Dr. Krzyzewski or: How I Learned to Stop Worrying and Love the Stall

[pfrduke]

This was a prime example of the "stall" almost blowing the game. Duke had 3 plays in the last minute that if any of them had gone the other way they could have lost (the offensive rebound of Dre's missed 3, Kyrie's floater and Morris' missed shot in the lane).

There's no question we didn't execute particularly well in the last 6:25. But that's not a condemnation of the stall as an overall strategy - no end of game strategy is 100% successful. We could just as easily have tried to force offense earlier in the shot clock, resulting in missed shots/turnovers, leading to even more opportunities for Michigan to score.

As it was, we limited Michigan to 9 offensive possessions in the last 6 1/2 minutes of the game. They executed extremely well and we defended very poorly - they got 15 points in those 9 possessions. We had 8 offensive possessions in that same stretch, during which we executed poorly - scoring just 5 points. That's 6 1/2 minutes of really outlying performances by both teams on both ends of the court and it still wasn't enough to get Michigan all the way back.

Certainly the game was closer than we would have wanted, and much of that was due to poor execution of the closing strategy down the stretch. But again, I think execution (and, to a certain degree, randomness - there are simply going to be times when a strategy doesn't work) more than strategy is responsible for that. And in the end, the fact that there weren't more than 8-9 possessions in the last 6 1/2 minutes could have been the difference maker - another 3-4 trips for each team might have allowed Michigan the window it needed to fully complete the comeback.

[InSpades]

Yea, BUT the problem with what if's is you can play that game on both sides. What if AD's three goes in? What if Kyle had gotten the 4 points he left on the foul line earlier, what if they miss one or just two of the three's they made...

Kyle missed his free throws in the 1st half, what do they have to do with the discussion to go into stall mode? Michigan was 7 of 21 on 3s... it's not like they were on fire or something. Obviously Duke had chances to extend their lead in the last 8 or so minutes and failed on most of them. That was unlucky... but they didn't exactly make their own luck there either.

I think the discussion is better if we don't apply it to any one game, but look at it over time and measure the results. Any one game can be pointed to and argued either for or against the stall ball strategy, but then once again you get into the what if's. Look at the record, 900 wins and counting. Speaking for myself, I just do not feel that I am in a position to question the tactics deployed by the soon to be winningest coach in the history of the NCAA...

How do you measure the results of stall ball? I mean... if you're in the position to go to stall ball then you have an extremely high chance to win the game no matter what strategy you employ the rest of the way. The idea that "it's worked a lot, it must be a good strategy" is flawed. It's like the guy who goes all-in pre-flop with 7-2 off-suit and hits quad 7s. Sure it worked but that doesn't make it the right play.

It's not like there's a universal consensus that stall ball is the way to go. A vast majority of coaches use it a lot less than Coach K. Are they all wrong? I'm not saying you should never go to stall ball, I just think Coach K went to it too soon against Michigan and against precisely the wrong defense. I'm sure he had his reasons... maybe Kyle's 4th foul. Maybe just his sixth sense on how the momentum was turning. Duke usually executes far better in the stall offense and I think it's great. If you can get just as good a shot in 30 seconds as you can in 15 seconds then why not go for it? However when you get a significantly worse shot in 30 seconds than in 15 seconds I think the benefit of stall ball needs to be weighed. I would've been much happier if Duke spent 30 seconds attacking the zone. Potentially getting fouled. Potentially finding a big open under the basket. Potentially getting a clean look at a 3 for a good shooter. Instead of pounding the ball against the floor for 15 seconds.

[SCMatt33]

I don't think that stalling can be characterized in general. It all depends on how good the team is at running it as opposed to our normal offense. The last two years, the team has been as good in the stall-fense as I've ever seen. I think about that Maryland game where Duke had two great stall possessions in the last two minutes. On the first one Nolan drove into the middle for a layup. On the second, Nolan and Kyle ran a perfect pick and pop at the top of the key. Both defenders stayed with Nolan honoring what he just showed in the last possession leaving Kyle wide open for a three to ice it.

As for Sunday, I just watched Duke's last several possessions (the last 5 minutes) on offense, and while they were hit and miss, they weren't all terrible like I had thought Sunday (probably had something to do with the nerves of the game).

1. Looks like it was supposed to be stall-fense, but Dawkins took a corner three off of a pass from the wing from Kyrie with 15 left on the shot clock. I'm not sure that Kyrie expected him to take that shot, but Dre had to turn 90 degrees to take the shot, and a long rebound led to a transition layup. This was probably the worst possession of the bunch because it was a quick shot without any legit attempts to get the ball inside the three point arc at any point.

2. Nolan had a great bounce pass in the middle of the shot clock off of dribble penetration that was not caught by either Miles or Ryan (I forget which one). Again, this was only a semi stall, because we ran time off without doing much, but didn't wait until 10 seconds to do something. This was a good possession that just didn't get finished.

3. After a missed three, this was the first true stall possession. There wasn't much going on, but at the end, Nolan passed from about the top of the key to Kyrie who took one dribble for an 18-foot jumper. It wasn't a great possession but was saved by Ryan Kelly with an easy put-back.

4. After Michigan got a layup in a timeout, Duke had what I thought was a pretty good stall possession that unfortunately ended with a badly missed shot. Kyrie drove into the lane and kicked to Nolan for three. It was like the Dre three where he had to turn. With Kyrie getting into the lane, Nolan was able to catch it facing the basket, but he just missed the shot badly, and didn't even give anyone a chance for a rebound.

5. Duke had actually responded to the last possession well, but Michigan made a tough shot clock buzzer beating turnaround jumper to wipe it away. This possession ended up with the shot clock violation, but I thought it was a well run play until Kyrie made a mistake near the end. Kyrie drove into the lane, but Michigan stepped in to take a charge. Kyrie had just picked up a charge on the possession before the first one mentioned here, and instead of taking a Nolan style floater, he dumped it to Kyle on the baseline, who really wasn't in good position to make a move. By the time he passed it back and Kyrie tried to get a shot off, it was too late.

6. Duke's final stall possession. This was another one where Dre took a less than ideal three, but given the shot clock, it wasn't nearly as bad as the earlier one. Duke got lucky that Michigan couldn't control the rebound and got it out of bounds. Instead of worrying about the defense setting up like the possession before, Kyrie made a nice little floater and the rest is history.

All in all, I saw two semi-stall possessions and four true stall possessions. I have to say, 4 of them were executed well, but only one ended with points. Ironically one of the two not-so-great possessions actually got points as well. I think one of the things that got lost is that Duke let the lead get whittled down from 15 to 8 in the normal course of the game. From that point Tim Hardaway Jr. made two consecutive tough shots during those stall possessions while Duke missed some open ones and you wind up with a two point game instead of an 8 point game. In this way, however, the stall-fense worked to perfection, because instead of being down 1 with 3 minutes to go after a run, there was only a minute left and Duke was able to hold on.

[Kedsy]

If only a game played out that way.

Other scenario:

-- we lose the initiative on offense, and end up with a bad shot.
-- other team pushes, gets a lay-up or a three point play. Crowd reacts.
-- they press, we turn the ball over. Crowd gets louder.
-- other team scores again. Crowd erupts.
-- we walk it up, come away empty on a long outside shot.
-- they come down and score on the run-out. K calls time out because momentum has clearly shifted.

Suddenly, the 12 point lead with 5 minutes to go is now a 5 point lead with 3 to go. Back to a two-possession game.

And, all that took was two empty trips and a turn-over.

And we still have a five point lead plus the ball. K draws up a good play during the timemout, we score to go ahead 7 with under 3 minutes to go, and the other team has to start fouling.

Assuming we hit most of our free throws, even in your worst case scenario we probably win.

Of course, if we draw up a second worst case scenario, where we come out of the timeout and miss a three, giving up a runout and now it's a one-possession game with under 3 to go, then we have to go out and play straight up and try to regain the momentum. Assuming we deserved the initial 12 point lead because we're the better team, we still have the advantage, although obviously not as much because they have the momentum. We may have to go down to the last possession, like we did against Michigan.

So, it seems to me, using this unscientific analysis, that going into the stall has the following chart:

(A) stall goes well, we win;
OR
(B) disaster (your scenario), leading to timeout, Duke;

(B)(1) we get the first score after the timeout; they start fouling;
OR
(B)(2) they score first after the timeout, and it's a one possession game;

(B)(1)(a) we hit most of our free throws, and we win;
OR
(B)(1)(b) we miss our first couple free throws and it's a one possession game;

(B)(2) & (B)(1)(b) lead to us playing the last two and a half minutes straight up, with our opponent having the momentum and us having a three point lead.


If every branching on this decision tree has a 50/50 chance, we still win 81.25% of the time. If you assign a 70% probability to (A) (stall goes well); a 60% probability to (B)(1) (we score first, with us starting with the ball); and a 67% probability to (B)(1)(a) (we hit most of our free throws), and a 50/50 probability to the last few minutes (although that might be low since we have a three point lead), then we win 91% of the time. And even though I just made those percentages up, I think my estimates are pretty conservative.

What would our winning percentage be if we played straight up every time we had a 12 point lead with 5 minutes to go? Would it be more than 91%? I don't know, but in your hypothetical, the other team is scoring 2.33 points per possession. Five minutes to play would generally give each team 8 or 9 possessions, although obviously teams might hurry up in the last few seconds. If they score 2.33 points per for 9 possessions that's 21 points. If we score our average of 1.2 point per for 9 possessions that's 11 points, which would leave us with a 2 point win. If we only score 1.0 point per and it's overtime, and any less than 1.0 and we lose in regulation. To me, that doesn't sound appreciably better than a 91% probability to win. Or even 81%.

[Starter]

Mathematical equations or no mathematical equations, if Duke had lost, should we still stop worrying and love the stall? I know, I know -- they didn't lose. But they very well could have. And they have lost in the past decade in this exact scenario, in some very high profile situations I'm sure we're all quite familiar with, when the sand just didn't go out of the hourglass quickly enough. Not to mention, our notable size advantage was stripped when the Plumlees were glued to the bench (and they've been ballin' lately).

I'm a huge Krzyzewski fan. I think he's one of the four top coaches in America on any level, next to Phil Jackson, Doc Rivers and Bob Hurley. I feel we're blessed to have him, and I'm proud to have gotten to know him a bit. But like any other human being, 900 wins or not, he's eminently fallible. I don't buy into this embargo against evaluating, or even criticizing, a strategy he chooses to employ. No, I've never won a single game as a coach. But that doesn't mean I can't formulate opinions based on years of watching thousands of games.

Meanwhile, everyone's still defending him for employing a strategy he claims we weren't mindfully doing in the first place, which doesn't even make sense, since we clearly were. His 900th win and rationales following the game might have been shakier than any of the 899 that came before it. The important part is indeed that they did win, and move on with Kyrie in tow. But that doesn't mean we can't discuss and even question decisions and elements of the game that got us to that point.

[COYS]

I think one of the things that got lost is that Duke let the lead get whittled down from 15 to 8 in the normal course of the game. From that point Tim Hardaway Jr. made two consecutive tough shots during those stall possessions while Duke missed some open ones and you wind up with a two point game instead of an 8 point game. In this way, however, the stall-fense worked to perfection, because instead of being down 1 with 3 minutes to go after a run, there was only a minute left and Duke was able to hold on.

Thank you very much for pointing this out. This is how I remembered the game as well, but I had not had the time to check the play by play. Michigan actually gained the momentum BEFORE we went to the stall. If anything, the slow-down game was a reaction to the game getting away from us as we forced our offense against the 1-3-1 and started playing poorly on the defensive end. The argument has been that the stall almost cost us the game. Your post actually indicates that the stall may very well have saved us the game, delaying Michigan's comeback until it was just too late.

[-bdbd]

Originally Posted by InSpades:
"Did you watch the game? Do you think this is really true? Do you think when Kyrie and Nolan were dribbling and passing back and forth 30 feet from the basket they were "confused" by the zone?"

Well, I watched the game, and I think it was both. We may have been running the slow down game, but unlike the usual situation when we do so, when it came time to run an offense in the last 10-15 seconds of the shot clock, we (and especially Nolan and Kyrie) looked like we had absolutely no idea how to get a decent shot against the 1-3-1. I think there was a lot of uncertainty about how to attack that defense.

Good discussion lead-in by JK. I've done the (very similar) math in the past and don't generally have a problem with stall-ball if executed properly. But it requires a disciplined team with clear roles, which K usually has, that can resist the temptations of lost momentum and which knows exactly where it is going down the shot-clock stretch.

If you can hold the ball for 20+ seconds and then still get off quality shots/cuts, thereby minimizing the number of possessions in the game, that absolutely serves the team with the lead. However, versus Mich there WAS confusion -- but I blame this confusion a little less on the 1-3-1 and more on the Duke team chemistry/roles in the post-KI-return situation. It just seemed we didn't know who was going to take charge and run the play or make the cut or take the shot. But having a shot-clock violation and another possession with a poor (low-percentage) shot in the last 2.5 minutes of a close game is just unacceptable. (Thank goodness we play great D this year!)

The marginal benefit of running off 35 seconds versus 28 seconds is almost nil. So often we seem to START running our offense with only 7-8 seconds left on the shot clock (and the ball 35+ feet from the bucket). I'd much rather see us start at the 13-15 second mark, as the improved probability of making points certainly outweighs the "cost" of an extra 7 seconds.

I'm sure that the staff is working a lot this week on those end-of-game situations. Having the guys spend more practice time together over this Mon.-Thurs. certainly will help firm up the 'roles' questions a bit as well. I expect that we'll see clear improvement. 'just hope that we're in a position to run the stall come late Thursday night!! (And make those FT's - always helpful to stall-ball success!!)

[Kedsy]

Mathematical equations or no mathematical equations, if Duke had lost, should we still stop worrying and love the stall?

Sorry, but yes. The whole point of the stall is the math. No strategy gives 100% success. If you've watched thousands of games, like you say, and in 100 of them the stall was employed, how many resulted in a loss? Probably only a couple. Sure those stand out in our memories, but if our success rate with any strategy is 98% (and I'm not saying it is; obviously I just made those numbers up), then it's a pretty good strategy.

Now, if we'd lost 30 of those 100 games, then it doesn't sound like such a good strategy. But the number of games we've lost after employing a stall is a lot closer to 2 per 100 than it is to 30 per 100, and that's why the mathematical equation is a better judge than a one game sample.

[Starter]

Originally Posted by InSpades:


If you can hold the ball for 20+ seconds and then still get off quality chots/cuts, thereby minimizing the number of possessions in the game, that absolutely serves the team with the lead. However, versus Mich there WAS confusion -- but I blame this confusion a little less on the 1-3-1 and more on the Duke team chemistry/roles in the post-KI-return situation. It just seemed we didn't know who was going to take charge and run the play or make the cut or take the shot. But having a shot-clock violation and another possession with a poor (low-percentage) shot in the last 2.5 minutes of a close game is just unacceptable. (Thank goodness we play great D this year!)

The marginal benefit of running off 35 seconds versus 28 seconds is almost nil. So often we seem to START running our offense with only 7-8 seconds left on the shot clock (and the ball 35+ feet from the bucket). I'd much rather see us start at the 13-15 second mark, as the improved probability of making points certainly outweighs the "cost" of an extra 7 seconds.

Terrific points here. When you wait too long to start an offensive set during stall ball, you basically create a late-game scramble for yourself on every single possession. If you run some clock, shorten the game but still get off a good shot, that's probably the ideal scenario.

But against a tricky, junk defense, I want to be able to have enough time to figure out how to get the best shots I possibly can -- which is why our offense was so much more effective when they were playing straight-up (with a bigger lineup) earlier in the game. Taking shot-clock-affected lousy shots only helps Michigan get numbers on the other end anyway.

[UrinalCake]

2. Nolan had a great bounce pass in the middle of the shot clock off of dribble penetration that was not caught by either Miles or Ryan (I forget which one). Again, this was only a semi stall, because we ran time off without doing much, but didn't wait until 10 seconds to do something. This was a good possession that just didn't get finished.

I'm pretty sure it was Kelly that bobbled the pass, and I remember thinking at the time that this is why I don't like stall ball. The problem is that you have asked your guys to stand around for thirty seconds doing nothing (except for Smith and Irving, who basically lobbed the ball back and forth). Then, all of a sudden you expect them to make a "winning" play. Mentally you've put them in a mindset of complacency and of "let's not lose" rather than continuing to attack and be aggressive. So Kelly wasn't really ready to receive the pass, perhaps because at some level he's more concerned about the clock. On top of that, you're allowing the defense to rest for thirty seconds, which can be a factor at the end of a game.

Great mathematical analysis by the OP, but I have two quibbles. First off, the opposing team generally doesn't start fouling until there's less than two minutes left. In theory you could say that if you're down ten with four minutes left it might make mathematical sense to start fouling then, but I've never seen it happen. Secondly, the last five or so minutes of a game constitute a small sample size, where a single empty possession or turnover can drastically affect your average points per possession. If your opponent hits a three, steals the inbounds pass and hits another three, all of a sudden they've scored six points in what should be one possession. Just ask Maryland about this.

[OldPhiKap]

But against a tricky, junk defense, I want to be able to have enough time to figure out how to get the best shots I possibly can -- which is why our offense was so much more effective when they were playing straight-up (with a bigger lineup) earlier in the game. Taking shot-clock-affected lousy shots only helps Michigan get numbers on the other end anyway.

This.


If the team was confused as to how to attack, why not call a time out and tell them?

Again, just asking because it doesn't make sense to me.



(Great thread, btw -- thanks to the OP for posting it, even if we are on different sides of the question).

[Starter]

Sorry, but yes. The whole point of the stall is the math. No strategy gives 100% success. If you've watched thousands of games, like you say, and in 100 of them the stall was employed, how many resulted in a loss? Probably only a couple. Sure those stand out in our memories, but if our success rate with any strategy is 98% (and I'm not saying it is; obviously I just made those numbers up), then it's a pretty good strategy.

You're leaving out the human ability to adjust when a tactic clearly isn't working. If Duke had squandered this game, with a totally loaded team that has a golden opportunity to compete for a championship, you can tell me all you want about how your made-up numbers indicate doing this usually works. It would not have mattered. We went away from what was working for something that didn't, and left it that way. I can't possibly get on board with that.

Morris missed, we move on. I'd imagine it goes down as a learning experience.

[Starter]

(Great thread, btw -- thanks to the OP for posting it, even if we are on different sides of the question).

I completely agree, by the way. I absolutely respect both arguments, especially when this much hard work went into the mathematics behind them. Kudos to all.

[oldnavy]

Kyle missed his free throws in the 1st half, what do they have to do with the discussion to go into stall mode? Michigan was 7 of 21 on 3s... it's not like they were on fire or something. Obviously Duke had chances to extend their lead in the last 8 or so minutes and failed on most of them. That was unlucky... but they didn't exactly make their own luck there either.



How do you measure the results of stall ball? I mean... if you're in the position to go to stall ball then you have an extremely high chance to win the game no matter what strategy you employ the rest of the way. The idea that "it's worked a lot, it must be a good strategy" is flawed. It's like the guy who goes all-in pre-flop with 7-2 off-suit and hits quad 7s. Sure it worked but that doesn't make it the right play.

It's not like there's a universal consensus that stall ball is the way to go. A vast majority of coaches use it a lot less than Coach K. Are they all wrong? I'm not saying you should never go to stall ball, I just think Coach K went to it too soon against Michigan and against precisely the wrong defense. I'm sure he had his reasons... maybe Kyle's 4th foul. Maybe just his sixth sense on how the momentum was turning. Duke usually executes far better in the stall offense and I think it's great. If you can get just as good a shot in 30 seconds as you can in 15 seconds then why not go for it? However when you get a significantly worse shot in 30 seconds than in 15 seconds I think the benefit of stall ball needs to be weighed. I would've been much happier if Duke spent 30 seconds attacking the zone. Potentially getting fouled. Potentially finding a big open under the basket. Potentially getting a clean look at a 3 for a good shooter. Instead of pounding the ball against the floor for 15 seconds.

Well, the opposite argument would be that it has worked a lot and therefore we need to abandon it. Not sure I follow that line of logic, but I do get your point. In any ONE game we can look at stall ball and pick it apart or praise it depending on the execution of the team and the results. In this case, your arguments have merit, but in the overall scheme of things, I think going to stall ball the way coach K has done in the past is the way to go. And yes, I do think that his winning 900 games is relevant to the strategy he employs and makes it extremely difficult to criticize his methods when they result 900+ wins. Have we lost some games because of stall ball, yes, but how many versus how many were won? Others have made better arguments for using stall ball in previous posts than I can make.

My example of Kyle’s' missed front ends, was just to illustrate that with 4 more points things are not as close at the end. Having a 11 point lead with 3 minutes is a world of difference than a 7 point lead. I used that example with a couple of others in an attempt to point out the problem with citing specific plays or execution of plays to support or decry stall ball.

Listen, I have fussed and cussed about slowing the game down for years, and I do not think that it is a perfect strategy either, but it is one that has won a lot. Sure you can point to coaches that don't use it, but the problem with that is they are not Coach K and he is soon to be the all time leader in NCAA wins including NCAA wins, so that might not be such a good example to use in support of your argument. Give me an example of a coach who has won more consistantly who employs a different strategy if you can find one. Coach K isn't perfect, but he is the best.

[NSDukeFan]

I guess what I would like to see is a statistic on offensive efficiency when running our non-stall offense versus when we stall. Because my sense is that we are basically allowing the other side to cut the margin and betting that time runs out before the margin shrinks to zero. If so, all stall ball does is bring losing into the equation by giving us less margin.

If what we've been running has jumped us out to a big lead, why go away from it (absent foul problems)?

Especially when the other team is in a zone, so we cannot run our foul-line-extended spread?

In theory, and probably in practice, using less time to find a good shot probably results in a decreased offensive efficiency. The point, however, is that it doesn't matter. As long as offensive efficiency is not atrocious and defensive efficiency remains high, the team will still win, with the opponents getting fewer offensive possessions.

I think this may be the key point. We looked really bad against the zone. Against man to man, I think the foul-line-extended spread has generally been one of our best offenses. In fact, spreading the floor and letting Nolan go to work was how we jumped out to our big lead against Michigan. If Michigan continued to play man and we just waited for Nolan to attack at the 10 second mark on the shot clock, I think we would have had similarly excellent results.

I agree and would argue that this would be the case with a regular 2-3 and maybe even a 3-2 matchup zone as well. In those cases, the team can set a high screen for Nolan and he has been very efficient at getting an efficient shot in a short period of time. With the 1-3-1 having an extra guy up top, I just think it makes it a bit more crowded and difficult for penetration without any ball movement.

If Duke was so "confused" by the zone, wouldn't K have called a timeout and given them something to run against it? Duke called a timout at 9:20 left in the game (to halt a Michigan 7-0 run that cut the lead from 15 to 8). The next timeout called was at 1:18 after Hardaway hit the 3 to cut the lead to 1.

This was a prime example of the "stall" almost blowing the game. Duke had 3 plays in the last minute that if any of them had gone the other way they could have lost (the offensive rebound of Dre's missed 3, Kyrie's floater and Morris' missed shot in the lane).

I have been brought over to the side of the stall over the past couple of years and find the biggest argument against it typically is that the team "almost lost the game." On the way to the pre-season, ACC, regional and national championships, Duke almost lost a bunch of games, but they won a high enough percentage using stall ball that I think it worked very well. I don't know that it has worked as well this year, but Nolan is a greater weapon in getting a good shot in very little time than any of the big 3 were last year.

There's no question we didn't execute particularly well in the last 6:25. But that's not a condemnation of the stall as an overall strategy - no end of game strategy is 100% successful. We could just as easily have tried to force offense earlier in the shot clock, resulting in missed shots/turnovers, leading to even more opportunities for Michigan to score.

As it was, we limited Michigan to 9 offensive possessions in the last 6 1/2 minutes of the game. They executed extremely well and we defended very poorly - they got 15 points in those 9 possessions. We had 8 offensive possessions in that same stretch, during which we executed poorly - scoring just 5 points. That's 6 1/2 minutes of really outlying performances by both teams on both ends of the court and it still wasn't enough to get Michigan all the way back.

Certainly the game was closer than we would have wanted, and much of that was due to poor execution of the closing strategy down the stretch. But again, I think execution (and, to a certain degree, randomness - there are simply going to be times when a strategy doesn't work) more than strategy is responsible for that. And in the end, the fact that there weren't more than 8-9 possessions in the last 6 1/2 minutes could have been the difference maker - another 3-4 trips for each team might have allowed Michigan the window it needed to fully complete the comeback.

Very well put.

... (great discussion of the final possessions of the game)

Thanks for the recap on what actually happened at the end of the Michigan game.

Stall ball certainly isn't pretty and I think Kedsy brings up good points about % chance of winning in stall vs. not, but I think that it does give the team a better chance of winning if the team executes poorly down the stretch and if the team executes well, it doesn't take long for the game to be out of reach.

I wasn't a fan of the 3-point shot as a big weapon until I learned a bit more about efficiency numbers and now am drinking the coach K Kool-Aid on both using and defending the 3-pointer and stall ball. I agree with others that say that does not necessarily mean it is right and there can't be dissenting views, which is why I am enjoying this thread very much.

[Kedsy]

You're leaving out the human ability to adjust when a tactic clearly isn't working. If Duke had squandered this game, with a totally loaded team that has a golden opportunity to compete for a championship, you can tell me all you want about how your made-up numbers indicate doing this usually works. It would not have mattered. We went away from what was working for something that didn't, and left it that way. I can't possibly get on board with that.

You can't judge a strategy after the fact. If historically and mathematically stalling gives us the best chance to win, then it makes no sense to deviate from it.

And I assume you read SCMatt33's analysis, but we didn't "[go] away from what was working." Our lead had dropped from 15 points to 8 points before we went into the stall. If we'd played the next five minutes like we played the previous 5 minutes, it would have been a one possession game. Which it was anyway, but it would be much more accurate to say we went away from what was NOT working for something else that didn't work any better.

[OldPhiKap]

In theory, and probably in practice, using less time to find a good shot probably results in a decreased offensive efficiency. The point, however, is that it doesn't matter. As long as offensive efficiency is not atrocious and defensive efficiency remains high, the team will still win, with the opponents getting fewer offensive possessions.

We disagree on this (although your entire post was very well-stated and reasoned).

1. Our decreased offensive efficiency often leads to a decreased defensive efficiency, because the other team can push the missed shot upcourt and not face our set defense. Michigan had several lay-ups right down the pipe when we got beat on run-outs.

2. From the "best defense is a good offense" file -- if we have increased offensive efficiency in our regular set, and we get more possessions, we keep pushing the score out to the point that the other side has to start fouling us to keep the game in reach. Which gets us to the double bonus, and a better chance of setting up the defense after a made free throw.

3. Stall ball allows a squeezing of the lead margin in a race to run out the clock. As the game gets closer, though, the fans get back into it. In Charlotte (UNC country), that made the end very much like a hostile away game and engergized the opponent.

[superdave]

And I assume you read SCMatt33's analysis, but we didn't "[go] away from what was working." Our lead had dropped from 15 points to 8 points before we went into the stall. If we'd played the next five minutes like we played the previous 5 minutes, it would have been a one possession game. Which it was anyway, but it would be much more accurate to say we went away from what was NOT working for something else that didn't work any better.

Bingo. If our defense had held up and gotten us a couple of stops, this thread probably would not have happened either! We played poorly on both ends to close the game out.

[4decadedukie]

-jk: I appreciate your fine analysis, and I agree with it as far as it goes (in fact, I had also independently ascertained a Coach K "rule of thumb" of two points per minute in, roughly, the game’s final eight minutes). My concern, however, is that this approach’s essential underpinnings simply does not go far enough, do not incorporate the psychological as well as the arithmetic.

By "stalling," we obviously make the clock our ally and we reduce the possibility of a Duke player either fouling-out or being injured (sometime, as we all know and have observed, emotions are especially elevated during a highly contested game’s final minutes). Unfortunately, what I fear we additionally do is reduce the individual Duke player’s and the team’s aggregate aggressiveness -- going “for the throat” and putting the opponent away with certainty and finality. I am not at all sure that the average twenty-year-old student-athlete (even those with superb talents, intelligence and coaching like our guys) is able to modulate ferocity and “killer instinct” with the time-precision and exactness of activating and deactivating a light switch. Further, I particularly worry that allowing the game’s initiative to migrate to our adversary risks generating an antagonist’s run at the worst time for Duke.

The leadership skills we developed, learned and practiced in the Navy, which do not differ materially from those Coach K mastered at West Point, strongly suggested that the exact time one wishes to be unremittingly aggressive is when the opposition is in extremis (badly behind with time expiring).

The foregoing are my feelings re “stall ball.” Representing the converse position -- that, used with exactitude and judiciousness, it is an effective tactic -- is (IMHO) history’s greatest, most proven, and most enduringly successful collegiate basketball coach, Michael William Krzyzewski. I have VERY little doubt (or concern) that Coach K knows exactly what he is doing, but I must admit my cardiovascular and pulmonary rates are less convinced than is my brain.

[InSpades]

And I assume you read SCMatt33's analysis, but we didn't "[go] away from what was working." Our lead had dropped from 15 points to 8 points before we went into the stall. If we'd played the next five minutes like we played the previous 5 minutes, it would have been a one possession game. Which it was anyway, but it would be much more accurate to say we went away from what was NOT working for something else that didn't work any better.

We went away from something that had worked for 29 minutes and didn't work for what... 90 seconds? There's no reason to think it wouldn't have continued to work for the last 8 minutes. We went to something that absolutely didn't "work" for the last 8 minutes. It may have ended up winning the game, but we were grossly outplayed during those last 8 minutes.

As a fan I'd much rather see Duke lose playing their game than lose trying to play the clock. I know that's entirely irrelevant to Coach K and maybe everyone else. There's a chance we lose either way, but one of them would leave a terrible taste in my mouth.