Saturday, February 20, 2010

Big East Differentials -- Two Weeks Out

The Composite With 2 Weeks Left...
I have continued to scrape Ken Pomeroy's web site on a regular basis (many thanks to KP for the raw data on D1 ball...his renewed blog posts are a terrific addition to the analysis this season) for his possession-based, game-by-game, results. I thought, with about 74% of the conference schedule in the books, I would revisit the conference-only games and take another look at the state of the Big East at this point...

RecordPoints Per Poss.
West Virginia940.6921.141.000.14
Notre Dame680.4291.131.130.00
Seton Hall670.4621.051.09-0.04
South Florida670.4621.031.08-0.05
St. John's490.3080.951.02-0.07
Std Dev.

The six tiers that were obvious in the earlier part of the conference season have largely rounded out into a curve as teams, both high and low in the standings, are beating each other up. The St. John's at South Florida game earlier this afternoon should have (according to Ken Pomeroy's log5 calculation) been a six point loss for the Johnnies, not the 16 point win that will put them one loss out of a four-way tie for 9th place. The separation between teams #1 through #7 are well separated from teams #8 through #16. Notre Dame, despite the 0.00 differential, is actually slightly to the negative (hardly surprising anyone who has followed the Irish for the past four seasons), about -0.003. Interesting because that indicates that teams with positive differentials (offense scores more points per possession than the defense yields) -- teams #1 through #7 -- have records >0.500, while teams that have negative differentials (defense gives up more points per possession than the offense scores) have records < 0.500. It is fairly common for a conference to have a team (or two) that has a positive differential and a winning percentage < 0.500, or a team with a negative differential and a winning percentage &ft; 0.500.West Virginia and Marquette continue to be the outliers of the upper division, as both have fewer wins than their differentials suggest their play should have earned them. According to Pythagoras, both are short two wins apiece. The Mountaineers' game against Seton Hall will probably shave their differential a bit. Syracuse along with Villanova have spent February jockeying for the spot below the Mountaineers. The Wildcats are into the more difficult stretch of their schedule, and have had difficulty getting ahead of the Orange. DePaul and Rutgers have engaged in a similar, spirited competition, albeit at the other end of the conference standing. Providence has only recently fallen into position #14, the end result of a stretch in their (February) conference schedule that includes Syracuse, Marquette, Georgetown, Villanova and West Virginia.

According to Pythagoras...
Using the differentials to project a final record is hardly reliable, but for a number of teams, seem interesting. According to the projection for example Georgetown and Marquette will close out their conference slates without dropping another game. For the Hoyas that means a 4-0 run, while the Golden Eagles would finish with a 5-0 run (a 9-1 run if you include the five previous games) -- that should be warrant an invitation to the NCAA. Connecticut and Notre Dame would finish with 3-1 runs (actually for UConn, given their win on Saturday, a 2-1 finish), while Louisville would finish with a 4-1 run (3-1 counting the DePaul win Saturday afternoon). For the Cardinals that may be a very tall order as they still have Connecticut, Syracuse, Marquette and Georgetown. That has to be one of the toughest finishing slates (though Villanova's Pittsburgh, Syracuse and West Virginia run might be a close second).

No comments: