A Look Back
02.24.06
Back on January 5th, I jotted down some of the more unusual developments in our little Pythagorean baby. Let’s go back and look at how some of the surprising conference frontrunners have fared now that the regular season is drawing to a close. All preseason predictions cited come from Blue Ribbon.
A-10: Xavier.
Nice collapse by Xavier. You have to consider their season a disappointment.
Big Ten: Illinois, with Indiana a close second.
Illinois has struggled of late, but I wouldn’t give up on them just yet. But Ohio State has made a good case as the Big Ten’s best. And Indiana is right up there with Xavier in terms of collapses. Another point is that Pythag was down on Michigan State and they haven’t consistently shown themselves to be of Final Four quality.
Big XII: Texas by a lot, but Colorado is a comfortable second.
Sheesh. Well, Colorado was picked for 10th and should finish no worse than 6th. So the stats will do at least as well as the preseason prediction.
Big East: Villanova, with Pitt second.
Up to that point Marcus Williams had not suited up for UConn. Pitt was picked for 6th and I think has proven to be the 3rd best team in the conference.
Big South: Winthrop is so far ahead of its competition, they should go undefeated.
And they did go undefeated. Except for three games. I think this is partly explained by the gravitational pull that a conference has. Any team that is a strong outlier, either good or bad, tends to come back to the pack during the season.
CAA: George Mason.
Chalk one up for Pythag. GMU was picked for 5th, and much of the press was still going to ODU in early January. A win on Saturday against lowly JMU, and Mason gets a share of the CAA title.
Horizon: Butler.
Another winner. Even Butler fans were selling their team short in early January. If they win at Detroit on Saturday, they tie preseason-favorite UWM for the regular season title. Butler was picked for 6th.
MAC: Miami (East), Toledo (West).
And back to the crapper. Actually, Miami is 12-3 and has outplayed everybody’s favorite in the East, Ohio. Toledo is another story. The Rockets started 1-6, partly due to a front-loaded schedule. They’ve clawed back to 7-8 and have a realistic shot at finishing 10-8.
MVC: Missouri State.
It took the Bears a while to play up to their numbers, but Mo State gets Creighton at home on Saturday to try and finish with a share of 2nd in the conference. They were picked for 6th in the preseason.
Mo. West: Air Force, with BYU a surprising second.
San Diego State will probably take the conference, but they added the services of a key transfer in the 2nd semester. Air Force should finish 2nd, and BYU is currently tied for 3rd with a favorable remaining schedule after being picked dead last by Blue Ribbon. Not bad.
Pac-10: Arizona, with Washington State 3rd?
I recently had a nightmare that I said some really good things about Washington State at some point.
SEC: Florida (East), LSU (West).
LSU was picked 3rd in the preseason, but now has a three game lead in the West. Florida has been dusted by Tennessee in the East, yet still leads in Pythag.
WAC: Utah State.
Nevada was still getting the press in early January, but the Aggies - the spurtiest team in the country - have a home date with Wolf Pack on Saturday for all the conference marbles.
Overall, efficiency wasn’t without its tremendous failures, but I think on balance it was insightful early on. There will still be some more judgments to make in the coming weeks as we see how the current outliers perform.
Line o’ the Night
FG 3pt FT Reb
Min M-A M-A M-A O-T A F S TO BLK Pts
Randy Foye 37 8-17 3-6 6-9 6-14 3 2 0 3 2 25
Result: Win. Villanova 74, Cincinnati 72.
Art Bell and College Hoops Don’t Mix
02.13.06
So this whole thing about Duke getting all the calls has been beat to death repeatedly over the past week. But I’m still going to chime in with a few tidbits on how the Devils stack up statistically in the fouling department.
Let’s start by looking at where Duke ranks nationally in a couple of categories. First up is what I’ll call “foul surplus”, or the difference in the fouls called on a team’s opponents as a percentage of the team’s fouls.
1. UConn .389 2. Alabama .362 12. Duke .232
There’s nothing too unusual here. Duke is one of the best teams, and great teams typically draw more fouls from their opponents for a variety of reasons. The Blue Devils’ numbers are not off the charts and pale somewhat to UConn having 38.9% more fouls called on its opponents.
The thing that stands out more is how often Duke gets to the line compared to its opponents. Again, this is in terms of a percentage surplus compared to a team’s opponents’ FTAs.
1. UConn .676 2. Alabama .642 4. Duke .541
The reason Duke gets to the line more often than their foul surplus would indicate has to do with the fact that their games are faster paced than most, and therefore more fouls are called compared to other teams. Their foul surplus results in a bigger total foul difference than other teams with a comparable percentage surplus, thus more trips to the line when in the bonus.
If I was going to argue for Duke getting more than its fair share of the calls, I would look at their style of play. UConn and Alabama have obvious elements in their style that could explain part of their advantage. UConn doesn’t force turnovers - their defensive philosophy encourages the opponent to get in confrontations with the UConn shot blockers. A confrontation which the opponents will frequently lose. UConn also shoots very few threes and thus you would expect their constant attacking of the hoop to generate more fouls on the opposition.
In the case of Alabama, much of their foul advantage could be explained on the defensive end. Their opponents shoot a ton of threes, and like UConn, the Tide rarely force turnovers.
But for Duke, just about every aspect you look at would encourage fouls. They shoot more threes than the average team, opponents hardly shoot any threes, and Duke forces a lot of turnovers. Without knowing anything about college hoops, they aren’t a team I’d expect to have a big foul advantage. But there are always exceptions. A bigger exception than Duke is revealed if we break things down by offense and defense:
Fouls committed per possession
1. West Virginia .198 2. Alabama .206 3. UConn .209 28. Duke .237
While most teams that commit few fouls also force few turnovers, WVU is a major exception, ranking 13th in turnovers forced. It makes Duke’s ranking of 65th in that category look less suspicious. Here’s how it stacks up on the other side of the ball…
Opponents’ fouls per possession
1. Texas A&M .322 52. Duke .291 57. UConn .290 104. Alabama .281
As you can see, UConn and Alabama derive most of their fouling advantage from not fouling rather than drawing fouls. This is consistent with what I would have thought about Alabama.
Well, I am sort at a loss for how to wrap this up. I looked at a few other things not shown here, but I didn’t find the smoking gun. The bottom line is the perception is there, and possibly being sensitive to this, the ACC took action either formally or informally against the officials in the Boston College and Florida State games.
Perhaps it’s a coincidence that in the two games since then, Duke has had a free throw deficit. What happens to Duke’s free throw advantage from now on might be something more worth tracking.
Move the Line Back
01.19.05
Wake Forest missed but one free throw in 53 attempts over the past two games against North Carolina and Florida State. Yet that miss cost them a win. Taron Downey clanked one with 4 seconds left in regulation that would have broken a 76-76 tie against FSU. The miss ended an NCAA record streak of 50 consecutive free throws made by Wake Forest. It adds further evidence that we're in an era where free throws are like layups. Really, they need to move back that line.
(By the way, there's been a discernible negative effect of playing UNC. Call it The Carolina Hangover. Each of UNC's four ACC opponents have lost their next game. Overall, UNC opponents are 4-12 in their next contest.)
Back to the free throws. This incarnation of my blog essentially started with a post analyzing Kansas' poor free throw shooting in the 2003 national championship game. So in that tradition, let's look at the astronomical odds against Wake Forest making 50 in a row. It's easy to do. Say a player shoots 75% from the line. The chance of him making two in a row is .75 x .75 or .5625 (56.25%). The chance of that player making four in a row is .75 x .75 x .75 x .75 (31.64%) and so on.
A rough approximation of Wake's streak could be arrived at by using the Deacs' team free throw percentage and muliplying it by itself 50 times (math-type people like to call this taking it to the 50th power). However, the team doesn't shoot free throws, players do. So with that in mind, let's look at the chances of each of the individual streaks involved in making up the new team record, using each player's FT% after the FSU game. Here's a breakdown of who made what free throws during the streak:
Chance of
Season personal
FTM FT% streak(%)
Chris Paul......19 86.1 5.84
Taron Downey.... 7 84.2 30.03
Justin Gray..... 6 73.8 16.22
Vytas Daneluis.. 6 70.8 12.63
Jamaal Levy..... 4 45.8 4.41
Chris Ellis..... 4 73.7 29.48
Eric Williams... 4 60.0 12.96
There's a couple of interesting things you can take away from this analysis. (1) Levy's four consecutive free throws were less likely than Paul's 19 in a row. (2) Overall the bad free throw shooters (Levy and Williams) didn't participate much in this streak. Trent Strickland at 56% didn't participate at all, despite playing 23 minutes during the streak.
The individual streaks were improbable, but what was the chance of the stars aligning - having all of the streaks happen simultaneously? We can multiply together the percentages in the right-most column to get the true odds of the overall event. The result is .000605%*. In other words it would happen once in every 1,651,987 sequences of 50 free throws. (In this case, each free throw starts a new sequence of 50.)
Q: So what does this number mean?
A: This kind of thing doesn't happen very often. Assuming Wake Forest shoots 25 free throws a game, you would expect this event to happen to the Deacons once in every 66,000 games...2,200 seasons...110 generations.
If I had bet you before every free throw ever shot in the history of college basketball that the team of the shooter wouldn't make their next 50, I would have been right every time - until this week. I don't know how many free throws have been shot in the history of college hoops, but I'm guessing somewhere in the neighborhood of five million. So maybe this kind of streak was overdue, or maybe there's some other factor that limits the probability of a streak like this more than the true odds would indicate, or maybe my estimate of five million is too high, or maybe this record has only been tracked over the last 30 years. All of these explanations are possible.
Regardless, there's probably a grumpy old Wake fan out there who is saying that if they had just been better from the line, they'd wouldn't have lost. And sometime around the year 4205 another grumpy old Wake fan will be saying the same thing.
*By comparison, if you use Wake's team FT% of 69.6, you get the chances of .000013% or about once in 76,000,000! So clearly the distribution of free throws in this streak being skewed more than usual towards the good shooters made the record possible. (The weighted average FT% of shooters during the streak was 76.2)