Tuesday, February 21, 2012

Is South Florida The Worst 10-4 Team In Big East History?

South Florida 56, Pittsburgh 47
South Florida has had an excellent season - by far their best since joining the Big East. They're a borderline Top 50 team and getting close to the NCAA Tournament bubble. But that said, the fact that they are now 10-4 in Big East play is mind-boggling when you look at some of the stats. South Florida is 1-6 against the RPI Top 50 and 4-7 against the RPI Top 100. They are rated 69th in the Pomeroy ratings. To put that in comparison, Illinois is 4-7 against the RPI Top 50, 6-10 against the RPI Top 100, 63rd in the Pomeroy ratings.... and 5-10 in the Big Ten. It shows you the vast difference in quality between those two conferences this year.

That said, South Florida's won-loss record is only partially due to the overall strength/weakness of the Big East. A big part of it is their schedule within the conference. According to Pomeroy. USF has played the easiest conference schedule so far of any Big East team. They haven't had to play Syracuse, Louisville, West Virginia or Cincinnati yet. And I have bad news: those are their four remaining opponents. I think that South Florida needs to get a split in those two games. If they lose three of four then they'll be 11-7 in Big East play heading into the Big East tournament with work left to do. They'll end up playing in the second round of the Big East tournament against a team like UConn or Seton Hall, and I won't like their chances. Their next game will be tomorrow night at Syracuse.

This loss should finally put to bed the Pitt at-large talk. They had responded to their 8 game losing streak by winning 4 straight. But they've now replied to that winning streak by losing 4 straight. Too many things have gone wrong this year. You can see the frustration on the face of every Pitt player. I never thought I would see a team with a coach as good as Jamie Dixon have a season like this, but I guess anything is possible. They are now 4-11 in Big East play and will play next on Sunday at Louisville.

#14 Baylor 77, Texas 72
The brutal luck for Texas in close games continues. Texas is now 1-8 in games decided by six points or less. I understand that casual fans tend to buy into the mainstream media's post hoc rationalization on clutch play, that somehow elite teams will "find ways to win" close games, but it's all nonsense. The stats say that when games are very tight, the best and worst teams will all win around 50% of the time. Plot Pomeroy Luck factor vs Pomeroy rating - there is no correlation. Last night I recommended this post on luck, which I wrote back in December.

Speaking of Pomeroy Luck factor, it's no surprise that Texas is by far the unluckiest team in the Pomeroy Top 50. The next unluckiest teams, though? Ohio State and Memphis, two other teams that are very much underrated by the media and the human polls. The luckiest team in the Pomeroy Top 50? Baylor, interestingly enough.

What is the future for Texas? I looked for other teams that were borderline Top 25 squads that ended up on the bubble because of horrific luck and came up with two decent comparisons from the past decade. One was the 2006-07 Mississippi State team that went 1-7 in games decided by five or less. Despite being a Top 25 team in Pomeroy, they ended up in the NIT. Another comparison is the 2004-05 NC State team that was rated 20th in Pomeroy but 1-6 in games decided by five points or less. They made the NCAA Tournament as a 10 seed and went to the Sweet 16 (winning a game by 3 points along the way). Obviously Texas will be a serious Sweet 16 threat should they make the NCAA Tournament (and there's a good chance that as a 10 or 11 seed they'll be favored in Vegas against whichever 7 or 6 seed they are matched up against), but they're going to need two more wins to avoid the NIT. They'll need to take care of business at Texas Tech (on Saturday) and at home against Oklahoma (next Wednesday).

This is a huge upset win for Baylor to turn around the momentum of their season. After getting swept by Missouri and Kansas, they were starting to get the reputation as a team that couldn't beat elite opponents. A home loss to Kansas State coming into this was particularly worrying. Now they're back to a comfortable 10-5 in Big 12 play with a couple of easy home games (Oklahoma and Texas Tech) coming up next. They'd need to have a terrible collapse in the final few games to end up with anything worse than a 4 seed in the NCAA Tournament.

Oregon 68, Stanford 64
This game just typified everything Pac-12 for me. Particularly down the stretch, the play was just so sloppy that both teams looked like they were trying to blow the game. Here are the stats for both teams combined in the final 2:45 of regulation: 0-for-6 shooting, 3 turnovers, 3 fouls, 2 points made on free throws. Yuck. Stanford was actually the better team for most of the game, but they couldn't do anything right late in the game. In the final eight minutes they had four turnovers and two made field goals.

And this is more than a demoralizing loss for Stanford. It's also a near-fatal blow for their at-large hopes, which had looked pretty good a month ago. They've lost six of nine and are only 8-7 in Pac-12 play and 7-9 against the RPI Top 200. Their RPI is 102nd and their Sagarin ELO_CHESS is 89th. Hard to see any way that they can get an at-large bid now. They currently sit 7th in the Pac-12 standings. They'll want to finish there or in 6th so that they can stay in the bottom half of the Pac-12 tournament bracket and avoid California, the one legitimately good team in the conference. Their next game will be Thursday at Colorado.

Oregon moves to 10-5 in Pac-12 play with this win. Their Sagarin ELO_CHESS is up to 62nd, but they're only 2-7 against the RPI Top 100. I think that if they can win their final three regular season games, it should be enough to get them on or very close to the bubble. But anything less than that and they're probably playing for the automatic bid or bust in the Pac-12 tournament. Their next game will be Sunday at Oregon State. That will be their toughest remaining regular season game.

23 comments:

Anonymous said...

The odds of South Florida playing a home and home against three of the Big East's bottom four teams AND playing those six games within the first fourteen games of the conference season are about 1 in 320. They've been the recipient of one of the most front-weak imbalanced schedules I've ever seen. They're 9-0 vs the bottom six and 1-4 vs the top 9. Their current conference record is an anomaly. I wouldn't read too much into it.

Anonymous said...

Oh please. Don't give me the "casual fan" nonsense. Stats don't always tell the entire story in sports. It's a game played by humans, not robots. Do you think it's a coincidence that year after year, no team seems to do less with more than Rick Barnes' Longhorns? Last year, they lost in the second round by a point to Arizona. The year before, by a point in the first round to a Wake Forest team that was vastly overseeded as a 9. The most famous example of Texas underachieving was when they started Kevin Durant, DJ Augustin, AJ Abrams and a young Damion James. I remember watching them choke away a game against Tom Izzo's Michigan State team early in the year. That talented Texas team got ousted in the second round by a USC team led by Taj Gibson and Nick Young. Come on. This isn't bad luck, this is Rick Barnes' inability to coach in late game situations. It's plagued his entire coaching career and is the reason why he's never made a Final Four despite being blessed with some of the most talented teams of the last 5-10 years. That's like telling me that Karl Malone had a better career than Tim Duncan, because there's no such thing as clutch play/intangible winning ability, and that Malone was simply the recipient of "bad luck" over and over again. Go ahead and pencil in Texas into your Sweet 16 or Elite 8 as a sleeper. History proves that they don't have bad luck. They're classic underachievers. Stop poring over statistics and start watching basketball before you dismiss the observations of others as "nonsense".

Anonymous said...

Ugh. Totally forgot about Texas' Final Four run back in either '02 or '03.

Apologize for the tone of the last post, let some personal stuff make it seem a little more high-strung/intense than I intended. I don't mean to insult, only to debate. Again, my apologies.

DMoore said...

"elite teams will "find ways to win" close games...it's all nonsense"

I agree with that statement.

However, I also believe that a flawed team won't find ways to win close games.

I can easily see how a number of close losses can be bad luck -- key shots hit or missed can be very random. But I can also see how many close losses might not be luck based at all -- the wrong player on the team frequently taking the shots in the late game, a coach overcoaching and consistently giving the opponent a chance to set up their defense, a team that simply isn't well drilled in late game situations.

I just don't think it's a given that good and bad luck is symmetrical.

Jeff said...

Well Anonymous, you are free to believe that good teams win close games and bad teams don't. You can also believe that the best hitter in baseball is the guy with the most RBIs, even if he has a .700 OPS. But the statistical community has studied this over and over again, in every major sport.

Yes, clutch play exists, but it's a minor factor. In games decided by five or less, luck is the dominant factor.

The facts don't change just because it's a team that you love or (in your case) hate. Go plot Pomeroy Luck vs Pomeroy rating and see how much correlation there is. Do it for yourself.

Jeff said...

One of the silliest things, by the way, is the idea that certain coaches don't know how to coach in the NCAA Tournament. The NCAA Tournament is, by design, a tiny sample size. Weird things will happen.

So you have coaches like Jim Boeheim and Lute Olsen and Bill Self that everybody decides "can't win in the Tournament". They're all "great regular season coaches", who "can't figure out how to win" in the Tournament. Until they do, of course. Then the narrative changes. If Kansas had lost that game to Northern Iowa six years ago, the story would have been Bill Self's failure to win close Tournament games. But now? You didn't hear that at all.

The media builds and changes these narratives all the time. Something is true... until it isn't. It's an ad hoc, random form of analysis with no attempt to understand why anything is happening.

Anonymous said...

How does pomeroy calculate his luck rating?

Jeff said...

It's a Dean Oliver creation. It's the difference between actual record and expected record. You can think of it as looking at baseball standings and taking the Pythagorean record minus the actual record, though the calculation of expected record in Pomeroy/Oliver's ratings is a bit more complicated than the Pythagorean method.

Anonymous said...

How about when you extend that sample size over four years? Texas has had an average rank of 284 in luck over the past 4 years, which I have to imagine is among the worst in the country. You really don't think coaching has anything to do with that? Texas' proven inability to win close games has everything to do with bad luck and nothing to do with how Rick Barnes has prepared his players? Yes, while luck definitely plays a factor, when you extend that sample size over 120 games and that bad luck continues to persist, I start to wonder.

I never said Rick Barnes couldn't coach in the NCAA Tournament. I said he was bad at preparing his teams for late game situations. I also think it's ridiculous to claim that there isn't a difference between regular season play and postseason play.

That's why I think it's silly to continue to seed Texas as a 7. That's basically implying that you think Texas will blast Kansas at the Phog and pick up at least one other good win in the Big 12 Tournament. At this stage, do you really consider those to be good odds?

By the way, I am completely indifferent towards Texas. Just because I think you're severely overrating them and don't like the potshot you took at all people who don't stay current on statistics doesn't mean I hate Texas. I hate that you're saying basketball is a game largely decided by data. At the end of the day, forget computer numbers, forget luck, forget ELO Chess, forget kenpom. It comes down to 10 kids on a court. Some predictive models may be able to give better insight into what will happen than others, but you can't tell me that what it comes down to is luck.

Jeff said...

I have said multiple times that I am dropping Texas to around a 9 seed because they got upset by Baylor. I drop any team after they get upset in a game I don't expect them to lose. Despite what you think, I don't have some irrational love for Texas - I treat them like any other team.


When I have more free time over the summer, without much college basketball to write about, I will put together plots on Pomeroy Luck factor correlation. I can assure you what I will find, which is that there is basically no correlation.


Clutch play exists. To take a baseball example that I know well, the New York Yankees have finished ahead of their "expected" record for about 15 straight years. They have had the best closer in baseball and a whole bunch of clutch hitters, so you'd expect them to do better than random in close games. But how much better than their expected record do they finish each year? Between 3 and 6 games. Out of 162.

And in a typical year, no major league team finishes more than 6 games away from their expected record. The greatest I remember was a wacky season about 7-8 years ago where one team (I think it was the Angels) finished 9 or 10 games ahead of expected. But they regressed right back to the mean the next year.


And to give a hilarious example that people forget now: go read anything about Michael Jordan written between 1988 and 1990. The media consensus was that Jordan was a flashy player who couldn't win clutch games, and was always doomed to fail in the playoffs. The reason? Because he had always flamed out in close playoff games... up to that point. And now he's known as the greatest clutch player ever. Kids now, who only know Jordan through Gatorade & Nike commercials, seem to think that Jordan never missed a game-winning shot.

Anonymous said...

Just explain to me why Rick Barnes' teams continually underachieve. I don't care about your other examples or your MJ stories. Just tell me why luck is deciding Rick Barnes' games and not Rick Barnes.

Jeff said...

I just took a quarter out of my pocket, flipped it 20 times, and got 15 heads. I insist that my coin is unbiased and that I should only have 10 heads and 10 tails.

Explain to me why I'm wrong, and why the coin is actually weighted.

Disgruntled about Texas said...

Jeff, have you ever played basketball? Do you recognize the human element to the game? Do you understand that Rick Barnes is not a coin that can be flipped, but a person who has the inherent HUMAN ability to influence the outcome of a basketball game. Do you seriously think it's just random coincidence that Texas has had horrible luck based on kenpom's own ratings in EACH of the last four years? If you have ever played basketball, or any sport for that matter, for any real period of time, which I doubt you have, you'd understand that you can't quantify the human element of the game. You can't quantify the experience MJ gained by failing in the playoffs before he won. He wasn't the same MJ that lost; his game manifested itself and that coincided with Scottie Pippen truly coming into his own in the '91 season. Stats can't dictate what changed in MJ that year. Stats can't explain why Wilt won 2 titles and Bill Russell won 11 (maybe they can; Wilt only cared about his numbers while Russell only cared about winning; or maybe Russell was just "lucky"). Stats can't explain why Karl Malone never won a ring despite being a great player on perennially great teams throughout his career. People's actions can't be decided by the flip of a coin. Not in sports, not in any facet of life.

For clarity's sake (because I'm the guy who had the initial gripe with Texas, not the most recent anonymous) I will now be posting under the moniker "Disgruntled about Texas".

Jeff said...

You are free to believe that. It's the "I don't care how much evidence you have saying evolution is happening, I know in my heart that God created Earth 5700 years ago" argument.

I'm not going to tell you what to think. But this is my website and I get to believe in whatever I want, and I choose to study advanced metrics and apply them as best I can.

Jeff said...

Hm. So if he was better than .500 early in his career but worse than .500 in the past four years, does that mean that he suddenly got really bad at coaching close games?

I would argue it's Apophenia.

Disgruntled about Texas said...

That application process isn't working all that well in terms of selecting the field, considering your current standing in the Bracket Matrix over the last few years is pretty lackluster.

Jeff said...

Once it gets to Selection Sunday, we're all just guessing whether the random biases of the Selection Committee will make a team a 7 seed or 8 seed. You'll notice how large the variance is for everybody - we've all been above average and below average in different years.

What this website specializes in is projecting months ahead of time what the bracket will be, rather than producing a "if the season ended now" snapshot. There's obviously nothing much to project once Selection Sunday rolls around.

Anonymous said...

There's at least 2 "Anonymous's" left. We will not be defeated.

Anonymous said...

It's pretty much an accepted fact that Rick Barnes is a great recruiter but a very so-so coach. That has nothing to do with "luck". His track record bears that out.

Jeff said...

I totally agree that Barnes is a great recruiter and a mediocre in-game coach.

And... apparently he's won 49.5% of close games over his entire career. About as close to 50-50 as possible.

Anonymous said...

If you had your way all games that were tied after regulation would end by coin flip instead of playing overtime.

Jeff said...

Nope. It's the randomness in sports that makes it so fun. Why do so many more people love the NCAA Tournament than the NBA Playoffs? Because in the NBA playoffs, with a best-of-7 series, the best teams tend to grind their way to the title. In the NCAA Tournament, every night could be your last, and teams like UConn, Butler and VCU can make miracle runs to the Final Four.

It's boring when the best teams always win. Understanding the randomness of sports makes it more enjoyable, not less.

Anonymous said...

Duh. That's why we watch.