Thursday, January 25, 2007

Clemson Preview

Tonight's game presents a potential matchup problem for the Devils. Duke's turnover problems are well documented, and Clemson has forced opponents to turn the ball over an astounding 28.46% of all possessions in ACC play. When Clemson has the ball, they're much more careful (17.73 TO%). Clemson likes to press, but they're also good at forcing turnovers in the half court. This will likely be the bigger problem for Duke - any Blue Devil (except Zoubek and maybe Thomas) can bring the ball up against the press. The key for Duke will be half court execution - Clemson's numbers suggest they play a gambling defense - while their turnover numbers are very high, they have the worst field goal defense in the conference (.579 EFG%) in conference play.

On offense, Clemson can present some matchup problems. Booker and Mays are both a little undersized, but extremely athletic, and they relentlessly attack the glass. As a team, Clemson is very good on the offensive glass (39.33% in conference play) and will put Duke's defensive rebounding - which has been absolutely superb so far - to the test. Mays has turnover problems that Duke may be able to take advantage of, particularly since McClure and Thomas (both good at forcing turnovers) will be guarding him. On the perimeter, Hamilton (12th in 3pt%), Hammonds (5th in 2pt%), and Rivers (4th in 3pt%) are all good shooters, and it's likely that the Tigers will try to exploit the matchups with Paulus, since he'll be at a size disadvantage on whoever he's guarding.

When Duke is on offense, the key is to maximize possessions with scoring opportunities. For a team with a shot-blocker as good as Booker down low, Clemson's interior D has been dreadful. They're not much better on the perimeter, letting opponents hit 41.8% from outside. If Duke can A) hold on to the ball reasonably well (20-23 turnover %) and B) hit the offensive glass (33-38 offensive rebounding%) they should be able to score a lot of points - enough to make up for the handful of easy baskets Clemson is bound to get off turnovers.

Around the ACC

Wake Forest hung in there against UNC for the first 26 minutes, trailing by just 7 after an Ishmael Smith layup with 13:50 to play. From that point, the Heels went on a six minute 18-2 run, and finished the game on a 32-11 run overall. For the first time in conference play, Ellington and Lawson had good games in the same game, combining for 33 points and 6 assists in just over 40 minutes of play. All in all, the Heels had a well distributed offensive attack, with 5 players in double figures. Still, they've shown a tendency to turn the ball over a lot in conference play (23.8% last night). Wake Forest couldn't find the bottom of the basket, even when no one was guarding them (6-16 from the line). (Random UNC note - their last 3 opponents have made only 17 of 45 free throws. Did someone change the rules for the Heels and let them guard free throw shooters? Yikes!). Wake's second half shooting was truly dreadful - 9 of 35 from the field and 2 of 11 from the line for a 0.58 PPWS and a .300 EFG%. Only Visser played well, with 16 and 5 in 31 minutes.

UVA took care of business on the road, leading to the following extremely long statistical accomplishment: first non-neutral-court road win outside of the state of Virginia for Dave Leitao's Cavaliers. Singletary and Reynolds lit it up again, scoring 56 of the team's 71 points - when you factor in the 10 minutes they sat on the bench, they scored 90.1% of the 'Hoos points while on the floor. Someone (ok, me) mentioned that Singletary was simply too good to keep shooting as bad as he had in the pre-conference season. His ACC numbers - 1.31 PPWS (tied for 4th), .601 EFG% (7th), and .429 3pt% (7th). State had a very rough day shooting against the conference's leading field goal defense. Only Trevor Ferguson shot better than 43%, and he took just 1 shot. Atsur returned for the Pack and posted 9 points and 3 assists. Atsur has done his best Wes Miller impersonation in the two conference games he's played - 10 threes attempted against just one 2 point attempt.

Georgia Tech continued their home/road Jekyll and Hyde show. Someone (ok, me) said there was no way their hot shooting would continue - in their last two games they've hit just 6 of 34 threes, including an epicly bad 1 of 17 last night. Thaddeus Young continues to be a beast, scoring 21 on 9 of 15 shooting (including Tech's only made 3) and pulling in 7 boards, 4 of those offensive. For Maryland, James Gist had a career performance - scoring 26 on 10 of 11 from the field, hitting all nine of his 2pt attempts. The Terps forced Tech to turn the ball over just about once every 4 possessions, and coasted to a 15 point victory.


Anonymous said...

Not really related directly to this post, but I wanted to run a thought past you. Pomeroy's new correlation data is pretty interesting. In most cases in confirms the idea that eFG has the highest correlation to offensive and defensive efficiency, but it sometimes shows other correlations for some teams. For Duke, its opponents' Opp Reb %age is fairly highly correlated to Duke's offensive efficiency, with a 95% confidence. I guess that means one of two things (or a combo): (1) Duke's O does better when its D gets a bigger share of the boards, perhaps making its O better off the rebound; or (2) it is an indicator of overall effort and engagement in the game's flow. But the second explanation would not explain why that one stat is correlated so high, and Duke's own O Reb %age is not, which would actually have a higher direct relation to O eff.

So my point is, does Duke's O do better after getting the defensive rebound? Do your stats/data lend themselves to analyzing efficiency after a defensive board versus after a made basket or turnover?

Here's the connection to the Clemson game: Duke grabbed 78% of the available defensive boards. It averages 71.5% and Clemson's O averages allowing 67.4%. Duke's O did better at 103.5 than the 98.4 the kenpom prediction would have given it. Coincidence? Or is Duke's defensive rebounding a big key to its offense?

Just curious and thought you might be too. Thanks again for the great stuff.


Paul Rugani said...

I may look at the full season on this over the weekend, but here's your answer for the Clemson game - it made absolutely no difference.

Clemson ended 24 possessions with a missed shot, and Duke scored 25 points on their ensuing possessions. The Tigers ended 29 points with a made shot, and Duke scored 31 points on their ensuing possessions. Finally, Clemson committed 12 turnovers, and Duke scored 10 points on those ensuing possessions. (I know that adds up to 66 points, not 68, but Duke scored the first two on the opening possession of the game). Clemson got offensive rebounds on only 4 possessions (two possessions featured two ORBs), and Duke scored 4 points on the ensuing possessions, compared with 62 points in 61 possessions where Clemson did not record an offensive rebound.

Obviously this is just one game, but there was no direct, per possession correlation last night between Clemson not getting offensive rebounds and Duke offense scoring points.

Anonymous said...

Thanks for the research! Thats very interesting that it doesn't matter. Do you have any insights, if the full-year similarly plays out, to explain this statistical correlation?