Shot Attempt Difference
01.27.05
The stats page has been updated through Wednesday's action. The bonus stat is Shot Attempt Difference. It represents the number of additional shots a team gets per 100 possessions as compared to its opponent. It's essentially a measure of how much of an advantage a team has in the turnover and rebounding departments.
A shot is defined by FGA + .42 x FTA, so possessions spent at the free throw line are also included.
I haven't had the time to scrutinize the numbers, but the top of the SAD list has a diverse group of teams. Houston's big advantage in shot attempts comes entirely from the three point line, where they make less than a third of their shots. Hence, their offensive efficiency is very low. Arizona has fewer turnovers than the opposition and has a rebounding advantage, but not as much in either category as Kentucky. Air Force and Virginia Tech are poor rebounding teams, but benefit from the turnover difference. We saw last week that Temple and Illinois were the best in the nation at taking care of the basketball, and not surprisingly, they fare well in this metric also.
Notable teams that shoot fewer shots than their opponents: Kansas, Georgia Tech and Utah.
I'm taking a long weekend, so see you Tuesday.
Possession Stats 101
01.26.05
If you ever hear the level of success of a team's offense or defense described in terms of points scored or allowed per game, you should ignore it. Among stats published by the NCAA, field goal percentage is much more valuable. But among stats not endorsed by the NCAA, points per possession is better still. A couple of things I saw yesterday used points per game in a misleading way to make a point.
Example 1) Tuesday night at halftime of the Illinois/Wisconsin game, a stat was posted in defense of Doug Gottlieb's claim that Wisconsin was the most overrated team in the Big Ten. It compared the points per game they have given up this season (60.5) to last season (57.2). (If memory serves, the numbers used were during Big Ten play only, where the gap was larger than the overall numbers I am using.)
To which the viewer was to draw the conclusion that the Badgers' D is much worse than last season. But a little research indicates that Badger games are averaging five more possessions than last season. The pace of the games allows the opponent five more opportunities to score, through no fault of Bo Ryan's defense. In raw terms, the Badgers were giving up 90.7 points per 100 possessions before last weekend's action. Last season, they gave up 92.3 points per 100 possessions. The 90.7 will increase during Big Ten play, but overall there isn't much difference between their defensive play in 2004 and 2005.
Example 2) On Monday night, Utah came within one made field goal of tying the NCAA single-game record for field goal percentage. They made 24 of 30 shots (80%) against Air Force in a 63-51 win. The wire story in your morning paper on Tuesday stated this:
The Utes shot a school-record 80 percent from the field (24-of-30) against the nation's No. 1 scoring defense.
To which the reader was to conclude that not only was Utah red hot, but amazingly they accomplished this against the best defense in the nation. If you see an Air Force game on TV, this is the "go-to" stat. One of the announcers will inevitably boast about how it's very difficult to score against the Falcons. And make no mistake, they are good, but not great. The pace of Air Force games is so slow that opposing offenses don't get as many opportunities to score as they do against conventional teams.
The very short Air Force team ranked 87th nationally in defensive efficiency before the weekend, allowing 95.8 points per 100 possessions. Given their lack of size, that's still an impressive ranking. I am a little surprised that they are ranked that high this season, because they have given up some huge FG% numbers to the high profile teams they have played - Georgia Tech (60%), Iowa (68%), and Utah (80%).
So folks, next time you see someone back up an opinion using a team's points per game, keep in mind that it's not very useful in assessing a team's effectiveness.
Rules Experiment Report: ‘05 Edition
01.25.05
Today, I bring you an exciting presentation about rules. (For those of you looking for hype on the Illinois/Wisconsin game, you should know the drill by now. Big Ten Wonk will not only have coverage, but coverage of the coverage, along with a steady diet of acronyms, all wrapped in a college-level vocabulary.)
The data is in, and in what should not be a surprise to anyone, the rules experiments for the exempt games of the 2005 season were once again a failure. The numbers for games through December 31st, which include all exempt games for this season, speak for themselves:
Exempt Regular
Games Games
(169) (1438)
Poss/40 min 69.1 69.5
Points/Poss .969 .986
Points/40 min 133.9 137.1
3P FG% 33.7 34.8
2P FG% 47.3 48.0
3PA/FGA (%) 31.7 32.8
Fouls/Poss .272 .266
As a review, the experimental rules for this season included an expanded lane, a three point line at 20' 6", and a silver-dollar sized "charge-free" arc under each hoop. The results are similar to what happened with the experimental rules for exempt games during the 2004 season. The disparity in offensive production between the exempt and regular games is less than last season. But that makes sense, because the larger international lane was used for the exempt games last season.
I expanded on the reasons for a reduction in offensive production under the experimental rules in this post over the summer. It's simple really, the expanded lane forces offensive players further away from the basket, thereby producing a reduction in two point field goal percentage. Extending the three point arc naturally reduces three point accuracy. In addition, the game slows down as the offense works longer because it is more difficult to find an acceptable shot.
Putting this in the context of the current trend in college hoops of declining pace and scoring (2005 is even slower and more low-scoring than 2004), and it makes it obvious that implementation of these rules would be foolish.