Why lie?

He who recognizes that all men are born liars best knows all the Truth that needs concern him, because he will then know himself and have an advantage over others.

–  from The Handbook of Lies (1554) by Bartholomaeus Ingannevole

The questions we wish to address with regard to bridge are these: what constitutes lies and do they pay off in the long run? With regard to liars, St Augustine cast a very wide net indeed. Subsequent saints accused him of over-fishing, recognizing that some lies are like marriage, sometimes a necessary means of reaching a desirable end. Today the general public, attuned to moderation in all things, agrees that some lies are forgivable, yet the attitude of Bartholomaeus is still condemned on the basis of insufficient proof. For the purpose of this article, let’s consider it a working hypothesis.

Some cynics claim that language is the stuff that lies are made of. A bidding system constitutes a language, so of its very nature must engender a pack of lies.  However, to paraphrase Sir Francis Bacon, bids are the images of matter, and to fall in love with a bid is to fall in love with a picture. So we mustn’t put too much credence in what a bid purports. It’s our own fault if we do. Perhaps Bacon’s imaginary was inspired by a famous historical lie, Holbein’s flattering portrait of Anne of Cleves that was brought to Henry VIII then considering re-marriage. In the context of bidding, a lie is a call that grossly misrepresents reality and is intended as a push in the wrong direction.

Misrepresentation is relative to a predefined structure limited by the HCP content and the distribution of cards. The more vague the definitions the less information the bids contain. Various hands can be included within the definition of a call, and the greater the variety the greater the uncertainty and the less the information content. The rare exception does little to change the information content, as the players involved will assume initially that the most likely hand is the one being represented, a view in which they, like Henry VIII, may be sadly mistaken. A lie occurs as an outlier far beyond the expected limits of variation. Is there a possibility of defining lies in terms of statistical variation? No. Even a miracle can be judged to be within finite mathematical bounds of probability. What about intent? Hard to judge. Most would conclude that although we can’t give an exact definition of a lie, we can recognize one when we see one.

Here is a simple example from my experience to illustrate the difficulty. A weak 2♠ bid is defined as showing a 6-card spade suit within the range of 6-10 HCP. Would you say this hand falls within the definition: ♠AJ9765 ♥4 ♦JT986 ♣9? Many would agree 2♠ is a legitimate bid in third seat but would condemn it in second position on the grounds that partner may hold a good hand and that 6♦ may be the best contract. This argument is based on the principle of self-interest. Next suppose that the RHO has spent some time considering her hand before passing in first seat. Well, that is as good as an opening bid in my book, and the possibility of game or slam in diamonds has diminished to the vanishing point. So I make a lead-directing preempt of 2♠, and it works. The information content of my 2♠ bid has been little affected because it is still much more likely that my shape is 6-3-2-2 or 6-3-3-1 than it is 6-5-1-1, however, my partner and the opponents are much more likely to assume the former cases than the latter one. I do not consider this bid to be a lie under the inadequate definition, but the risks were judged to be greater for the opposition than for my partner, and my intent was to take advantage of the uncertainty created.

Bridge World Views Quite often problems arise where there is no bid that quite conveys what one wishes to tell partner. The Master Solvers Club thrives on situations where there are at least 3 viable alternatives to a bidding problem. This tells us immediately that we are close to a condition of maximum uncertainty where 3 bids can be chosen with equal probability. It is a consequence of the inadequacy of the communication system. There is no real solution to such problems, even long-time partners often disagree, so the MSC may continue on for years to come without resolution.

For Problem B of the March 2009 issue the top score was awarded to an overcall of 1NT with a hand that contained a singleton. BBO commentators have informed us that bidding 1NT with a singleton is common practice in China. Does that mean most Chinese players are liars? No. Lies through custom lose the name of falsehood. The experts’ answers to Problem B tell us that the definition of a 1NT overcall is being altered in America, because, presumably, it works. For a while non-experts will be bamboozled by professionals promoting the interests of their clients until the dupes catch on and try the same thing at the local club. Eventually it may become a part of ‘Standard American.’

Here is Problem E from the same issue.  Matchpoints, None Vulnerable:

1♦   (1♥)    ???    You hold:  ♠A86 ♥942 ♦T93 ♣AQT2.     Your bid?

Here we are faced with a situation where we might bid 1NT without a stopper in the overcaller’s suit. Playing Precision where 1♦ is nebulous and limited to at most 15 HCP, I just might try that and term it a ‘pre-balance NT’. Today so many overcalls are made on worthless suits that it is not guaranteed that I’ll get a heart lead. Also, as we saw in the recent Vanderbilt Cup, ♥942 can act as a stopper opposite ♥QJ tight, nevertheless I would classify 1NT without a stopper as a lie, but not a lie that experts yet condone.

Only 2 experts could bring themselves to pass, because the modern game demands immediate action. This meant the other Master Solvers had to ‘lie’, as no bid was pre-defined in such a way as allow this hand to be included. Of course, because they bid, their definitions changed through usage. In this case the biggest change was to the meaning of double, which previously promised 4 spades. The lesser fib was a bid of 2♣ which would now no longer promises 5. Because the definitions were changed by their actions, their bids can no longer be classified as lies, but as established agreements. The hand certainly looks like one best suited for 1NT, stopper or no stopper.  As noted by Anders Wirgren, a Swede, there is a classical solution that solves Problem E in a pristine manner: a double denies as many as 4 spades. To avoid lying should we redefine our doubles?

Players continue to play systems in which they are forced to go against prior agreements. Most bidding systems, like Bridge World Standard are ‘kludges’, consensual patchworks that have evolved to combat changes in practices as they occur. It is dishonest to continue to maintain we play the old definitions designed in gentler times when we don’t. It is hypocritical to interpret one’s own agreements freely, then to demand that the opposition have unassailable agreements in rare situations and to appeal an unfortunate result on the grounds of being misinformed. Fans who watched the Vanderbilt Cup on BBO know to whom I refer. It is permissible to be inexact when there is no specific agreement, as inexactitude is the normal condition of the game. An admission of uncertainty does not constitute a lie, in fact, it may be the most truthful description in many situations. We should never classify reasonable departures from expectations as ‘lies’, but as a normal part of the game. The convention card presents only a vague outline with questionable limits, but the idealistic proposition that an opponent should know everything a long-term partner knows is unsound. Suppose my partner were to inform an opponent, ‘Bob sometimes bids 2♠ with 6-5-1-1 shape.’ That would be a true but misleading statement because it is most unlikely to happen again.

Is Honesty the Best Policy in Bridge?

All that one gains from falsehood is not to be believed when speaking the truth

– Aristotle (384BC – 322BC)

Aristotle put his finger on the pulse of poker where bluffing is an essential element of the winning strategy. Of course, lying is pleasurable in and of itself, otherwise, I assume, it would not have merited the status of a sin and poker would not be as popular as it has become. Poker is essentially a psychological game of percentages where the percentages include not just the probability of the card distributions but also the probability that an opponent is bluffing or can be bluffed.

Poker analysis is reduced to card combinations when an opponent invariably bets on his good hands and folds on his bad hands. The information gathered from a predictable opponent influences the probability of the card distributions, and so makes it easy for an expert to prevail on that basis alone. Players who go strictly by the book are easy victims of those who can translate their actions correctly. When only experts are involved, masters of the odds and good bluffers to boot, psychology comes to the fore, and it becomes a clash of wills. One has to create an atmosphere in which yours is the dominant personality. You set the agenda by bluffing. Not only that, but it is to your advantage to be caught bluffing early to show you are not afraid. To fold with a good hand is equally deceptive, but it is very bad psychology. So you lie when you hold nothing of value.

The strategy of bluffing applies to bridge insofar as bridge can be considered in the context of psychology. It appears that some leading players have adopted the idea that lying is beneficial in the long run. That has always been a minor element in bridge, usually restricted to bids in third seat on the grounds that passing partner is less likely to act on an optimistic misrepresentation of one’s assets. The Drury Convention testifies to that practice. The difference today is that experts will bluff in first and second seat before partner has had a chance of express his own values.

Does bluffing (lying) pay off at bridge? It is a hard question to answer. The bluff and its resulting loss are sure to be revealed, so that is not of concern to the practitioner. He expects to make up for the loss in many following hands where he tells the truth but the opponents don’t believe him. For example, this may affect an opening lead against a close game. So, the current loss is obvious, but the future gains are hidden.

The trouble with applying poker strategy to bridge is that it is not expensive for an opponent to call the bluff – he simply bids 3NT. That is a good bet as only seldom will he miss a slam, and the cost of going down may not be great. The upside of 3NT is that it may be slated for defeat but actually makes on poor defence based on the misinformation provided by the bluff bid. It may even be doubled and making when partner takes the bluffer seriously or when declarer is able to read the distribution. We encounter this phenomenon frequently on BBO. Here is an example from the 2009 Vanderbilt semi-finals.

Dealer: South Vul: None		North – Moss
		♠	AKJT32
		♥	K7
		♦	T6
		♣	762
West – Zia				East – Hamman
♠	Q4			♠	875
♥	T2			♥	AQ9
♦	742			♦	K983
♣	QJT543			♣	AK9
		South – Gitelman
		♠	96
		♥	J86543
		♦	AQJ5
		♣	8

West	North	East	South
			2♥
Pass	3♥	3NT	All Pass

This was Board 11 of a 64-board match, early enough to try a bluff and reap the rewards immediately, always a hope, or make up for the loss on subsequent sessions. A lawyer might argue that Gitelman’s 2♥ bid is within the letter of the law, but to me it is a gross misrepresentation because the HCP are concentrated in diamonds and the hearts are weak. I conclude that Gitleman may have been influenced by co-residents of Las Vegas, the poker capital of the world, and that deception was the obvious intent.

Zia had nothing to contribute to the auction and Moss gave a typical unsound raise, leaving Bob Hamman with a tough decision. He followed the rule named after himself, namely, when in doubt bid 3NT, thus reaching a poor contract that could be set by the taking of at least 8 tricks off the top. But something happened on the way to the bank. Gitelman led an attitude ♥8, but it was a bit late to show a dislike of hearts. Now it was Hamman who had 8 tricks off the top, and in the time honored manner he ran the long suit in an attempt to pressure the opponents into giving him a 9th on a misreading of the cards. It worked! On the last club Gitelman discarded his life line to Moss’ spades and was endplayed into giving up a trick in either diamonds or hearts. In the confusion signaling again proved to be a weakness in the expert’s game.

What would be the effect of a pass in the South seat? We know the answer, because in the other room Nickell passed and Freeman bid 2♠, making 3, for +140 NS. So Gitleman’s brilliant bid was slated to pick up 3 IMPs; instead it lost 11 IMPs. That represents good odds in favor of Hamman’s Rule. Furthermore, it reduced the deficit from 21 IMPs to 11 IMPs giving new life to the Nickell team – to no avail as it happened.

Oh What a Tangled Web!

The Nickell team is justly famous for its tremendous late-stage comebacks. (Perhaps one should consider more carefully how they get so far behind in the first place.) Meckstroth was in a desperate situation by the time Board 31 was played in the Vanderbilt semi-finals with his team down by 47 IMPs. The best lies are those that have some element of truth. In bridge that translates to bidding good suits especially on bad hands. The main advantage is that when one defends it won’t cost to lead the suit from either side of the table. When one opens light on a garbage suit, the consequential lowering of expectations on the number of total tricks due to poor trump quality makes it all the safer for the opponents to stop in 3NT …. and then to make it, because, unlike poker, there is always a partner there to mess it up.

Dealer: Vul:		North – Meckstroth
		♠	6
		♥	K86532
		♦	9432
		♣	AJ
West – Greco				East – Hampson
♠	42			♠	AKQ95
♥	QJ74			♥	AT
♦	KQJT8			♦	5
♣	63			♣	Q9752
		South – Rodwell
		♠	JT873
		♥	9
		♦	A76
		♣	KT84

West	North	East	South
			Pass
Pass	1♥	2♥*	Pass
2♠	Pass	3♣	Pass
3NT	Pass	Pass	Dbl

Meckstroth’s 1♥ opening bid was deceptive except with regard to distribution. Once Hampson was able to show a strong hand, Greco applied Hamman’s Rule, and it was up to Meckwell to defeat a 3NT contract that wouldn’t be bid in the other room. It was a bit late to get the necessary close co-operation that was not part of the opening bid strategy.

Meckstroth got off to a good spade lead. Greco set about establishing 4 diamond tricks in the hidden hand without having an entry to cash them all. On winning the ♦A Rodwell returned the top spade from ♠JT8 towards dummy’s ♠Q95. Greco cleverly ducked and Rodwell had to find the right switch. Those seeing all 4 hands saw the winning defence was to lead a club to Meckstroth’s ace, win the club return with the ♣K and lock declarer in dummy.  The setting trick would come either from a black suit winner in the South hand or from Meckstroth’s ♥K which would at long last prove its worth by snatching victory from the jaws of defeat. Unfortunately, Justice was served when Rodwell returned a heart giving Greco 5 undeserved tricks in the red suits. This nail-in-the-coffin cost 12 IMPs as the Nickell teammates, EW at the other table, played undisturbed in 2♦ going down 1.

Once again defensive signaling proved inadequate. Meckstroth had the opportunity of making 2 discards with which to make the Ultimate Signal, which is, ‘forget what I told you before’, much needed in current state of bridge bidding.  His discards, upside-down attitude, count, and suit preference, were the ♥5 and ♥2. The clear message of ‘clubs, not hearts’ was not emphatic enough to avoid the losing play. Obviously, if Meckstroth’s hearts had been as good as ♥KJ8652, there would have been no bad play for Rodwell.

Finally we look at Board 4 in the Vanderbilt semifinals, a deal that illustrates a real gamble at bridge, the redoubling of a contract which may not make. Like a raise in poker, the potential for a gain is substantially increased, as real pressure is placed on the opposition to make a costly move. Is it surprising that Zia was the perpetrator?

Dealer: Vul:		North – Moss
		♠	Q92
		♥	Q97542
		♦	—
		♣	QJ95
West – Zia				East – Hamman
♠	KJT5			♠	A63
♥	83			♥	A
♦	K876432			♦	Q95
♣	—			♣	KT8732
		South – Gitelman
		♠	874
		♥	KJT6
		♦	AJT
		♣	A64

West	North	East	South
Pass	Pass	1♣	Pass
1♦	2♥	Pass	3♥
4♦	Pass	5♦	Dbl
Rdbl	All Pass

Gitelman followed normal practice by not overcalling 1♥ over Hamman’s natural 1♣. On the next round he gave a gentle raise of partner’s weak jump overcall despite his prime support: 2 outside aces and 3 heart honors.  Some might admire his restraint at this vulnerability. Imagine his surprise when given room Hamman raised Zia to game at the 5-level. This seemed too good to be true, so he doubled with good prospects for 3 tricks outside the heart suit. So far the auction had taken a normal course. In the other semifinal match both teams played in 5♦* (Versace and Katz being the successful declarers), but here Zia added to the excitement by redoubling. If Greco and Hampson at the other table had also reached 5♦*, Zia’s redouble would gain 6 IMPs if the contract made, and lose 5 IMPs if it went down 1 at both tables. If NS had escaped to 5♥, the upside for Zia would be considerable, but Gitelman had no inclination to take what would be an expensive phantom sacrifice. He probably felt his plan had worked. It didn’t.

The action at the other table could be taken as an illustration of where previous bluffs pay off. The fear engendered without evidence is that one is being talked out of something. It is difficult to measure the effectiveness of previous deceptions, because even the victim may not be aware of the full extent of the influence on his decision.

Greco	Freeman	Hampson	Nickell
Pass	Pass	2♣*	Pass
Pass	2♥	Dbl	3♣
3♥	Pass	3♠	Pass
4♠	All Pass

The auction began poorly for the Precision pair in the EW seats. 2♣ was natural with 6+clubs, 11-15 HCP. Precision players may open with a long club suit and not much else, as Sontag did to his regret in an earlier round, and Meckstrorth did later in this match. On this hand it looks wrong to balance in the North seat. One shouldn’t suspect skullduggery at this vulnerability. South holds an opening bid, but he took no action even though his clubs must be weak. West might have given a weak raise to 3♣ on 3 poor clubs, so it doesn’t appear the opposition has a fit. No controls and QJ95 in the opposition’s only guaranteed long suit are bad omens. Who has the diamonds? Ah, there’s the clue, as West may have them and couldn’t bid because the Precision system doesn’t provide a natural nonforcing diamond bid. These considerations scream ‘PASS’, but what if one suspects a bluff? What are the chances of missing game?

Many feel that in today’s game one has to get into the bidding on what previously would have been considered inappropriate values. We have discussed 2 examples from the MSC. The most convenient change involves the double, and some would suggest that Nickell should have doubled on his hand in the Italian style. It would have been counter-productive here, as it gives West a chance to enter the auction freely. Whether one is sympathetic or not, bidding 2♥ represented an expensive failure to take advantage of a weakness in the opponents’ bidding system. All was not lost as Greco and Hampson still couldn’t find their diamond fit, but ended up playing in a 4-3 fit in 4♠, a contract that could be defeated. So maybe bidding 2♥ was not so bad after all. Nickell led a heart to Hampson’s ♥A. The ♦9 followed and it was up to Nickell to cover with a low honor. Instead he went up with the ♦A and gave Freeman a ruff, removing his nuisance trump from ♠Qxx. Understandable, but the real damage had been done before, perhaps long before.

Afterthought:  Beauty and Bridge

Accuracy is essential to beauty. – Ralph Waldo Emerson (1803-1882)

I do not know what the philosopher was thinking of when he wrote the statement above – perhaps of the fancy clock on his mantelpiece that couldn’t keep time, or of a coloratura’s faulty rendition of the Queen of the Night aria still ringing in his ears from the night before, or maybe of a shot curving past a goalkeeper’s outstretched arm only to deflect off the left-hand goal post.  Whatever the case, a bidding sequence can be judged beautiful if and only if it accurately reaches the optimum contract. It is most annoying when a beautiful sequence of bids results in failure due to an ‘unfair’ distribution of the cards. Accuracy is a higher form of neatness, a characteristic we all admire in others, but it represents a cool kind of perfection that appeals to the part of the brain capable of reasoning, a part that has a depressingly small role to play in our everyday decision making.

The realm of top-flight bridge is a small one with the same few players competing against each other repeatedly for the highest prizes after the vast majority of lesser talents have fallen by the wayside. Superior cognitive powers are enough to get one to the territorial boundary, but once admitted to the tribe, psychology comes to play an increasingly important role as each player strives to achieve or maintain a dominant position. Applying pressure through bluffs becomes common practice against other experts. Logically, this shouldn’t work.

Some idealists, and many commentators fall within the category, wish the game would provide a refuge within which what might be termed ‘the higher functions’ can freely roam. They treat a deal primarily as a puzzle capable of being solved by an application of logic. They write books embodying the concept that there is a right contract and a right way of playing the hand. They would be happy if the bidding process was strictly regulated so that the information provided would provide a sound and familiar basis for reasoning from the known to the unknown. To those idealists I say, control your rancor, embrace uncertainty. The closer the world of bridge reflects the real world, the more amusing it becomes.

(I highly recommend a new book, How We Decide by Jonah Lehrer (2009), a terrific read. He has a section on poker, but none on bridge. Don’t wait for the paperback edition.)

Nothing is beautiful viewed from every aspect. – Horace (65BC – 8BC)