The Monroe-Drummond Tandem

Another negative against hack a blank is your allowing them to set up there defense better. I can't remember once that it worked against Ben and the Pistons, it might have closed the game abit.... But did we ever lose a game because of it?

I'm actually wondering if it is technically correct to foul the other team in a tie game when they have the last possession and can take the last shot.

If you figure that their odds of winning the game at that point are as follows:
40% chance of making the game winner with no time to respond.
50% chance of winning in OT if they miss.

Their odds of winning the game would be: 80%.


If you foul them... say you can target a 50% FT shooter like Ben Wallace or Drummond, then:

a) There is a 25% chance they miss both
b) Thee is a 50% chance they make 1 and miss the other
c) There is a 25% chance that they make both


In scenario a, you would then have the last shot. So, your odds of winning would be 80% (same analysis of their last shot odds including OT)
In scenario b, you would have the last shot, but a miss would be a loss. So, your odds of winning would be 40%.
In scenario c, you would have the last shot to tie... or take the lead with a 3. In the tie scenario, you would have a 40% chance of tying and then 50% chance of an OT win, so 20% odds of winning. In a go for it strategy, you would have say a 30% chance of winning, so you would opt for that.

If you combine those three scenarios and probability weight them, you have

.25 x .8= .2
.5 x .4 = .2
.25 x .3 = .075

Combined, your odds of winning if you foul them first would be approximately .475 vs .200 if you let them take the game winner while you "trust" your defense.


With the assumptions I just picked, it isn't even close. So why don't teams ever do this?


Also, it is pretty clear to me that when a team is down 2 points on a last possession, they should go for 3 unless the defense gives them the 2 pointer. i.e. not just when you're the road team, but always. The threshold would be that if you have a 2 point opportunity where the odds of conversion is above 60% (or more specifically 2 times your expected FG% on a game winning 3), you should attempt it. Many teams when they are at home will go for 2 and probably take a 40% type shot that is heavily contested.


If you could hook up an NBA video game to a neural net, it would be interesting to see what sort of non conventional end game strategies it would settle on through strategy evolution. In backgammon and other games, the non intuitive solutions that have become standard practice for top level players originated through that method.

Monroe and Stuckey for a disgruntled Kevin Love, maybe throw in a future first. Wouldn't work this year, but could when his option is on the horizon.

If Kevin Love is still holding a grudge, then I think that's teh perfect deal for both teams next year. STuckey will be an expiring and we'll be deciding whether or not Greg Monroe is worth investing in via a 5 year $65M contract, we'll know the answer by the trade deadline next year.

Throw Shabazz and Love on this team, and they could make a run at Miami. They would be one of the best shooting teams in the league, and THE best rebounding team. Love's threeball will come back. The kid can shoot.

If the team misses the shot and the game goes into overtime, which would happen 60% of the time:

* The non-shooting team would win the game: 50% * 60% = 30%
* The shooting team would win the game: 40% + 50% * 60% = 70% (NOT 80% as you state)

Quite right. I screwed up by doing (1 - (.4*.5))

Still, seems like the play, doesn't it (assuming I don't have any other errors)?


Many complicating factors, but with such a big buffer in the back of the napkin calculation, seems like detailed analysis couldn't overcome the spread.



By this same logic, if the above strategy checks out under serious scrutiny, then it leads to one more strategic issue.

If you are down 2 points with less than 24 seconds left on the clock, shouldn't you make sure that you take the last shot of the game by doing it very late in the shot clock?

I would think that if most teams are down 2 points with 24 seconds left, they will try to go for 2 and do it rather quickly so that if they fail, they can still foul and have a chance of winning. However, if they succeed in scoring quickly, which is only say 40% likely, the other team will have the final possession with a similar odds of winning outright. So, you'd have to get your .4, survive their miss, and then pull it out in overtime (.4*.6*.5 = .12?). If you can make sure to hold for a really late 3-point try, you'd only have to find a shot with a greater than 12% likelihood of going down for it to be the better alternative. That is probably what Charlie V's odds are of hitting a shot from half court. That strategy is never seen to be purposefully used however.

If this strategy checked out as valid and caught on, is there a counter strategy to beat it? i.e. fouling.

If the team ahead by 2 points fouls early on the clock, then let's say that they are forced to foul an 80% FT shooter because a PG will likely be controlling the ball then. There is a 64% chance that the team that is down will tie you and close to a 0% chance that they will pull ahead from the line. If they tie you, you have about a 40% chance of beating them with a last second game winner. So, your simplified odds of winning are .36 (odds they miss at least 1 FT) + .64 * .4 (odds they make both and you hit the dagger) + (.64 * .6 * .5) (odds they make both, you miss the dagger, and then you win in OT). = .808. I guess that would mean that if you think the other team can muster up a last second three pointer with greater than .192 odds of dropping, you should foul them as soon as possible.

This ignores the chain reaction where if they only hit 1 FT to pull within 1 that they will immediately foul you back and then at least have a last second 3 for the tie, etc. That whole thread probably knocks 10% off the total odds. Or of course, they miss both, then foul and you miss at least 1, then repeat.

Similarly, if you are down 3 with 24 seconds to go, then waiting for the last shot to make sure that the other team doesn't get a chance to give you the dagger back seems like the right play.


There are similar parallels to this in football where you often don't want to score quickly at the end of a game to tie or take a small lead because you are legitimately concerned that the other team will beat you after you do. You have to weigh the cost of delaying your opportunity against the other team's odds of daggering you.


There has to be some serious analysis on this subject that has been done by the NBA stat geeks. Because nobody seems to employ it, I'm open to the possibility that I'm missing some key factors.

Coaches in FIBA use intentional fouls when the game is tied and the other team have the final possession.

Really? Is it pretty standard? Would be interesting to see the win rate when used.

I don't watch FIBA but I remember there was alot of writing about it on the advanced stats forums and blogs where they concluded that this was the mathematically optimal strategy and the american coaches sucked because they didn't want to implement new stuff. No idea how common it is but a did a quick search and found that Argentina used it in the 2006 WC semifinals.

They even fouled Calderon!

A look at coaching strategies at end of Serbia-Spain - TrueHoop Blog - ESPN

Old school coaching says that you trust your defense to make one stop instead of giving your opponent free points. If you were Lawrence Frank, would you intentionally foul and give away the lead so that you could go isoTay for the last 20 seconds?

I'm honestly not sure that the absurd example you give is so absurd.

Tay is 14 for 35 in isolation shot attempts this season (40%) and drew fouls 9.3% of the time. He has no and 1's. He is an .829 FT shooter. So here is how his FT success would shake out (to be used in the calculations later):
- miss both: .093 * .171 * .171 = .003 of the time.
- make both: .093 * .829 * .829 = .064 of the time
- make only one: .093*.284 = .026 if the time

Overall on the season so far, our defense gives up a 44% FG percentage against and the other teams average 0.87 points per possession. They score at least 1 point 41.4% of the time.

So, if we trust our 11th ranked defense in this category, we would outright lose about 41.4% of the time. Then we would lose 50% of 58.6% of the time in OT. Combined, we'd lose .414 + .5 * .586 = 70.7% of the time and of course win 29.3% of the time. So, 29.3% is the number we have to beat in an alternate strategy.

If we foul, and then go Iso with Prince, here would be our odds of winning:

.833 = the 15th best starting PG FT% in the league. We'll use that as our base case to see if it works against really good FT shooters and a poor option for us with iso Tay.

Step 1: We foul their PG, who is a really good FT shooter.
They either get:
a) 0 points: 2.8% of the time. If that, then we iso Tay for a 2-point attempt late on the clock. He will win the game about 40% of the time. He will get fouled an hit at least 1 FT 9.0% of the time.
b) 1 point: 27.8% of the time. If that, then we iso Tay for 2-point attempt again and we win the game about 40% of the time. 2.6% of the time, Tay will hit only 1 FT and we will go on to win in overtime 50% of those times. 6.4% of the time, Tay will get fouled and drain both for the win.
c) 2 points: 69.4% of the time. Let's assume coach is "clueless" and goes for 2 to tie with an iso Tay. We will tie the game about 40% of the time. Then we will win the OT 50% of the time. We would also get the tie through FT's 6.4% of these times and win OT 50% of those times.

a) .028*.4 + .028*.090 = .014
+
b) .278*.4 + .278*.026*.5 + .278*.064 = .133
+
c) .694*.4*.5 + .694*.064*.5 = .161

Total odds of a Pistons win = .308 (slightly better than trusting the defense).



The coach could drastically improve our odds of winning by looking for an open 3 pointer in scenario c instead of opting for an iso Tay 2 pointer for the tie.

In addition, if we fouled a 70% FT shooter like Rondo instead of an 83% shooter, then our fouling strategy would draft the defense strategy.

I would rather lose than watch another isoTay.

The 2 usually go hand in hand.

So, anyone wanna give a guess as to when (or if) Andre gets past the sub 20 mpg ceiling on minutes this season?

Last ten games he's almost 23 and ahalf mins a game.... so roughly 3 more games at that pace will put him over 20 per game.

So a week from thur against the Knicks is my guess... but won't be suprised if its Utah on Sat night.

2014

Those number might be skewed because of some ot games where he played some extra minutes.

There is a pattern with a promising trend:

October: 13.0 mpg
November: 17.3 mpg
December: 22.1 mpg
January: 23.0 mpg

how many games has he played in feb?or is that a really small sample size?