The Underbidder's Angle

[kfay]

Partner is having an off day when it comes to slam bidding, as you can see from this hand:

 

[hv=d=s&v=n&n=saxxhkj87d10xxxcxx&s=skxxha10965daqcakq]133|200|Scoring: IMP[/hv]

 

Your auction went:

 

2♣-(P)-2♦-(Dbl); 2N-3♣; 3♥-4♥

 

The lead is a low diamond, which you win cheaply (RHO played the K).

 

You cash the ♥A--all follow--and play a heart up, LHO following small.

 

Hook or drop?

 

Does it matter?

[awm]

 

 

If the table you're comparing to plays 4H also, then playing to the heart king is very slightly better than finessing (percentage line in 9-card fit).

 

If the other table plays 6H, then the only time it matters what you do is where they play 6H= and you can score up +510 to lose only 10 imps (in all other cases you lose 11 imps when 6H makes and win 11 imps when 6H fails regardless of whether you get +480 or +510).

 

The other table is likely to play to the king if they are in 6H, as this is (very slightly) the percentage line with a nine-card fit. So it seems as if you should take the same line, since you were basically given a diamond trick (DK appears offside) which suggests that if the heart queen is falling you will make +510 and they will make +980 (lose 10) whereas if the heart queen is not falling you will get +480 and they get -50 (win 11, but +510 vs. -50 also win 11).

 

[kenrexford]

I agree that someone being in slam is irrelevant. If they play it right and make, they beat your score. If they play it wrong and go down, you beat their score. So, what difference does it make what they would do?

 

On the flip-side, something seems way off with playing 8-ever-9-never in the first place, if that was the thought. RHO doubled 2♦ missing the Ace, Queen, and 10 of diamonds?!?!? RHO must logically have diamond length. The most has has is a six-card suit, but he surely has at least five.

 

The non-diamond cards, therefore, are split with East having 8, or perhaps only 7, non-heart cards. West, on the other hand, and at least 11 and maybe 12 non-diamond cards. Let's assume only five diamonds, as the six-diamonds scenario is even stronger indication of what I am thinking.

 

West starts with 11 non-diamonds, East with 8 non-diamonds.

 

Both played a non-diamond on the first round of trumps, so now West has 10 non-diamonds and East 7 non-diamonds.

 

West just played another non-diamond.

 

So, West still has 9 non-diamonds to East's 7 non-diamonds.

 

So, it seems to me that the hook is a 9:7 favorite to work.

 

If East has 6 diamonds, then the hook is a 10:6 favorite to work.

[kfay]

I agree that someone being in slam is irrelevant.  If they play it right and make, they beat your score.  If they play it wrong and go down, you beat their score.  So, what difference does it make what they would do?

The whole premise of this thread is actually based on the reasonable assumption that your conterparts will be in slam.

 

Although this may not be the ideal hand, this type of play is one of my favorites since it highlights what I believe to be the best aspect of the IMP table. Problems like this come up frequently, though you'd rather not be in the position to have to deal with them. That's just my opinion.

[Little Kid]

I would hook on the basis that I judge RHO to have at least 5 ♦s to the KJ for his double, making finesse the % line. He surely doesn't have less than 4♦s, in which case hook and drop are same.

[kenrexford]

I agree that someone being in slam is irrelevant.  If they play it right and make, they beat your score.  If they play it wrong and go down, you beat their score.  So, what difference does it make what they would do?

The whole premise of this thread is actually based on the reasonable assumption that your conterparts will be in slam.

 

Although this may not be the ideal hand, this type of play is one of my favorites since it highlights what I believe to be the best aspect of the IMP table. Problems like this come up frequently, though you'd rather not be in the position to have to deal with them. That's just my opinion.

What difference does it make that the field, you assume, will be in slam?

 

Actually, I do think it matters, the more I think about it. If I was convinced that everyone in the room would be in the slam, then here's my line:

 

I'd win the diamond and play the Ace. If that survives, I'll cash three clubs, ditching a spade. I'd then play two spades, ending in dummy, and ruff a diamond with the Ace. I'd then ruff a spade on dummy with the Jack. If that all works so far, I'd then play the last diamond from dummy, ruffing with the 10.

 

If this survives, I'm almost home. I'll lead the trump 9 toward dummy. When LHO plays low, I go into a very short tank, sweating if I can, before calling for the trump King.

 

This way, I might get a good result on the next hand, because my opponents will think my play is horrifying and may double a making contract or may overbid.

[billw55]

This way, I might get a good result on the next hand, because my opponents will think my play is horrifying and may double a making contract or may overbid.

meheheh

 

so Kfay .. what are you talking about with "this type of play is one of my favorites since it highlights what I believe to be the best aspect of the IMP table" ?

[mcphee]

A BBO dble of D is completely meaningless. AS an example yesterday in a Cayne match S splintered in C and the player dbled with KJxx. With the rush to be involved in the bidding or be a macho man these dbles are random. Hook or drop to me on this hand is a waste of mental energy and the 8-9 thing has never been high on my list. That being said I finesse, because if it loses the opps may believe the partners at the other table are going down in slam and if it were to win they will be wondering if their partners made the same play.

[Free]

This is not a B/I problem imo. It's a combination of chances, and if you have to play percentage or anti-percentage. I hope my calculations are correct. :rolleyes:

 

Slam can't go down, so there are a few things to consider:

- opps are in 4 making 6 = 510-480 (480-480) = 30 (0) -> 1(0) imps

- opps are in 4 making 7 = 510-510 = 0 (-30) -> 0(-1)

- opps are in 6 making 6 = 510-980 = -470 (-500) -> -10(-11)

- opps are in 6 making 7 = 510-1010 = -500 (-530) -> -11(-11)

- opps are in 7 going down 1 = 510+50 = 560 (530) -> +11(+11)

- opps are in 7 making = 510-1510 = -1000 (-1030) -> -14(-14)

 

So whatever we do is irrelevant if opps bid 7, or make 6♥+1. If we can assume our opps will make the percentage play in whatever contract they are, we can only gain an imp if we play anti-percentage.

 

We make 12, they make 13 tricks in 4♥: we lose 1 imp

We make 13, they make 12: we win 1 imp if opps are playing 4♥ or 6♥

Other situations don't make a difference

 

Everything depends on how much we think opps are in slam (P(A)), and how successful the percentage line actually is (P(:D). We have to add the following:

P(A)*(1-P(B)) = chance opps are in slam and make 12 tricks

(1-P(A))*P(B) = chance opps are not in slam and make 13 tricks

(1-P(A))*(1-P(B)) = chance opps are not in slam and make 12 tricks

 

So playing anti-percentage wins if:

P(A)*(1-P(B)) - (1-P(A))*P(B) + (1-P(A))*(1-P(B)) > 0

so

P(A) > (2*P(B)-1) / P(B)

 

Taking Ken's numbers: P(B)=0.5625 => then P(A)>0.222222

 

If you suspect opps will bid slam at least 22.2222% of the time, then the anti-percentage play is the winning option.

 

Take Ken's 2nd odds (10 to 6), then P(A) must be bigger than 0.4!

 

Since the hook is the better play, the drop will gain in the long run if opps bid their slams enough, take the percentage play, and when the percentage play is not too successful (example with P(B)=0.8 you need P(A)>0.75).

[bid_em_up]

..

[mikeh]

I don't think there is a clear answer here, in part because we can't know whether our teammate will be given the opportunity or will take the opportunity to double 2♦.

 

Absent the double, declarer in slam will at least consider cashing the heart K, cashing the blacks, ruffing a spade and then playing a heart towards hand: if rho follows low, insert the 10 and if we lose to the Qx offside, he'll be endplayed. I think that is higher than simply playing top hearts and then falling back on the diamond finesse, but I haven't crunched the numbers. I think we need to play the hook in order to try to save an imp.

 

With a double, the odds are that declarer will hook LHO on the missing spaced principle. So we play for the drop to cut our losses by 1 imp.

 

Without a double, but with a diamond lead, declarer will play for the drop as the percentage play. So we need to hook.

 

Sounds to me as if the double was, to be kind, questionable, so I think I'd assume no double. That leans me towards the hook.

 

My thinking may be all wrong, but if it is even close to being on point, this is not a B/I problem: I doubt many B/I players would play slam as I suggested, on auctions without a double and without a diamond lead.

[kfay]

Well I suppose if there is an actual solution, that it could get to be complex.

 

I was mostly looking for:

 

Assume the other table is in 6 making 6. We save 1 IMP if we make 7, even though we're only in game.

 

So, between finessing and dropping, I should do what I think the other table will NOT do. Since if they make 7 and I make 6, it makes no difference.

 

 

These situations where you might be inclined to believe 'well we missed a cold game, there goes X IMPs,' are not altogether infrequent. Often you can turn it into 'well i'm only going to lose X-1 IMPs.' Several conditions must be met, however.

 

1) You are stuck in a partscore/game when there's a cold game/slam available.

 

2) The other table MUST HAVE BID THAT GAME/SLAM

 

3) There is a play available to you that you believe the other table won't choose in their higher contract that could provide you extra tricks while not placing your own contract in jeopardy.

 

4) You are aiming for:

 

i) Vulnerable major-suit games where you are in a partscore, making 11 tricks to their 10

ii) Vulnerable NT games where you are in a partscore, making 10 tricks to their 9

iii) Small slams where you are in game, making 13 tricks to their 12.

 

Way outside shots include:

 

iv) Minor-suit games where you are in a partscore, taking 13 tricks to their 11

v) Non-vul major-suit games where you are in a partscore, taking 12 tricks to their 10

vi) Non-vul notrump games where you are in a partscore, taking 11 tricks to their 9

 

 

 

Sorry for the long list... I'm sure anyone could figure it out easily just staring at an IMP table.

 

For those who scoff at 1 IMP, apparently you haven't lost that many matches by this margin... although with such an attitude it's more likely that you would lose by many. Why not try to pick up an IMP for free? For my part I appreciate it when my partners never say die, especially when they're put into what appears to be a losing situation, since that must certainly mean good things for us when we're in a winning position.

[gwnn]

I hope you don't mind me asking: why do you think this is the best aspect of IMP scoring?

 

I think this is a relatively uncommon situation because I will almost never be sure that the other table is in slam.

[kfay]

I hope you don't mind me asking: why do you think this is the best aspect of IMP scoring?

 

I think this is a relatively uncommon situation because I will almost never be sure that the other table is in slam.

I mean it just makes sense to me.

 

Why do people have favorite colors? :blink:

 

A lot of the other scores just seem random to me, although maybe they wouldn't be if I looked deeper. But the lines here are obviously drawn at the game bonus-50 and it shows. I just like it.

[billw55]

For those who scoff at 1 IMP, apparently you haven't lost that many matches by this margin... although with such an attitude it's more likely that you would lose by many.

I know what you mean. When I land in a cold game, it is really easy to lose focus and just lazily make it. But an overtrick is an IMP, you should always be on the lookout for safe ways to get them.

[cherdanno]

Nice point Kevin, even though I don't think this hand is the ideal one to make it.

[hanp]

I'd guess diamonds are 6-1 and finesse. I didn't get awm's point about the finesse being the percentage play, is that because the double was ignored or because the heart queen makes a double more likely?

 

If I was in slam on a black suit lead, I might strip the black suits and finesse hearts into west.

[bab9]

For those who scoff at 1 IMP, apparently you haven't lost that many matches by this margin... although with such an attitude it's more likely that you would lose by many.

I know what you mean. When I land in a cold game, it is really easy to lose focus and just lazily make it. But an overtrick is an IMP, you should always be on the lookout for safe ways to get them.

I was always told by more experienced players at my original club, forget overtricks and just make sure you make the contract you bid.

 

In situations where we have lost/won by about 1 IMP, we have generally done badly overall because at other tables teams have had a greater IMP differences.