
Hvordan spiller du Ekn8 i bordet mot D763 på hånden?

How would you play:
AJ8
Q763
for three tricks, no entry problems?
Eric Kokish
My instinct is to cash the Ace, play one to the Jack, then run the 8 if the 9 or the 10 came from West on the second round.
Boye Brogeland


Navn: 
Michael Rosenberg 
Dato: 
30.1.2008 
Epost: 
Ingen epost 
Hjemmeside: 
Ingen hjemmeside 
Kommentar: 
First instinct was this: If I lead low to the J, then run the Q (hereafter called LJQ), I will maximize, except against an opponent who ducks with Kmiddle fourth. That appears to be the case. Subsequent analysis appears to confirm this. Of the other plays I looked at, low to J, then A (LJA) seems to beat out low to A then low to J (LALJ) (which gains against stiff K, but loses to middle sing  since you cannot reasonably finesse on second round). Also, leading the Q on the first round cannot be as good as LJQ, since it loses to stiff K or middle, and gains only against a 60 break.
So, assuming for the moment that I play "pure" (playing for 109x if the win the K) the 2 plays (LJQ and LJA) are almost equal. Each picks up five 51 breaks, LJA beats LJQ on three 42 breaks (basically Kx) and LJQ beats LJA on three 33 breaks (109x). Since each 33 break is more likely than each 42 break (approx. 1.8% to 1.6%), LJQ wins.
But that is theory. In practice, it is far more clear, and better to go with the original plan of LJQ, and finessing on the way back if they win the K and a middle drops (losing to 109x). There are 6 holdings you gain on, while losing to only 3. So unless they would duck more than half the time, finessing is better. While they might like to always duck, in practice they will not. First, with K10xx, they will fear ducking since, if partner does not have the 9, they might be making things easy for you (yes, you might come back and play the Q with Q9xx, but not all hands go that way due to transportation). And with either holding (K10xx or K9xx) they will fear you were all along going to play LJA  and they have taken away whatever guess you might have had. Furthermore, with either holding, he might fear he needs to win the K for tempo reasons.
To sum that up, I would only play for 109x under very precise conditions  maybe 6N with plenty of entries, obviously nothing else to the hand, and a strong RHO. In practice, I'm going to tend to pay off and assume he won't duck from K10xx. There may well be game theory issues here  I should make either play a certain percentage of the time  but understanding that is a weakness for me. I prefer the practical aspects. Phil Martin and/or Jeff Rubens (and others) can probably tell you if game theory is involved here.
_*The Defense
*_Since it seems that you would like to be able to know when to duck with Kmiddle fourth, it would be nice if partner could help. It looks as if when the 7 is played by LHO, that could have a meaning. In other words, the 7 must be from 107 or 97  now partner KNOWS to win (since declarer no longer has the option of coming back to hand and playing the Q). If he ducks, it must go ace next, giving no chance.
So a thought here is don't play the 7 with 7x. But maybe you can carry it even further, playing the 9 or 10 from these holdings (again forcing partner to win). So now the 7 would have the opposite meaning  denying the higher (and implying you want partner to win the trick)! Instead, maybe the SIX promises the higher (although it's arbitrary  you could preselect any small spot), or implies you want partner to duck. So you must play the 7 from 76 doubleton (if you want him to win).. Presumably with 109x , you can play anything (except the two high ones first), but perhaps should avoid the 7 on the first round with 1097. So, in a perfect world, RHO watches carefully, and has a better idea of when to duck.
_*The Composer
*_It is nice to be as clear as possible when presenting a problem.
Specifically, I would have preferred the holding to be given as Q765 or Q762. That way, a lunatic such as myself wouldn't spend any time on considering whether the trey could be a relevant card.
_*Confidence
*_I tend not to be especially confident that I am right about things. I would not be surprised to find I made mistakes in any of the above analysis. My confidence stems more from the wrongness of others than from my own accuracy. So I apologize if it was inappropriate of me to write this.


Navn: 
Bart Bramley 
Dato: 
30.1.2008 
Epost: 
Ingen epost 
Hjemmeside: 
Ingen hjemmeside 
Kommentar: 
Start with low to the jack.
1. If the jack wins and low cards appear, then against perfect defenders, play the ace next.
2. If the jack wins and RHO plays 9/10, return to hand and lead the queen.
3. If the jack loses, play the ace next and finesse on the third round when LHO plays 9/10.
This works against 17 of 20 33 splits (losing to LHO’s 109x, 3 occurrences), 1 of 15 42 splits (RHO’s 109), 12 of 15 24 splits (all but RHO’s K109x, which cannot be picked up by any play), 5 of 12 51 or 15 splits (stiff K, 10 or 9 with LHO, and stiff 10 or 9 with RHO). Total is 35 layouts out of the 57 that allow some play to work. (RHO’s K109x, K109xx and K109xxx allow no winning play.)
Against imperfect defenders, who may not duck as RHO with K10xx or K9xx, then if the jack wins it is better to return to hand and lead the queen next. This loses to Kx onside (3 cases) but gains against K10xx or K9xx onside (6 cases). Therefore, if you judge that RHO will win the king from K10xx or K9xx at least 50% of the time, this is the best play, and would be my pick for all but special situations.
Starting with the ace and then leading toward the jack picks up only 34 holdings. Leading the queen first is even worse.


Navn: 
Bob Hamman 
Dato: 
30.1.2008 
Epost: 
Ingen epost 
Hjemmeside: 
Ingen hjemmeside 
Kommentar: 
Low to J and then cross and lead Q if you think defender will win 1st trick with K9xx or K10xx. Assume optimum defense then defender with K9xx or K10xx will duck low to J and give up on Kxx facing 109x restricted choice. So he picks up on 6 combinations and loses on 8 combos by winning trick He also picks up when declarer crosses to dummy and leads Q.. I have concluded that best is to lead low to J then the Q in both the real world and theory. This loses to sing K offside but picks up sing 9 or 10. There is some element of what is best by defense.



