[kenberg]

 nullve, on 2018-July-24, 08:09, said:

Idea:

After winning the first trick with the ♦A, play a spade to the A and another spade. If East wins the second spade, then blah blah. If he ducks, play a third spade and discard the ♥Q from hand. If East wins the third spade, then blah blah. If he ducks again, cash the ♦K and play three rounds of clubs and discard a spade or small heart from dummy. If clubs are 3-3, 12 tricks are guaranteed. If clubs are 4-2, in which case West is almost certainly the one with four, play a fourth club and make sure to keep ♥AJ and the remaining diamond in dummy, forcing West to give you two heart tricks unless he can exit with a third diamond.

Indeed, that might be the best. This is an interesting hand!

[mikeh]

 nullve, on 2018-July-24, 08:09, said:

Idea:

After winning the first trick with the ♦A, play a spade to the A and another spade. If East wins the second spade, then blah blah. If he ducks, play a third spade and discard the ♥Q from hand. If East wins the third spade, then blah blah. If he ducks again, cash the ♦K and play three rounds of clubs and discard a spade or small heart from dummy. If clubs are 3-3, 12 tricks are guaranteed. If clubs are 4-2, in which case West is almost certainly the one with four, play a fourth club and make sure to keep ♥AJ and the remaining diamond in dummy, forcing West to give you two heart tricks unless he can exit with a third diamond.

We are of course discussing this hand on the assumption that diamonds are not coming home. In a sense, we are really discussing how we'd play the hand had West followed to the first diamond with a small card, given that the carding of the 3 by East and the Jack by West makes it appear odds-on that the suit is coming home (even a bot, or especially a bot, would not play the J from Jx....caveat: I have never played against a bot but have watched a few times as my wife has).

On that basis, there is, I think, zero chance that any competent West would ever duck the spade K from Kxx or Kxxx...anyone with any understanding of the game would win and return a spade (or, possibly, a diamond if the play was timed out to allow that).

Meanwhile, I don't understand the 'blah, blah' should East win the 2nd or 3rd spade.

The following may strike some as esoteric, but on this hand East should always switch to a heart if he wins the third spade (and he should never duck from Kxxx). Why?

If he has the King, he knows you hold the Queen, and he can count 12 winners for you should you take the finesse. So he has to hope that you have some other route to 12 tricks, and that you will not want to take a losing finesse at this point in the hand.

But what, you may say, if he lacks the King? Won't he be worried that the switch will be into declarer's K10(x)(x), or, if he has the 10 himself, into declarer's K9(x)(x)?

Here is the esoteric part, and I wouldn't expect any but an expert West to get this right.

If West owns the heart King, he needs to tell partner, and he does this by the cards he plays in spades.....he should play from the top down, regardless of how many he holds. Count is obviously irrelevant once declarer plays low from dummy so giving a count signal is not only unnecessary but known by both defenders to be unnecessary. A very good rule is to never give count unless count may help partner.

If West lacks the King, then he should STILL signal for hearts! Why?

From his perspective, if East has either the heart K or Queen, he will have no other meaningful card, and declarer is going to eventually take the winning finesse. Since declarer cannot go wrong in hearts in that case (especially if West owns the 10), make declarer guess, and to add a little spice to the mix, announce you have the heart King by your signals.

If East has nothing in hearts, then a heart switch cannot cost. So heads the heart switch may win, by causing declarer to refuse the winning line, or tails it breaks even by giving nothing away.....bear in mind that if declarer has KQx in hearts, and only a 15 count, he is probably going down so long as we defend passively.

The more I have looked at this hand, the more interesting it would be, if diamonds lay differently, say 10xxx on lead and Jx in West

[jammen]

 mikeh, on 2018-July-24, 11:51, said:

We are of course discussing this hand on the assumption that diamonds are not coming home. In a sense, we are really discussing how we'd play the hand had West followed to the first diamond with a small card, given that the carding of the 3 by East and the Jack by West makes it appear odds-on that the suit is coming home (even a bot, or especially a bot, would not play the J from Jx....caveat: I have never played against a bot but have watched a few times as my wife has).

On that basis, there is, I think, zero chance that any competent West would ever duck the spade K from Kxx or Kxxx...anyone with any understanding of the game would win and return a spade (or, possibly, a diamond if the play was timed out to allow that).

Meanwhile, I don't understand the 'blah, blah' should East win the 2nd or 3rd spade.

The following may strike some as esoteric, but on this hand East should always switch to a heart if he wins the third spade (and he should never duck from Kxxx). Why?

If he has the King, he knows you hold the Queen, and he can count 12 winners for you should you take the finesse. So he has to hope that you have some other route to 12 tricks, and that you will not want to take a losing finesse at this point in the hand.

But what, you may say, if he lacks the King? Won't he be worried that the switch will be into declarer's K10(x)(x), or, if he has the 10 himself, into declarer's K9(x)(x)?

Here is the esoteric part, and I wouldn't expect any but an expert West to get this right.

If West owns the heart King, he needs to tell partner, and he does this by the cards he plays in spades.....he should play from the top down, regardless of how many he holds. Count is obviously irrelevant once declarer plays low from dummy so giving a count signal is not only unnecessary but known by both defenders to be unnecessary. A very good rule is to never give count unless count may help partner.

If West lacks the King, then he should STILL signal for hearts! Why?

From his perspective, if East has either the heart K or Queen, he will have no other meaningful card, and declarer is going to eventually take the winning finesse. Since declarer cannot go wrong in hearts in that case (especially if West owns the 10), make declarer guess, and to add a little spice to the mix, announce you have the heart King by your signals.

If East has nothing in hearts, then a heart switch cannot cost. So heads the heart switch may win, by causing declarer to refuse the winning line, or tails it breaks even by giving nothing away.....bear in mind that if declarer has KQx in hearts, and only a 15 count, he is probably going down so long as we defend passively.

The more I have looked at this hand, the more interesting it would be, if diamonds lay differently, say 10xxx on lead and Jx in West

Only if you believe that the bot will always lead 4th best vs 6nt, which I don't. As I've posted above I once had a bot lead J from Jxx, spurning his partners 1♦ opener [url=" (http://www.bridgebas...__1#entry886232)./url%5D
GIB system notes also say "How GIB Defends: It's difficult to describe precisely how GIB defends. It doesn't use rules and guidelines, like humans often do."

However, I'm certain that I misplayed the hand and should have maintained my squeeze possibilities as well as played the spade suit correctly.

[smerriman]

 jammen, on 2018-July-24, 15:34, said:

Only if you believe that the bot will always lead 4th best vs 6nt, which I don't. As I've posted above I once had a bot lead J from Jxx, spurning his partners 1♦ opener [url=" (http://www.bridgebas...__1#entry886232)./url%5D
GIB system notes also say "How GIB Defends: It's difficult to describe precisely how GIB defends. It doesn't use rules and guidelines, like humans often do."

It still uses rules for deciding which card to lead, just that simulations can override it. Thus the reason it makes the "normal" lead 95+% of the time. Some of the other 5% are due to simulations; the rest developers wouldn't respond to.

[kenberg]

The hand is presented as being played against the bots and so the bot leading choices are relevant. But it is also interesting to imagine the bots are just normal opponents. Here I am not sure it matters much but in fact I think a human is slightly more apt than a bot to deceptively lead fifth best from T8xxx. Bots are programmed and, as has been noted; they follow rules unless over-ridden by simulations. But I do not think that the simulations take into account the human factor, thinking that it might be wise to make a fifth best lead when a fourth best is promised. Their discards can be "deceptive", almost random. And they do seem to know such things as "play the card you are known to hold". But I don't think that I have ever seen them lead, at least at T1, a J from QJx for example. There are some standard false-card situations that can be employed near the end of a hand, but at T1 if they are leading from T8xyz then I think that they lead the y. As would most humans. If they lead that suit at all.

It is probably wandering a bit off topic but I would like to see an exploration of the Gib mind on defense. I very much believe that when I get a bad result against the Gibs it is far more likely that it is because of an error of my own rather than because of some great Gib cleverness or duplicity, but I still think it would be fun to explore the Gib mind.