Tempo And Efficiency, Before And During The NCAA Tournament | Stat Geek Idol
March 22, 2012 - by Howard Hamilton
This is a Sweet 16 submission in our inaugural Stat Geek Idol contest. It was conceived of and written by Howard Hamilton of Soccermetrics. The opinions and statements below are solely those of the author, and do not reflect the thoughts of TeamRankings.
One of the most important contributions of Dean Oliver’s Basketball on Paper has been the concept of offensive and defensive ratings. (As Oliver generously admits, the concept predates his own book in Frank McGuire’s Defensive Basketball.)
Offensive and defensive ratings are the ratio of the average number of points scored or conceded by a team, divided by the number of possessions, then multiplied by 100 to normalize the result. These two metrics along with the number of possessions have become useful measures of a team’s playing style and efficiency.
In this brief study, I seek to answer the following questions:
– How does the efficiency of a NCAA tournament qualifier differ from that of a non-qualifier?
– Is there a difference in regular-season team efficiency among teams that advance to later stages of the NCAA tournament?
How I Looked At Tempo And Efficiency
I used team averages from the last five NCAA basketball regular seasons (2006-07 to 2010-11, 345 DI teams) as well as team statistical data from the last five NCAA basketball tournaments (2007-2011). For the regular season data, the possessions/game, offensive efficiency, and defensive efficiency metrics were precomputed by TeamRankings, but for the NCAA tournament data, I computed the estimated possessions using Oliver’s formula:
- Estimated Possessions = FG_ATT – 1.07*OFF_REB/(OFF_REB + OPP_DEF_REB)*(FG_ATT-FG_MADE) + TURNOVERS + 0.4*FT_ATT
I cluster the regular season tempo and efficiency in four groups:
– Non-NCAA tournament qualifiers
– The entire NCAA tournament field
– Teams that won at most two games in the NCAA tournament (First Round to Sweet 16)
– Teams that won at least three games in the NCAA tournament (Elite Eight to National Champion)
NCAA Tournament Teams Vs. Non-Qualifiers
The group tempo and efficiency averages are displayed in the figure below.
In terms of tempo and efficiency, there is not much difference between the overall tournament field and the rest of Division I. To be sure, there are significant differences between the very worst teams and the elite, and the averaging smears away these differences.
There are fewer differences between those bubble teams that made the tournament and those 15-20 who were on the borderline yet failed to win an at-large bid. Tournament teams do appear to play at a slightly higher tempo than non-tournament teams, at least in the five-year window that I am examining.
When we divide the NCAA field further into teams that won zero to two games in the tournament, there remains little difference in terms of tempo and efficiency. Offensive and defensive efficiencies are roughly the same — if anything, defensive efficiency has degraded slightly! — and the tempo of these teams is almost unchanged. One could say that these averages are being affected by those one-and-done teams, but a separate analysis for those teams shows the following:
Yr Poss/Game O-Eff D-Eff 2007 68.330259 1.004074 1.008390 2008 68.965363 0.989128 0.993327 2009 68.007443 0.976624 0.984731 2010 68.657583 0.974729 0.987496 2011 68.860426 0.978479 0.996361
Elite Teams Have Higher Offensive Efficiencies
When we look at the elite of the NCAA field, we finally start to see some differences in their regular-season performance. The number of possessions per game is lower than non-tournament teams or teams that only advanced as far as the Sweet 16, with the exception of the 2011 tournament.
The really significant number is the team’s offensive efficiency in the regular season. In almost every season, the offensive efficiency of these elite teams — those that advance to the Elite Eight or further — are significantly higher than their peers.
Moreover, the differential between offensive and defensive efficiencies are the most positive of the four groups (the full field combines the latter two groups, of course).
There are some further questions that are motivated by this study.
First of all, what explains the drop in defensive efficiency among the Elite Eight teams in the 2010 tournament?
Do these elite teams, which give indications of their status during the regular season, play at a higher level during the tournament itself?
The following chart, which ranks the offensive efficiencies of the 2007 tournament teams in terms of Z-scores, gives some indication that the elite teams do indeed play at such a level. I’d like to explore this topic further in the future.
(click to enlarge)