Saturday, March 09, 2013

In Which I Defend Florida

Kentucky 61, #11 Florida 57
I spent a lot of time on twitter this afternoon talking about Florida. And I do know that some of my regular readers don't use twitter, and I also think it's somewhat limiting to try to explain myself 140 characters at a time, so I want to flesh out in more detail here why the panic over Florida is silly. Let's first establish the facts. First, Florida has only been clearly outplayed once this season (Arkansas). They have had five games that were decided by six points or less, and have lost all of them. Every other game has been a blowout victory. The conclusion that the media makes from this is that "Florida stinks in the clutch" and "Florida can't win close games", and so they are going to tumble to near 20th in the polls. Florida fans have been tweeting me and saying that Florida is a "Sweet 16 team at best". Yet at the same time, they're going to remain either first or second in Pomeroy and the Sagarin PREDICTOR.

So what's the deal? Let me first say that clutch play does exist - it's a common strawman that the stats community doesn't believe in clutch play. The problem is that clutch play is just one small factor in the results of close games. A really clutch team can lose 5 or 6 straight close games, while a really unclutch team can win 5 or 6 straight close games. Statistically, you need at least 30 or 40 final possession games (and probably closer to 100) before you can start collecting statistical evidence for how clutch a team is. The problem is, most people in the media are incapable of telling the difference between clutch play and luck. If a team wins a bunch of close games then they're "clutch" and if a team doesn't then they "can't win close games". If a team gets a wide open shot and misses it then the media hammers their "execution", while teams that screw up their final possession and win games on off-balance 30 footers are praised for their "execution". It's nonsensical.

I'll give you a perfect example of this silliness. Heading into March 2011, Kentucky was 0-6 in games decided by four points or less during the 2010-11 season. They were outside the Top 20 in the human polls, and the media had a tidy narrative: that Kentucky was just too young and too inexperienced to win close games (the same argument to explain Miami's luck during their big ACC winning streak this season). What happened then? Not only did they reach the Final Four, but three of those four wins were close in the final minute, with two of the wins combining by two points apiece. And in case you think I'm not being consistent about this, go read what I was writing about Kentucky at the time (here and here for two of many examples).

So do Florida's late collapses in several close games matter? Of course they do. Nobody is saying that they don't. What I'm saying, though, is that in five close games, you'd expect Florida to win 2 or 3 of them, and they easily could have won all five. So let's say Florida had gotten lucky and won all five of those close games. They'd be 29-1 and the unanimous #1 in the polls. And what would I be typing here instead? I'd be saying that Florida has had several troubling collapses and has been very lucky to avoid 2 or 3 losses. I'd still be telling you that Florida is only the third or fourth best team in the country, and is overrated by the media.

In short, it's absurd that Florida is viewed more negatively because they've been losing games by 2 or 3 points instead of losing by 15. If they were losing games by 15 then they'd still be the SEC champ and a top contender. But because their losses have been close and unlucky, they "can't win close games" and thus are prevented from making a Tournament run because of unicorns and pixie dust. I'm sorry, it's just not true. Florida remains a legitimate National Title contender, and they are not significantly less likely to win their next tight game than Miami or Florida State or any other team arbitrarily designated "clutch" by the media.

Oh, by the way, this result is why I said Kentucky wasn't done for after the loss to Georgia. An upset here was unlikely, but they pulled it off and finish the regular season 12-6 in SEC play, with wins over Florida, Missouri, Ole Miss and Maryland, along with bad losses to Texas A&M, Alabama and Georgia. They have a 7-9 record against the RPI Top 100, with a Sagarin ELO_SCORE that is sitting near 40th. If the season ended now, they'd more likely than not be a Tournament team. To stay in the Field of 68, they need to win at least one SEC tournament game, and possibly two. In the quarterfinals, they will play the winner of Arkansas vs Vanderbilt.

Purdue 89, Minnesota 73
Purdue has really been playing well the past few weeks. They beat Wisconsin, nearly beat Michigan, and then crushed Minnesota here. Their Pomeroy rating will move up into the 60-65 range with this win, which means that if they were in another conference and had a little bit of luck, they might actually be a bubble team this season. But they're not, so... they're not in at-large contention.

The biggest spark for Purdue the past couple of weeks has been their shooting. They hit 50% of their threes against Wisconsin and 47% here. They finished with an eFG% of 60% in both games. DJ Byrd has been the biggest spark, hitting 6-for-9 behind the arc against Wisconsin and 4-for-8 here. Ronnie and Terone Johnson have also been a lot more effective attacking the rim (that was particularly true against Michigan, where Terone Johnson went for 32 points on 12-for-17 shooting).

This loss doesn't drop Minnesota to the bubble, but it puts them in danger. They finish the regular season 8-10 in Big Ten play and 12-8 against the RPI Top 100, with wins over Indiana, Wisconsin, Michigan State, Iowa, Illinois and Memphis, along with bad losses to Northwestern, Nebraska and Purdue. Their RPI is still 26th and their Sagarin ELO_SCORE should stay inside the Top 30. And the Selection Committee will love their strength of schedule, which was arguably the best in the nation (#2 in Pomeroy and Sagarin, #3 in the RPI). They need to avoid flaming out in the Big Ten tournament first round, though - ending the season on a three game losing streak (with all three losses coming to non-Tournament teams) would be really bad. I think they need one win to keep themselves in the Field of 68. Their first round opponent isn't certain yet, and it actually could end up being a rematch with Purdue.

Speaking of Purdue, even after this win their RPI is still outside the Top 100, though their Sagarin ELO_SCORE will be close to the Top 50. Either way, they need to win the Big Ten tournament to make the NCAA Tournament. Their first round opponent is not yet set.

#16 Saint Louis 78, La Salle 54
La Salle had a huge opportunity here in St. Louis. With a win they might have won the Atlantic Ten regular season title and would have made the argument that the Atlantic Ten's Big Three (VCU, Butler and St. Louis) should be a Big Four. Alas, it was not to be. La Salle couldn't get an open shot all day, and St. Louis did a nice job of limiting three-point opportunities as well. In all, the Explorers finished with only a 42.1 eFG%, and only 0.81 PPP. They shot worse against Charlotte on January 9th (a 35.1 eFG%), but overall it's La Salle's worst offensive efficiency this season. To be fair, St. Louis does that to a lot of teams. La Salle is the 7th team to have their single worst offensive efficiency of this season against St. Louis.

Defensively, La Salle has had success this year eliminating the three (they lead the A-10 in 3PA/FGA and 3P% against this season). St. Louis neutralized this by attacking the rim, which is where the La Salle defense can be pretty soft at times. Rob Loe was 7-for-7 from the field, Dwayne Evans was 7-for-13, and Kwamain Mitchell was 7-for-10 on two-point attempts.

With this win, St. Louis finishes the season 13-3 in the Atlantic Ten with the 1 seed in the A-10 tournament locked up. They are 5-2 against the RPI Top 50, with a Sagarin ELO_SCORE in the Top 25. If the season ended now they'd be looking at a 5 or 6 seed in the NCAA Tournament, with a chance to move up to a 4 or even a 3 if they win the A-10 tournament.

La Salle's final position in the standings is not locked in yet, but they will very likely earn a bye to the A-10 tournament quarterfinals. They finish the regular season 21-8 overall and 11-5 in the A-10, with a 5-7 record against the RPI Top 100. They have a win over Villanova, as well as bad losses to Charlotte, UMass, Xavier and Central Connecticut State. Their RPI is 41st, but their Sagarin ELO_SCORE is narrowly outside the Top 50, and I think their RPI is inflated by the fact that they have only played one team with an RPI worse than 265th, and none outside the Top 300. If the season ended now, La Salle might sneak into the NCAA Tournament, but only barely. They will have to win at least one game, and probably two, in the A-10 tournament to stay in the Field of 68. At this point, we do not know what their draw will be.

No comments: