Phase III (2011/12)

[superdave]

A lot of this thread seems to be focusing on Dr. Dawkins. If he plays well enough to earn more minutes, that means he is most likely scoring the ball incredibly efficiently and that is good for Duke. Here's a look at Andre's stats over his first two years:

2010 season - .379 on 3s, averaged 1 2-point shot per game. 4.4 points in 12.6 minutes. 2 DNPs.
We know the personal issues he had about mid-season with his sister. His minutes and scoring really tailed off in January. He only scored 51 points in the final 26 games and had no double digit scoring games during that span.

2011 season- .427 on 3s, averaged 1.5 2-point shots per game. 8.1 points in 21 minutes.
Andre scored in double digits in 10 of the first 17 games last year. Then he only scored 101 points in the final 20 games and had two double digit scoring games over that span. Over that stretch he shot 22-67 from 3, less than the break-even .333 rate. He also had three games in that stretch in single-digit minutes.

Here's a quick look at all his games where he's played 30+ minutes. I know this is a chicken-egg argument. If he's hot early, he's going to play big minutes:

2010 season - No games with 30+ minutes.

2011 season
Butler - 32 minutes - 3-5 fgs, 2-4 3s, 10 points. Win.
Bradley - 31 min - 10-17 fgs, 8-14 3s, 28 points. Win.
Elon - 32 min - 5-9 fgs, 3-6 3s, 17 points. Win.
UAB - 32 min - 3-4 fgs, 2-3 3s, 8 points. Win.
Maryland - 30 min - 3-6 fgs, 2-5 3s, 8 points. Win.
UVa - 32 min - 5-11 fgs, 3-8 3s, 14 points. Win.

2012 season
MSU - 38 minutes - 8-15 fgs, 6-10 3s, 26 points. Win
TN - 32 minutes - 4-7 fgs, 2-4 3s, 10 points. Win.
Michigan - 35 minutes - 5-12 fgs, 4-9 3s, 14 points. Win.
Kansas - 34 minutes - 2-5 fgs, 2-4 3s, 6 points. Win.
Washington - 32 minutes - 5-13 fgs, 2-9 3s, 17 points. Win.

So Duke is 11-0 in games where Andre averages 30+. Some of those games he scores big, some of those games he's right around his average.

Here's a quick look at Andre's stat lines in Duke's losses the past 3 seasons:

2010 season
Wisco - 22 min - 4-4 fgs, 4-4 3s, 12 points.
GT - 12 min - 1-3, 0-2, 2 points.
State - 9 min - 0-0, 0-0.
Gtown - 13 min - 1-4, 1-4, 5 points.
Maryland - 9 minutes - 2-3, 2-3, 6 points.

2011 season
FSU - 29 min - 2-9, 1-8, 8 points.
St Johns - 27 min - 3-8, 1-6, 7 points.
VT - 18 min - 2-4, 2-4, 6 points.
Unc - 12 min - 0-1, 0-1.
Zona - 22 min - 3-5, 1-2, 9 points.

2012 season
OSU - 19 min - 0-1, 0-0.
Temple - 14 min - 0-3, 0-2.
Scoreless in both our losses this year.

I am not entirely sure what all this means. If I were the coach ( a big stretch, thanks) I'd run a few plays for him early and get him 2-3 goods looks and decide from there if it's going to be a 15 minute night or a 30 minute night. This of course assumes he's busting his tail on defense. I would also make sure he knows he must channel Rip Hamilton coming off screens to get open. No loafing. But if he's playing well and earning big minutes, we're going to win. That's awfully enticing.

[Kedsy]

I am not entirely sure what all this means. If I were the coach ( a big stretch, thanks) I'd run a few plays for him early and get him 2-3 goods looks and decide from there if it's going to be a 15 minute night or a 30 minute night. This of course assumes he's busting his tail on defense. I would also make sure he knows he must channel Rip Hamilton coming off screens to get open. No loafing. But if he's playing well and earning big minutes, we're going to win. That's awfully enticing.

Very interesting analysis. Thanks. Although I do think there's some chicken and egg going on here. I agree with you that running a few plays early for Andre would be good for both his game and the team.

[Saratoga2]

A lot of this thread seems to be focusing on Dr. Dawkins. If he plays well enough to earn more minutes, that means he is most likely scoring the ball incredibly efficiently and that is good for Duke. Here's a look at Andre's stats over his first two years:

2010 season - .379 on 3s, averaged 1 2-point shot per game. 4.4 points in 12.6 minutes. 2 DNPs.
We know the personal issues he had about mid-season with his sister. His minutes and scoring really tailed off in January. He only scored 51 points in the final 26 games and had no double digit scoring games during that span.

2011 season- .427 on 3s, averaged 1.5 2-point shots per game. 8.1 points in 21 minutes.
Andre scored in double digits in 10 of the first 17 games last year. Then he only scored 101 points in the final 20 games and had two double digit scoring games over that span. Over that stretch he shot 22-67 from 3, less than the break-even .333 rate. He also had three games in that stretch in single-digit minutes.

Here's a quick look at all his games where he's played 30+ minutes. I know this is a chicken-egg argument. If he's hot early, he's going to play big minutes:

2010 season - No games with 30+ minutes.

2011 season
Butler - 32 minutes - 3-5 fgs, 2-4 3s, 10 points. Win.
Bradley - 31 min - 10-17 fgs, 8-14 3s, 28 points. Win.
Elon - 32 min - 5-9 fgs, 3-6 3s, 17 points. Win.
UAB - 32 min - 3-4 fgs, 2-3 3s, 8 points. Win.
Maryland - 30 min - 3-6 fgs, 2-5 3s, 8 points. Win.
UVa - 32 min - 5-11 fgs, 3-8 3s, 14 points. Win.

2012 season
MSU - 38 minutes - 8-15 fgs, 6-10 3s, 26 points. Win
TN - 32 minutes - 4-7 fgs, 2-4 3s, 10 points. Win.
Michigan - 35 minutes - 5-12 fgs, 4-9 3s, 14 points. Win.
Kansas - 34 minutes - 2-5 fgs, 2-4 3s, 6 points. Win.
Washington - 32 minutes - 5-13 fgs, 2-9 3s, 17 points. Win.

So Duke is 11-0 in games where Andre averages 30+. Some of those games he scores big, some of those games he's right around his average.

Here's a quick look at Andre's stat lines in Duke's losses the past 3 seasons:

2010 season
Wisco - 22 min - 4-4 fgs, 4-4 3s, 12 points.
GT - 12 min - 1-3, 0-2, 2 points.
State - 9 min - 0-0, 0-0.
Gtown - 13 min - 1-4, 1-4, 5 points.
Maryland - 9 minutes - 2-3, 2-3, 6 points.

2011 season
FSU - 29 min - 2-9, 1-8, 8 points.
St Johns - 27 min - 3-8, 1-6, 7 points.
VT - 18 min - 2-4, 2-4, 6 points.
Unc - 12 min - 0-1, 0-1.
Zona - 22 min - 3-5, 1-2, 9 points.

2012 season
OSU - 19 min - 0-1, 0-0.
Temple - 14 min - 0-3, 0-2.
Scoreless in both our losses this year.

I am not entirely sure what all this means. If I were the coach ( a big stretch, thanks) I'd run a few plays for him early and get him 2-3 goods looks and decide from there if it's going to be a 15 minute night or a 30 minute night. This of course assumes he's busting his tail on defense. I would also make sure he knows he must channel Rip Hamilton coming off screens to get open. No loafing. But if he's playing well and earning big minutes, we're going to win. That's awfully enticing.

Probaby a lot of Duke fans are looking for more consistency as Andre's experience grows. That hasn't been the case as he still can have a big game or two followed by a disappearing act. Andre has been around long enough to expect him show up every night and at least contribute in a meaningful way. We all are rooting for him, but he has to bring his game all of the time.

[NSDukeFan]

Probaby a lot of Duke fans are looking for more consistency as Andre's experience grows. That hasn't been the case as he still can have a big game or two followed by a disappearing act. Andre has been around long enough to expect him show up every night and at least contribute in a meaningful way. We all are rooting for him, but he has to bring his game all of the time.

I agree with you in theory and in practice he has to make sure he brings his 'A' defensive game every game. Offensively, however, his best trait is his fantastic 3-point shooting and that depends on how many open looks he gets, which may not be completely under his control, depending on how well the team is moving the basketball, who is guarding him and how well, and the defensive strategy and focus of the other team. My impression is that Andre has been looking to take the ball to the basket a little bit more lately and he tends not to turn the ball over very much, but his best offense is his spot-up shooting, which is in part determined by his movement without the ball, but also in part on elements outside of his control.

[Kedsy]

I agree with you in theory and in practice he has to make sure he brings his 'A' defensive game every game. Offensively, however, his best trait is his fantastic 3-point shooting and that depends on how many open looks he gets, which may not be completely under his control, depending on how well the team is moving the basketball, who is guarding him and how well, and the defensive strategy and focus of the other team. My impression is that Andre has been looking to take the ball to the basket a little bit more lately and he tends not to turn the ball over very much, but his best offense is his spot-up shooting, which is in part determined by his movement without the ball, but also in part on elements outside of his control.

Also, part of Andre's value on offense is the threat of his taking a three-point shot. Even if he doesn't get many shots, the opposition has to guard him very closely. Obviously this value is not reflected in his stats, but it makes it more difficult for opposing defenses to help on Austin, Seth, Ryan, and Mason, and that's a good thing.

So I agree that the key for more minutes for Andre is his defensive effort.

[Saratoga2]

I mentioned in another thread, that while playing Cook at the point solves one problem (pure point, assists up, TO's down), it comes with a price, as it puts Seth and Austin on the wing and Andre (or MG) on the pine. That trade off hurts our wing defense and also puts a lethal offensive weapon in Andre in a more limited role. That is the current issue K faces with the various combinations he is tinkering with. At the end of the day, if K sticks with the Cook/Thornton duo at the point, he may have to consider putting Seth in a 6th Man role, and start a perimeter of Cook/Rivers/Dawkins. Yet, again, that comes with a price. Seth is a vet who has shown the ability to take over on the offensive side and hit dagger after dagger. We play much better when Seth is on his game too. But, in 07-08, Jon Scheyer excelled in a 6th Man role, and played starters minutes doing so. It gave us an offensive punch coming off the bench. I have no idea if putting Seth in that role would help or not, but I would be interested to see it attempted.

Yes, there are difficult decisions for the coaching staff. We seem to have in Quinn a primary ball handler who has court vision and the ability to get players the ball in scoring position. He is very good in ball security and can to some degree get past his first defender. The problem is that he is also short as guards go and is a freshman learning to play Duke defensive basketball. When you compare him to Tyler, there is a big difference in ball handling and offensive fluidity, yet while Tyler is nearly the same size, he is aggressive, has some leadership skills. He is very aggressive, but runs up the fouls at a high rate. Putting either of these two into the game forces a decision on other players. We want to keep Austin in the game for 30 plus minutes. He has reasonable size for a shooting guard and can put a lot of pressure on the defenses, even though he is still quite turnover prone. That leaves one starting/high PT minute position left. If you play Seth, you get a versitile scoring guard who plays reasonable defense, but the backcourt is small and appears to have difficulty with large guards. If you put Andre in, we get a little more size, but less scoring versatility. He provides a little more size but it hasn't been showing up on the scoreboard. If you put Michael in, you get a lot more size in another Freshman. Maybe it has happened, but I don't remember coach K starting 3 freshmen.

Maybe one answer would be to make Quinn and Seth the two headed point guard monster and place Tyler further back on the bench. That way we could have Andre and Michael share the wing duty while perhaps Andre could sub for Austin to keep the lineup size up while still maintaining significant scoring punch. It is a little different view of the solution and one in which coach K seems unlikely to implement.

[Kedsy]

Maybe it has happened, but I don't remember coach K starting 3 freshmen.

1982-83. Actually four frosh started a lot of games that year (Johnny Dawkins/Mark Alarie/David Henderson/Jay Bilas). On the other hand, our record that year was 11-17.

I also believe we started three freshmen on December 21, 1999 (Jason Williams, Carlos Boozer, Mike Dunleavy -- only game that year), but since 1983 that may be the only time.

[superdave]

We're #50 in Adjusted D on KenPom right now. But as I look over a little more to the right, I see that our Opponents’ average adjusted offensive efficiency (OppO) is #2.

Am I reading this correctly to see that part of why our defense is ranked fairly low for a Duke team is that we've played a solid run of offensive teams?

Here's where our opponents rank in AdjO according to KenPom:
8. Kansas
9. Belmont
10. Michigan St.
12. Ohio St.
14. Davidson
20. Colorado St.
30. Michigan
41. Temple
62. Washington
96. Virginia
97. Tennessee
120. W. Michigan
171. Penn
175. Clemson
178. GT
212. PC
257. UNC-G

That's three games vs. Top 10, six vs. Top 20, and 11 of 17 games vs. Top 100. That's a pretty impressive schedule (we're also #1 in Overall Strength of Schedule (Pyth)).

Here's where future opponents rank -
220. WFU
113. FSU
95. Maryland
181. St John's
65. VPI
43. Miami
19. UNC
25. NCSU
278. BC

So our next "really good" offensive opponents are UNC (twice) and State (once). Everyone else is ok to bad. It looks like our AdjD could rise somewhat significantly just because the 2nd half of the season is against teams with significantly less offensive efficiency than the 1st half of the season.

Everyone has known our season was front loaded with Maui, Ohio State and Michigan State, but when you see the relative AdjO rankings of Belmont, Davidson and Colorado State, you can really appreciate the path we've been through already. The mid-majors we've played are pretty solid.

Of course we all know this team has to play better defense, but the fact that we've played such good offensive opponents and are 15-2 is pretty encouraging. We will get better defensively and I look forward to seeing how high we can rise in AdjD considering future opponents.

[Kedsy]

Am I reading this correctly to see that part of why our defense is ranked fairly low for a Duke team is that we've played a solid run of offensive teams?

Theoretically, I don't think you're correct. That's why it's "adjusted," to take into account the quality of the opponent.

I suppose it's possible that the constant pressure of playing great offensive teams has artificially depressed our defensive rating. But I imagine it's also possible our poor D has artificially inflated the offensive rating of the teams we've played. We need to improve on D, either way.

[superdave]

Theoretically, I don't think you're correct. That's why it's "adjusted," to take into account the quality of the opponent.

I suppose it's possible that the constant pressure of playing great offensive teams has artificially depressed our defensive rating. But I imagine it's also possible our poor D has artificially inflated the offensive rating of the teams we've played. We need to improve on D, either way.

Some follow up questions -

Is Pomeroy adjusting for quality of opponent in OppO/AdjO or is he adjusting for pace? I thought he was adjusting for pace to show relative efficiency, by taking a team's level of efficiency (points per possession) and applying it to 100 possessions.

If the average AdjO of the opponents in your first 20 games of the season is 50 and the average AdjO of the opponents in your second 20 games of the season is 150 then your AdjD ranking should steadily improve over the course of the second half of the season, right? You should be less efficient on D vs. above-average AdjO teams and more efficient on D vs. below-average AdjO teams, right? That was the point I was trying to make above. If I'm off base on this point, I'm still not quite sure why.

[NSDukeFan]

Some follow up questions -

Is Pomeroy adjusting for quality of opponent in OppO/AdjO or is he adjusting for pace? I thought he was adjusting for pace to show relative efficiency, by taking a team's level of efficiency (points per possession) and applying it to 100 possessions.

If the average AdjO of the opponents in your first 20 games of the season is 50 and the average AdjO of the opponents in your second 20 games of the season is 150 then your AdjD ranking should steadily improve over the course of the second half of the season, right? You should be less efficient on D vs. above-average AdjO teams and more efficient on D vs. below-average AdjO teams, right? That was the point I was trying to make above. If I'm off base on this point, I'm still not quite sure why.

As Kedsy was saying though, your offensive and defensive number are adjusted for the teams you have played, meaning they are not raw efficiency numbers, but adjusted based on the strength of the team's opposition.

[superdave]

As Kedsy was saying though, your offensive and defensive number are adjusted for the teams you have played, meaning they are not raw efficiency numbers, but adjusted based on the strength of the team's opposition.

OK - this makes sense. Thank you.

[Kedsy]

Some follow up questions -

Is Pomeroy adjusting for quality of opponent in OppO/AdjO or is he adjusting for pace? I thought he was adjusting for pace to show relative efficiency, by taking a team's level of efficiency (points per possession) and applying it to 100 possessions.

If the average AdjO of the opponents in your first 20 games of the season is 50 and the average AdjO of the opponents in your second 20 games of the season is 150 then your AdjD ranking should steadily improve over the course of the second half of the season, right? You should be less efficient on D vs. above-average AdjO teams and more efficient on D vs. below-average AdjO teams, right? That was the point I was trying to make above. If I'm off base on this point, I'm still not quite sure why.

What Pomeroy calls "raw" defensive efficiency is points allowed per 100 possessions, so that's where the pace comes in. What he calls "adjusted" efficiency is adjusted for the quality of opposing offenses, the site of each game, and when each game was played (recent games get more weight).

Thus, although you're right we should be less efficient against top offenses, Pomeroy claims to take this into account. So the only way your observation is correct is if he's not adjusting enough for the quality of your opponent. Which is possible, but unlikely to be too big a factor.

[Greg_Newton]

So obviously the Gbinije ship has sailed at this point... but dare I say, Rivers is looking more and more like our lead guard on offense. Don't look now, but he's dished out 5 assists in each of the last two games, along with a few more kick-outs for threes that didn't go down. 6 TOs during that span, so he hasn't been perfect, but could we be starting to see a breakthrough in his offensive approach? He's certainly looking up one dribble earlier on most drives - rather than waiting until he's triple-teamed, he's anticipating the help and dishing it off at the correct time.

Our defensive issues are obviously priority 1-A, but this could quietly become an important storyline to follow.

[Kedsy]

...but could we be starting to see a breakthrough in his offensive approach?

Austin said as much in a Herald-Sun article linked from the main page:

“Sometimes you look at things you are doing well and things you haven’t been doing well,” Rivers said. “I’ve been looking at it. I know how I can create. Everybody is going to try to come at me because I can score so well. These past couple of games I’ve been trying to distribute more.”

Hopefully he really means it and is not just paying lip service. He and the team will both be much more effective if he hits the open man after drawing the double- or triple-team.

[MCFinARL]

Austin said as much in a Herald-Sun article linked from the main page:



Hopefully he really means it and is not just paying lip service. He and the team will both be much more effective if he hits the open man after drawing the double- or triple-team.

Well, there is a middle ground here--probably no reason to think he is just paying lip service, even though he is obviously pretty media savvy. But he might mean it in his head and still not have it so ingrained that he looks to do it instinctively in games. That being said, recent trends in his play suggest he is getting there.

[Kedsy]

Well, there is a middle ground here--probably no reason to think he is just paying lip service, even though he is obviously pretty media savvy. But he might mean it in his head and still not have it so ingrained that he looks to do it instinctively in games. That being said, recent trends in his play suggest he is getting there.

Sorry, that's what I meant. I don't think he was saying it just to say it. But I'm also sure he knows it's the right thing to say. And the right thing to do. And he's done it for two games now. I just hope going forward he expends his mental energy (while on offense) trying to recognize the double team and to distribute more. If so, our offense will be very difficult to stop.

On the other hand, he also has to spend his mental energy on defense. If we can only have one thing, that should be it.

But I'm hoping for both.

[gam7]

Is there a Phase IV in the works? If not, I'd be willing to take a crack at it.

[Kedsy]

Is there a Phase IV in the works? If not, I'd be willing to take a crack at it.

I believe SilkyJ is working on it.