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

[Indoor66]

You're being melodramatic. Duke won by twelve. That's more than "a couple of more [made] shots" for carolina and a few more missed free throws for Duke. The closest they got was 8, after Paige's four point play. At that point there were 2:49 left. Without Carolina fouling there probably would have been 5 or 6 more possessions. I suppose there is a way they could have won, but you have to construct some pretty unlikely scenarios for that; mathematically possible, but very unlikely. And all of this is with Duke turning it over 6 times in the last 5 minutes!

I also dispute the notion that players forget how to play offense just because of stall ball. I think that's silly. You just don't see Duke have to come out of the stall very often... because the strategy works.

The fact is, the issue is how many and not how. Thus far, K has come out on the long end of the how many a lot more than on the short end. The strategy works.

[TampaDuke]

I'm not sure this would be a fair comparison. If we're in the stall, the players have also presumably been instructed not to foul on defense and simply to try and make the other team take some time before they score. It's a lot harder to play D if you're not allowed to foul, and really the point of the stall is take the opponent's offensive efficiency out of the equation.

No doubt they're not supposed to foul when we're protecting a large lead late in the game. No disagreement there. Not sure what you mean by "taking the opponent's offensive efficiency out of the equation"? I know we're trying to shorten the game by limiting possessions, but are you saying that it doesn't matter if they score more per possession? That seems counterintuitive to me, but I might be misunderstanding you.

In any event, my only point here is that if the use of the stall on offense nearly always coincides with the tactic of playing less intense defense to avoid fouls, then that factor (the loose defense) should surely be part of the calculus in deciding whether the stall is good strategy or not. Perhaps looking at an opponent's offensive efficiencies is not the best way to account for it, but it should be accounted for in some fashion.

The assumption that we can make the other team use 15 seconds of clock on average when we are in stall mode seems inaccurate to me, particularly when we are almost always also trying to avoid fouling on defense.

[Kedsy]

Before the other team begins immediate fouling, we can burn about 30 seconds in each stalled possession. In the other direction, we try to force the other team to use at least 15 seconds per possession with a soft, low-risk full-court press and solid half-court D. That gives us a 45-second exchange of possessions, on average.

If we limit them to netting (pun intended) less than 1.5 points per 45-second exchange, we'll win.

If we average just over half a point per possession and hold them to average just under 2 points per possession, the math works. We should be able to average half a point per possession, even if we occasionally (or even three times in a row) get no shot off. On the other hand, it takes an extraordinary performance for a team to average 2 points per possession over multiple possessions. If we hold them to under 1.5 points per possession and 45 second exchanges then we don't even have to score to hold on. We need to value the ball, make occasional shots, and play smart defense. No turnovers. No fouls.

If the other team does start immediate fouling so that we go to exchanges every 15 seconds instead of 45, we need to hit 75% free-throws (shooting 2; 1-and-1 won't last long) to get 1.5 points per possession while still holding them just under 2 points per possession. We must inbound and pass to the best free throw shooters, and make sure everyone can shoot adequately.

No doubt they're not supposed to foul when we're protecting a large lead late in the game. No disagreement there. Not sure what you mean by "taking the opponent's offensive efficiency out of the equation"? I know we're trying to shorten the game by limiting possessions, but are you saying that it doesn't matter if they score more per possession? That seems counterintuitive to me, but I might be misunderstanding you.

In any event, my only point here is that if the use of the stall on offense nearly always coincides with the tactic of playing less intense defense to avoid fouls, then that factor (the loose defense) should surely be part of the calculus in deciding whether the stall is good strategy or not. Perhaps looking at an opponent's offensive efficiencies is not the best way to account for it, but it should be accounted for in some fashion.

The assumption that we can make the other team use 15 seconds of clock on average when we are in stall mode seems inaccurate to me, particularly when we are almost always also trying to avoid fouling on defense.

It is part of the calculus. If you look at the math as -jk outlined it (quoted above), you'll see that he's expecting the opponent to score almost 2 points per possession. Since in general the best offenses in the country score 1.2 ppp, and UNC, for example, scores on average 1.09 ppp, being OK with almost 2 points per possession assumes we're going to play porous defense. Our only goals on D presumably are to avoid fouling and make them run a few seconds of clock. So, no, it doesn't matter if they score more per possession, that's already assumed and added in. And that's what I mean about taking the opponent's offensive efficiency out of the equation.

[TampaDuke]

It is part of the calculus. If you look at the math as -jk outlined it (quoted above), you'll see that he's expecting the opponent to score almost 2 points per possession. Since in general the best offenses in the country score 1.2 ppp, and UNC, for example, scores on average 1.09 ppp, being OK with almost 2 points per possession assumes we're going to play porous defense. Our only goals on D presumably are to avoid fouling and make them run a few seconds of clock. So, no, it doesn't matter if they score more per possession, that's already assumed and added in. And that's what I mean about taking the opponent's offensive efficiency out of the equation.

Thanks for the clarification. I can see what you're saying. I don't necessarily disagree with the conclusion, but how does this calculus compare to the average non-stall/normal defense? What are the average possession times/efficiencies in a non-stall offense versus what you would see in the stall? If the tradeoff for running more clock is that we are less efficient on offense and at the same time the opponent is more efficient on offense, it seems the number of reduced possessions for the other team would need to be pretty high for us to start employing the stall above the 5-minute mark as we seem to do. I'd like to see how those numbers compare to the average game pace. Is the tradeoff worth it?

BTW, I'm not necessarily against the stall, but I do question going to it a little too early at times.

[rsvman]

I don't like stall ball, but not just because it sometimes loses a game or makes it a lot closer than it should be.
I don't like watching it.

Watching Duke basketball, it turns out, is entirely voluntary.

If all I cared about was that Duke wins the game....

I can assure that that's all many people care about. Even in the worst-case scenario, Duke is only playing stall ball for 4 or 5 minutes of playing time, right? So you can be thoroughly entertained by watching the first 3/4 of the game. And, since stall ball usually works, and you don't like to watch it, you could then turn the game off and check the result on SportsCenter or in your morning paper. Best of both worlds.
For me personally, I'm much more entertained when the team not only plays well, but goes on to win the game. And I prefer to watch to the end, even if the game is played at a slower tempo. To each his own, I guess.

I watch Duke basketball games because I want to
see our guys playing basketball well and hopefully better than the other team. Watching a guy stand there
dribbling around for 30 seconds and then chucking up a bad shot is not what I tuned in to watch.

So we've heard that it's not pretty. True. But a basketball game is not a painting on a museum wall or a symphony. It's a contest. The point of the contest is to win. If stall ball helps us to win, I'm good with it.
Now we are hearing that it's not entertaining enough. Well, OK, but a basketball game is not a movie or a play. It's a contest, and the point of the contest is to win.
If you all you want is entertainment, there are plenty of other options available.

[-jk]

Thanks for the clarification. I can see what you're saying. I don't necessarily disagree with the conclusion, but how does this calculus compare to the average non-stall/normal defense? What are the average possession times/efficiencies in a non-stall offense versus what you would see in the stall? If the tradeoff for running more clock is that we are less efficient on offense and at the same time the opponent is more efficient on offense, it seems the number of reduced possessions for the other team would need to be pretty high for us to start employing the stall above the 5-minute mark as we seem to do. I'd like to see how those numbers compare to the average game pace. Is the tradeoff worth it?

BTW, I'm not necessarily against the stall, but I do question going to it a little too early at times.

From observation, I think the usual time to start the stall it is when we're more than twice as many points as minutes left in the game.

-jk

[Clay Feet POF]

In my appropriately humble opinion, this has been a great thread on Stall Ball. I learned a lot. Thanks everyone!

My summary take:

1] Practice Stall Ball enough so that it becomes intuitive
2] Helps to have a really good PG and several very good ball handlers [duh, I know -- making a full list here]
3] Don't start it too early re the game clock [a function of extent of lead and possible remaining possessions]
4] Do it with good FT shooters
5] Execute it with a "want to win" mentality, not a "don't lose the lead" mentality
6] Have to remember to transition to high intensity defense -- different mindset each transition, not that easy to do
7] Against a good zone, start your final shot sequence earlier than 10"
8] Trust Coach K's infinite wisdom :-)
9] Now and then, maybe every third possession, or once or twice during the Stall Ball phase, attack on offense, to gain surprise and keep defenses wondering whether they will see Stall Ball or attack mode.
10] Care more about winning than athletic aesthetics [this is the one I have to become more mature about]

I can watch Stall Ball now with a more nuanced, less frustrated and desperate perspective. Thanks!

Go Duke!

What a thoughtful post! Really nice to see

[Kedsy]

From observation, I think the usual time to start the stall it is when we're more than twice as many points as minutes left in the game.

-jk

Using very rough math that makes sense -- if there are two possessions a minute and the opponent scores 2 points per possession, that gives them 4 points per minute. If we can score 1 ppp, that gives us 2 ppm, and gives our opponent a net 2 points per minute. So if we're ahead by more than twice as many points as minutes left when we start the stall, we ought to be ahead at the end of the game.

While it's true we might not score 1 ppp, it's also true our combined possessions might add up to more than 30 seconds per exchange and thus create fewer than two possessions per minute. So hopefully that evens out. If it doesn't, then to win we'd have to do something like get a couple of stops to keep their ppp under 2. All in all, if you start at the point you suggest, it seems you should have a really good chance to win.

[TruBlu]

From observation, I think the usual time to start the stall it is when we're more than twice as many points as minutes left in the game.

-jk

If we play UNC again, here's hoping we start stall ball at halftime.

[tele]

The fact is, the issue is how many and not how. Thus far, K has come out on the long end of the how many a lot more than on the short end. The strategy works.

I agree. To me running this endgame stall is no different than playing for the last shot, repeatedly. Hard to see how this would hurt the team and might just be good practice for those cclose game on the line last shot situations. It is a little boring, I suppose.

The interesting part of it to me is seeing when opposing coaches decide they need to start fouling. I think most miscalculate this and the game is over before they know it. I have my own idea of how you determine this, but won't post it here where future opponents might see it. I will say that this miscalculation may be why the stall ball works more often than not, as much as what the players do or don't do on the court.

[Newton_14]

I hear all the folks saying it is bad for this team and understand where you are coming from but I will point out we have not lost a single game this year using Stall Ball.

There are valid points on both sides, but like I said earlier, I see teams lose by not employing it (when they should) a lot more times that I see Duke or anyone else lose by employing. K's winning percentage in games where he went to the strategy is incredibly high. If Duke lost often while employing it, the anti-stall ball argument would have much more reason. With the really high success rate I just cannot argue against using it.

[-jk]

Pitt tried to give it away. Could really have used up a little more clock a few times. Phew!

GTHC!

-jk

[Kedsy]

Pitt tried to give it away. Could really have used up a little more clock a few times. Phew!

GTHC!

-jk

I was thinking the same thing as I watched the game -- if only Pitt could have employed a decent stall. Of course as bad as they shot free throws, even the stall might not have helped them.

At least they managed to win in the end.

[rsvman]

I was thinking that they need to decide whether they want to run some clock or attack the basket. On several occasions they were caught in a sort of "in between" scenario and it didn't work very well for them at all.

[1 24 90]

Did anyone watch the Nebraska blown lead against OSU? I was wondering if stall ball came into play in that game.