PDA

View Full Version : K and the short bench



m g
09-01-2010, 02:03 PM
"Well, it's a long tournament, and we need to develop our entire team," U.S. coach Mike Krzyzewski said. "My position is based on running a marathon and not a sprint. That's something I've learned over the last five years. We need 12 guys to win. As we keep going along, it's not one guy. Not one guy."

Before reading this, it had seemed to me that K has been moving away from having as short a bench as he had in the past, but I've ascribed a lot of it to the players on the team. This quote makes me think that maybe he's had a change in his coaching philosophy. I don't know that Ryan Kelly or Andre Dawkins would have appeared in as many games as they did last year if this had been an earlier edition of Duke. (More true of Kelly than Dawkins, obviously, because we were shorter on Dawkins' position).

Wander
09-01-2010, 02:43 PM
Doesn't make much sense to me - wasn't last year one of the shortest benches K has ever played? (obviously, that's not a complaint, given the results)

Kedsy
09-01-2010, 03:21 PM
Doesn't make much sense to me - wasn't last year one of the shortest benches K has ever played? (obviously, that's not a complaint, given the results)

We had eight guys on last year's team who played 12+ mpg. That isn't a short bench at all, and certainly not shorter than a lot of Duke teams, like 1999-2000, when we only had six guys with more than 8.5 mpg, or 2005-06 when we only had seven guys who played more than 6 mpg.

Bluedog
09-01-2010, 03:33 PM
"Well, it's a long tournament, and we need to develop our entire team," U.S. coach Mike Krzyzewski said. "My position is based on running a marathon and not a sprint. That's something I've learned over the last five years. We need 12 guys to win. As we keep going along, it's not one guy. Not one guy."

Before reading this, it had seemed to me that K has been moving away from having as short a bench as he had in the past, but I've ascribed a lot of it to the players on the team. This quote makes me think that maybe he's had a change in his coaching philosophy. I don't know that Ryan Kelly or Andre Dawkins would have appeared in as many games as they did last year if this had been an earlier edition of Duke. (More true of Kelly than Dawkins, obviously, because we were shorter on Dawkins' position).


Doesn't make much sense to me - wasn't last year one of the shortest benches K has ever played? (obviously, that's not a complaint, given the results)

Uh, you guys know that he's referring to the US national team, right? Not Duke...

Olympic Fan
09-01-2010, 03:36 PM
Before reading this, it had seemed to me that K has been moving away from having as short a bench as he had in the past, but I've ascribed a lot of it to the players on the team. This quote makes me think that maybe he's had a change in his coaching philosophy. I don't know that Ryan Kelly or Andre Dawkins would have appeared in as many games as they did last year if this had been an earlier edition of Duke.

Sigh ... I thought last year would have killed the myth that K doesn't develop his bench.

Deja vu all over again. I can't wait for the threads about how his teams shoot too many 3s or doesn't play enough zone or doesn't play enough true road non-conference games or is too short to be a good quarterback ... oops, wrong sport (although I see where Cut pre-empted his one by jumping Renfree from 6-3 to 6-5 on the latest roster. Quite a growth spurt!).

HaveFunExpectToWin
09-01-2010, 03:38 PM
Doesn't make much sense to me - wasn't last year one of the shortest benches K has ever played? (obviously, that's not a complaint, given the results)

I could be wrong, but I thought we had a pretty deep bench last year. 8 guys averaged over 10 min/game and Ryan averaged 6.5 min. 2005-06 was a much shorter bench with only 7 guys getting a good amount of playing time.

Wander
09-01-2010, 03:43 PM
We had eight guys on last year's team who played 12+ mpg. That isn't a short bench at all

We only had nine scholarship players. We had three guys who more or less played every single important minute that was available to them. Kenpom puts Duke in the bottom 10% of 300+ teams for 09-10 in bench usage. There's no way to argue we didn't play a short bench. Again, that's not a complaint or a criticism. As Bluedog said, I imagine K's comments were only in reference to the Olympic team.

Bob Green
09-01-2010, 03:45 PM
This quote makes me think that maybe he's had a change in his coaching philosophy. I don't know that Ryan Kelly or Andre Dawkins would have appeared in as many games as they did last year if this had been an earlier edition of Duke. (More true of Kelly than Dawkins, obviously, because we were shorter on Dawkins' position).

Coach K and the short bench is an Urban Legend formulated and perpetuated by the anti-Duke crowd, specifically, the inhabitants of Chapel Hill.

To debunked the short bench myth, one needs only to review the minutes played by past Duke teams:

Take a look at the 1997-98 (http://goduke.statsgeek.com/basketball-m/seasons/season-stats.php?season=1997-98) team with nine players averaging double digit minutes per game with a 10th player at 9.9 mpg and an 11th player at 8.0 mpg.

Or take a look at the 1990-91 (http://goduke.statsgeek.com/basketball-m/seasons/season-stats.php?season=1990-91) team with nine players averaging double digit minutes, or take a look at 1989-90 (http://goduke.statsgeek.com/basketball-m/seasons/season-stats.php?season=1989-90) with nine players averaging double digit minutes.

There are additional years where Coach K played a deep bench. The bottom line is when Coach K has talent on his bench, he plays it.

Wander
09-01-2010, 05:15 PM
Coach K and the short bench is an Urban Legend formulated and perpetuated by the anti-Duke crowd, specifically, the inhabitants of Chapel Hill.

It's not a myth. Look at the kenpom numbers for the years the data is available (I think the last five or so). There are many coaches - probably the vast majority - who would play Andre Dawkins and Ryan Kelly more than K did this year.

The part that's a myth - and I strongly agree that it's false - is that playing a relatively short bench is inherently a bad thing. And I'd argue 2010 proves it.

-jk
09-01-2010, 05:31 PM
I'd like to look at the data, but I'm on my BB. Do you have a link?

-jk

sagegrouse
09-01-2010, 05:46 PM
Coach K and the short bench is an Urban Legend formulated and perpetuated by the anti-Duke crowd, specifically, the inhabitants of Chapel Hill.



I don't like to argue with the ever-knowledgeable Bob Green, but now that he is no longer an East Asian mystic, maybe some of the mysticism has been washed off on the wet side of the Cascades.

IMHO, where the H is most often silent, the "short bench charge" comes from overstoked DBRers who want to see their favorite recruit on the floor as soon as they join the team. And who can blame them -- er, us -- after we get so excited with the new recruits and want to see them in action.

sagegrouse

Olympic Fan
09-01-2010, 05:54 PM
It's not a myth. Look at the kenpom numbers for the years the data is available (I think the last five or so). There are many coaches - probably the vast majority - who would play Andre Dawkins and Ryan Kelly more than K did this year.


The myth is because people pick selective data and use it to support a theory based on a few seasons.

Coach K has coached at Duke for 31 seasons. He's one of the most adaptive coaches in college basketball. He's had teams with short benches ... teams with deep, long benches ... with medium benches. He's had teams that extend pressure. He's had teams that pack it in (even if they are still playing a version of man-to-man). He's had big teams (last year) and small teams (1991). He's had teams that shoot a ton of threes and teams that work it inside.

The myth that "he doesn't develop his bench" or that he's wedded to a certain style is part of the anti-Duke and anti-K propaganda spread by the Duke haters. I hate to see Duke fans buy into the big lie.

K has had teams where he didn't feel comfortable with his bench -- his 2000 team might be the most obvious example of that. He'[s had teams that go as deep ay any team anybody has ever had -- his 1998 and 1999 teams fit that bill.

Last year, Duke had nine players average double figure minutes. How does that work out to a short bench?

Well, four teams did have more players get double figure minutes last year -- BC, Miami and Virginia had 10 each and UNC had 11 players ... and all four teams just happened to finish among the bottom five in the ACC.

Two teams had shorter benches -- Maryland had eight players get double figure minutes and tied Duke for the regular season title. Virginia Tech had just seven players with double figure minutes -- and they won 10 ACC games and finished third.

Everybody else in the league -- including Duke -- had nine players average double figure minutes.

It's like the Rob Mahoney story that DBR linked to yesterday. The US opens with 60 points and 19 rebounds from its bench against Croatia, then follows with 46 points and 26 rebounds from the bench against Croatia. ESPN's experts use a graphic to show how K is using his depth to crush opponents.

Then comes Brazil and the bench plays like $%#@. 6 points (on 1-7 shooting) with 5 rebounds, no assists and two turnovers. Suddenly Mahoney's point is that K isn't using his bench enough. Well, he might have given them more than a combined 39 minutes if just one of them was contributing.

Today, the bench played well 55 points and 23 rebounds. They got plenty of minutes.

The question is, did K all of a sudden shorten his bench against Brazil because he didn't know how to use it ... or because he didn't like what he saw from the subs on the floor?

When K doesn't play his subs at Duke, it's because they haven't earned the time. A guy can be buried on the bench and still earn himself a major role because of his play in practice -- Elliot Williams is a recent example of that ... to some extent, so was Brian Zoubek late this year.

THAT is the one constant of Coach K's coaching philosophy -- the best players play. Period.

OldPhiKap
09-01-2010, 06:02 PM
There are many coaches - probably the vast majority - who would play Andre Dawkins and Ryan Kelly more than K did this year.

The vast majority of coaches don't have the talent ahead of Andre and Ryan that K had last year.

Their time will come, hopefully this year. But what coach would have sat Scheyer and Nolan? Or Lance and the Plumlee Bros.?

To be clear, I love Andre's heart and I think both of them will be great players. Better each season. But I can't see how anyone can really say that K should have played Scheyer (senior captain), Nolan (most explosive player), Singler (could've gone pro), Lance (senior, best defender), Zoubs (mountain), or the Plumlees (solid bodies and performers) any less. Dawkins came in and spelled the guards, and Ryan was at a position in which there were a lot of bigger, stronger players ahead of him. This year, it's a new ball game and a new opportunity.

Wander
09-01-2010, 06:14 PM
The myth is because people pick selective data and use it to support a theory based on a few seasons.
...
The myth that "he doesn't develop his bench" or that he's wedded to a certain style is part of the anti-Duke and anti-K propaganda spread by the Duke haters. I hate to see Duke fans buy into the big lie.


I'm using the data available to me, which is the past five seasons or so. I'm also aware there are certain seasons one can point to where K has played a long bench. And, though Coach K is pretty versatile, like every coach in the history of every sport he does have certain trends and preferences. One of these is to play man-to-man a lot. Another is to use a shorter bench relative to what other coaches would use. I have no problem with either of these things - in fact, I prefer them.



Last year, Duke had nine players average double figure minutes. How does that work out to a short bench?


Two things. First of all, one of those nine players is Olek Czyz at 10.2 minutes per game. Second of all and more importantly, the number of players averaging 10 minutes or more a game is a totally arbitrary stat. What kenpom tracks is the not perfect but much more sensible percentage of available minutes bench players play - and Duke was in the bottom 10% of that category last year.

Wander
09-01-2010, 06:18 PM
I'd like to look at the data, but I'm on my BB. Do you have a link?

-jk

Sure. The data is the percentage of available minutes that bench players play. Duke has ranked, out of 300+ Division 1 teams, the following places in that category:

09-10: 315th
08-09: 144th
07-08: 220th
06-07: 307th

Unfortunately, it doesn't go back further, but I don't think anyone on here is going to argue that 2006 or 2005 were deep teams.

basket1544
09-01-2010, 06:29 PM
I love how everyone has to chime in on what USA basketball could be doing better. If Kevin Durant doesn't play alot of minutes (against Slovenia he played only 23 minutes), then Coach K doesn't know what he's doing and obviously should play Durant more. If Durant plays almost the whole game (Brazil and Spain games), then Coach K doesn't know how to use his bench effectively.
I think the commentators need to stick to reporting on the game and not on how they would substitute players better. Obviously Coach K knows what he's doing or he wouldn't have all the wins he has.

Acymetric
09-01-2010, 06:58 PM
I think that the Kenpom data is pretty uninformative if you're comparing Duke to every other team in D-I (most of which aren't that good at all). It would make more sense to isolate the top teams each year and compare bench usage for them. How Depaul or Hofstra uses their bench doesn't mean anything to me. If you want to look at how elite teams should use their bench compare them solely to other elite teams (defining "elite," of course, becomes the issue here). Being in the bottom 10% sounds low, but may not be if other top teams have low bench usage too.

For my money, high bench usage implies that your best players aren't good enough to command more minutes. Is that a good thing?

Plus, the difference in percentage of minutes played by the bench between us (315) and Kansas(182) is only 7.3%. Do the math and it basically means that on average our starters are playing slightly less than 3 minutes more than the starters at Kansas (which is just inside the top 50% for bench usage).

Other top teams from 2010:

Kentucky: 233
Syracuse: 279
Ohio State: 347
Baylor: 316
Kansas St: 213
West Virginia: 290
Wisconsin: 285

I don't really see how you could argue that Duke's bench was really that much shorter than several other top teams. Keep in mind we were even closer in bench minutes to all of these teams than we were to Kansas. Where's the issue?

Bob Green
09-01-2010, 07:21 PM
I don't like to argue with the ever-knowledgeable Bob Green, but now that he is no longer an East Asian mystic, maybe some of the mysticism has been washed off on the wet side of the Cascades.

sagegrouse

I definitely got a chuckle out of your post. I'm not sure I've ever been mystic but I will concede living on the wet side of the Cascades will rust :cool: your senses.

OldPhiKap
09-01-2010, 08:24 PM
I definitely got a chuckle out of your post. I'm not sure I've ever been mystic but I will concede living on the wet side of the Cascades will rust :cool: your senses.

As the great Canadian Mystic Neil Young noted, "rust never sleeps." So be careful with that.

Dukeface88
09-01-2010, 09:07 PM
Numbers about K's bench:

I looked at the top 15 teams accordin to Pomeroy's efficiency figures from the past 4 years. Plugging into Excel, the average minutes played by the bench for those teams is 28%, with a standard deviation of 4.8%. The average ranking of the teams was 217, with an SD of 89. Over the same period, Duke averaged 26.7% and a ranking of 246. That's slightly below average, but easily within 1 SD, so not unusual.

Last year's top 15 averaged 25.8% (SD 4.3%) and a ranking of 267 (SD 61), compared to Duke's 23.4% and 314. Again below average, but well inside the expected range.

I haven't run the national team's numbers, but glancing at the stat sheet, the bench appears to be playing more minutes than a Duke team over the last four years. I would submit that K is going deeper on to the bench because the talent disparity between the bench and the starters of the national team is probably smaller than the last few year's at Duke.

However, that's only looking at the total bench time. A better metric might be the number of minutes for each recruited scholarship player (since we don't know the number of minutes played by walk-ons in blowouts). This could also account for last year's below average numbers (that is, K could have been making more use of his available bench players, but had fewer of them). However, I don't know of a convenient source for those numbers.

Newton_14
09-01-2010, 10:53 PM
I agree with Olympic fan and thought his post was a solid fact based argument. Well done.

I do see where Wander is coming from and he also has stats to back up his argument. Job well done there also.

Looking at this years team on its own, K played the only cards available. We will never know for sure how Olek would have fared so taking him out of the equation is fair in my view.

I think there are 2 points regarding this year that get overlooked by some pundits. Point 1 is the fact that K had very limited options. He had no reserve PG on the bench (Love you Jordan but counting recruited players only). The only real back up PG to Jon was Nolan who was also the starting SG. This meant 1 of Jon and Nolan had to be on the floor at all times unless the score was like 82 to 50 or something:).


K also did not have a true small forward on the bench to back up Kyle. So for 3 positions he basically had 1 bench player to spell those guys and that was Andre, who oh by the way came to the party late, missed the entire summer, and was supposed to be dominating in High School. Andre could spell at shooting guard and a little bit at small forward. He started out well, but the death of his sister combined with missing the summer hurt his progress badly.
K did have depth at the 4 and 5 and he used it well. He had 5 guys for 2 positions and 4 of them were used game in and game out. Ryan was the odd man out but he was simply the 5th best in a 5 man pool. That will change this year though!

Point 2 is K believed (and he stated this in interviews) that he had to prepare the Big 3 to play heavy minutes in big games by playing them heavy minutes night in and night out. Many would say resting them in the season would have "saved their legs" for tourney time. As a former basketball player, I agree with K's approach. If he would have had 10 to 12 capable, ready players, he would have played it differently.

I feel he played the bench this year to perfection and really it was his only hope to get the results he got.


All that said I would bet anyone right now the rotation this year will be vastly different from last season as K will have far more options especially at the 1 thru 4 positions..

jacone21
09-01-2010, 11:26 PM
... unless the score was like 82 to 50 or something:).

...


Well played, sir. *golf clap*

COYS
09-02-2010, 09:40 AM
Numbers about K's bench:

However, that's only looking at the total bench time. A better metric might be the number of minutes for each recruited scholarship player (since we don't know the number of minutes played by walk-ons in blowouts). This could also account for last year's below average numbers (that is, K could have been making more use of his available bench players, but had fewer of them). However, I don't know of a convenient source for those numbers.

This is also a really good point. In recent years, Duke has frequently been under the maximum in terms of the scholarship players on the roster (someone more knowledgeable will have to tell me if this has been the case through much of K's tenure). After Olek left, this season was no different. Ryan was the only scholarship player who didn't always play important minutes down the stretch . . . and as has been mentioned by numerous posters, he was the weakest defender/rebounder on a team with 5 bigs that needed outstanding rebounding/defense from its front line. This is not a noteworthy event, in my opinion. Even going back to teams like 2006 that are often exhibit A for those who claim that K doesn't develop his bench . . . who on the 2006 roster could have spelled JJ or Shelden for significant minutes? Rewind to 2008 and you can easily see that if coach has the players to spell his stars, he will use them.

uh_no
09-02-2010, 09:56 AM
Many would say resting them in the season would have "saved their legs" for tourney time. As a former basketball player, I agree with K's approach. If he would have had 10 to 12

I was hoping that after last season I would have never heard that point even mentioned again....I know you don't think that way, but if anyone does, I will personally drop a

DUKE 2010 National Championship

on their head

JasonEvans
09-02-2010, 10:18 AM
One important thing that stats cannot factor into bench usage is how the minutes get distributed.

For example, what if I showed you a team that had 11 players who averaged double-digit minutes. You would probably think they had a deep bench and their players never got tired, right? But, what if those minutes broke down as 3 players who played 40 minutes each (every minute of every game) and 8 guys who each played 10 minutes a game.

Obviously, what I am talking about above is an extreme example but it points out some of the folly in looking at statistics to measure bench usage. Another example would be a situation where one guy starts (perhaps a senior) but there is a sub who comes in for him fairly quickly and plays 25 or 30 minutes a game as a stud "6th man" -- sorta like Vinnie Johnson (for those of you old enough to remember the great Pistons teams of the late 80s and early 90s).

Anyway, I will conclude with the point many others have made -- K plays the guys capable of helping the team and winning games. Some years that means using more of his bench players, some years it means using less. He adapts to each given season... and he adapts damn well!

--Jason "I suspect we will see a longish bench this year -- partially because we will have leads that allow more bench usage" Evans

OZZIE4DUKE
09-02-2010, 10:30 AM
1) a) I don't need to read this thread, 'cause the HPR says it all. :cool:

1) Koaching Koncerns
a) Krzyzewski’s management of the rotation will bite us/has bitten us in the I'm a real wanker for saying this.I'm a real wanker for saying this.I'm a real wanker for saying this..

OZZIE4DUKE
09-02-2010, 10:34 AM
1) Koaching Koncerns
a) Krzyzewski’s management of the rotation will bite us/has bitten us in the I'm a real wanker for saying this.I'm a real wanker for saying this.I'm a real wanker for saying this..1) a) I don't need to read this thread, 'cause the HPR says it all. :cool:
OH MY GOD! I can't believe the friggin filter on this &#@$%* software actually wankers out the HPR now. Very disappointing.

Now the HPR doesn't say it all. It's a wanker.

Scorp4me
09-02-2010, 10:57 AM
One of the major factors I think we're forgetting is who is sitting. As someone else mentioned earlier much of this myth is perpetuated by the anti-Duke crowd and mentioned that they hated to see Duke fans buying into it. I think the Duke fans who buy into it though do so for a different reason.

As someone else mentioned we have our favorite recruits and we want to see them play. So when Andre hits a dry spell and we don't see him for a while we grow frustrated that K isn't "using his bench". But the point is who do we sit just so he can play?

It's a combination of having great players on the bench who aren't being played and often even better players on the floor who you can't afford to sit. The Duke fans lament the players on the bench. The opposing fans lament the talent already on the floor. And out pops complaints about K using a short bench for different reasons.

Except for when Pete Gillen complains. Then out pops great quotes about building statues to players like Taymon Domzalski