Sunday, February 19, 2012

San Diego State's Luck Is Evening Out

Air Force 58, #15 San Diego State 56
One thing I talked about yesterday with regards to New Mexico was the fact that over time, team resumes will tend toward team quality. If a team has a Sagarin ELO_CHESS (i.e. their won-loss resume) and a Sagarin PREDICTOR (i.e. the team quality) that are way out of whack early in the season, they will get closer to each. And much more often than not, it's the ELO_CHESS that catches up to the PREDICTOR. Luck will only carry a team so far. And while New Mexico was a team that the computer loved but that just didn't have the results to back it up, San Diego State was the opposite. They had a slew of tremendous wins that put them into the Top 15 in the polls, even though the computer hated them.

To be fair, I feel like Pomeroy and the Sagarin PREDICTOR somewhat underrate San Diego State. The Aztecs are a young team that has improved throughout the season, and the computer ratings don't really have a way to take that into account. But there's no question that it was ridiculous for them to be ranked as high as 12th. At one point this season they were 8-1 in games decided by five points or less, which meant that they were tremendously lucky in close games. And as I always say, luck evens out - there is almost no correlation between past and future performance in tight games. And lo and behold, San Diego State has lost their last two games decided by five points or less, including this one.

It's always seemed so silly to me that pollsters and media analysts will move teams up in their ratings for narrow wins over bad teams, but drop them like a rock for losing close games to elite opponents. You have to play much better to narrowly lose to a good team than to narrowly beat a bad team. San Diego State is a perfect demonstration. Three weeks ago they beat Boise State by two points and beat TCU by ten. The pollsters moved them up by inertia - they moved from 17th to 13th in the AP Poll. The Pomeroy ratings, however, dropped them from 53rd to 59th. But now that San Diego State has lost by two points to a bad Air Force team? They'll plummet in the human polls. Pomeroy only dropped them from 57th to 61st. Pomeroy simply didn't see a big difference between losing by two to Air Force and beating Boise State by two.

It's been a tremendous week for Air Force, knocking off both Wyoming and San Diego State. They play a slow, defensive style that grinds on teams. And with two straight wins they have pulled themselves out of the Mountain West cellar and into a tie with Boise State. They'll play TCU on Wednesday. San Diego State will play Wyoming and Colorado State at home next week. They should at least split those games, but if they don't then it's not inconceivable that San Diego State bubble talk could start up.

Nebraska 80, Illinois 57
It's been a brutal two weeks for Illinois basketball, and it bottomed out with this debacle against Nebraska. Nebraska's Bo Spencer, who is quietly one of the better guards in the Big Ten, had 27 points on 9-for-11 shooting. Nebraska as a team had a 64.8 eFG%, their best shooting in Big Ten play this year. Illinois was an ice cold 6-for-23 behind the arc, finishing with 0.83 PPP. To put that offensive futility in perspective, the only teams this season to score worse than 0.83 PPP against Nebraska are Alcorn State and South Dakota.

Illinois is now 5-9 in Big Ten play with a road game at Ohio State looming (they also will have to close the regular season at Wisconsin). They've lost five straight, and things are not looking good. But my one word of calm to panicking Illinois fans is that we all tend to overrate teams when they're hot and underrate teams when they're cold. Every team goes through streaks (though it seems like Illinois has been an awfully streaky program throughout the entire Bruce Weber era). I think they'll improve and play some better basketball in their next two outings. But the question will be if it's too late. If they fall below 8-10 then they're going to need multiple wins in the Big Ten tournament as well as a lot of help from a very weak bubble. And remember that at this point in the year the conference tournament seedings are starting to take shape. Illinois is looking very likely to be stuck, should they win a Big Ten tournament game, playing either Ohio State or Michigan State in the Big Ten tournament quarterfinals.

Nebraska is now 4-10 in Big Ten play. And to be fair, they're not as bad as most people think. Their Sagarin ELO_CHESS moved up to 104th with this win, which means that their resume rates as stronger than either Villanova or Butler, among others. They have a brutal finishing stretch to the regular season, though, so it's almost impossible to see them finishing higher than 11th in Big Ten play. Either Iowa or Nebraska will end up in last place in the Big Ten. Their next game will be on Wednesday at Purdue.

Cincinnati 62, Seton Hall 57
Seton Hall has struggled to shoot the ball all season long, and it ended up costing them this battle of bubble teams. They had a 37.8 eFG%, the sixth time in 15 Big East games that they've had an eFG% of 41.0 or worse. They're dead last in the Big East in 2P% shooting in conference play (42%). Cincinnati is not a spectacular offensive team, but they take care of the ball and have enough weapons to score a decent amount. Cincinnati had 1.01 PPP here, and their 1.03 PPP average in Big East play is tied for fourth best.

Cincinnati is 9-5 in Big East play, which in many other years would mean that they'd already clinched a spot in the NCAA Tournament. But this year, they still have plenty of work left to do. They are only 5-5 against the RPI Top 100, with three RPI 100+ losses (including that awful loss to Presbyterian). Their Sagarin ELO_CHESS is 54th. If the season ended now, I think Cincinnati would be a Tournament team, but they'd be one of the last teams in the field. In my opinion, they will need at least two more wins before Selection Sunday, and probably three, to earn their spot in the NCAA Tournament. Their next game will be Thursday night against Lousiville.

Seton Hall is much more popular with the computers than Cincinnati is. Their Sagarin ELO_CHESS is 40th and their RPI is 34th. The problem is that they're only 7-8 in the Big East, and only 6-8 against the RPI Top 100. I think they need to get to 9-9 in conference play, and then need to win at least one game in the Big East tournament to earn an at-large bid. They will play Georgetown on Tuesday, which will be a great chance for a big scalp. But even if they can't close the deal there, they'll get to finish the regular season against Rutgers and DePaul.

2 comments:

Mike said...

Nice writeups. I think you meant to write either Nebraska or Penn St will finish last in the Big Ten. Not Iowa

Jeff said...

Yes, you are right. Sorry for the error.