Friday, March 4, 2011

Defensive Score Sheet: Oakland vs. Southern Utah

In two meetings this season, Oakland held Southern Utah to an average of just 0.898 points per possession, significantly below the Thunderbirds' conference average of 1.056. Oakland used this advantage to win both games by double digits, even after shooting just 39% in the contest a week ago. The match in Cedar City, then, was marked by a solid defensive effort by both teams. Though we are able to understand team defensive efficiency fairly well from box score data, there has never been a sure-fire way of doing the same for individuals. However, a Defensive Score Sheet, introduced by Dean Oliver in his seminal book Basketball On Paper, can help fill this gap.

The Defensive Score Sheet is basically a way of tabulating what a player does on the defensive end of the court. The recorder marks down each time a player allowed a field goal, forced a miss, committed a foul leading to a miss or make, or forced a turnover. To see which players were most effective in that defensive meeting a week ago, I created one of these sheets after I watched a replay of the game. Assigning credit and blame for defensive stops or lack thereof is a fairly subjective matter, but for the time being let's assume my results are more or less consistent with reality. To see what this looks like, let's go to Will Hudson's line (click to enlarge)
Hudson was very engaged on the defensive end of the court. Southern Utah rarely takes three-pointers (and is in fact last in the nation in percentage of shots taken from beyond the arc) and often goes to its plethora of big men for scoring opportunities. As a result, Hudson faced a lot of scoring possessions - and he did fairly well in denying them points. Now the next stage of the analysis utilized by Oliver is much more complex. It involves a series of detailed calculations that include everything from these raw numbers to items like defensive field goal percentage and rebounding percentage. I first learned about this through the work of David Hess who has been doing this for Kansas all season long. For a brief introduction into some of the terms, you can check out Hess' outline here. I've read the work of Hess and Oliver's chapter on individual defensive ratings many times over, and a lot of it is still over my head. But after working through the calculations a few times, I feel comfortable enough posting the results. Please note any errors are my own.

(I also have to give special thanks to J. Leverenz of the Kentucky blog A Sea Of Blue. He has been doing the Defensive Score Sheets for UK this season and was kind enough to pass along some of his tools and tips for this project. While the tables might appear nice and orderly, a lot of work goes into making these, so many thanks to him for helping make it a bit easier.)
I'm leaving a bit out of the tables for now, but these three columns represent well the areas from which we can draw some conclusions. Thoughts on the Oakland side:
  • The lower the defensive rating, the better. Larry Wright comes out with the best defensive rating from this game. He managed to stop 84% of plays against him while he was in the game whether by forcing a miss, a turnover, or grabbing a defensive board. Drew Valentine, who had four steals in this game, also comes out with a stellar defensive rating despite the fact that he only accounted for 10% of team defensive possessions while he was in the game. Either Southern Utah didn't try to attack him, or he did a great job of denying his player a shot attempt (which then wouldn't register on this chart).
  • I highlighted Keith Benson's numbers across the board for a reason. While he was in the game, he faced a team high 24% of SUU's scoring attack, and he stopped them 63% of the time. The average we would expect a player to face is 20% (100% divided by five players on defensive), so Kito's mark is likely higher for two reasons: 1) SUU went to its big men a lot, and 2) Kito is an eager help-defender. His stop percentage is likely affected by the subjective nature of this task. For example, Matt Massey hit a few 15-foot jumpers that Kito did not directly contend, but that is because his role in the defense is to guard the basket. So how does one assign blame for such makes? That is why this is such a tricky experience. Still, Benson ends up with a great defensive rating.
  • Everyone else on the team posted a solid defensive rating. Will Hudson, despite his great looking raw numbers, ends up with a 91.5 rating. This is still a great rating, but it lags behind his teammates. This is where the math involved in calculating these percentages goes over my head as one would think Hudson's rating would have been better because of his strong raw numbers shown above. I think this results from the fact that Hudson is involved on so many defensive plays, including those where he runs out to contest a shot from another player's man. Also, Hudson took the brunt of SUU's offensive attack early on as he guarded two guys, Kyle Davis and Ramell Taylor, who hit a few fading jumpers that were well contested. This is perhaps where the numbers do the defender a disservice.
  • Finally, while Ilija was in the game it appears as if Southern Utah went at him quite a bit - 29% of the time. He had a pretty weak stop percentage and defensive rating as a result.
Now let's turn to Southern Utah's defensive percentages.
  • As expected, SUU's starting frontcourt duo, Kyle Davis and Matt Massey, faced the brunt of the offensive attack by Oakland. They had to contend with both Keith Benson and Will Hudson, to mixed results. Davis does a decent job while he's on the court by stopping 55% of defensive scoring possessions. Massey, on the other hand, struggled to get stops. I can also tell you that his figures were hurt by the fact that his fouls lead to four made free throws versus two misses, and he also failed to record a defensive rebound.
  • Tyson Koehler held his ground in the paint for short spurts, but when scoring opportunities abounded for the Golden Grizzlies, he largely came up short.
  • Jackson Stevenett, an underrated forward, is a bit of a do-everything man for the Thunderbirds. He ends up looking pretty solid by this standard, especially considering the amount of minutes he played.
  • Finally, Damon Heuir did a fine job in his limited minutes. He did not allow a single point to be scored against him, stopping two plays and helping on another. He also had a defensive rebound and a forced turnover. Because of this, Heuir is rewarded with a low defensive rating.
The defensive rating charts tell us that this game was very much played in the paint, and the big men were very active on the defensive end as a result. Additionally, the numbers show that Larry Wright did a particularly great job on the defensive end, while SUU received nice defensive contributions from role players in short minutes. It is very likely that the quarterfinal match on Saturday will again feature a lot of scoring opportunities for the big men. Knowing what we know of last weekend's meeting, it will be interesting to monitor which side ends up with the defensive advantage. If time and video permits, I will attempt to track these totals again for the rematch.

No comments:

Post a Comment