A Tricky Trump Split

[OliverC]

Sometimes you encounter a truly horrendous trump split in what ought to be a fairly straightforward contract. This was an interesting hand from a Declarer-play perspective.

NS Game, Dealer North

You are North and open 1 . Everyone passes.

East leads the Ace of Hearts and this is what you see:

South (Dummy)
  5
  J542
  A1092
  Q976

A led

North
  AK843
  107
  KJ3
  A42

East cashes the AK of Hearts and continues with the 3 on which West plays the 6, 8 and then the Queen and you ruff the 3rd Round. You try the AK of Spades and East shows out on the second round. Not good! How do you plan the play?

Clearly West started with 6-card Spades and exactly 3 Hearts (they would definitely have played the 9 rather than the Queen on the 3rd round if they started with 4). That gives them only 4 cards in the Minors. Clearly you cannot avoid losing 6 tricks in the Majors so you have to try to avoid losing any other tricks in the Minors.

Another strand to this is as follows: East has already shown up with the AK of Hearts. They are unlikely to have too much else or they would surely have made a takeout double over 1 so either the Queen of Diamonds or the King of Clubs (if not both) will probably sit with West.

A decent plan, therefore, is to cash the King at trick 6 and then (assuming West plays low) either to play a Diamond to the Ace and try to drop the Queen in the West hand or to play a small Diamond to the 10 and play East for the Queen, since they surely have the greater Diamond length. Whichever you choose, if you now lead the Jack of Hearts, West is stuffed if they started with 2 Diamonds.

If West ruffs the Jack of Hearts you simply discard a Minor suit card (if you cashed the King Ace of Diamonds, and the Queen didn't appear then discard a Diamond, but if you opted for a finesse of the 10 rather than trying to drop the Queen the obvious play is to discard a Club). West can cash all of their Spades, but will eventually have to lead a Club (probably away from the King, as explained above).

If, on the other hand, they discard a Diamond, then they still have a Club left and you can lead a Club to the Ace. If they discard a Club, then definitely discard a Diamond from your hand and lead a Diamond from Dummy (now, whether they follow or ruff high in front of you, you must come to another trick in the end).

That line works as long as West doesn't have 4-card Diamonds and a void in Clubs, but that would give East at least an 11-count with KJ10xxx in Clubs and they would surely have overcalled when they have a singleton Spade.

What doesn't work is what Partner tried, which was to play a Diamond to the Ace (without cashing the King first) and then leading the Jack of Hearts, because West still has a safe exit card (their 2nd Diamond) when they've taken all of their trumps. That line works when West has a singleton Diamond, but if you're going to play for that, you must cash the King first and then run the Jack (or play a small Diamond to the 10).

[brian_m]

Not in any way wishing to comment on what you say about playing this specific hand, this sort of bidding problem is something about which Georgi and I had extended discussions.

I've been brought up on the idea of dropping the bidding at the 1-level on North's hand, as happened here, but Georgi tried to persuade me that the odds in fact favour a 1NT response on North's hand. I ran a bunch of simulations to try to cover the various possibilities, and without having a double-dummy solver available, it did appear to me that Georgi was right (although not overwhelmingly so), and the 1NTF on  1=4=4=4 shape (0=4=4=5 is even better of course!) does seem to pay off, even if only slightly, in the long run.

I will try to find the simulations I ran (it needed a whole bunch of them to cover the various permutations), and if I can, will take another look at them and post the results if I still think they support 1NTF with North's hand.

[OliverC]

Part of the problem, brian, is the (fairly sacrosanct for very good reasons) guidance on Opener's rebids after 1M-1NT-??. Showing a second 4-card or longer suit has a higher-priority than rebidding a 6-card Major. Similarly, when Opener rebids 2m, they might only have a 3-card suit when they're 5332. That gives South on this hand a bit of a dilemma when the bidding starts 1 - 1NT - 2 and an even worse problem if they happen to have 5-cards in a red suit and only 3-card Clubs, because they really will have no idea whether to leave Opener in Clubs, give preference to Spades (which will really mislead Opener about their hand) or even to insist on their 5-card red suit.

Lots of the time Opener will, in fact, rebid 2 over 1NT so now you're obliged to pass that but you've given up all of the advantage that Precision naturally has of being able to play this hand at the 1-level rather than at the 2-level like every other system will end up doing.

On many other hands when Responder passes, Opps feel they are obliged to protect, so they will end up playing in one of your 4-card suits instead.

Give South another couple of high-card points and I'd agree with you that the potential benefits in bidding something outweigh the possible dangers, but not with a bare 7-count. 1 making is a plus IMP score (only about 2 IMPs), which is always good enough on a part-score hand. Several people were making 7 tricks in Spades. Nobody was making 8 tricks. I rest my case

[brian_m]

Part of the problem, brian, is the (fairly sacrosanct for very good reasons) guidance on Opener's rebids after 1M-1NT-??. Showing a second 4-card or longer suit has a higher-priority than rebidding a 6-card Major. Similarly, when Opener rebids 2m, they might only have a 3-card suit when they're 5332. That gives South on this hand a bit of a dilemma when the bidding starts 1 - 1NT - 2 and an even worse problem if they happen to have 5-cards in a red suit and only 3-card Clubs, because they really will have no idea whether to leave Opener in Clubs, give preference to Spades (which will really mislead Opener about their hand) or even to insist on their 5-card red suit.

I would disagree that this is a problem, Oliver. If you bid 1NTF on the above hand you pass opener's minimum rebid, end of story. If opener rebids 3H, then you might well wish to reconsider - but if opener rebids 2 , then you made your choice when you bid 1NTF. You pass it, for the reasons you state so clearly.

Lots of the time Opener will, in fact, rebid 2 over 1NT so now you're obliged to pass that but you've given up all of the advantage that Precision naturally has of being able to play this hand at the 1-level rather than at the 2-level like every other system will end up doing.

Agreed - but you only play 2 on at least a 6-1 fit. In exchange for playing 2 rather than 1 on a 6-1, you often get out of playing 1 on a 5-1. You pays your money...

On many other hands when Responder passes, Opps feel they are obliged to protect, so they will end up playing in one of your 4-card suits instead.

Sure, if you think opps will rescue you, that changes things. I don't have that faith in my opponents. If 2nd hand is sandbagging, it's quite possible that 4th hand has no good rescue bid, with the possible exception of a protective double.

Give South another couple of high-card points and I'd agree with you that the potential benefits in bidding something outweigh the possible dangers, but not with a bare 7-count. 1 making is a plus IMP score (only about 2 IMPs), which is always good enough on a part-score hand. Several people were making 7 tricks in Spades. Nobody was making 8 tricks. I rest my case

As you know full well, one hand proves nothing.  I will see whether I can find those simulator files. If I can, I will run them again and post some numbers. You may not change your mind, but having extra data on which to base a decision is seldom a bad thing.

[OliverC]

Brian, you say you might get out of playing in a 5-1 fit but, as I've pointed out, some of the time you will end up in a 4-3 fit at the 2-level instead of a 6-1 fit (which might play better) at the 1-level. Here, Responder was 1444 exactly, but they will not always have that 4-card support for all of the other suits and might well up playing in a 3-3 Minor-suit fit. Sometimes, no question, you will end up in a better 4-4 or even 5-4 fit.

As you say, you might still scramble the required number of tricks even in a poor fit (Burns Law violation), but often you won't. Against that, the pressure on Opps to balance when the opposing Spade split is not quite so extreme is considerable. Here it was the defender in the protective position who had the 6-card Spades and so could happily pass. If it was the other way around, 4th-in-hand might double and 2nd-in-hand would pass that for penalties, but now Responder still has a bid left and any move by them must be showing a complete lack of tolerance for Spades, which at least makes the position clear to Opener whether Responder bids 1NT, or an SoS redouble, or whatever.

OCP doesn't mandate what 1M-1NT shows, merely what the most common hand-types are that use it and what the normal continuations with each of those responding hand-types will be. If you feel like bidding 1NT on this kind of hand, feel free.

As you say, you pays your money and takes your choice. I know what my money is on, though... In my experience, sod's law says that if I am ever tempted to bid 1NT on this sort of hand (but a little stronger, as I said), Partner invariably rebids their Major and so we just play 1 level higher than we might have. Most of the time we're still making 8 or maybe even 9 tricks (because I will normally only do it on about a 9 or 10-count), but sometimes we only making 7 tricks. Occasionally we even find a better fit somewhere else, but my luck is not that good that it happens very often

[brian_m]

<snip>

As you say, you pays your money and takes your choice. I know what my money is on, though... In my experience, sod's law says that if I am ever tempted to bid 1NT on this sort of hand (but a little stronger, as I said), Partner invariably rebids their Major and so we just play 1 level higher than we might have. Most of the time we're still making 8 or maybe even 9 tricks (because I will normally only do it on about a 9 or 10-count), but sometimes we only making 7 tricks. Occasionally we even find a better fit somewhere else, but my luck is not that good that it happens very often

As you're aware, I have faith in the statistics classes that you skipped to learn Greek and Latin.    I admit that I was of your opinion until Georgi tried to convince me otherwise sufficiently often that I ran the numbers in order (so I thought!) to disprove what he said, but it was me who was in for the surprise. If I can't find the scripts, then I will re-create them. It may take me a little while because I have PC problems at the moment, I think a motherboard fault as the whole shebang is locking solid every 60-90 minutes or so, requiring a big red button recovery. I've eliminated the graphics card and drivers as a source of the problems (happens with both NVidia and Radeon based cards on two distros with very different parentage) so I've got a new barebones setup on the way, but I have to cannibalise the current PC and obviously do a full reinstall, that burns up a few days worth of my spare time these days.

[kiwi_2o1]

I'm voiced my opinion before on bidding with North's type of hand and dislike Oliver's approach. It's his system so if I played with him or standard OCP I'd have to stick with that. In my experience it pays to get active and to try and play the contract at the 1 or 2 level. To pass on that North hand just gives up the battle.

Recently I purchased a quality article on bidding over weak and strong NT's with DONT. I wanted to compare the action with the methods I play locally. I'm a bit staggered on the weak hands opponents now days are prepared to bid on over a 15-17 NT. Hands that I wouldn't have previously considered over a weak opening. It's not exactly the same but if they are prepared to bid just to grab the contract or disturb the 1NT opening, then I'm willing to support partner more than OCP recommends when we open with 11-15 bid to exert more pressure at the table. Like Brian and Georgi, I think, on balance a more liberal responding approach brings in a higher share of positive boards. If the point spread between the two sides is even or close to it, as the 15 + 7 is here, then I don't want to announce weakness and allow them a lead directing or easier balancing bid. 11 + 7 is fine for me too. Fewer pairs now days have doubles that can catch you for a penalty at the two level. For me this penalty loss swings the balance in being more active as responder.

To contrast I remember the Roth-Stone Systems', keep standards up policy; supported by liberal balancing back in. IMO they got shut out of auctions too often. To me one of the main advantages of limited openings is that responder can bid freely knowing that opener very rarely will make an embarrassing response. In standard one needs to sometimes make a bid on junky values with the fear of game or GF bids from opener. But North's hand here isn't junk. Its worth one bid.

Likewise Oliver's point about bidding an outside 4 card suit with 6M is debatable IMO too. With say 6 spades I would rather have 1S - 1NT, 2S than 1S - 1NT, 2C/2D giving the opponents an easier ride. Also it means the openers rebid is likely to be at least 3+ if the M is only 5 cards. That's more comfortable and certainly better than say Acolities bear, who are sometimes required to bid 1S -1NT, 2S on 5 cards. Although its true the only 6 cards are known compared to 8 with say 1S - 1NT, 2D on balance I would rather know that opener is 6+ rather than 5 with an outside 3+ suit.

[OliverC]

LOL, Ash: It's been one of the fundamental tenets of playing Precision Club since the late 1960's. It's not my approach. It's Precision's approach.

As I said before to Brian, OCP doesn't say you must pass with this kind of hand. It's simply a matter of where you, Brian or I, personally, draw the line between passing and bidding. For myself, I tend to pass with less than an 8-count (and maybe even some poor 8-counts), and take some action (eg: 1NT) if I have a decent 8 or more but, as always, I take each hand as it comes rather than giving myself hard and fast rules set in concrete.

If someone else sets their line lower (or even higher), that's absolutely fine with me. OCP doesn't set a line. All the system says is that a hand that would typically bid a limit 1NT over a Major Suit opening, playing Acol or Standard American, will normally pass when playing Precision. I may be wrong, but I rather think that is almost identical to what Eric Jannerstern says somewhere in his excellent book on Precision (which is what I read when I first really started to learn Precision).

Similarly with the order of priority with Opener's rebids over a Forcing 1NT response. Those are definitely not my original ideas. Admittedly, the reasoning behind them may originally have had less to do with making the bidding as constructive as possible and more to do with  keeping the bidding as low as possible for times when when Responder has the "long weak suit" variety of response (which, if you're not playing Asking Bids, is always available in Precision [and 2/1]). I don't know.

Once again, if you decide to rebid 2M on a 6-card suit and hide a 4-card Minor rather than showing it, that's a matter for you to decide (you're playing the hand, after all), but that approach definitely is at odds with Brian's strategy of bidding 1NT on a 1444 6 or 7-count in the hope of Opener showing a second suit, because when you're 6-4 you are not going to be showing the second suit and will simply end up playing in 2 rather than 1 .

[brian_m]

Just by way of information, my approach to this mirrors what I thought was Oliver's approach (again, I think it was an e-mail discussion which I vaguely remember).

1) A real (4+) card side suit almost always takes precedence. 'Almost' because I admit I might be tempted to rebid  AKQJxx in preference to  5432! 

2) If you've no 4+ side suit then show the 6+ spades, if you have them.

3) If not 1) or 2) then bid the lowest 3 card suit.

Back to the original hand which started all this, we should note that things may be different after a 1 opener rather than a 1 opener, because of the possibility of opener rebidding a doubleton minor after 1 with a minimum hand and 4=5=2=2 shape. That could quite possibly be enough to change the percentage action to a pass, as Oliver advocates. With a 1 opener, obviously opener must have at least 3 cards in any new suit.

[brian_m]

OK, I recreated the scripts (e-mailed on request). To be included, South must have any 4441 or 5440 shape with short spades and 0-8 HCP. North must have a 1 opener (2NT opener excluded) and it's assumed that North will rebid any 4+ side suit in preference to rebidding a 6+  suit.

Given those criteria, South will hit a 4-3 or better with North holding only 5 card  on 52,965 deals out of 100,000,000.  North will rebid 2 with a 6+ card suit and no 4+ side suit on 19,252 deals out of 100,000,000.

Yes, it surprised me when Georgi got me to run these scripts for the first time, but there you go! 

Now, before Oliver points it out, this takes no account of what contracts may or may not be makable, nor does it take any account of opponents rescuing you. The only question answered is if opponents stay silent, are you better off responding 1NT or not. By a factor of about 2.5 to 1, it would appear than the answer is "Yes".

As I said upthread, the question is significantly more complicated when the opener is 1 , and the scripts will need a bit more thought.