[AlexJonson]

If you polled me on this hand I would pass with no second choice ( auction 2NT - 3NT natural with a penalty double).

The bidding side may need to call the TD, and the doubler may well want to call the TD anyway to get his double removed.

Easy auction for me (after ignoring partner's explanation) because I don't see an LA to pass in this or any reasonably closely related universe.

[aguahombre]

Enjoying this thread.

One would think our agreements on 1N-3N, and 2N-3N, would not be in question. It is so unusual for the two sequences to be anything but a NT raise that we had better know what it means, if not that.

And, yes, those two sequences are not normal raises for us. But if we don't alert them correctly as to what they are, or if we forget our agreement, we should first be shot....then we should suffer whatever happens.

Pass by responder is 100 % regardless of any noises by partner (UI or otherwise). Opener sat.

[gnasher]

 aguahombre, on 2012-April-12, 19:50, said:

One would think our agreements on 1N-3N, and 2N-3N, would not be in question. It is so unusual for the two sequences to be anything but a NT raise that we had better know what it means, if not that.

Our agreements weren't in question - I just momentarily forgot them.

[iviehoff]

 gnasher, on 2012-April-12, 04:26, said:

But isn't it "affected by UI from another source"?

Indeed. But you are always perfectly entitled to know the explanation itself - ie what is the meaning of their double over an artificial 3N. What is UI is (among other things) the fact that the opponent has (or should have) a hand consistent with that explanation.

[FrancesHinden]

 AlexJonson, on 2012-April-12, 18:19, said:

The bidding side may need to call the TD, and the doubler may well want to call the TD anyway to get his double removed.

He can try, but as the double was totally insane whatever 3NT meant, I don't think he's going to succeed.

[AlexJonson]

 FrancesHinden, on 2012-April-13, 14:36, said:

He can try, but as the double was totally insane whatever 3NT meant, I don't think he's going to succeed.

May versus you know the hand. You win.

[FrancesHinden]

 AlexJonson, on 2012-April-13, 14:51, said:

May versus you know the hand. You win.

Sorry, I was aware I had additional information about the hand. As the UI question seems to have been solved, I'll tell you what the doubler had:

KQ10x
J10xx
-
AKxxx

[nige1]

 FrancesHinden, on 2012-April-13, 14:36, said:

He can try, but as the double was totally insane whatever 3NT meant, I don't think he's going to succeed.

 FrancesHinden, on 2012-April-14, 03:08, said:

As the UI question seems to have been solved, I'll tell you what the doubler had: KQ10x J10xx - AKxxx

Is the doubler "totally insane"? He "knows" that the prospective declarer has at most three ♠ so he hopes to indicate a ♠ lead rather than a ♦ lead.

[lamford]

 nige1, on 2012-April-14, 21:06, said:

Is the doubler "totally insane"? He "knows" that the prospective declarer has at most three ♠ so he hopes to indicate a ♠ lead rather than a ♦ lead.

I can easily visualise layouts where a spade lead is needed, and I think double does ask for a spade. However, I expected 3NT to be making far more often than not. But when I plugged it into to bridge analyser, using DF, I found the opposite. Assuming the responder has 4-7 with four spades, 3NT is favourite to be beaten with the right lead. Originally I had a different result, because I wrongly had this hand on lead, and often a spade lead is from the wrong side. The figures are now 67% 3NT beaten, 33% making. I have no idea how many of these require a spade lead, but the computer changed my view dramatically, and I now think double is a huge winner, and certainly not "totally insane"! How long will it be before computers are better than humans at bridge? In chess and backgammon they are already streets ahead.

And, for completeness, the doubler does not get to remove the double whether or not it was SEWoG because we are told the explanation was correct and gnasher misbid. If the explanation had been wrong, I would allow the double to be withdrawn.

[blackshoe]

DF?

[lamford]

 blackshoe, on 2012-April-15, 16:51, said:

DF?

Deep Finesse, which most simulation programs use to assess the success or otherwise of contracts

[blackshoe]

Ah. I suppose I should have recognized that.

[FrancesHinden]

 lamford, on 2012-April-15, 15:46, said:

I can easily visualise layouts where a spade lead is needed, and I think double does ask for a spade. However, I expected 3NT to be making far more often than not. But when I plugged it into to bridge analyser, using DF, I found the opposite. Assuming the responder has 4-7 with four spades, 3NT is favourite to be beaten with the right lead. Originally I had a different result, because I wrongly had this hand on lead, and often a spade lead is from the wrong side. The figures are now 67% 3NT beaten, 33% making. I have no idea how many of these require a spade lead, but the computer changed my view dramatically, and I now think double is a huge winner, and certainly not "totally insane"! How long will it be before computers are better than humans at bridge? In chess and backgammon they are already streets ahead.

I don't understand your argument at all.
Given that double asks for a spade lead, it is a winner if either
(i) partner leads a spade and otherwise would not have done, and that beats the contract which otherwise would make
(ii) the size of the penalty we do get from the times when we beat the contract on a spade lead outweighs the losses when it makes

saying that 3NT goes of 67% of the time doesn't tell you anything useful about the merits of doubling, because you have 'no idea' how often a spade lead is right and you miss out on every time you are running the club suit (or declarer has one club stop and a card to knock out) partner has led a spade. As a very simplistic counterargument, it's possible that on all 67% of the hands when 3NT is off, it's off on a club lead, and now the double is a huge loser.

Also, doubling loses (i) when they redouble and (ii) when opener was about to bid 4S (iii) when they run

[nige1]

 FrancesHinden, on 2012-April-23, 14:42, said:

I don't understand your argument at all. Given that double asks for a spade lead, it is a winner if either
(i) partner leads a spade and otherwise would not have done, and that beats the contract which otherwise would make
(ii) the size of the penalty we do get from the times when we beat the contract on a spade lead outweighs the losses when it makes

saying that 3NT goes of 67% of the time doesn't tell you anything useful about the merits of doubling, because you have 'no idea' how often a spade lead is right and you miss out on every time you are running the club suit (or declarer has one club stop and a card to knock out) partner has led a spade. As a very simplistic counterargument, it's possible that on all 67% of the hands when 3NT is off, it's off on a club lead, and now the double is a huge loser.

Also, doubling loses (i) when they redouble and (ii) when opener was about to bid 4S (iii) when they run

Earlier, I mistakenly wrote that the doubler knew that declarer had less than 3♠. I now realise that is not true.

We are told the doubler had ♠ KQ10x ♥ J10xx ♦ - ♣ AKxxx
I don't think the double is insane, although there is the frightening downside of possible redouble/overtricks .
Suppose you judge that

• Partner is likely to lead a red suit unless you double for a ♠ lead and
• A ♠ lead is more likely to defeat the contract than a red suit lead.

Then, I think you are in a kind of Lightner double context i.e. If you don't double the contract is likely to make but If you double it will often go down.
You double, not to increase the penalty, but in the hope of defeating a contract that you judge you are unlikely to defeat otherwise.

Most games bid by good players have a less than 50% chance of making. Hence, in these circumstances, the arithmetic seems to be on the side of the doubler

With this understanding, you may need to double more often because, when you don't double partner is even less likely to make the directed lead.
Paradoxically, that may be a hidden bonus. When you don't double, there are useful negative inferences available to partner.

In the brave new world of Puppet Stayman, opponents often bid suits they don't have. Doubling a suit may be safer than doubling 3N. But, in general, I don't think that players double enough. Depending on your hand, the form of scoring, and your probability estimates, the case for doubling must sometimes be overwhelming

Here, Frances thinks the doubler's judgement is faulty and she may well be right but that does not undermine the underlying principle.

[WellSpyder]

 nige1, on 2012-April-23, 20:57, said:

Most games bid by good players have a less than 50% chance of making.

Are you sure????

I know good players will bid games with less than a 50% chance of making, especially when vulnerable and playing for imps, but they will also usually manage to bid the solid games with an easy 10 or 11 tricks, too.

[nige1]

 WellSpyder, on 2012-April-24, 02:43, said:

Are you sure???? I know good players will bid games with less than a 50% chance of making, especially when vulnerable and playing for imps, but they will also usually manage to bid the solid games with an easy 10 o11 tricks, too.

Sorry. Thank you, WellSpyder. "Most" was wrong. I should have have written "Many games bid by good players have a less than 50% chance of making" .

This thread poses interesting questions on on several levels.

• In theory, is it sound to play Lightner doubles of games?
• Could a Lightner double be appropriate on this auction? on this hand?
• Suppose the doubler claimed damage from misinformation. Because, for the sake of argument, Gnasher's actual agreement is that 3N is natural, to play and nothing to do with ♠. Would the director treat the double as a serious error?
• Presumably, Paul Lamford (after his simulation) would rule NO. Frances Hinden (who rates the double as "insane") would rule YES. It seems that neither would have much doubt about their ruling. They are mavens of Bridge and Bridge Law. Most directors lack their expertise in either area. Does this underline how daft SEWOG law is?

[AlexJonson]

 lamford, on 2012-April-15, 15:46, said:

I can easily visualise layouts where a spade lead is needed, and I think double does ask for a spade. However, I expected 3NT to be making far more often than not. But when I plugged it into to bridge analyser, using DF, I found the opposite. Assuming the responder has 4-7 with four spades, 3NT is favourite to be beaten with the right lead. Originally I had a different result, because I wrongly had this hand on lead, and often a spade lead is from the wrong side. The figures are now 67% 3NT beaten, 33% making. I have no idea how many of these require a spade lead, but the computer changed my view dramatically, and I now think double is a huge winner, and certainly not "totally insane"! How long will it be before computers are better than humans at bridge? In chess and backgammon they are already streets ahead.

And, for completeness, the doubler does not get to remove the double whether or not it was SEWoG because we are told the explanation was correct and gnasher misbid. If the explanation had been wrong, I would allow the double to be withdrawn.

Of course the potentially damaged player can try to have an unsuccessful double removed.

If the explanation is correct, there remains the possibility of fielding. We have already seen an insane double redefined as almost mandatory. Still plenty of room for our TD to see beyond MI.