Friday, February 6, 2009

Updating the Big East Aerial

I'm still in the throes of an amazingly hectic schedule, but I thought I'd post some non-recap content to prove I'm still a contributing member of society.

First, some housekeeping:
  • Most of the stats on the upper right sidebar are now updated (all except individual net points, which isn't really conveying what I'd like, so I'll let that sit).
  • The Big East HD box score project has obviously fallen by the wayside. I'll try to get it rolling again this weekend, but I can't make any promises as I don't know what my schedule will look like more than a couple of days ahead.
  • I have promised some additional functionality in my play-by-play parsing program, but that will likely have to wait until the off-season due to my current amount of free time.
-------------------------------------------------------------------------------------------------

About a month ago I trotted out an aerial for all Big East teams, to take a look at where everyone ranked as the conference schedule got underway. I thought I'd update that graph with stats through yesterday (Feb. 5th) [click on graph to enlarge].


This graph, like most I produce, will take some explanation.
  • Adj. offensive and defensive efficiencies come from KenPom.com, and are through games played 5-Feb-09. These are the large circles that are also color coded. The old off. and def. efficiencies (from 31-Dec-08) are the small blue circles.
  • The team labels are next to their current marker, while the thin black line connects the label to the previous position. This allows one to judge the change over the last 5 weeks - the Hoyas have taken a precipitous decline in both offense and defense, Marquette had improved its offense while maintaining its defense, West Virginia had fallen off defensively while its offense has shown only a slight decrease.
  • A couple of teams have their old positions obscured: Syracuse from 12/31 is under G'town's current spot, and Rutgers from 12/31 is under S. Florida's current spot.
  • Note that the ordinate axis (defensive efficiency) is scaled backwards, from high to low, so that better defensive teams are at the top of the graph
  • I've color-coded the current team markers, from black to yellow, indicating the nominal Pythagorean Rating, using Rating = AOEff^11.5 / (AOEff^11.5 + ADEff^11.5)
  • The solid grey line is a sort of 1:1 line, indicating where a team would be equally capable at offense and defense. Teams to the left or above the line are better defensively, while teams below or to the right are better offensively.
  • The diagonal grey dotted lines are isopleths. Teams that are the approximately same distance from a line (e.g. Marquette and Villanova) should be equally good, although they may do so by different means (here, Marquette by offense, Villanova is equally good on both ends).

An important consideration here is that the isopleths can be a bit deceiving, since the scaling there is not linear - the difference in "goodness" of teams between the two upper right isopleths is much smaller than the difference between the two lower left isopleths. This is a result of the way the Pythagorean rankings are calculated. So, the most improved team over the last 5 weeks is S. Florida, rather than Marquette.

Here's a table for all teams' change in the ratings between 31-Dec and 5-Feb, in Pythagorean change (remember, Pomeroy's ratings scale 0-1):
Team            Rating    Change
South Florida 0.652 0.129
Marquette 0.935 0.030
Providence 0.793 0.007
Seton Hall 0.714 0.003
Villanova 0.941 0.003
Connecticut 0.977 0.002
Louisville 0.957 -0.008
Pittsburgh 0.969 -0.008
West Virginia 0.956 -0.029
Syracuse 0.905 -0.034
Rutgers 0.592 -0.044
St. John's 0.644 -0.054
Georgetown 0.936 -0.055
Notre Dame 0.860 -0.064
Cincinnati 0.798 -0.088
DePaul 0.430 -0.173

From the table, we can see that a number of teams have been holding steady the last 5 weeks (Providence, Louisville, etc.) with only S. Florida and Marquette making strong improvements. A number of teams have declined since December, and Georgetown's huge drop on the graph isn't nearly as severe in real terms (e.g. Notre Dame has declined more). DePaul is serving as a good example of a team where the bottom really has fallen out.

3 comments:

  1. This is incredibly awesome!

    Do you mind if we showcase this on our blog (http://www.bigeasthoops.com)?

    I especially like the yellow and blue balls. Wouldn't it be neat to have this for the whole season (perhaps plotted in lines where the intensity of the color gets stronger as dates become more recent)?

    ReplyDelete
  2. Feel free to reproduce what you'd like - just add a credit or link back, please.

    I'd love to track this in more detail as the season progresses, but unfortunately Pomeroy doesn't archive his ratings as far as I know.

    ReplyDelete
  3. No Pomeroy does not archive his ratings, but he does make the dailies available as .csv files. I have been using them for the Villanova previews, etc. for a while now.

    ReplyDelete