In the first half of the ACC season, Duke scored 693 points and surrendered 573 points on 597 possessions - that's 1.16 points per possession scored and 0.96 points per possession surrendered for an efficiency margin of an astounding 0.20. At this point, we as Duke fans were justifiably excited about our season. At the end of this stretch of games, we were 8-0 in the ACC, 20-1 overall and performing at an extremely high level.
At some point (I like to look at the UNC game) and for some reason (read on for some hypothesis), Duke's effectiveness on both offense and defense dropped by about the same amount each. In the 2nd half of the ACC season, Duke scored 647 points and surrendered 617 points in 600 possessions: scoring 1.08 ppp, surrendering 1.03 ppp for an efficiency margin of only 0.05. Well, what went wrong? I have no answers, but I am fairly confident it's NOT a few things:
MYTH: Duke's shooting woes started because of fatigued legs or something similar (reason why Duke's offense got worse)
Basically, I'm not going to argue that Duke shot just as well in the last 8 games as they did in the first 8 games. In the first 8 Duke shot 55.6% eFG, and in the second 8, they shot 49.6%. It's obvious that they shot much worse. But the assumption in this statement is that Coach K didn't play enough players, they got tired and the long distance shooting (which was essential to Duke's success - sub-myth #1) suffered.
The truth is actually that Duke shot better from 3 (37.7% to 36.7%) and shot 3s slightly more often (41.4% of the time against 40.3%) in the 2nd 8 games of the season. The huge difference came in Duke's 2 point shooting. Duke shot 55.9% from 2 in the first half of the ACC season, a number that dipped down to 44.7% in the second half.
The biggest question is why that is? Was it a function of something Duke was doing wrong, or the function of something the opponents were doing differently? In other words, was it scouting that exposed a previously unknown weakness or did a weakness arise as the season went on? My tentative and unsatisfyingly conservative answer is: a bit from column A, a bit from column B. My sense just from watching the games was that Henderson's injury really hindered his offensive effectiveness. He was obviously still a very useful player, but he was also a very versatile offensive player near the beginning of the year.
It would be stupid of me just to suggest that Henderson's injury was 100% of it though. In the first 8 games, Duke's opponents blocked 28 of Duke's shots. In the final 8 games, Duke's opponents blocked 51 of Duke's shots. Obviously the big contributers to that total that were absent in the first 8 games were Boston College (21st best block rate, and with Tyrelle Blair, the player with the 9th highest block rate in the country at 14.2%) and Wake Forest (the 14th best block rate, with James Johnson, Chas McFarland and David Weaver all sporting 5-9% block rates). And of course, the biggest contributing game was the North Carolina game in Durham where UNC blocked 15 of Duke's shots behind tremendous defensive performances of Danny Green and Deon Thompson.
As I said, I think part of this was scouting. It's easier to block the shot of a 6'3" or 6'4" perimeter player driving to the basket than a 6'9" or 6'10" player in position for a similar shot. It's possible that teams played less to take away the 3 and more to block or contest the 2s in the second half of the season. A team as dependent on guys creating good 2 point shots off the dribble as Duke was required either excellent 3-point shooting every night or hitting the 2s - especially considering how average Duke was on the offensive glass all season.
This was the significantly most important difference between Duke's offense in the two halves of the ACC season. Duke did turn it over slightly more often (17.5% vs. 16.9%) and rebounded slightly worse offensively (32.5% vs. 34.8%), but the one huge glaring difference was Duke's struggle in converting their twos in the second half of the ACC season.
MYTH - In the 2nd half of the season, Duke was abused more heavily by big men or quick guards scoring easy baskets (reason why the defense got worse)
Actually it appears that in two key areas related to defensive play on the interior, Duke improved in the second half of the ACC season. Teams shot 51.2% inside the arc against Duke in the first half of the season, and 47.7% inside the arc in the 2nd half. Additionally, Duke grabbed 65.5% of possible defensive boards in the 2nd half as compared to 63.2% of defensive boards in the first half of the year.
The decline in defensive effectiveness appears to come chiefly from 3 factors: teams shot much better from 3, teams turned it over a bit less often against us, and teams hit their free throws.
Uniformly throughout the ACC season, teams barely took 3s against Duke. Whether this was a result of Duke's defense or because teams perceived more of a weakness on the interior defense is ultimately unknowable and irrelevant, but Duke is consistently among the top of Division I in fewest 3 point attempts allowed per field goal attempt. Now, in the first half of the season, Duke opponents took only 101 3s, and made just 27 of them (26.7%), whereas in the 2nd half of the season teams made 42.1% of their 3s (45 for 107). Obviously good 3 point shots have to be created. I saw the Wake game and the Miami game - quick guards breaking down Duke's defense and requiring help left shooters open and guys hit their shots. That was, without a doubt, a contributing factor. Though, I am not sure it was the only factor. Sometimes, teams just shoot better against you.
The turnovers were noticeable but not a huge reason for Duke's defensive decline. In the first half, Duke caused 149 turnovers in 597 possessions - thus, opponents lost 25% of their possessions to turnovers. In the second half, Duke caused 144 turnovers in 600 possessions, leaving 24% of opponent possessions empty. The difference is there, but it's not huge. Duke continued to turn teams over late into the season, just slightly less often.
Finally, free throws. Whether by luck or providence, teams shot a fairly abysmal 61.7% against Duke in the first half of the ACC season, making 116 of 188 free throws. In the second half of the season, Duke's opponents shot fewer free throws (156) but made just 2 less. I'll readily admit that the first half of the year, Duke probably benefited from opponents shooting poorly from the stripe. This wasn't always just dumb luck (see: Clemson's season long woes from the stripe - minus the one mystical game against Duke in the ACC semis of course - who's bitter?), but if teams shot as poorly in the 2nd half of the season from the line and from 3 as they did in the first half, I estimate Duke would have surrendered about 0.97 ppp - negligibly worse than the first half (probably due to fewer turnovers). More realistically, you should expect your opponents to shoot about 70% from the stripe. If that happened consistently all season - Duke would have had about a 0.99 ppp in the first half of the season, and 1.02 ppp in the 2nd half - not such a huge difference and still successful (frankly not on a Kansas or UCLA level of defensive dominance, but pretty good nonetheless).
So what's the point of all this? Well, clearly Duke played worse down the stretch. Clearly the team did not perform up to the expectations either of the program or that the team held for itself up to and including the first North Carolina game. But it's not always doom and gloom, it's the end of the world, Duke was exposed, it's smoke and mirrors, it's this, it's that. Sometimes the explanations are pretty rational. It seems to me that teams became more adept at attacking Duke's defense, but the biggest problems in the 2nd half of the ACC season came offensively. Teams scouted the players well, one of Duke's best inside-the-arc offensive players was fairly badly injured and for whatever reason, they were unable to finish around the basket like they did earlier in the year.
It's not all terrible - Duke hit over 50% of their 2s in 4 of the final 8 games (including 2 of the losses, oddly enough), had a fantastic offensive game against Georgia Tech in the ACC tournament (47.4% from 3, 54.3% from 2, 40% offensive rebounding rate). I am not sure you can say that the Clemson game was a poor performance as much as it was Clemson playing very well and hitting free throws. The one weekend Duke stayed in the NCAA tournament was a nightmare through and through. Were the players sick? Perhaps. It's ultimately irrelevant. The point is really that the way Duke performed that weekend is incongruent with Duke's performance even the previous week and certainly "down the stretch." Often fans remember only the last and worst memories of a season and no doubt many will walk away from 2007-08 remembering how close Duke came to joining a none too illustrious club of #2 seeds to lose to a #15 and how Duke lost a relatively comfortable lead against West Virginia and were embarrassed to the media by the West Virginian players. Fair enough, but it's pretty unfair for a team that played extremely high levels of basketball deep into the season before one unfortunate weekend derailed the team.
I leave this post with a bit of a stat dump: herein are the stats of each major player in ACC play only for 2006-07 and then 2007-08 including Duke's team performance for both years. There is no doubt the team and all the constituent players improved dramatically from 07 to 08 and there is no reason why fans should not be excited to see these same players improve once again from 08 to 09. In my estimation, Duke fans have a little too often called for a reboot and should appreciate the quality of players Duke has right now. Perhaps Duke in 08 is not as good as UCLA in 08 or North Carolina in 08 or Kansas in 08, but it's a luxury as a Duke fan to believe that we should always be among the elite teams, but a foolish one when one consider just how difficult it is to be that good year in, year out. Enough of the didactic nonsense, here's the goods.
2006-07
| Player | %Minutes | Ortg | %Poss | Floor% | 3P% | eFG% | TO% | FTRate | OR% | DR% | Ast% | Stl% | Blk% |
| DeMarcus Nelson | 78.9% | 99.1 | 24.8% | 47.1% | 34.5% | 51.5% | 19.6% | 20.9 | 5.4% | 13.7% | 14.6% | 2.8% | 1.1% |
| Josh McRoberts | 88.9% | 103.8 | 21.1% | 51.2% | 20.0% | 53.2% | 21.5% | 38.6 | 7.0% | 20.6% | 15.6% | 2.2% | 5.7% |
| Jon Scheyer | 83.8% | 120.5 | 19.0% | 53.3% | 34.7% | 50.3% | 12.1% | 47.4 | 4.2% | 8.2% | 9.3% | 2.3% | 0.4% |
| Greg Paulus | 84.3% | 105.0 | 22.2% | 45.6% | 41.1% | 54.7% | 23.9% | 21.1 | 1.4% | 7.2% | 23.7% | 1.7% | 0.2% |
| Gerald Henderson | 49.0% | 99.2 | 22.5% | 48.7% | 35.7% | 52.4% | 22.4% | 20.8 | 5.1% | 11.6% | 13.1% | 1.5% | 0.9% |
| David McClure | 55.4% | 105.6 | 11.6% | 53.5% | 50.0% | 52.0% | 22.7% | 39.2 | 10.1% | 17.7% | 3.9% | 2.8% | 3.0% |
| Lance Thomas | 33.9% | 84.0 | 16.4% | 44.8% | 0.0% | 61.3% | 39.6% | 67.7 | 11.4% | 12.2% | 0.8% | 1.9% | 0.9% |
| Brian Zoubek | 13.1% | 68.3 | 19.2% | 37.4% | 0.0% | 50.0% | 47.2% | 110.0 | 11.8% | 19.5% | 4.0% | 0.0% | 3.4% |
2007-08
| Player | %Minutes | Ortg | %Poss | Floor% | 3P% | eFG% | TO% | FTRate | OR% | DR% | Ast% | Stl% | Blk% |
| DeMarcus Nelson | 80.2% | 116.0 | 24.2% | 57.4% | 45.7% | 59.6% | 16.1% | 62.3 | 5.3% | 13.3% | 17.4% | 3.3% | 0.6% |
| Kyle Singler | 77.7% | 107.9 | 23.0% | 48.3% | 34.8% | 52.7% | 17.9% | 24.7 | 7.0% | 14.7% | 8.4% | 2.4% | 2.1% |
| Gerald Henderson | 69.4% | 107.0 | 22.4% | 55.0% | 25.0% | 51.2% | 17.7% | 62.6 | 6.6% | 13.1% | 11.6% | 2.3% | 2.1% |
| Jon Scheyer | 71.4% | 129.3 | 16.9% | 58.8% | 37.7% | 54.9% | 11.8% | 51.3 | 4.0% | 9.8% | 15.0% | 2.9% | 0.8% |
| Greg Paulus | 71.9% | 124.1 | 18.8% | 51.0% | 41.8% | 55.0% | 13.2% | 30.5 | 1.5% | 7.4% | 19.6% | 3.7% | 0.4% |
| Nolan Smith | 34.4% | 103.7 | 22.9% | 47.1% | 40.7% | 57.3% | 27.1% | 29.0 | 4.1% | 11.0% | 16.9% | 1.0% | 0.4% |
| Taylor King | 19.1% | 93.4 | 21.6% | 36.9% | 30.0% | 40.8% | 13.0% | 14.3 | 7.4% | 12.9% | 5.9% | 1.9% | 3.1% |
| Lance Thomas | 45.3% | 96.3 | 13.8% | 52.1% | 0.0% | 48.9% | 18.5% | 80.0 | 10.9% | 14.2% | 2.3% | 2.6% | 2.9% |
| Brian Zoubek | 9.2% | 108.1 | 11.3% | 52.9% | 0.0% | 60.0% | 17.0% | 20.0 | 9.6% | 16.5% | 3.0% | 1.9% | 6.3% |
| David McClure | 20.9% | 81.7 | 10.1% | 39.1% | 0.0% | 35.3% | 29.3% | 5.9 | 8.4% | 15.5% | 7.2% | 1.7% | 2.1% |
| Duke 06-07 | Duke | Opp. | |
| eFG% | 52.6% | 50.0% | |
| TO% | 22.2% | 21.8% | |
| ORB% | 34.0% | 29.1% | |
| FTRate | 22.1% | 33.7% | |
| Tempo | 66.7 | ||
| Offensive Efficiency | 103.8 | ||
| Defensive Efficiency | 99.3 | ||
| 3-point FG% | 36.1% | 37.4% | |
| 2-point FG% | 52.0% | 48.2% | |
| Free Throw % | 68.9% | 74.7% | |
| Block % | 9.2% | 10.6% | |
| Steal % | 10.4% | 10.9% | |
| 3PA/FGA | 30.8% | 22.5% | |
| A/FGM | 49.6% | 51.6% | |
| Duke 07-08 | Duke | Opp. | |
| eFG% | 53.6% | 50.9% | |
| TO% | 17.2% | 24.7% | |
| ORB% | 33.3% | 34.9% | |
| FTRate | 29.6% | 37.6% | |
| Tempo | 70.3 | ||
| Offensive Efficiency | 113.2 | ||
| Defensive Efficiency | 99.1 | ||
| 3-point FG% | 37.4% | 34.0% | |
| 2-point FG% | 51.8% | 50.9% | |
| Free Throw % | 68.8% | 66.6% | |
| Block % | 7.5% | 11.7% | |
| Steal % | 13.1% | 7.2% | |
| 3PA/FGA | 41.1% | 22.3% | |
| A/FGM | 49.9% | 45.7% |
Finally, I leave with this - to help us remember the good times in the coming months of darkness!
2 comments:
Interesting commentary and stats. To me it seemed Duke didn't get the same production from the bench late in the season that they had earlier. I wonder if your numbers support this. I think a number of factors contributed to the decline. Despite the way things ended I really enjoyed this team and am looking forward to next season (shame it is such a long wait!)
I think that is a fair assessment. Scheyer continued to produce off the bench, but Nolan and Taylor really stopped offering the kind of production they were giving mid-season by season's end. For Nolan's sake, I wonder how badly that hyperextended knee set him back and Taylor just seemed to be ice cold as the season went on (12 for 40 from 3 in the ACC season). We were a young team depending on freshmen for a fair amount of production. It's not unusual for young teams to have these kinds of troubles.
Post a Comment