[lmilne]

Scoring:
IMP

---

Playing against high-level opponents, you bid this collection 1♦-3NT (this isn't a bidding problem).

West leads the K♠ and you win the Ace.
You lead the 2♦ to the 5, Ace and 6, unblock the top two clubs from dummy, then lead the J♦ back towards hand. East follows to this trick with the 7.

Do you finesse or play for for the drop?

[gwnn]

LHO probably doesn't have 6 spades. He probably has either 4 or 5. If he has 5 then there is a 2 card discrepancy between slots and if he has 4 there is none. I read Frances' post here a while ago that you should deviate from '8 ever 9 never' only if there is a known 2 card difference which there is none here. So I'd play for the drop.

here it is:

http://forums.bridge...showtopic=24578

[gnasher]

I wouldn't have started from here.

I duck the first spade, partly because seeing two rounds of the suit may tell me more about how they break, and partly because LHO might have KQJxxx xxx x xxx

Then I cash two clubs. If LHO shows out, I'll play him for ♦Q

Otherwise, unless I think spades are 6-2, I play a diamond to hand and cash the other clubs. With a bit of luck I'll get some idea of how the hearts are divided too.

[dellache]

gwnn, on Nov 3 2009, 11:32 AM, said:

LHO probably doesn't have 6 spades. He probably has either 4 or 5. If he has 5 then there is a 2 card discrepancy between slots and if he has 4 there is none. I read Frances' post here a while ago that you should deviate from '8 ever 9 never' only if there is a known 2 card difference which there is none here. So I'd play for the drop.

No this reasoning is not right. You need to evaluate your "probably doesn't", and your "probably has" properly, otherwise the result of your estimation will be random.

Even if we don't know the leading agreements of oppos, in the worst case WC for finessing (i.e. you know by agreement of the WC that LHO has KQ♠, and NEVER the Jack), the finesse would be a strict 50% bet (at the time you have to make up your mind, you'd know KQ♠ by west, J♠ by East, x♦ by West, and xx♦ by East, so exactly the same number of cards). And... OK, if their card says "K from KQx, x from KQxx+, and Q from KQJx+", I'll change my mind (definitely not the case where I play), but then I wouldn't even cash the ♦A !

In other cases, the Finesse is the winner. Depending on what we know from the lead, here is the percentage for finessing :
- KQJ+ : 56%
- KQJx+ : 56%
- KQJxx+ : 59% (in my partnership, K would have shown KQJxx+)
Here for vacant places calculus, it's the spade honors and the situation in diamonds which is important.

Actually you could try to infer something from the non Heart lead and the spot plays in clubs, but I think it's really secondary.

Odly enough in this case, even if you knew the parity of the spade suits at trick one (which may be the case against some oppos), the %age when they play KQJx+ don't change.

Clear case for finessing a priori, but let's have a look at their CC.

[dellache]

gnasher, on Nov 3 2009, 12:31 PM, said:

I wouldn't have started from here.

I duck the first spade, partly because seeing two rounds of the suit may tell me more about how they break, and partly because LHO might have KQJxxx xxx x xxx

Then I cash two clubs.  If LHO shows out, I'll play him for ♦Q

Otherwise, unless I think spades are 6-2, I play a diamond to hand and cash the other clubs.  With a bit of luck I'll get some idea of how the hearts are divided too.

Hi Andy, this is interesting !

Actually I didnot see a way to get enough information from cashing the Clubs without restricting my play possibilities too much :
- I could overtake in Clubs to check for a doubleton in the West hand, but then I loose all the cases when East has a Diamond void. This does not work practically because of that ;
- play as you suggest :
-o- let's suppose you know spades are 5-3 at trick 2. Then your play offers the choice of 2-2, or stiff Q, or finesse against West, the latter being odds on if Clubs are 2-5. The problem is that when Clubs are 4-3, getting a Heart count may be of value only if West is 5233, and you can read East small Heart discard as showing 5(♥) for sure. Even in that case, when you add all the possible cases you "only" get 51%.
-o- so now your table presence tell you that spades are not 5-3. But how would you guess 4-4 against 6-2 ? (this is needed imo).

[gwnn]

dellache,

I was under the mistaken impression that LHO has passed over 1♦. Since that is not the case, my reply was nonsense and I agree with your post. With the caveat that I'm not familiar with your use of 'WC'.

[phil_20686]

This feels like a finesse to me.

lho would not lead K form KQxx(+), he would lead a low one in case parther has Jx or Ax. Thus he could have KQx, or KQJ(+). It feels like having KQJx is less likely thank KQJxx, as having KQ and J in 5 cards is much more likely than KQJ in 4 cards. Moreaover the 5-3 break is a priori more likely. My memory of the % is that taking the finesse when the spades are 4-4 is a smaller loss than the gain when spades are 5-3, so even if 4-4 and 5-3 were equally likely, then the finesse is better. (2% loss 5% gain?).

[KQJx in 4-4 is 4/8*3/7*2/6 = 7%]
[KQJxx in 5-3 is 5/8*4/7*3/6 = 18%]

Since we are talking about only one 5-3 break, i guess its 24%, and 4-4 33%, so maybe KQJx is about half as likely as KQJxx.

Also, why didnt i duck trick one and watch the spade pips for their count - rho very unlikely to lie on the lead.

[dellache]

phil_20686, on Nov 3 2009, 02:37 PM, said:

Also, why didnt i duck trick one and watch the spade pips for their count - rho very unlikely to lie on the lead.

Yes ducking seems mandatory.
But as I said in a previous post, the odds for finessing are *globally* the same in both odd/even cases if you cannot distinguish 6-2 from 4-4.

Maybe you can infer something from spade continuation *tempo* at trick 2 though (when West has no outside honor and 6 spades, he may think about a switch).

[gnasher]

gnasher, on Nov 3 2009, 01:31 PM, said:

I wouldn't have started from here.

I duck the first spade, partly because seeing two rounds of the suit may tell me more about how they break, and partly because LHO might have KQJxxx xxx x xxx

Then I cash two clubs. If LHO shows out, I'll play him for ♦Q

Otherwise, unless I think spades are 6-2, I play a diamond to hand and cash the other clubs. With a bit of luck I'll get some idea of how the hearts are divided too.

Now that I've counted my tricks properly, I've changed my mind: I overtake the second club to cash a third one, thereby keeping flexibility in the diamond suit.

[gnasher]

dellache, on Nov 3 2009, 02:29 PM, said:

- I could overtake in Clubs to check for a doubleton in the West hand, but then I loose all the cases when East has a Diamond void.

You mean West? With a diamond void in East, there is no problem.

Quote

- play as you suggest :
-o- let's suppose you know spades are 5-3 at trick 2. Then your play offers the choice of 2-2, or stiff Q, or finesse against West, the latter being odds on if Clubs are 2-5. The problem is that when Clubs are 4-3, getting a Heart count may be of value only if West is 5233, and you can read East small Heart discard as showing 5(♥) for sure. Even in that case, when you add all the possible cases you "only" get 51%.

Yes, I get more out of the discovery play if I do overtake in clubs. Now if I find out about 5xx4 on the left I can finesse against West. I think that more than makes up for losing when diamonds are 0=4.

Quote

-o- so now your table presence tell you that spades are not 5-3. But how would you guess 4-4 against 6-2 ? (this is needed imo).

I wasn't planning to use my table presence (I have none), so much as the opponents' spot cards. If forced to guess whether they were 4-4 or 6-2, I'd assume 4-4, because it's more likely.

Edit: If they are 4-4, one of the opponents may throw on on the fourth round of clubs.

[dellache]

gnasher, on Nov 3 2009, 03:53 PM, said:

dellache, on Nov 3 2009, 02:29 PM, said:

- I could overtake in Clubs to check for a doubleton in the West hand, but then I loose all the cases when East has a Diamond void.

You mean West? With a diamond void in East, there is no problem.

Yes I meant West of course. That's 6% of the cases and you need them.

[Phil]

phil_20686, on Nov 3 2009, 09:37 AM, said:

This feels like a finesse to me.

lho would not lead K form KQxx(+), he would lead a low one in case parther has Jx or Ax. Thus he could have KQx, or KQJ(+). It feels like having KQJx is less likely thank KQJxx, as having KQ and J in 5 cards is much more likely than KQJ in 4 cards. Moreaover the 5-3 break is a priori more likely. My memory of the % is that taking the finesse when the spades are 4-4 is a smaller loss than the gain when spades are 5-3, so even if 4-4 and 5-3 were equally likely, then the finesse is better. (2% loss 5% gain?).

[KQJx in 4-4 is 4/8*3/7*2/6 = 7%]
[KQJxx in 5-3 is 5/8*4/7*3/6 = 18%]

Since we are talking about only one 5-3 break, i guess its 24%, and 4-4 33%, so maybe KQJx is about half as likely as KQJxx.

Also, why didnt i duck trick one and watch the spade pips for their count - rho very unlikely to lie on the lead.

I love summaries like these.

Soon, we'll hear about the likelihood of the exact shapes of the opposing hands.

Agree with the spade duck however. I'll go with Gnasher's line. Winning only feels right if RHO has specifically ♠Jx and the ♥AK.

[dellache]

Phil, on Nov 3 2009, 04:25 PM, said:

phil_20686, on Nov 3 2009, 09:37 AM, said:

This feels like a finesse to me.

lho would not lead K form KQxx(+), he would lead a low one in case parther has Jx or Ax. Thus he could have KQx, or KQJ(+). It feels like having KQJx is less likely thank KQJxx, as having KQ and J in 5 cards is much more likely than KQJ in 4 cards. Moreaover the 5-3 break is a priori more likely. My memory of the % is that taking the finesse when the spades are 4-4 is a smaller loss than the gain when spades are 5-3, so even if 4-4 and 5-3 were equally likely, then the finesse is better. (2% loss 5% gain?).

[KQJx in 4-4 is 4/8*3/7*2/6 = 7%]
[KQJxx in 5-3 is 5/8*4/7*3/6 = 18%]

Since we are talking about only one 5-3 break, i guess its 24%, and 4-4 33%, so maybe KQJx is about half as likely as KQJxx.

Also, why didnt i duck trick one and watch the spade pips for their count - rho very unlikely to lie on the lead.

I love summaries like these.

Soon, we'll hear about the likelihood of the exact shapes of the opposing hands.

Agree with the spade duck however. I'll go with Gnasher's line. Winning only feels right if RHO has specifically ♠Jx and the ♥AK.

This actually maximises your chances of making (59% for overtaking in clubs+counting versus 55% for simply finessing against East), but also will cost you an extra undertrick/overtrick when you are wrong/right.

Probably more than just elementary arithmetics is involved.

[nigel_k]

Agree with the others about trying to find out more information. Many players show count when partner leads the king and obviously I would be especially interested in that.

On the problem as given, I think the finesse has a slight edge. The honour sequence increases West's expected spade length just because there are more ways for the small cards to be distributed, ie there are 5 ways to have KQJx opposite xxxx and 10 ways to have KQJxx opposite xxx.

[Fluffy]

hey guys this was called SIMPLE drop or finese situation

[nigel_k]

Fluffy, on Nov 4 2009, 12:53 PM, said:

hey guys this was called SIMPLE drop or finese situation

Sorry I thought that was meant as a challenge.