[lmilne]

These are both from Kate McCallum's partnership checklist.

"Partner leads your unbid 5-cd suit vs. 3NT and you win the 1st trick (His view is that you could have only three.) Which card do you return?"

"Partner is ruffing & declarer is over-ruffing. Partner doesn’t know it. Which-card do you lead?"

[mikeh]

original 4th best.

I will be interested in the agreements that posters like Frances, Justin et al play here, if other than this. It seems to me that there is unlikely to be any better agreement. It will be relatively rare that partner will truly be in the dark to the extent of not being able to work out our relative length.


on the second one:

Since we are expecting partner to ruff, I assume also that dummy has the high card(s) in the suit.

It will, again, be rare that partner has no clue as to the count, given that one assumes some rounds of the suit have been played to get to this point, and the auction plus any other suits played will have informed partner to a great extent. However, there can undoubtedly be cases in which partner isn't aware of the situation.

I don't have any agreements with any partner. However: I wouldn't put partner into this position if I were concerned that he might blow his trump holding by ruffing...I am doing it either to kill winners in dummy or to hope for an uppercut, and partner would usually recognize this and play accordingly. I suppose one could agree that an obviously low card assures partner he can ruff without overruffing and a high one the reverse....or the other way around, but I don't see how that would work in a situation in which, by definition, we are assuming that partner thinks declarer may still hold one or more cards in the suit.

I think one relies on one's partner to be able to work this out. I don't recall this ever being a problem with a good partner opposite. Weaker players, yes, since they tend to woodenly ruff low all the time But you aren't Bowl bound with a weak partner

[wank]

in general return your lowest from 5. partner might think you've only got 4, but that's normally fine. you just want to make sure he continues them. if you return 4th highest, partner might take you for 3 with the t1 card, the 4th highest and the 5th highest and so switch.

2) i don't know myself.

[Phil]

1. Return 5th best IF we know partner led from 4. I don't know what we are supposed to do if we know partner led from 3, but probably original 4th best I guess. It might depend on context.

2. Do we want partner to uppercut? If yes, lead low. If it doesn't matter, and we are simply ruffing away dummy's winner, I suppose we lead high.

[mikeh]

wank, on 2015-July-21, 10:08, said:

in general return your lowest from 5. partner might think you've only got 4, but that's normally fine. you just want to make sure he continues them. if you return 4th highest, partner might take you for 3 with the t1 card, the 4th highest and the 5th highest and so switch.

I think that in reality the choice is context-driven.

The problem with returning original 5th best is that partner may place declarer with an extra, and important, card in the suit.

Say that dummy is irrelevant. I can't portray all possible holdings, but let's assume that it is something like:

---

We win the Ace and return our lowest card. Declarer plays the 9 then the Q. Partner is on lead. Clearly he can afford to return the suit, and indeed needs to do so if we hold our side's entry, and will often want to do so anyway, to avoid opening another suit, or to influence opener's line of play.

But what if declarer held our 8? So he had QJ98? Now, for partner to lead the 3rd round is picking off our 10 and is likely to be a disaster.

Were we to have returned the 6, then partner can see that the 3 is missing, and it is now impossible for opener to have QJ98. He might, I suppose, hold QJ1098, but we can probably assume that to be irrelevant, since we have likely blown the hand by this stage of the play, if our first two plays were to set up his suit

Thus if we are going to consider anything but original 4th best back, we need to be very careful to ensure that we are not going to leave partner with an insoluble guess.

[gnasher]

I'd give current count, so the highest one I could afford. It will usually be easier for partner to distinguish between 3 and 5 than between 4 and 5.

If I knew that pretending to have four would solve a problem for partner, I'd do that, obviously.

I don't think it's sensible to have an agreement about the second one. Anything you agreed to cater for this rare situation would mess up your suit-preference signals on the everyday hands.

[FrancesHinden]

I agree the second one is context-driven; you have to look very closely at the pips to work out (i) what you want partner to do and (ii) what will give the clearest signal.
Our general agreement is that if partner knows you have at least 4, then return the highest you can afford to give count; if partner knows you have at most 4 then bottom from 4 original, top from 3 original
If you are trying to distinguish between 3 and 5 then original fourth highest is usually your best bet; but if you want partner to know you have 'at least' 4 then bottom

Mikeh's example is interesting: why is partner leading from Kxx? If we know that declarer has a stop in the suit, and the layout is something like Jx in dummy opposite a presumed Qxx in hand then the easy way to get this right is to duck the first round (with Jx opposite Kxx declarer would rise jack at trick 1).

[mikeh]

FrancesHinden, on 2015-July-21, 15:27, said:

Mikeh's example is interesting: why is partner leading from Kxx? If we know that declarer has a stop in the suit, and the layout is something like Jx in dummy opposite a presumed Qxx in hand then the easy way to get this right is to duck the first round (with Jx opposite Kxx declarer would rise jack at trick 1).

I have certainly led from Kxx before, if I suspect that partner has length and most of our side's entries. I've even seen the K led, tho I would not have chosen to do so. But the real reason for my example was that I was trying, in a bit of a hurry, to illustrate my point about not creating a problem for partner, and the E-S hands were the ones I came up with on a first try, which made the West hand sort of automatic if there were to be any point to the example I think I could come up with other, more plausible, examples if I spent some time, but I see no reason to do so.

[gnasher]

Mike, your example illustrates why "original 4th best" doesn't work very well. From West's point of view, the 6 could be from A1086.

In this situation current count works much better. West knows that the 8 is from A10863 or A8x, so returning the suit can't cost a trick.

[mikeh]

gnasher, on 2015-July-21, 15:51, said:

Mike, your example illustrates why "original 4th best" doesn't work very well. From West's point of view, the 6 could be from A1086.

In this situation current count works much better. West knows that the 8 is from A10863 or A8x, so returning the suit can't cost a trick.

I confess I don't see the point you are trying to make.

It will be rare that declarer can afford to conceal his lowest spot on both the first two rounds of the suit and, unless he does, partner will always be able to read our original 4th best. For example, with A1086, you cannot give me a hand on which declarer, with QJ93, can afford the 9 then an honour at tricks 1 and 2. He has no meaningful false-carding available.

Take my example as a case in point...you assert that the 6 could be from A1086. I say: impossible. It gives opener a holding on which he could not afford the 9 at trick one...how can he know where the 10 is? If partner has K8xx and he plays the 9 at trick one, he has blown a trick for no good reason that I can see.

Meanwhile, partner, who has led from a surprisingly short suit, so we are told to assume, doesn't know that we may have figured it out. For all he knows, he may have guessed poorly. So leading back the 8 from both original holdings of A8x and A10863 won't help much, especially since opener, blessed with QJ9xx, can afford to falsecard more commonly than from QJ9x, and may have guessed the lead accurately due to the rule of 11. Partner may think we think he has a 4 card suit, hence our return of the 8 from A8x. While I agree that the damage in the suit has been done, there is a lot more to defence than just going passive....maybe, having blown a trick or two, partner needs to find a switch. I admit that it is usually too late in such cases.

[gnasher]

mikeh, on 2015-July-21, 16:25, said:

I confess I don't see the point you are trying to make.

It will be rare that declarer can afford to conceal his lowest spot on both the first two rounds of the suit and, unless he does, partner will always be able to read our original 4th best. For example, with A1086, you cannot give me a hand on which declarer, with QJ93, can afford the 9 then an honour at tricks 1 and 2. He has no meaningful false-carding available.

Take my example as a case in point...you assert that the 6 could be from A1086. I say: impossible. It gives opener a holding on which he could not afford the 9 at trick one...how can he know where the 10 is? If partner has K8xx and he plays the 9 at trick one, he has blown a trick for no good reason that I can see.

Meanwhile, partner, who has led from a surprisingly short suit, so we are told to assume, doesn't know that we may have figured it out. For all he knows, he may have guessed poorly. So leading back the 8 from both original holdings of A8x and A10863 won't help much, especially since opener, blessed with QJ9xx, can afford to falsecard more commonly than from QJ9x, and may have guessed the lead accurately due to the rule of 11. Partner may think we think he has a 4 card suit, hence our return of the 8 from A8x. While I agree that the damage in the suit has been done, there is a lot more to defence than just going passive....maybe, having blown a trick or two, partner needs to find a switch. I admit that it is usually too late in such cases.

Sorry, you're right about that layout. I should come up with my own examples instead of trying to adopt someone else's.