To finesse or not?

[OliverC]

Came across a hand which is particularly interesting at Matchpoints, but bears consideration in a Teams Match as well. Partner and I were playing OCP and with no opposition bidding we bid to 4 by South. I won't bore you with the bidding, but partner opened 1 and my 1 response showed a balanced hand. Partner had Spades, however, so I ended up as the Declarer in 4 .

North
  AK853
  K4
  QJ1043
  A

South
  J1072
  AQ105
  K7
  1084

LHO led the Ace of Diamonds and a second Diamond which ran round to my King. How do you play the hand? Specifically, how should you play the Spade suit?

I'll let you chew on this and give my analysis in a different post...

[kenberg]

T3, a small spade toward the K. If spades are 4-0 either way, I take 11 tricks via spade to the K, and, if someone did not follow,  small spade back toward my J.
If everyone follows with a spot, I am taking the K and then laying down the A.  It's true that the diamond lead suggests that the diamonds were split 4-2, with 4 on my right, and this ups the chances that there are 3 spades on my left, but it doesn't up them much. So I just play AK.

I suppose one could argue that since the opponents have 9 clubs and were not in the bidding this ups the chances that spades are 2-2 instead of 3-1, but I also don't put much faith in this. I just go with small to the K, if everyone follows I play the A.

What I don't do in these situations  is to worry that if spades are 2-2 then those who are in 6 are making a slam if they bid it. I can't do anything about those who are in slam. If they are there and make it then they get a good score against me no matter how I play the spades, and if they are in 6 and don't make it, I get a good score against them no matter how I play the spades. So I play the spades trying to do better than the others who are in 4, not fretting about those who are in 6.

A  eature that can sometimes influence play, but I don't see how it does here: At many other tables, N will open 1S. This will mean that N, not S, will be declarer. Sometimes when this happens, that the board is played the other way at many tables, you have to look and see if this led to the opponents being lucky with a lead that won't happen at the other tables, or unlucky with a lead that won't happen at other tables. But on this hand you can see that the opponents will be getting their diamond Ace sooner or later no matter who is declarer and what the lead is.


If there is a reason for doing something other than just playing of the top spades when everyone follows to the first spade. I am not seeing it. I look forward to your thoughts.

[OliverC]

The reason I set it as a problem at Matchpoints is precisely because (1) you ought to guard against those people who have pushed on to 6 , but mainly because (2) you need to try to beat the [majority of] people who are in game. Most people will test the Ace of Spades, cross back to hand with a Heart, lead the J and then plonk down the King  when LHO follows low, breathing "8 ever, 9 never". As you said, nothing you can do about the people who are in 6 if the Spades are 2-2, but this is Matchpoints and if the Spades are 2-2, then you're going to get a low average at best because fewer people will go off in the slam than make it.

At MatchPoints there is a much bigger incentive to take risks and go against the flow in an attempt to get a good score. If LHO has Qxx in Spades then now  taking the finesse will beat most of the people in game (who are plonking down the AK), because you're making +680 to their +650, and almost all of the people in the slam (because most of them will be going off) and you'll come out with a near top. It's true, you will come out with a poor score when the Spades are 2-2, but then you're booked for a low average anyway if that's the case.

At IMPs it's not really an issue, because it's only a matter of gaining or losing 1 IMP if the other table are also in game, and it makes zero difference if the other table are in a making slam (650 vs 1430 and 680 vs 1430 are the same IMP score), and zero difference if they are going off in the slam (650 vs -100 is the same IMP score as 680 vs -100). To my mind, though, it's almost like taking out insurance to take the finesse in the hope the other table are going off in slam.


At the table we were playing at IMPs and I lost 1 IMP taking the finesse, but even the people who were +680 were losing an IMP because several more pairs bid and made the slam than went off in it.

[kenberg]

An interesting hand.


Take board a match as an extreme form of matchpoints. Not only is there nothing that we can do against a making slam at the other table if the spades are 2-2, there also is nothing that we need to do against a failing slam at the other table if the spades are 3-1. In the first case you lose the board regardless of how you play the spades, in the second case you win the board no matter how you play the spades. So you may as well assume they are also in 4 at the other table, it's the only situation where your play makes a difference.

Now for imps. An imp is only an imp but for all that it's an imp.

First assume that they are in 4 at the other table.
Assume the other declarer plays the AK, but we play the A and then run the J.
If we are right, we make 680 to their 650. If we are wrong we make 650 to their 680. 1 imp our way if running the J is right, 1 imp their way if it is wrong.

Here the payoff/loss is equal.

Now assume that they are in 6S at the other table and, as is likely, declarer plays off the AK of trump.
Assume everyone is vul.

Assume that we  run the J on the second round of trump.  If we are wrong the score will be 1430 at the other table and 650 at ours.  That's 780 and it only would have been 650 had we dropped the Q. 13 imps either way though, so it didn't cost. Now suppose we are right. We run the J and RHO shows out. We get 680 and at the other table it is -100  so that's 780. Had we gone up, it would be only 650 at our table, still -100 at the other, still no imp swing at all.

So it is one of those things where we can hope that they are in slam off 1 at the other table, but if that is so then it doesn't matter at all what we do at our table. So we go with whichever seems  the most likely.

And what is the most likely lie? A few thoughts, largely inconclusive:

Of course playing for the drop has a slight a priori edge but often clues from the auction/play alter that.  Is there anything?  The silence of the opponents suggests that clubs are not 7-2, certainly not 8-1. That is not much. The lead of the A and small suggests he was hoping for  a ruff. Maybe that helps? If he held Qxx would he want a ruff? Perhaps. He might imagine taking the A, leading to his partner's hypothetical K, getting a ruff and then some miracle occurring that gives their side another trick. But then he might equally figure that if his partner has the K they are getting two diamonds anyway, that he might well be getting his trump Q in normal play, and that the chance of a fourth trick is better if he doesn't rush in.
So we might reason that he is more likely to have gone after the ruff holding trump xx than with trump Qxx.  But this has to be balanced against the fact that if he has only two diamonds he is less likely to have a second doubleton.

All in all, I went with the AK but on another day it would have been wrong.

[OliverC]

In BAM scoring (and at IMPs in a Team game and potentially at Barometer Pairs) a lot will depend on your estimation or knowledge of the state of the game. If I was or felt I was 10 IMPs down on the last board of a Team Match, I would always be playing for a swing, because I have nothing to lose. Same applies to being 1 Board down at BaM or needing to make up a percentage point in a Barometer Pairs event.


On a different hand (not really possible on this hand because transport was something of an issue) you might play some tricks in side-suits to get a better count on the hand before making the crucial decision in Spades. That a strategy commonly applied by experts, but unfortunately it's not always possible, as here. As it happens, I rightly gave West Axxx in Diamonds (because of East's UDCA play of the 2 and then the although it was also entirely possible they had started with Ax and was hoping for Kxx with partner. Although I could have established the count in Hearts (West would have shown out on the third round) that in itself would only have given me useful info if they ruffed (in which case I can just drop the queen regardless). I can never establish what the Club position was with any accuracy, because I don't have enough entries to my hand and in any case can only ruff 2 Clubs if I did.


You're right in that, a priori, the odds favour the drop (restricted choice permitting if we are missing the QJ to start with), because at the point where West follows low to the second Spade, we've already eliminated half of the nearly 50% of times when the cards are distributed 3-1 (ie: all the times when East has 3 and West has 1. This wasn't to do with probabilities, though. It was to do with taking a different line to most in the hope of getting a really decent score on the hand in spite of only being in the game, because we were booked for a below average score anyway if the drop succeeds, because that's what most people will play for.