[rogerclee]

Scoring: IMP

---

1♦ - (3♥) - Dbl - (P)
4♣ - (P) - 4♥ - (P)
4♠ - (P) - 4N -(P)
5♥ - (P) - 6N - All Pass

RHO preempts aggressively, and you have a really bad auction to 6NT.

LHO leads the ♥8, x, K, A.

Like it or not, you think for about a minute and then play a small diamond towards dummy, it goes ♦x, K, A, x. You play a small diamond from dummy, RHO plays the ♦8.

This is not really a full play problem, I am just looking for experienced declarers to tell me in their opinion how likely ♦KT is against various levels of adv+ opponents. Also feel free to abstain if you think it was ridiculous to think at trick one for that long.

[JLOL]

I was playing against Gavin at a regional and we had this suit combo and I insta played for 32.

Anyways, according to my calculations its about 53.6% to 46.4% without bidding. So if you think that 1 time every 15 they won't play the K from KT doubleton when they have it then you should be finessing. I would expect almost everyone to not play it 1 time every 15. If I knew my opp to be both very expert (truly world class), and also at least somewhat deceptive, and also one who is careful with stuff like this (there are many players who are world class who seem be lazy about things like this which rarely matter).

The other thing is you are never making a huge error by finessing even if they are one of the few people who always find the falsecard. If they are one of the many people who never find the falsecard though you have made a really really big error. In those spots is usually better to go for the big win play that sacrifices a little bit against perfect play.

All that theoretical stuff aside, on this hand we certainly have clues from the bidding. If hearts are 71 then lefty will have 2 diamonds rather than 1 more than 8 times as often! This means that if lefty will falsecard more than 12.5 % of the time you should play for the drop. I would assume any real expert player, or perhaps a youngish and/or tricky intermediate player or someone who reads forums a lot could find this with that frequency. I would still hook against most club players, but against a random in flight A I would definitely go for the drop. Even if RHO can have 6 hearts it's about 6 times as likely they have 2 rather than one diamond on our left and our cutoff point for their frequency of falsecarding has to be 1 in 6 which is not that often.

Edit: I think I did some math wrong, it's 1 in 4 not 1 in 8. That makes it closer and makes their age more important for the read

[kfay]

JLOL, on Feb 8 2009, 01:06 AM, said:

I was playing against Gavin at a regional and we had this suit combo and I insta played for 32.

Anyways, according to my calculations its about 53.6% to 46.4% without bidding. So if you think that 1 time every 15 they won't play the K from KT doubleton when they have it then you should be finessing. I would expect almost everyone to not play it 1 time every 15. If I knew my opp to be both very expert (truly world class), and also at least somewhat deceptive, and also one who is careful with stuff like this (there are many players who are world class who seem be lazy about things like this which rarely matter).

The other thing is you are never making a huge error by finessing even if they are one of the few people who always find the falsecard. If they are one of the many people who never find the falsecard though you have made a really really big error. In those spots is usually better to go for the big win play that sacrifices a little bit against perfect play.

All that theoretical stuff aside, on this hand we certainly have clues from the bidding. If hearts are 71 then lefty will have 2 diamonds rather than 1 more than 8 times as often! This means that if lefty will falsecard more than 12.5 % of the time you should play for the drop. I would assume any real expert player, or perhaps a youngish and/or tricky intermediate player or someone who reads forums a lot could find this with that frequency. I would still hook against most club players, but against a random in flight A I would definitely go for the drop. Even if RHO can have 6 hearts it's about 6 times as likely they have 2 rather than one diamond on our left and our cutoff point for their frequency of falsecarding has to be 1 in 6 which is not that often.

Edit: I think I did some math wrong, it's 1 in 4 not 1 in 8. That makes it closer and makes their age more important for the read

I don't think I quite fall under the blanket of 'experienced declarer' but I think you're giving a little too much credit to flight A randoms. Unless by flight A you mean 'flight A'.

Edit: I'm wrong wrong wrong.

0+
7
3+
1+

So yeah your numbers seem right to me.

Anyway if I hook and I'm wrong then I'm down. If I play for the drop and I'm wrong I can still hope for ♣H with RHO. This must shift some percentage points back in favor of playing for the drop whether or not LHO will falsecard. In fact I feel like a stiff honor isn't terribly unlikely. Probably a higher liklihood than the numbers would indicate. So playing for the drop really has a lot going for it on this hand, imo. If RHO has a void then that sucks I guess.

Playing against Clee I would play for the drop. Which I guess gives this post the same title as the whole thread. What I mean to say is that I think I could rely on an actually good player to play this falsecard consistently, but I hardly think that encompasses all of who I might consider actual Flight A players, at least in Michigan.

[jdonn]

kfay, on Feb 9 2009, 12:45 PM, said:

Playing against Clee I would play for the drop.

Playing against Clee I would play for the revoke.

[Edmunte1]

Considering hearts 1W-7E, then the probability for diamonds being:
a ) K ------ 10 xxx is 0.42%
b ) K10 ----- xxx is 1.54%
so you're 79-21 odds playing for the drop. If you consider that other factors compensate this rate (like West is a weak player or west is usually known as falsecarding player under a rate of 1/5), you should finesse. In any advanced+ field you should play for the drop.

[orlam]

I asked someone who knows how to compute these things, he said JLOL's numbers are wrongn fwiw. (I can ask him again in case it matter to anyone.)

[matmat]

orlam, on Feb 10 2009, 03:10 AM, said:

I asked someone who knows how to compute these things, he said JLOL's numbers are wrongn fwiw. (I can ask him again in case it matter to anyone.)

do you expect his numbers to come out differently a second time?

[gnasher]

orlam, on Feb 10 2009, 09:10 AM, said:

I asked someone who knows how to compute these things, he said JLOL's numbers are wrongn fwiw. (I can ask him again in case it matter to anyone.)

As appeals to authority go, that leaves something to be desired.

We're all bridge players, so we're all supposed to know how to compute these things. For those of us who're too lazy, there is also Richard Pavlicek's suit break calculator.

I (with Pavlicek's assistance) agree with Justin's figures, give or take some rounding errors. Regarding the impact of hearts being 1-7, that makes a 2-3 diamond break about 8 times as likely as 1-4, and K10 doubleton about 4 times as likely as stiff K.

What does your friend say is wrong with Justin's analysis?

[orlam]

gnasher, on Feb 10 2009, 04:32 AM, said:

orlam, on Feb 10 2009, 09:10 AM, said:

I asked someone who knows how to compute these things, he said JLOL's numbers are wrongn fwiw. (I can ask him again in case it matter to anyone.)

As appeals to authority go, that leaves something to be desired.

We're all bridge players, so we're all supposed to know how to compute these things. For those of us who're too lazy, there is also Richard Pavlicek's suit break calculator.

I (with Pavlicek's assistance) agree with Justin's figures, give or take some rounding errors. Regarding the impact of hearts being 1-7, that makes a 2-3 diamond break about 8 times as likely as 1-4, and K10 doubleton about 4 times as likely as stiff K.

What does your friend say is wrong with Justin's analysis?

I asked him again, he explained it very nicely:
Without bidding, you can use vacant spaces, LHO has 12 and RHO 10, so it is 12:10 or 6:5 in favor of KT versus K. He says this means the break-even point is where LHO fails to falsecard one out of every SIX times, not one out of 15. This seems very logical to me (that would it 5:5 not 6:5, right?), but I don't understand the flaw in JLOL's one-out-of-15 computation.
With the bidding, he basically agrees with JLOL's correction - I had asked him about it before reading the correction, and didnt notice JLOL had posted one.
He again uses vacant spaces, he says that now LHO has 11, and RHO has 3. Not sure I quite understand (isn't there a restricted choice problem here), but this would give 11:3, almost JLOL's 4:1.