Thursday, October 25, 2007

Slams from STAC week

There were some difficult slams presented in the hands from STAC week. This is one we got right.

Love all, dlr S (hands rotated)

♠ A J 4 2
♥ A 3
♦ Q 6
♣ A Q 10 9 5

[ ]

♠ K 9 7 6 3
♥ 6
♦ A J 8 4 3
♣ K 7

1♠    2♥    3♥   Pass
4♠    Pass  4NT  Pass
5♥    Pass  6♠   All pass

K♥ led.

3♥ was the usual limit raise or better, and 5♥ was 2 keycards without the Q♠. The play presented no difficulty when East produced the singleton Q♠ at trick 2. The South hand is clearly an opener, but North expressed surprise that I had accepted the game invitation with only 11 HCP. I pointed out that I only had 6 losers, and in fact I didn’t hesitate to go to game. Only about a quarter of the field bid this slam though, so perhaps people don’t follow the LTC well enough.

This one was tough, fortunately for the opponents, not us.

Game all, dlr W

♠ A K Q J 10 7 5 2
♥ 4 2
♦ 10 8
♣ J

[ ]

♠ 4
♥ A K 7 5 3
♦ A K 5
♣ 9 7 6 3

North opened 4♠ and played there. I have to admit that this is difficult. To make bidding the slam accurate, you need for North to show South both that he has an 8-card solid suit, and a singleton club. The opponents mentioned Namyats, but I doubt even that would do. Even if Namyats shows a solid suit, I don’t think you promise 8 cards. What is needed is something like a Gambling 3NT, but with a major suit instead of a minor. I don’t know if there is an opening scheme that includes such a weapon.

The opponents also had this one.

EW vul, dlr E

♠ A K 8 4
♥ A Q 3 2
♦ 8 7
♣ A 5 2

[ ]

♠ Q
♥ K 10 8 7 6
♦ K J 5
♣ K 9 8 6

South opened 1♥, North responded 2NT (strong ♥ support) and drove to 6♥. On the face of it, this isn’t a bad slam, really. If trumps are no worse than 3-1, you need to find the A♦ onside, or some other favorable diamond position combined with a good guess. In practice, you are pretty much bound to simply play to the K♦, but there is also the chance of an opening diamond lead. So it isn’t a great slam, but it’s by no means awful. And since the opening lead was a spade, and the A♦ is onside, they seemed destined to make it. However, declarer drew only one round of trumps before playing a diamond off dummy. East had passed as dealer holding seven diamonds to the ace, so he rose, and quickly played another for me to ruff. Down one: severe punishment, since the only other pair to bid the slam did make it.

Declarer didn’t anticipate the ruff, but he only needed one trump in dummy, to ruff the third round of diamonds for his 12th trick. So there was no reason not to draw trumps, and that was a plain error. Partner didn’t open 3♦ vulnerable because he felt he was too weak. We were trying to maintain some standards, even on pre-empts, so that was correct. I didn’t lead my singleton Q♦, because I feared I might be wasting its honor power. And North had bid spades at one point, so I hoped the J♠ opening might push through a tenace of some sort. That could be considered an error, though, because the Q♦ lead might win a couple of ways. Partner might have the A♦, and give me a ruff, or partner might have the K♦ and another trick. But I don’t see it as a serious error, because I could have been correct.

Friday, October 19, 2007

Roman Key Card

When I returned to bridge after my long hiatus, I found that Roman Key Card is all the rage, around here at least. At first, I was skeptical. I’ve seen a number of such conventions, almost all of which are consigned to the mists of history, and with good reason. But I have to admit, I’m impressed with the way RKC can make for halfway decent slam bidding with a complete stranger. I think that any convention needs partners to be in tune, to some extent, about the way that the convention should be used. For the daily grind of bridge in the real world, the really good conventions are the ones that show up as winners in a cost-benefit analysis when you are playing with someone you have never seen before, and about whom you have no idea whether you are in tune or not. Stayman is the one that everyone agrees about. I’m willing to argue that RKC is another winner.

Now, I’m not saying that RKC is a small thing, far from it. After a brief overview from someone, and a little experience playing it in half-assed fashion, I went looking and found Eddie Kantar’s book. I strongly recommend getting it and studying it if you want to play the convention to full power. And if you do, you will certainly agree that RKC is non-trivial. But most people play a watered-down version that loses some of the potential benefits, in exchange for losing most of the most dangerous areas for misunderstanding. I am willing to claim that such a watered-down version shows a big plus in slam bidding with strangers. And if I ever get a regular partner who’s willing to study Kantar, I bet the full-speed version really kicks ass.

The watering-down mainly consists of choosing either 0314 or 1430, always using 4NT as the asking bid, and forgetting about Exclusion Blackwood. Those three things eliminate probably 95% of the confusion potential. Kantar suggests playing both 0314 and 1430, and there are rules for figuring out which is in play for a particular ask – usually opener asks 1430 and responder asks 0314, but there are a lot of exceptions. Just picking one is a lot simpler, with little downside. Exclusion Blackwood is arguably not that complicated, but since we’re talking about a complete stranger, better not to go there. And always using 4NT as the asking bid is definitely simpler than figuring out what suit is available to use at the four level.

But that last choice has some serious downside. A problem for all forms of Blackwood is running out of space when the agreed trump suit is a minor. Something like Redwood (using 4 as the ask when aiming for ♣, and 4 as the ask when aiming for ) might be OK with a stranger. Kantar has some more specific (and therefore harder to remember) rules. An idea that I like, and I can’t remember if it’s in Kantar’s book or not, is to say that in a forcing auction, if you raise partner’s minor to the 4 level, you are not only setting the trump suit, but you are also asking. This isn’t too hard to remember for a strange partner, I think, and it has the potential to make some very smooth auctions when you run into a big hand that should clearly be going towards slam.

When the following hand came up at the Manhattan Bridge Club, nobody bid 7♣, and only a couple of pairs bid even a small slam.

♠ A x x x x
A x x
A
♣ A 9 x x

[ ]

♠ x
Q x
K Q J 10 x
♣ K Q 10 x x

How about this for a sequence?

1♠   2 (GF)
3♣   4♣ (RKC)
4   7♣

You can count 13 probable tricks: 5, 1, 1♠, 4♣ in hand and 2 ♠ ruffs in dummy.

Friday, October 5, 2007

A lucky/unlucky sacrifice

Here is a hand from the Manhattan. Let’s start by just giving my point of view.

♠ 7 3
K 8 7 6 5 2
K 9 6 5 4
♣ -

Hardly a stellar collection, but the bidding starts with a pass from partner and 4♣ on my right. Now, naturally I want to bid something, but is there any way to justify taking action? The game is matchpoint pairs, and they are vulnerable and we are not, so that all helps. With partner having passed, it seems reasonable to assume that they have at least game, and possibly slam. The pre-emptor is the most experienced player at the table, and his partner is the least experienced, so the opening bid probably indicates an 8-card suit, and the strongest hand at the table is certainly on my left. She is bound to raise, at least to 5♣ and maybe more.

Clearly, you have a simple decision: sacrifice or not. If you are going to sacrifice, you start to speak now, and if you aren’t, you keep quiet and hope they pick the wrong level. I haven’t had a good sacrifice in a while, so 4. This produces a look of consternation on my left, followed eventually by 5♣. That sounds like the wrong bid, so maybe I will pass and leave them in it. Double says partner, swiftly passed on my right. Well that won’t do, so 5, double on my left, passed out. The K♠ appears as the opening lead.

♠ J 9 5 4 2
9
J 8 7 2
♣ A Q 9
 
   [ ]

♠ 7 3
K 8 7 6 5 2
K 9 6 5 4
♣ -

I was hoping that the opening lead would be more useful, but dummy is excellent. After inspecting the dummy, West switches to the 3♣. That’s more like it – up with the Ace, discarding my other spade. If there are 8 or 9 clubs on my right, together with a couple of spades and at least 1 trump, there can’t be many hearts. But my only plan is to ruff hearts, hoping to make as many trump tricks as I can, and maybe even establishing the suit. So I pass the 9 - 4, 2, 10. East is back on lead and not too happy about it, which means she is looking at the AQ and maybe the 10 and 3 also. So she goes back to spades, and I ruff the A♠ as West follows with the 10♠. Now the 5 - 3, 2, 3. Blast the man. Only one heart, and he makes a trick with the 3. But now I get to ruff a spade and a club in hand and a couple of hearts on the table (West discarding clubs on the hearts, spades were 3-3).

When the smoke clears, I’m down 4 for -800. Could have been worse, and in fact it scores exactly average. Half the field stopped in game, but a few people tried 4 as a sac, and went for 1100 or 1400, and some others bid slam (6♣ and 6NT both make, of course). And I am left regretting that ruff with the 2. I was able to ruff 3 times in dummy, and if I had used the 7, 8 and J I would have escaped for -500 and a complete top, beating even the people who stopped in game.

Which of course poses a question: how do you stop in game with 20 points (seven controls) and doubleton support opposite an 8-card suit?

                 ♠ J 9 5 4 2
                  9
                  J 8 7 2
♠ A K 8          ♣ A Q 9           ♠ Q 10 6
A Q J 10 3                        4
A Q 10          [ ]              3
♣ 3 2                              ♣ K J 10 8 7 6 5 4
                 ♠ 7 3
                  K 8 7 6 5 2
                  K 9 6 5 4
                 ♣ -

Tuesday, October 2, 2007

The weak no trump

The strong 1NT opening is commonplace throughout US bridge, at all levels of competition, and regardless of system. However, for the average player, using a 12-14 opening NT presents no problems, and is generally a significantly better choice both in terms of causing problems for the opposition and in terms of simplifying the constructive bidding challenge. There are some differences of approach that you must adopt, if you are switching to a weak NT for the first time, and there are certain dangers that you must pay attention to. But if you are afraid of being doubled for a large penalty, rest assured that this rarely happens if you take reasonable precautions.

Generally speaking, the same systems of responses that you use with a strong NT will work just as well (or better) with a weak NT. Opener is about a king weaker than you are used to, so to take forward-going action, responder must be about a king stronger than you are used to. As a rule of thumb, responder should not be considering game unless he has at least 11 hcp. Most hands in the 6-10 point range will be passing, or seeking to settle on a suit part-score if that seems preferable to 1NT. With about 11 hcp, responder can consider his hand invitational, and with 12+ he should drive to game. This may seem a little aggressive, since I am suggesting that a balanced 12 opposite 12 constitutes material for game. But it is a fact that the hands generally play better when the strength is split. If you review some of your results, I’m sure you will find that 12/12 was easier to play, and more successful, than (for example) hands where the split was 15/9 or 18/6.

With less than about 6 points, responder should consider taking evasive action even before anyone starts doubling. It is a fact that many opponents who would double 1NT in a flash are much less certain when faced with a Stayman or transfer sequence. And rightly so, I should add, since the vagaries of distribution may nullify some of their high-card strength. And of course, if the auction should start 1NT – double, responder needs to be able to escape to a suit part-score when that seems advisable.

A fairly basic responding scheme that I would suggest would be based around 4-suit transfers. That's an old idea that is familiar to many (most?) people, and it still works pretty well. For example:
2♣ - Stayman
2♦ - transfer to ♥
2♥ - transfer to ♠
2♠ - transfer to ♣
2NT – transfer to ♦
3♣ - 5-5 in the minors, weakly invitational
3♦ - 5-5 in the minors, game forcing
3♥ - 5-5 in the majors, invitational
3♠ - 5-5 in the majors, game forcing
3NT – to play
4♣ - Gerber
4♦, 4♥ - Texas transfers to 4♥, 4♠ respectively
4NT – quantitative, inviting 6NT
5♣, 5♦ - to play
5NT - quantitative, forcing to 6NT, inviting 7NT

I'll expand (expound?) on some sequences in a later post.

Why do I say that the weak NT helps constructive bidding? Well, the argument revolves around balanced hands in the 12-17 range. Some of these hands you will open 1NT. Some you will open 1m and rebid 1NT, and the opened suit may be less than stellar. Playing a strong NT, I have had sessions where I seemed to be opening 1♣ on 10xx all evening. If you play a weak NT, this doesn't happen so often, for the simple reason that 12-14 point hands are much more common than 15-17 point hands. And the dynamics of the auctions for the 12-14 point hands is completely different. If you open 12-14 point hands 1m, the opponents don't know that you are weak, and you and your partner have scope to find a playable spot at the 1 or 2 level if he is weak also. The result is that you are safe from penalty doubles, except for really pathological distributions. But also, the opponents have plenty of scope to get involved in the auction, and since the opener is unlimited, responder may feel the need to get involved on fairly poor values while having little idea of either your strength or distribution. If you open these hands 1NT, you get the hand off your chest at once. Responder knows immediately whether you are looking at part-score, game or slam, or if the hand really belongs to the opponents. The response system means that you can identify suit fits as necessary and proceed with well-defined stop/invite/force sequences. The opponents meanwhile have been robbed of a whole bidding level, and while much is made of the possibility of them doubling for penalty, you shouldn't overlook the prospects of you doubling them. If they intervene at the 2 or 3 level and responder has 8-10 points and suitable distribution they may find they have “stepped in it”. And when you do open 1m, responder knows that either you have a real suit or you have 15+ points. Either case will make competitive auctions more tractable.