Monday, November 12, 2007

Duke 86 New Mexico State 61

New Mexico State, on paper, looked like a good test for Duke. Conventional wisdom thus far this season is that without a true big man (aside from Zoubek I guess), Duke might run into problems against bigger teams who might be able to muscle around the Devils on the interior and out-rebound us. Well, test passed, I guess. While Duke did not do especially well on the offensive glass (12 offensive boards vs. NMSU's 24 defensive boards), we absolutely controlled the defensive glass, grabbing 26 defensive rebounds and allowing NMSU only 6 offensive boards. For the record, that means Duke grabbed 81.25% of the available defensive boards, against a big, experienced team.

That all being said, NMSU did not look particularly sharp. Defensively they were really unable to deal with Duke's offensive versatility and quickness. In fact, the Aggies experimented with smaller line-ups some in the second half in order to adjust to Duke, with some success (though with Duke leading 61-30, any success the Aggies had was pretty much futile).

This was also Duke's worst offensive game of the season, scoring only 1.06 ppp. When I realized this after I got home, I was a little surprised. I had thought we looked pretty good offensively in general. We shot 56% inside the arc, but only about 35% from beyond the arc (with many good, open looks - I really can't think of any 'forced' 3s that I found painful to watch). We didn't turn it over too often (17 TOs in about 81 possessions - about 21% TO rate - not great, but we've had worse). I think offensive execution was, overall, very pleasing, but we missed an unnerving number of easy lay-ups and a few too many open 3 pointers. We really ought to have scored over 100 in this game given how we played. A win is a win, eh?

Player
Eff.
Usage
Floor%
Kyle Singler
125.5
16.5%
65.3%
Lance Thomas
93.1
14.7%
44.7%
Greg Paulus
117.6
20.8%
52.2%
Gerald Henderson
79
25.6%
40.3%
DeMarcus Nelson
125
18.9%
54.0%
Nolan Smith
45.8
24.4%
21.4%
Martynas Pocius
144.1
15.8%
71.2%
Taylor King
90.7
18.5%
35.7%
Jon Scheyer
132.2
21.7%
57.1%
Brian Zoubek
105
11.8%
51.1%

Team Stats

Duke
Opp.
eFG%
54.48%
49.09%
TO%
20.98%
32.03%
ORB %
33.33%
18.75%
FT Rate
19.40%
23.64%





Tempo
81.03

Offensive Efficiency
106.14

Defensive Efficiency
75.29






3-Point FG %
34.62%
23.53%
2-Point FG%
56.10%
55.26%
Free Throw %
68.42%
53.85%
Block %
15.79%
7.32%
Steal %
13.55%
2.47%





3PA/FGA
38.81%
30.91%
A/FGM
56.25%
48.00%

Player
Est. D-Rat
Real D-Rat
Est. Stop%
Real Stop%
%Dposs
Kyle Singler
66.55
49.45
84.53%
90.29%
42.70%
Lance Thomas
81.95
78.53
46.83%
26.73%
8.99%
Greg Paulus
69.12
55.18
78.23%
92.71%
31.31%
Gerald Henderson
86.7
78.90
35.21%
17.78%
7.28%
DeMarcus Nelson
76.68
74.20
59.73%
54.43%
18.06%
Nolan Smith
90.55
76.41
25.79%
7.72%
1.72%
Martynas Pocius
75.07
80.66
63.67%
35.26%
23.23%
Taylor King
71.54
69.01
72.30%
69.29%
21.96%
Jon Scheyer
74.31
70.87
65.52%
65.91%
18.80%
Brian Zoubek
57.81
62.46
105.93%
85.47%
24.10%

Hello huge block of numbers! Here are the highlights:
  • Some (mildly alarmist) Duke fans are semi-concerned over Kyle Singler's "bad" game. Let's recap. Singler shot 3 for 7 from the field, 4 for 5 from the free throw line for 10 points, 5 rebounds, 2 steals, 2 blocks, 1 assist, no turnovers, forced 2 other turnovers, and 5.5 forced misses. This was against a big and physical team. Most of his shots were an inch or two from going in and were good shots. He, thus, managed a 125.5 rating (though using only 16.5% of poss), had the highest floor% on the team (65% is very high - the highest I could find for a season was Brand in 1999 with about 67.9%), with a 49.5 defensive rating, over a 90% stop% (very, very high), while being involved in over 42% of NMSU's possessions defensively while he was playing. Singler showed why you don't need to fill the stat sheet to be an invaluable player.
  • So did Greg Paulus. One stat absent in the huge block of numbers above (but you'd find in a conventional box score) is the fact that Paulus had 8 assists and 3 turnovers. In Paulus' 26 minutes, he assisted almost half of the baskets made. Paulus may have been our 2nd most important defensive player tonight, himself causing 8 turnovers (4 credited as steals, 4 as other forced turnovers). He constantly harassed the Aggies' various ball handlers, showing a defensive aggression he lacked last year. He looks quicker, sure, but a big part of it is skill and attitude.
  • Smith had a pretty poor game, shooting just 1 for 6 with 3 assists and 3 turnovers. Freshmen will have growing pains. I like, though, that a guy can have a poor game, and other guys are able to step up. We may not have a single guy whose performance our team's fate depends on. When Singler sat with 4 fouls, King managed to come in and had 4 rebounds in 17 minutes (grabbing 29.4% of the defensive boards he could).
  • Zoubek contributed 11 solid minutes, grabbing 5 rebounds, blocking 2 shots (against a 7' center) and had only 1 foul. And zero turnovers.
  • Nolan Smith was touted as a defensive specialist when he was hyped in the months prior to the season. And yet, in two games, he has zero steals, and 1 forced turnover. Notably, though, Paulus has used many more defensive possessions than Smith has, even accounting for the greater number of turnovers he's caused. Defensive stats have limitations - what about situations where a player's defense causes a player who might otherwise look to drive or shoot to give up the ball and be unable to use a team's possession? These are assets that stats cannot really cover very well. From watching the games, Smith has looked like a very good positional defender. When he learns to step it up and show Paulus' aggression defensively (who ever thought that would make sense?), he'll be an exceptional defender.
  • Not all is well, though. NMSU shot over 55% inside the arc. Could that be a sign of a troubling interior defense, or could that be just a random fluke? It's hard to conclude anything at this point. The Aggies produced almost no points from the block, and a good few of their points inside the arc came from Hawkins driving to the basket or open mid-range jumpers after a Duke player gambled, and failed, defensively. Even so, as mentioned, we dominated the defensive glass, and caused 26 turnovers in 81 possessions. One interior defensive positive is that we managed to block 6 of the Aggies 38 2-point shots (15.8% - a ratio I cannot possibly expect for the season, but encouraging for the game).
  • The tempo belies the speed of the game, as Coach K decided to slow it down into a form of "stall ball" for most of the last 7 :30 or so. The Aggies started the game trying to run with Duke, producing terrible results. Had we continued to run, we would likely have passed 90 possessions.

3 comments:

Jeff P said...

Great Blog!

What does the Floor% mean?

What is the difference between the estimated and real defensive stats?

Do you have a past post where the definitions of all of these stats are outlined. I am sure linking to a page explaining what you are calculating would be very helpful to new reads, such as myself.

Thanks

Unknown said...

Second that.

Was going to post what jeff said verbatim. Great blog.

Michael said...

Sure, I think I briefly explained what the stats mean in the recap of the Central game, but it's probably buried in some over-wordy paragraph.

Basically, Floor% is a calculation of scoring possessions divided by total individual possessions. Scoring possessions are like made baskets and free throws essentially, whereas total possessions also includes missed shots and turnovers. The idea is it tries to indicate how reliable of a scorer a player is. For instance, if you look at, say, 2006 Duke, you see that the team is dominated by two players: JJ Redick and Shelden Williams. Redick had about 120.4 personal efficiency for the season (or produced about 1.2 points per personal possession), whereas Williams' had an efficiency of 117.1. The idea here is that Redick was a slightly more efficient player offensively because he took so many 3s and had so many possessions where he scored 3 points. On the other hand, Williams had a very high free throw rate (65.9), so a lot of his personal possessions were also spent at the line. Thus, his Floor% was about 61%, whereas Redick's was lower, around 53.2%. My hunch for this coming season is that we will need to find a couple guys who have fairly high Floor% if we want to be consistently excellent offensively. Usually big men have the best floor% because of trips to the foul line - can Singler do that for us this year?

As for the real and estimated defensive stats, I guess I'll stop showing the estimated stats because that's really just out of some interest to me. Dean Oliver created an insanely complicated formula to estimate defensive ratings assuming 20% of defensive possessions per player on the court based on the available stats (defensive rebounds, steals, blocks, personal fouls) and overall team efficiency (based on the assumption that defense is essentially a team endeavor anyway). I have estimated defensive ratings/stop% (which is basically the anti-floor% - how often a player stops the man he's guarding from scoring when that guy uses a possession) for Duke teams going back to 1987 and various other teams since then, but only this year did I start personally recording the stats needed to come up with real defensive ratings that require stats like forced turnovers (think charges or pressure a player applies to a player causing him to travel or throw the ball out of bounds), forced misses (when a player is in another player's face and he misses), and stats like that. This allows us also to see which players are most involved in defense. In the first game this season, Zoubek was. In this past game against NMSU, Singler and Paulus were.

I hope that made sense.