.
|
mountainred
Masters Student
Posts: 513
Age: 57
Loc: Charleston, WV
Reg: 04-11-10
|
02-18-12 01:08 PM - Post#121130
In response to Brian Martin
Brian, 24 hours ago I would have said "of course defenses affect opponent three point shooting percentages" for all the reasons you give. But, if true, there should be some correlation between a team's effectiveness there in the first half of a conference slate and the second half. The date Mike linked from Ken Pomeroy says just the opposite. Maybe 2011 was some kind of freak season, but the evidence is pretty compelling that, in 2011 at least, while a defense can reduce the number of three pointers attempted, once the ball is in air the defense is irrelevant.
As for "FT Defense," I don't think enough teams are actively fouling poor FT shooters to really change the data. Looking at the team I know best, Cornell, they are literally one of the worst in the nation in FT defense (#340). I can tell you that they foul indiscriminately: guards, forwards, centers, starters, subs, managers, mascots -- the Big Red will send them all to the line and they all hit their FT's. And it's not just a Courtney issue; no Cornell team has been above average in that stat since the 2003 team finished #64. There is nothing remarkable about the 2003 team.
Similarly, look at Harvard. They are #8 in FT defense this year, but were #331 last year. Or Princeton: #14 this year, # 309 last year. Those are pretty wild swings.
Not evidence, I understand, but curious.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-18-12 01:30 PM - Post#121133
In response to mountainred
I've been putting together a data set on FT defense, and at a top level it appears that teams have very little control over FTs made beyond what their opponents normally shoot weighted by how many FTs the team yields to each opponent.
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
02-18-12 05:52 PM - Post#121157
In response to mountainred
I looked up the two teams that I follow closely and their opponents 3-point shooting for the 1st half and 2nd half of conference play last season seems pretty consistent:
Princeton 1st 7 Ivy games: 30/105 28.6%
Princeton 2nd 7 Ivy games: 27/87 31.0%
Harvard, Brown, Columbia, and Cornell made a higher percentage of threes against Princeton the second time they played, but Penn, Yale, and Dartmouth shot better in the first meeting. Four of Princeton's opponents attempted less than 10 threes in their second meeting. The first time through no opponent attempted less than 11 threes. So it appears that the Tigers made a special effort to limit three point attempts later in the season.
Georgetown 1st 9 BE games: 40/123 32.5%
Georgetown 2nd 9 BE games: 48/154 31.2%
The Big East has 16 teams. Teams play 12 teams once and 3 teams home and home, so the opponents in the two halves of league play are different.
Four teams in the 2nd half of conference play - Louisville, Providence, Syracuse, and Cincinnati had games with more than 20 three point attempts against Georgetown. None of the first nine league opponents attempted more than 16 threes.
I am still surprised that Pomeroy's chart shows such a big difference between the first and second halves of conference play since the two teams I follow gave up very similar shooting percentages in the two halves of league play last season.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-18-12 05:55 PM - Post#121159
In response to Brian Martin
I am still surprised that Pomeroy's chart shows such a big difference between the first and second halves of conference play since the two teams I follow gave up very similar shooting percentages in the two halves of league play last season.
This is a joke, right?
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
02-18-12 06:07 PM - Post#121164
In response to mrjames
No. Not a joke. I don't see a wide variation among any teams. I haven't looked up the numbers for other teams, but Harvard appears to consistently guard threes. Duke does. Wisconsin does. And so on. And some teams like Craig Robinson's Oregon State team are consistently terrible at guarding threes because they play 1-3-1 and trap to try for turnovers but good teams beat the trap and get wide open shots for good shooters.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-18-12 06:10 PM - Post#121165
In response to Brian Martin
So, you're either saying that Pomeroy has made a mistake, or that your few data points trump his? I don't follow which argument you're making.
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
02-18-12 06:17 PM - Post#121167
In response to mrjames
I assume his data is correct but it is very surprising. Are Princeton and Georgetown the only teams that played consistent three point field goal defense? That seems hard to believe.
I guess I need to see the data on teams that held opponents to under 30% for one half of the conference season but gave up more than 40% in the other half.
I know that some teams are consistently good at defending threes and other teams are consistently bad. Sometimes players make contested threes and sometimes they miss open ones, but over the season the team that consistently contests shots will give up a lower percentage than the team that consistently allows open shots.
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
At Penn/Princeton
02-18-12 06:35 PM - Post#121168
In response to Brian Martin
Okay, now I looked up Harvard, and Harvard's numbers for last season are even more consistent than Princeton's or Georgetown's.
1st 7 Ivy games, Harvard opponents made 46 of 124 threes for 37.1%
2nd 7 Ivy games, Harvard opponents made 47 of 128 threes for 36.7%.
Edited by Brian Martin on 02-18-12 06:41 PM. Reason for edit: No reason given.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-18-12 06:39 PM - Post#121170
In response to Brian Martin
There are going to be a lot of teams that have consistently good and bad 3PT% allowed. The point is that if teams could really control 3PT% allowed, you'd at least see some weak correlation.
You can make an argument that teams who have poor 3PT% allowed make a more concerted effort to guard the 3PT line during the second half of conference play and vice versa. But even then, you'd still expect a weak correlation between first and second half. That there is ZERO correlation is totally striking. Especially when there's a high correlation of 3PT attempts between the first and second half. If teams changed their defense, you'd expect both the rate of makes to rise and fall and the rate of attempts.
This also meshes pretty well with the argument of why teams that are reliant on 3PT shooting might be high variance. If you can't control your 3PT percentage that much, and you rely on many of your points from there, you're going to get vastly different outcomes night after night. To that extent, it makes sense.
I find this to be fascinating, and I'm working up a new metric that incorporates this and FT defense. I think it has the ability to enhance the predictive nature of college basketball models.
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
02-18-12 08:37 PM - Post#121189
In response to mrjames
Please do me a favor. If you have the data handy, plot the Ivy League teams on their 3-pt fg defense in the 1st half and 2nd half of league play and tell me if your chart looks like Pomeroy's.
I have added up the numbers for four teams from last season and I have not yet found a team with a significant difference between the 1st half and 2nd half.
Harvard 37.1, 36.7
Dartmouth 38.1, 38.4
Penn 36.2, 36.6
Princeton 28.6, 31.0
Princeton's 2.4% difference is the greatest, but Princeton played 5 of the first 7 at home, so some of the better shooting by opponents in the second half may be due to shooting at home vs. shooting at Jadwin.
|
SRP
Postdoc
Posts: 4911
Reg: 02-04-06
|
02-19-12 06:21 AM - Post#121317
In response to Brian Martin
Here's an interesting question: Is 2-pt. FG% defense more correlated across season halves than 3-pt FG%, and if so, why? I mean, it would be odd for defense to have no consistent effect from three but to have a big effect from just inside the arc.
One hypothesis would be that FG% defense is genrally more consistent as you look at shots closer to the basket. That would be a very interesting thing, if true.
|
mountainred
Masters Student
Posts: 513
Age: 57
Loc: Charleston, WV
Reg: 04-11-10
|
02-19-12 11:16 AM - Post#121337
In response to SRP
Looking at the Penn, Princeton and Columbia games -- Cornell's 3pt D was 32.7% the first time through and 40% the second time. That's the difference between being top 100 and bottom 10 nationally if maintained for an entire season.
As for the Cornell/Princeton game -- it proved to be as bad a match-up as I feared. The Tigers shot the lights out and dominated the glass, at least until the game was completely out of reach. As I said on another forum that will go nameless, this isn't the first time, nor will it be the last, that Cornell ran into a buzzsaw at Jadwin. And it has happened to better teams than this year's edition.
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
02-19-12 04:32 PM - Post#121376
In response to mountainred
I looked up the last two Ivy seasons and 11 teams were consistent at defending threes in league play in the first half and second half of Ivy play (4% or less difference between 1st and 2nd half d3fg%).
Generally, those teams had a few games where they held opponents under 25% and a few in which opponents exceeded 45%, and more games in between, but the highs and lows were fairly evenly distributed between early and late games in the league schedule.
Of the five outliers, the biggest by far was the 2010 Brown team: 26% in the 1st 7 games and 45% in the last 7 games.
Brown 2010 had its three best or luckiest defensive games against threes in the first seven games: Yale 3 for 18, Penn 3 for 15, and Dartmouth 0 for 7. They had their three worst or unluckiest defensive performaces against threes in the last two weekends of the season: Harvard 12 for 19, Cornell 20 for 30, and Columbia 6 for 10.
Six of the seven Ivy opponents shot better from three in the second game against Brown than they had in the first game. I do not know what that proves. While their may have been some difference in Brown's defensive and opponents' offensive strategies, a fair amount of the difference between the two halves was the fluke of not having a hot opponent earlier in the season and not having a very cold one later in the season.
Seven games is a small sample that can be distorted by an extreme performance or two. If Cornell had made 20/30 in the first meeting and 7/25 in the second meeting instead of the reverse, Brown's splits would have been 35.8% and 37.0%.
This does not show that defense has no effect on opponents three point shooting. Just because there is high variability game to game does not mean that the season average is random. A team that holds opponents to an average of 30% from three is not just luckier than a team that holds opponents to an average of 40%.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-19-12 05:23 PM - Post#121387
In response to Brian Martin
You still don't have anywhere near enough data to come close to refuting KenPom's findings. I'm sorry. I don't know how to explain this any clearer.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-19-12 05:28 PM - Post#121389
In response to mrjames
I'll elaborate: if someone shows you a huge sample of data, you can't take a fractional subset of that data which implies a different result and deem that the latter trumps the former. You need to either take the same dataset or produce an equally or more robust dataset with a new previously omitted variable that explains away some of the findings from the first dataset.
|
mountainred
Masters Student
Posts: 513
Age: 57
Loc: Charleston, WV
Reg: 04-11-10
|
02-19-12 05:56 PM - Post#121400
In response to Brian Martin
I looked up the last two Ivy seasons and 11 teams were consistent at defending threes in league play in the first half and second half of Ivy play (4% or less difference between 1st and 2nd half d3fg%).
In 2011, the difference between a top 100 3 pt defense and a bottom 100 3pt defense was 2.5%. So, a 4% swing is actually fairly significant. The bell curve for that stat is pretty steep.
I'll be curious if KenPom runs the numbers on 2012 as well and gets the same results.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-19-12 05:58 PM - Post#121401
In response to mountainred
I'm going to (try to find the time to) run the numbers on Ivies historically going back as far as I can. We'll see what shakes out (and how far back I can get).
|
SRP
Postdoc
Posts: 4911
Reg: 02-04-06
|
02-19-12 05:59 PM - Post#121402
In response to mrjames
You shouldn't do data mining to find subsets that differ from the overall, but if you have a prior causal theory (such as "Ivy League teams are more likely to have consistent 3-pt% defense") then you can perform statistical tests to see if the subset is non-randomly different from the overall.
Our Bayesian friends have a whole other way of looking at it, of course...
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-19-12 07:00 PM - Post#121411
In response to SRP
That's absolutely true. But to test that point you need more than 16 data points. That's why I'm going to try to dig back into the past. If I can get 15 years or so, then we'll be on our way to at least a decently sizeable dataset.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-19-12 08:25 PM - Post#121427
In response to mrjames
I just ran this for all Ivy teams back to 1997. The R^2 I get on first half 3PTFG% vs. second half 3PTFG% is 0.02. The R^2 I get on first half attempts vs. second half attempts is 0.45.
That's pretty close to what KenPom got in his blog post, so I'm prepared to reject the hypothesis that there are reasons to believe the Ivy style of play doesn't lend itself to the same lack of control of 3PTFG%.
|
SRP
Postdoc
Posts: 4911
Reg: 02-04-06
|
02-19-12 10:55 PM - Post#121453
In response to mrjames
On reflection, the first-half v. second-half test used byKenPom is invalid for the purpose of the inference he wants to make. What we want to know is whether Team A is significantly (and importantly) different from Team B in 3 pt.% defense. A lack of correlation for the same team across season halves is suggestive but does not speak to that cross-sectional difference. Maybe a look at the correlation of the order statistics across season halves would be okay, but a test of whether two teams's performances are distributed with the same parameters would be more direct.
I also think you have to be careful about proving too much. If a similar correlation analysis appeared to show that overall FG% defense was random, that would strongly indicate a flaw in the procedure. Plausibility would go up if teams showed smaller cross-sectional variation in 3pt % defense than they did in 2pt% defense.
The initial finding is certainly provocative, though.
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
At Penn/Princeton
02-20-12 12:14 AM - Post#121461
In response to SRP
That was my original objection. How does 1st half of league play compared to 2nd half of league play prove that defense has no effect on 3 point percentage? It proves that 3 point shooting is highly variable. If you looked at offensive three point shooting percentage for teams you probably would find similar lack of correlation. Would that mean that the offense also has no effect on 3 point shooting percentage? No. It means that 3 point shooting is a highly variable statistic.
The question that matters is whether Harvard (30.5% in league play) is better at defending threes than Brown (55.0%)?
Does anyone really believe that Pomeroy has proven that Harvard is only much, much luckier than Brown and has had no effect on its opponents' three-point shooting?
Edited by Brian Martin on 02-20-12 12:16 AM. Reason for edit: No reason given.
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
Re: At Penn/Princeton
02-20-12 01:08 AM - Post#121463
In response to Brian Martin
Correction: Brown's league opponents shoot 44% from three, not 55%.
That was my original objection. How does 1st half of league play compared to 2nd half of league play prove that defense has no effect on 3 point percentage? It proves that 3 point shooting is highly variable. If you looked at offensive three point shooting percentage for teams you probably would find similar lack of correlation. Would that mean that the offense also has no effect on 3 point shooting percentage? No. It means that 3 point shooting is a highly variable statistic.
The question that matters is whether Harvard (30.5% in league play) is better at defending threes than Brown (44.0%)?
Does anyone really believe that Pomeroy has proven that Harvard is only much, much luckier than Brown and has had no effect on its opponents' three-point shooting?
Edited by Brian Martin on 02-20-12 01:09 AM. Reason for edit: No reason given.
|
Tiger69
Postdoc
Posts: 2814
Reg: 11-23-04
|
At Penn/Princeton
02-20-12 09:49 AM - Post#121480
In response to SRP
You guys could do some serious numbers-kicking in a sports bar. 
Let's have a good "Yo' Mama" algorithm joke.
Edited by Tiger69 on 02-20-12 09:53 AM. Reason for edit: No reason given.
|
1LotteryPick1969
Postdoc
Posts: 2274
Age: 73
Loc: Sandy, Utah
Reg: 11-21-04
|
Re: At Penn/Princeton
02-20-12 10:09 AM - Post#121485
In response to Brian Martin
I am hesitant to weigh in on this because I am no statistician. The observation is intriguing, but...
I agree with Brian on this: Pomeroy merely proved that he can't MEASURE three point defense, not that it doesn't exist. In other words, his premise was flawed.
Three point defense does not exist as a separate parameter. It is one component of team defense, and team defense has many aspects, most of which are difficult to quantitate and very interrelated.
Also, you are looking at a number (3-pt shooting percentage) which has a high coefficient of variabilty, and the sample size is small.
We have much the same problem in clinical medicine, trying to measure benefits of treatment on parameters which have high coefficients of variability. The solution is make more measurements before and after treatment, but that is not an option in this case.
|
1LotteryPick1969
Postdoc
Posts: 2274
Age: 73
Loc: Sandy, Utah
Reg: 11-21-04
|
Re: At Penn/Princeton
02-20-12 10:14 AM - Post#121486
In response to Tiger69
Let's have a good "Yo' Mama" algorithm joke.
Here's my lame attempt:
Yo' mama so ugly, her love life is a Bortkiewicz distribution.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-20-12 11:31 AM - Post#121496
In response to SRP
Now we're on the right track. The questioning of the data is ridiculous, but the questioning of why is important.
The problem that I have with saying that conference play is too small of a sample is that in the grand scheme of things, the whole season isn't that much larger of a sample. So, the question is whether even a team's full-season 3PT% allowed is actual ability or random variation around the weighted average of what their opponents have shot for the season.
The 3PT%/2PT% analysis is an important one, and I'll get to that next. It would make sense, to me at least, that 3PT% defense would be more controllable than FT% defense, but far less controllable than 2PT% defense. This just has to do with ability to defend. You can't defend the FT line. You can't always defend threes. But two pointers are easily the most defensible of the three.
I'm not trying to say that 3PT defense is completely uncontrollable, especially since forcing people off that line is an important element of defense. My point is that Pomeroy's data would seem to indicate that it isn't very controllable. Not just that it can't be measured, but that even the full season result might be a lie - or, otherwise stated, the best predictor of a teams 3PT% defense is the weighted average of its opponents' 3PT shooting percentages.
Most importantly this might be an important part of the variance piece that comprises a team's profile, so it's worth digging into deeper to understand.
|
Silver Maple
Postdoc
Posts: 3777
Loc: Westfield, New Jersey
Reg: 11-23-04
|
Re: At Penn/Princeton
02-20-12 11:53 AM - Post#121502
In response to 1LotteryPick1969
Let's have a good "Yo' Mama" algorithm joke.
Here's my lame attempt:
Yo' mama so ugly, her love life is a Bortkiewicz distribution.
Yo mama is so mean that she has no standard deviation.
or,
Yo mama is so fat that her derivative is strictly positive.
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
02-20-12 01:10 PM - Post#121514
In response to Silver Maple
To measure three point defense you need a new hybrid stat that combines percentage, rate, and maybe other variables.
Harvard made 4 of 9 threes against Princeton but I consider that a very good job by Princeton because Harvard usually shoots and makes twice as many. I would say 4/9 is a better defensive performance than 8/20 even though the percentage is worse. Of course it also matters if a team is not shooting threes because they are getting easy twos. So maybe effective fg defense is the better stat and we should not isolate threes.
Three point stats can be tricky because of the variability and the sample size. Friday night Columbia was 2/10 while the outcome was in question but made 3 of 4 at the end in between Princeton free throws to end up a respectable 5/14. Last weekend McNally made a three at the buzzer to make an 11 point loss an 8 point loss and make Harvard's percentage 44.4 from 37.5.
In the league stats Pomeroy used, one game might have 30 threes attempted by an opponent while two or three games have less than 10 attempts, so the 30-attempt game overwhelms the 7-game sample.
|
Tiger69
Postdoc
Posts: 2814
Reg: 11-23-04
|
02-20-12 01:13 PM - Post#121517
In response to Silver Maple
Thank you.
Any candidates for statisticians walking into a bar?
|
palestra38
Professor
Posts: 32812
Reg: 11-21-04
|
02-20-12 01:38 PM - Post#121523
In response to Tiger69
3 statisticians sitting at a crowded bar on a busy Friday night talking, as they will, about the probabilities of succeeding with the opposite sex.
Gerard Butler walks in. Leaves 15 minutes later with two models. The statisticians all look at each other knowingly and say, "That was a 99% probability."
Jeremy Lin walks in. Women are all over him from the minute he enters the bar. He is polite, but leaves alone a half hour later. Each statistician nods and says, "I could have predicted that."
Finally, Jeff walks in. Walks out with a cute hipster chick after a little while. The statisticians all look at each other and shake their heads. The first one declares, "That's why you have to play the game." The other two nod and say, "I never would have believed it unless I saw it for myself."
|
1LotteryPick1969
Postdoc
Posts: 2274
Age: 73
Loc: Sandy, Utah
Reg: 11-21-04
|
02-20-12 01:38 PM - Post#121524
In response to Tiger69
and say to the bartender:
"any chance we can get a drink?"
Bartender says...
"You tell me".
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-20-12 02:02 PM - Post#121528
In response to Brian Martin
A lot of those critiques are noise that would not obscure an effect, if one existed, over a large sample.
I am looking at 2PT% and eFG% now. Just taking a while because of work. Certainly seems like the next angle of attack here - specifically with the 2PT% to make sure there isn't a shift.
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
02-20-12 02:06 PM - Post#121530
In response to 1LotteryPick1969
Me: did you ever notice that if you drop a piece of toast it usually lands butter-side down?
Stats guy: No, it lands butter-side up just as often. You just remember the butter-side down times more because of the consequences.
Me: watch. (drops toast, it lands butter-side up)
Stats guy: see!
Me: obviously I buttered the wrong side.
|
1LotteryPick1969
Postdoc
Posts: 2274
Age: 73
Loc: Sandy, Utah
Reg: 11-21-04
|
02-20-12 02:12 PM - Post#121531
In response to Brian Martin
Very good!
|
TheLine
Professor
Posts: 5597
Age: 60
Reg: 07-07-09
|
02-20-12 02:20 PM - Post#121533
In response to palestra38
"I never would have believed it unless I saw it for myself."
That's not what yo mama's stat sheet told me.
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
02-20-12 02:29 PM - Post#121535
In response to Brian Martin
To me, 3-point percentage and rate are only useful for analyzing a team's offensive or defensive efficiency and effective field goal percentage and understanding the variability in their performance. Some perimeter-oriented teams live and die on three-point shooting. Some teams (habitual zoners for example) are vulnerable to hot shooting by opponents. The overall percentages often are not as important as the variability and the distribution of the highs and lows.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-20-12 03:02 PM - Post#121539
In response to Brian Martin
I ran through them all. Here's what I found...
As I said before, the R^2 on 3PT% allowed during the first half of the season and the second is 0.02. So, there's basically no correlation. The correlation between 3PT attempted rate is 0.42. So, teams allow opponents to shoot three pointers at pretty consistent rates, but don't show any consistent proficiency at affecting conversion rates.
Moving to 2PT%. The R^2 on first and second half 2PT% is 0.24. That's a decent, but not overly strong correlation, but shows that a team can affect its opponents' conversion rate from inside the arc to a decent extent.
Finally, FT% allowed. As we'd expect, the R^2 here is 0.02. We know teams have very little control over how their opponents shoot (other than who they send to the line, maybe), so it's no surprise that the results are pretty random. But at the same time, FTRates have an R^2 of 0.30, showing that teams are in pretty solid control of how many times they send opponents to the line.
I find it very hard to believe that 3PT% allowed being as inconsistent as FT% allowed and far less consistent that 2PT% allowed means little to nothing. Especially, when at the same time, the rates (which could potentially vary wildly) are solidly correlated as well.
The 2PT% and the rates are the clinchers for me. Not necessarily that opposing 3PT% is totally uncontrollable, but that there is a huge luck factor there beyond what the defense can force. I'm going to start testing an adjusted defensive rating (that awards teams their weighted opponents averages from FT and 3) to see if it has somewhat better predictive ability than the normal Pomeroy numbers.
|
Brian Martin
Masters Student
Posts: 963
Loc: Washington, DC
Reg: 11-21-04
|
02-20-12 03:09 PM - Post#121541
In response to mrjames
A lot of those critiques are noise that would not obscure an effect, if one existed, over a large sample.
I get that, but that is my complaint that Pomeroy discredits the results of a large sample because he cant resolve the substantial noise in 7 or 8 game samples.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-20-12 03:32 PM - Post#121548
In response to Brian Martin
I understand what you're saying, but if it was all noise then we wouldn't see anything.
The fact that I see correlations some places and no correlations others leads me to believe that if there was an effect, we'd still be able to discern it. At least slightly.
In other words, if everything showed up with an R^2 only trivially different from zero, then I'd agree that it's all noise. But that some factors are showing up with a correlation and others aren't leads me to believe that it's less of a sample size issue and more of an issue of there not actually being a correlation.
Often, I wonder if the resistance to the possibility of such findings is that people refuse to admit how much of the game is truly luck (and by extension, how much of life is really luck).
|
1LotteryPick1969
Postdoc
Posts: 2274
Age: 73
Loc: Sandy, Utah
Reg: 11-21-04
|
02-20-12 03:43 PM - Post#121552
In response to mrjames
Proving conclusively that an Ivy league education is a waste of time, since life is just luck. Damn, wish I had known. Could had more money and more fun at my local U, and maybe learned how to bet on Ivy basketball, which apparantly is very remunerative.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-20-12 03:47 PM - Post#121554
In response to 1LotteryPick1969
since life is just luck.
I never said THAT. I merely said that life is more dominated by luck than most people give credit for.
|
1LotteryPick1969
Postdoc
Posts: 2274
Age: 73
Loc: Sandy, Utah
Reg: 11-21-04
|
02-20-12 04:15 PM - Post#121563
In response to mrjames
Robert Pirsig left the natural sciences (so I have read) because he was frustrated by the thought that he could never prove anything, since he could never refute all hypotheses to the contrary, or something like that. So he moved into the philosophy of "quality".
We need a book like "Zen and the Art of 3-point Defense."
|
mountainred
Masters Student
Posts: 513
Age: 57
Loc: Charleston, WV
Reg: 04-11-10
|
02-20-12 07:18 PM - Post#121594
In response to mrjames
...
I find it very hard to believe that 3PT% allowed being as inconsistent as FT% allowed and far less consistent that 2PT% allowed means little to nothing. Especially, when at the same time, the rates (which could potentially vary wildly) are solidly correlated as well.
The 2PT% and the rates are the clinchers for me. Not necessarily that opposing 3PT% is totally uncontrollable, but that there is a huge luck factor there beyond what the defense can force. I'm going to start testing an adjusted defensive rating (that awards teams their weighted opponents averages from FT and 3) to see if it has somewhat better predictive ability than the normal Pomeroy numbers.
Thanks Mike. My other message boards would never understand how interesting I find this stuff. Please don't tell them, or it will be middle school all over again.
|
IvyBballFan
Masters Student
Posts: 479
Age: 77
Loc: Central Florida
Reg: 11-19-09
|
02-20-12 10:53 PM - Post#121615
In response to mrjames
since life is just luck.
I never said THAT. I merely said that life is more dominated by luck than most people give credit for.
but... "Chance favors the prepared mind." - Louis Pasteur
and... it sure is nice that when you are lucky enough to have the guy watchin' you slip and hit the deck right in front of you 21 feet from the basket, that you know have made 25 in a row from that spot every day in practice for the past four months.
|
SRP
Postdoc
Posts: 4911
Reg: 02-04-06
|
02-21-12 09:38 PM - Post#121736
In response to IvyBballFan
Let's make the simplifying assumption that each team has only a "true" season-long 3pt% defense parameter tau. Then one could test the hypothesis that tauA and tauB were statistically indistinguishable at some confidence level by performing a difference-of-means t-test. You'd have nA+nB-2 degrees of freedom, where nA is the number of treys shot against A and nB for B, which should be enough to accept or reject the joint hypothesis that the modeling assumption is true and the means are the same.
More complex models, with game-specific shocks to tau, say, would have fewer d.f.and might not work with the data available.
|
mrjames
Professor
Posts: 6062
Loc: Montclair, NJ
Reg: 11-21-04
|
02-24-12 04:52 PM - Post#121956
In response to SRP
This is certainly the right thought if we're worried about true randomness - that over a seven to 10 game span the posted 3PT% will differ from what is "true."
I wonder if proving out whether the subset's mean is truly different from the full set's mean is necessary though (or, somewhat by extension, that there is no "true" 3PT shooting mean). The point seems much simpler than that. Three-point field goals are somewhat random, and they can be the catalyst for strange results.
I do agree with the general point of the test, though, if I were seeking to prove the point of whether the means are truly different over that small of a sample.
|
