Ivy League Efficiency Margins: 2011 Week One (Updated with SOS Adjustments)

By John Ezekowitz

Last year, I tracked Efficiency Margins during the Ivy League basketball season. The stat is simplistic in concept, yet it can reveal quite a bit about team performance that is not evident from simply looking at records. For those who may not remember, Efficiency Margins simply track three things: points scored, points allowed, and possessions. A team that wins a game 66-60 over 60 possessions is said to have an EM of .1 for that game (1.1 points per possession – 1 point per possession allowed). On the flip side, the team that lost would have an EM of -.1.

John Gasaway at Basketball Prospectus has done a huge amount to popularize this stat through his Tuesday Truths column (definitely worth a click through), but understandably focuses on the top conferences. This year, however, the Ivy is deeper than it has been since at least 2001-2002, and ranks 15th in Ken Pomeroy’s conference ratings (14th in the RPI). So what do the EMs say about the first two weeks of Ivy play?

Unsurprisingly, Harvard leads the way here. They have the second best offense and by far the best defense in conference play. To be fair, however, these numbers may overstate any advantage Harvard may have over Princeton. Harvard has played three of its four games against the two worst teams in the league so far: Cornell and Dartmouth.

Even just two weeks into the Ivy season, it looks like the conference is shaping up into a two-horse race. While Penn is undefeated, their EM through two games does not compare to Harvard’s or Princeton’s (that 80-78 win over Brown in 81 possessions did not help). This may change further into the season, but right now Penn is firmly in the 2nd tier.

Finally, these numbers show how much the league has changed from last year. Last year, Cornell posted extremely good efficiency numbers, while Dartmouth was historically bad in conference. This year, there is less spread within the middle of the very competitive league, and Dartmouth, while last, has improved by .1 PPP so far.

Be sure to check back in every week as the Fourteen Game Tournament progresses for further updates.

(UPDATE: I was able to adjust for strength of schedule and home/away. See below):

This year, Ken Pomeroy has been keeping track of his predictions on his FanMatch pages. This has allowed me to estimate Expected Efficiency Margins using the best possible data: pythagorean expectation at the time of the game. Below is a table of the expected performance for Ivy League teams: 

As you can see, the big movers are Columbia and Cornell (in opposite directions). Cornell’s poor non-conference win-loss record hid the fact that the Big Red actually lost quite a few very close games. On the other hand, Columbia has a fantastic record in close games this year. Call it luck or whatever you want, but so far the Lions are vastly exceeding expectations in conference play. The 20-point drubbing of Dartmouth on Saturday certainly helped.

Finally, here is a helpful plot of the expected vs. actual EM’s so far. The line represents playing exactly to expectations. Teams above the line have outperformed expectations, while teams below it have underperformed.

This entry was posted in NCAA Basketball. Bookmark the permalink.

10 Responses to Ivy League Efficiency Margins: 2011 Week One (Updated with SOS Adjustments)

  1. Tarek Kamil says:

    Actually I think the EM in the example is 0.167 (70/60=1.167 – 60/60=1.000)

    • jezekowitz says:

      Very interesting. Is that just the average of the opponent pythag expectations for the SOS calculation? Home-Road adjustment is 3.5 points.

      • Brian says:

        I haven’t done one of those in a while, but I recall that it’s the full-blown Pyth. calc, using averages of KenPom’s adjusted off. and def. efficiencies , corrected for venue (+/- 1.4%).

        I think last year’s post is a bit more explicit in explaining the process.

        I was planning on looking again later in the week, now that half of the BE schedule is complete, and would probably be more lucid after that. VBTN has more on this as well.

  2. Pingback: Tweets that mention Ivy League Efficiency Margins: 2011 Week One | Harvard Sports Analysis Collective -- Topsy.com

  3. Brian says:

    I think your plot looks great – it speaks very well for Ken’s predictions. I actually call your stat [actual eff. margin – expected eff. margin] Performance. Dan Hanner is about the only other basketball blogger to notice, though.

    I also like that you’ve used his actual team stats on the day of each game. That’s too much work for me, so I just poll his website on today’s stats, and apply them retroactively.

    Now that I think about it, the Ivy League schedule makes your work a lot easier than for me, dealing with the monstrous Big East and with conference games at least 5 days a week.

    Now I’m just bitter.

  4. Pingback: Ivy League Adjusted Efficiency Margins: Week Two | Harvard Sports Analysis Collective

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s