More USBC Slams

Thanks to BBO reportage the 120-board matches played in the USBC Trials give us a good sample of deals played at a high level of competence with which we can test our ideas and distinguish between faulty assumptions and brilliant insights. Recently commentators have got away from the idea that the team that gets the slams right wins the match. It seems competitive tactics have gained prominence, because there are more part score deals than slams. That affects how one goes about bidding a hand. However, as there are several competitive deals bid under uncertain circumstances, the law of averages tells us that in the long run gains and losses may balance out on the hands which can be played randomly by either side. This was the case in the 2011 USBC Final where 2 ‘young’ teams with similar tactics duked it out.

There was a contrast of styles in evidence in the semi-final in which the veteran, value-driven Welland team (Welland – Bramley, Schermer– Chambers) faced an active Diamond team that featured 2 Precision partnerships. The swings on 6 hands on which a slam was bid accounted for 48 of the 50 IMP winning margin, Welland gaining on just one hand. This strikes me as odd, as one would expect solid citizens who always have their bids to have the advantage over those who by design open on garbage. What happened?

In the previous blog we introduced the idea of an information filter. Bidding in a natural setting is a ‘broad band’ filter that is designed to express general values within the context of the individual’s hand. Bids are selected in a setting in which the general nature of the hand is taken into account. Usually good minor suits take a back seat in the bidding schemes geared to finding major suit fits. This makes sense on the basis of the probable outcome, but minor suits slams, being improbable a priori and may be missed on the rare occasions when they do arise. Bidding according to the Law of Total Tricks, a statistical rule that reflects normal conditions, may prove faulty when unusual distributions are in play. A slam exploration should provide specific information that separates exceptionally fortunate conditions from what is most probable.

Normally accurate slam bidding in a suit requires an approach in which the hands being bid are considered in a narrow sense with regard to a mutually agreed trump suit. The bidders must consider how their values fit into a particular requirement for 12 tricks for which tricks can be generated through judicious cross-ruffing. Asking bids and cuebids act as ‘narrow-band’ filters that pass exact information that is not subject to speculation based on probability. That’s the theory – let’s see how the results support this view.

When More is not Enough

		Bramley	Welland
♠ AKJ98	♠ T43	2♣	2♦*
♥ AQ	♥ K4	2♠	3♠
♦ A6	♦ 954	4NT	5♦ (Dbl)
♣ AK93	♣ Q8754	5♥ ?	6♥
		6♠	Pass

Here we see 2 balanced hands with wasted values in the heart suit. How much nicer it would be if the ♥K were the ♠Q, for then 6NT would largely depend on the clubs coming home. Bramley and Welland got it wrong when Welland bid 3♠ rather than signing off in 4♠, and Bramley, not ruling out a Grand Slam with 10 controls and stuffing in his long suit, launched an inquiry with RKCB. The double of 5♦ directed the killing lead. Welland did altogether too much holding a weak, balanced hand with weak support – yet another case of the tail wagging the dog. I think they are wrong those who argue that Bramley should have jumped to 6♠ over 3♠. Over a weak 4♠ raise, yes!

The Precision auction at the other table illustrates that a HCP evaluation worked better when there was no ruffing potential to add to the number of available tricks. The bidding was restrained, and the club suit was entirely lost – a potential disaster one might think.

		Platnick	Diamond
♠ AKJ98	♠ T43	1♣*(16+HCP)	1♦* (0-7 HCP)
♥ AQ	♥ K4	1♠ (F)	1NT (0-5 HCP)
♦ A6	♦ 954	3NT	Pass
♣ AK93	♣ Q8754

Basically all Platnick knew was that the partnership didn’t hold more than 30 HCP, less than the usual requirement for 12 tricks without ruffs. There is not much to recommend in the jump to 3NT other than it was to the point, but the final pass was good as declarer made just 10 tricks while gaining 12 IMPs.

Fortune Favors the Shapely

Next we look at a deal where vague description common in a natural auction, but made here by a Precision pair, worked in their favor by getting them to a slam missing 2 cashable losers. Less information resulted in more IMPs. In contrast to the previous deal, the 2 hands were balanced in HCP values (13 opposite 14 HCP) but one hand contained a known singleton in a side suit. The ‘blind’ opening lead proved critical.

		Greco	Hampson
♠ AKJ	♠ Q3	2♦(< 16 HCP)	2NT (asks shape)
♥ QT97	♥ AKJ864	3♠ (3=4=1=5)	4♠ (RKC in ♥)
♦ 5	♦ A86	5♣ (1 key card)	5NT (pick a slam )
♣ QJ984	♣ T2	6♥	Pass

The bidding makes a lot of sense until you look at Greco’s hand which he described as a maximum when he responded 3♠ instead of 3♣, the response that shows a minimum. His concept of a maximum greatly differs from mine. One important feature is trump solidarity, another, controls, especially aces. Bidding should be geared towards those features. With distributional hands there is less need to transmit a complete description of the partnership’s holding to guide an opponent to a killing lead. Given a tempo, inevitable losers may disappear. The opponents may not attack a side suit for which one has shown length, and with the AK missing the chances of the opening leader holding both top honors is low. That doesn’t mean one should count on split honors, but it does mean one shouldn’t be too concerned when the killing lead is found with little guidance.

In the previous blog we noted that in the final Greco-Hampson had a similar auction based on a Precision 2♦ opening bid in which Greco again promised a maximum he didn’t have. Hampson didn’t bid the making slam on that occasion and lost 12 IMPs. If he had remained consistent he would have gone on. Nonetheless, the evidence indicates this concept of ‘maximum or minimum’ is not working well. If one is asked to describe one’s hand as being ‘above average or below average’ the expectation is that half the time one would categorize one’s hand as being ‘above average’. Such a process doesn’t provide much in the way of distinction as it merely conforms to expectations. A better response structure would distinguish between ‘slam suitable and not.’ Such a distinction should include information concerning the number of controls held.

Welland – Bramley suffered a lack of resolve in a similar situation where one hand was known to be shapely, but information on the total number of controls was lacking.

		Welland	Bramley
♠ 5	♠ QJ742	1♣	1♠* (GF, asking)
♥ K62	♥ A74	2♥* (6 clubs)	2♠ (shape ask)
♦ AK4	♦ 8	3♠ (1=3=3=6)	4♣ (max or min?)
♣ K98754	♣ AJT2	4♦ (minimum)	5♣

Welland was able to give a complete description of his shape, but to him 5 controls represented a minimum. Recall that Greco with 3 controls thought he possessed a maximum under similar conditions. The answer to a question one asks should provide the information one needs. If Welland were able in response to 4♣ to disclose that he held 5 controls, a subjective description would be transformed into hard facts and Bramley could easily have bid 6♣ and tied the board. One sees that a vague, general description relating to the expected HCP total for an opening bid is not suitable when both hands feature shortage. Even if Bramley were 2-2 in the reds, slam would have been biddable if he knew the necessary controls were in place. I think with 5 controls one should encourage slam, whereas with 3 controls one should let it go.

Splintering with an Ace

Shortage in a side suit adds tricks when ruffs can turn losers into winners. That can be taken into account easily enough, but the bidding becomes more difficult when the shortage is a singleton ace or even a singleton king. Partner shouldn’t discount his high honors in that suit as the shortage is no longer a liability opposite a strong holding. That may be obvious, but look at what happened on a hand where at both tables a player in a 2/1 auction, according the BBO commentator, splintered with a singleton ace.

		Chambers	Schermer
♠ KQJ32	♠ 975	1♠	2♦
♥ A	♥ KQT8	3♥ (splinter)	3♠
♦ KQ876	♦ AJ94	4♥	4♠
♣ 63	♣ A4	Pass

Looking at the 2 hands one feels it should be easy to reach a contract of 6♦ once responder bids 2♦, whether that constitutes a game force or not. If the opening bidder could discover that responder held 5 controls, the ♣A, ♦A, and ♥K, he would surely be prepared to take his chances in 6♦. How can he go about obtaining the information without getting too high if conditions aren’t ideal?

Mashall Miles in his book ‘It’s Your Call’ reasoned through examples that one’s choice of bids should be directed towards reaching the most probable good result as one sees it at the time. This constitutes filtering information so as to steer towards a suitable conclusion. Those without shortage assume the world is flat, whereas those with shortage know it isn’t. Let’s speculate in this light after the start 1♠ – 2♦.

Chambers: This has become a slam exploration in diamonds.

Schermer: This is a 3-card game forcing raise in spades, and not a great one at that.

The major task at hand for Chambers was to convince partner that he liked diamonds a lot, and he used a splinter to 3♥ conveys that message. Schermer continued to think that a contract in spades was where the bidding was headed,

so he bid accordingly. With a flat hand he thought probabilistically in terms of the a priori odds, and a splinter on a singleton ace is extremely rare (therefore ill-advised).

So what is the solution? Opener should raise diamonds directly. If 2♦ is a game force, so much the better, as 3♦ can be the start of an informative sequence. Responder must resign himself to these facts of life: 1) if you can’t describe your hand with one bid, it may require several, 2) in a natural sequence ambiguities will arise, and 3) the more information exchanged, the better the position when making a final decision.

		Bob1	Bob2
♠ KQJ32	♠ 975	1♠	2♦
♥ A	♥ KQT8	3♦	3♥
♦ KQ876	♦ AJ94	3♠	4♣
♣ 63	♣ A4	4♥	4♠
		6♦	Pass

So, over a raise to 3♦, responder should bid a descriptive 3♥. Some might suggest 3♥ is a ‘torture bid’ as it leaves open too many options, but isn’t flexibility good? It is especially good when one holds a balanced hand with 5 controls and the bid coincidently saves space. Let’s see what Bob2 has in mind.

Bob2: I am not sure where we are headed. For the time being I’ll accept partner’s suggestion of diamonds as trumps and bid accordingly. Partner knows as well as I do that any game is preferable to a diamond game, so spades can wait while I describe my features. As my best suit is hearts, I’ll bid them next. I won’t be surprised to hear 3NT.

Bob1 bids 3♠ to indicate club weakness, therefore, strength elsewhere. The bid of 4♣ indicates interest in slam without being fully committed to it. With a stronger hand for spades responder could have launched into RKCB over 3♠, so a bid of 4♠

shows a willingness to stop in game. Bob1 has to look back over the entire sequence and resolve whatever ambiguities exist. Looking back and guessing with more information is better than looking forward and guessing with less. Responder’s

Goldilock’s approach was just right: too weak for slam, too strong for merely signing off in game. Bob1 bids what he thinks can be made, preferring the sure 9-card fit to the probable 8-card fit. 6♠ would have made as well as 6♦, but getting to the safer slam is always commendable at IMPs.

I don’t intend to suggest that natural bidding is the best approach, rather that it is the exchange of information that is the key. It is pseudo-scientific to use bids that have the appearance of science without actually providing the necessary information. Numbers are the basis of science, so scientific bidding is quantitative in nature rather than qualitative.

Let’s consider the Ogust 2NT asking bid. The primary objective is to determine whether the preemptor’s hand is game-suitable, so simple good-bad distinctions are sufficient for that task, but not for slam exploration. Compare this to the Precision 2♦ auction where opener responds 3♠ to a 2NT ask. As the partnership is already committed to game, it is now a question of whether the hand is suitable for slam. So a further 4♣ ask must be geared to slam, and the responses must be designed with that objective in mind. Similarly with the Welland-Bramley hand where 4♣ was the asking bid, so game was not in question. The response to 4♣ must be specific to the question at hand: slam or no slam?

On the last hand, after 1♠ – 2♦, the partnership is forced to game, so any questions must be related to slam-suitability. Who is going to be doing the asking? If a 3♥ bid is a normal splinter, a ‘telling’ bid, then responder should take the initiative. With the actual deal it would be better if 3♣ could be used as a slam-try relay to diamonds, asking responder to bid 3♦ so that opener can cuebid his aces up-the line. This allows for some flexibility on the part of an apologetic responder who could bid 3♠ or 3NT as a negative response to a slam try. A simple raise to 3♦ by opener would show a lesser, balanced hand without slam ambitions. It seems that we need some improvements over the present state in which good players feel they are obliged to make splinter bids on singleton aces.

So what if the bidding starts: 1♠ – 2♣? Please, one problem at a time.