Friday, January 21, 2011

Conference Differentials, Part 11/01/21

About 30% of the Has Been Completed
All of the teams in the conference have between four and six games in the record books, so an early look at offensive and defensive differentials is in order. The offensive and defensive efficiencies measures the number of points a team scores (offensive) or allows it's opponents to score (defensive) per possession. I have taken the (raw) offensive and defensive ratings from Ken Pomeroy's site for each of the Big East teams, extracted the conference games only, and averaged their efficiencies for conference games only. Any team's efficiency differential is simply the team's offensive points per possession (ppp) minus their defensive points per possession. The differential is a shorthand way to measure the team's offense versus it's defense. A team with a winning record usually has a positive differential, there can be exceptions, often worth investigating. The table below reflects conference games through January 19, and for each team shows their Big East record, and offensive & defensive efficiencies and their differential. I have also included the team's Pythagorean Winning Percentage (again conference games only). Like the differential, the Pythagorean often restates the Win/Loss percentage, but there are (more) often exceptions worth reflecting on.

TeamWLPct.OffenDefenDiff.Win %
West Virginia320.6001.
Notre Dame430.5711.001.03-0.030.426
St. John's430.5711.001.04-0.040.378
South Florida150.1671.031.13-0.100.268
Seton Hall250.2860.921.02-0.100.227

I have sorted the data by efficiency differential. The conference divided evenly into groups of eight (an upper and lower division?) based on positive and negative differentials -- if the differential is positive, the team's offense scores more points per possession than it allows. The undefeateds and winless behave, sorting to the top and bottom of the data set as expected. Things in between however get a bit murky. If sorted by won/loss percentage, Syracuse would appear after top ranked Pittsburg, followed by Syracuse, with Villanova and Louisville clustered (1/2 game) behind. In the next group however any semblance to won/loss order breaks down. Notre Dame and St. John's fall into the lower half of the table, but have won/loss records that suggest they ought to be ranked higher. Are these two "over performing" relative to the conference at this point and due to settle downward? Syracuse's differential was "low" compared to their Pythagorean Winning Percentage and their won/loss percentage. Has the Orange benefited from an easy section of their conference schedule? I decided to create a back-of-the-napkin SOS to determine what (if any) impact the strength of schedule may have had on any school's record. The table below shed some light on why some of the team's are performing as they are.

Notre Dame0.5680.193
Seton Hall0.5370.300
South Florida0.5350.281
St. John's0.4790.214
West Virginia0.3700.278

I also took the standard deviation of each schedule to get a sense of the "consistency" of the Pythagorean Winning Percentage. A high standard deviation would suggest there was a large degree of dfference (less consistency) between a team's opponents. I left this list in alphabetical order to make it easier for those interested in a particular school to locate and find the team's information. Notre Dame's schedule so far has been "difficult", and the standard deviation would suggest slate has been uniformly difficult. The low (relative to Big East teams with a similar record) PWP reflects the degree of difficulty of their schedule. Color me sceptical when the topic turns to a Notre Dame resurgence this season, but the Irish have a 4-3 record after playing (on the road) Marquette, Syracuse and St. John's (twice) to go with Connecticut, Cincinnati and Georgetown (at home).

Syracuse's schedule appears to have been a bit lighter than the "average" for the conference, Providence, Seton Hall have struggled. Notre Dame and St. John's have above 0.500 records, but have not established strong winning margins in those victories.

Villanova has an "average" schedule with a typical mix of teams, Rutgers and South Florida, conbined with Connecticut, Louisville and Cincinnati.

No comments: