Bid Outside the Box. Jan Eric Larsson

Transcription

1 Bid Outside the Box Jan Eric Larsson Today s bidding systems have grown from the early days of contract bridge. Most of the time, you try to exchange information about high cards and distribution while naming plausible final contracts. West East 1h 1 1s 2 2s 3 4s 4 1) I think I can make one heart. 2) I have spades. 3) If so we can make two spades. 4) Then we can make four spades. Of course, many conventions and artificial bids have been invented, but the basis remains. Most bids show extra length or strength and name a potential final contract. In the sequence above, 1s is the only forcing bid. What would be the result if we tried first to formulate general principles for what a bidding system should accomplish, and then designed the system from scratch, based on those principles? This question has been asked many times before. Here are my suggestions. Bidding Principles What do we want a bidding system to be like? Here are some simple criteria: Precise in constructive bidding. Aggressive in competition. Easy to remember. This is probably so obvious that no one will object. But let us look at these criteria at work in different types of situations. The situations I have in mind are: When our side is stronger, (and the deal is reasonably balanced). When our side is weaker. When the deal is fairly equal in strength. When there are wilder distributions. Let us look at each of these situations and see what first principles we would wish for in a bidding system. Our Side is Stronger First assume that our side holds most of the high card strength, and that the distributions are reasonably balanced, (or at least not extreme). In this case we want to conduct precise and constructive bidding to find the correct part score, game, or slam. At the same time, we do not want to give away more information than necessary to the defense. I propose the following principles, (nothing new here). Keep Low with Strong Hands In constructive bidding, we need to preserve bidding space and keep the level low. Each extra bid gives us twice as many sequences to use in the search for the best contract. The Weaker Hand Should Tell When a stronger and a weaker hand meet, the weaker hand should describe itself to the stronger one. It is easier to tell about fewer features, and it is easier for the one who sees most of the strength to make good decisions. The Stronger Hand Should Declare It is also good if the stronger hand becomes declarer. The lead will come towards the hand with more high cards, and it is more difficult for the defense if less of the strength is visible in dummy. If the weaker hand has described itself in the bidding, it is also better if it ends up as dummy. Analysis Natural systems are only half good according to the above principles. One 1 (10)

2 advantage in Standard American and similar systems is that an opening bid promises a fairly strong hand, and that this hand immediately grabs a potential trump suit. We want the stronger hand to declare. Five-card majors may have a small advantage over four-card majors, since the bidding starts lower with some balanced hands. But there are also several weaknesses. First, we open too high with many strong hands. For example, the opening bid 2N is bad. With points, the bidding is already at the three level, and the strong hand is describing itself. The same goes for a forcing 2c, especially in the sequence 2c 2d; 2N. There are more examples of bad sequences in the second round, for example, 1c 1d; 2N. A jump in a suit also tends to violate the principle of saving space, for example, 1h 1s; 3h and we have preempted ourselves. Recently, the so-called 5542 system has become popular, especially in Europe. In this system, strong balanced hands, (17-19), are opened 1c, and after that responder uses transfers on the one level. In this way, you can avoid raising the level with a strong balanced hand. Another advantage with this method is the use of transfers after 1c, which increases the chances that the stronger hand becomes declarer. In my opinion, Culbertson s most brilliant invention was to make 1-over-1 and 2-over- 1 forcing. At the same time, these bids also are a major weakness in most natural systems. If opener is strong (16-20) and responder is weak (6-10), it would be much better if responder described his hand and opener only listened, and if responder could avoid naming good suits first. For example: s A-Q-5-3 s K h K-8 h 7-5 d K-Q-7 d A-2 c A c In a natural system, the bidding could go: West 1c 3s East 1s 4s East grabs the spades and the contract may go down with a heart lead. In 5542, the bidding could go: West East 1c 1h 1 3s 4s 1) Transfer to spades. West becomes declarer and the contract is clearly better. In a strong club system, the bidding may go: West East 1c 1 1d 2 1N 3 2h 4 2s 3c 4s pass 1) 16+ 2) 0-7 3) balanced 4) transfer If we could manage to build a working system around the idea, it seems that a weak responder should use transfers, or bid shorter suits first. An interesting idea would be a structure where the responses mean either 6-12 and a transfer, or 13+ and a natural suit. 2 (10)

3 One possible solution is to use circular transfer responses at the one level. It works as follows: 1c 1d 6+ four or more hearts 1h 6+ four or more spades 1s 6+ balanced 1N 6+ four or more diamonds 1d 1h 6+ four or more spades 1s 6+ balanced or clubs 1N 6+ four or more hearts 1h 1s 6+ balanced or minor 1N 6+ four or more spades 1s 1N 6-10 negative The subsequent bidding is the same as in a natural system, but after opening 1c or 1d, opener shows a minimal balanced hand by completing the transfer, while bidding 1N shows balanced. Using these responses, responder avoids grabbing his own four-card suit as trump, and does not immediately wrong-side a contract where opener may be strong and responder very weak. Another idea along the same lines is to use combined weak and strong jump shifts, where the strong version is natural and the weak is a transfer. For example: West 1h East 2s Here, East shows either 12+ and a good sixcard spade suit, or 4-10 and a long club suit. Opener will complete the transfer with a minimum opening had. Strong 1c For strong hands, I am quite convinced that Precision and other strong club systems are superior. The opener starts with 1c and responder describes himself. If you use the Greek alphabet asking bids, opener asks (and grabs trump suits), while responder describes his hand (with mostly artificial step responses). Precision has one weakness, the natural suit responses, but that is easy to fix, by using positive transfer responses: 1c 1d 0-7 1h 8+ five or more spades 1s 8+ no five-card suit 1N 8+ five or more clubs 2c 8+ five or more diamonds 2d 8+ five or more hearts After a positive response, opener s relay is a Beta asking bid, that is, asking for controls. In our own strong club system, the Cottontail Club, responder uses transfers in a circle between hearts, spades, and balanced hands after 1c 1d. If opener bids hearts or spades, it is taken out of the transfer circle. For example: West East 1c 1d 1h? Now, 1s by responder shows a balanced hand, while 1N shows spades. We have also switched the positive transfer responses so that 1N shows hearts and 2d shows clubs. In this way, responder avoids grabbing notrump when holding a club suit. A Polish Idea However, 1c is not the lowest opening bid. We forgot pass! The pass gives us twice as many sequences up to a certain level, when compared to 1c. In the eighties, a number of strong pass systems proliferated, and many of them came from Poland. The theoretically most interesting Polish weak opening system is Delta. The opening bids are as follows: 3 (10)

4 pass 13+ all distributions 1c 8-12 no shortness 1d 0-7 all distributions 1h 8-12 short h 1s 8-12 short s 1N 8-12 short d and five in a major 2c 8-12 short d and no major 2d 8-12 short c and no major 2h 8-12 short c, 5+ hearts 2s 8-12 short c, 5+ spades 2N 8-12 short c, 55+ in majors Short means singleton or void. After an initial pass, responder uses the same method, but with the point limit 6-10 or 11+ instead. We will not look at the continuations, but just note that there is a relay system where the stronger hand bids the nearest bid and the weaker hand describes itself, using artificial methods. The two creators of the Polish weak opening systems, Lukasz Slawinski and Stanislaw Ruminski, have a clear philosophy, where pass is used as the strong opening bid, because it is the lowest bid, and a weak responder describes his hand by bidding short suits, so as not to grab trump suits. Likewise, a weak opener bids short suits. Delta really follows our stated principles. Tips for the Defense Against strong club and pass, it is a good idea to overcall on weak hands, to remove bidding space. But since responder gets extra bids (pass and double), you have to get in with at least 1s versus a strong club and 1d versus a strong pass. In my experience, it is good to overcall with two of a minor. It is low enough that the opponents often prefer to bid, rather than look for a penalty double, but high enough to steal valuable bidding space. A Drawback with Artificial Methods There is a drawback with the methods suggested so far, namely that they give the opponents a chance to double the artificial bids to show a suit or recommend a lead. This is, without doubt, a negative effect that impacts all conventional bids. My opinion is that the advantages of having the strong hand as declarer and the weak, sold hand as dummy, outweighs this drawback. As a counter, you may decide that partner s pass (or redouble) invites our side to play in the doubled denomination, while other bids continue the constructive exchange. Our Side is Weaker Let us now look at the case when our side is the weaker one. Here, the objectives are completely different. The two things we want to achieve are to find a fit and bid this as high as possible, and to preempt the opponents in general. Again, we do not want to give away more information than we have to. I propose the following criteria. Open as Often as Possible It is a clear advantage to be the first side to tell something about our distribution, in order to find a potential fit. East dealer, East-West vulnerable s Q h J-9 d J-4 c J s s J-9 h A-K-2 h Q d K d A-Q-9-5 c A-7-4 c 9-8 s A-K-10-5 h d c K-Q-5 West North East South pass 1s pass 2s pass pass pass South opens a four-card spade. West s distribution is not ideal for a take-out double, and partner has passed, so West 4 (10)

5 passes. East is too weak to enter at the three level. West leads the ace of hearts and East- West wins two hearts, two diamonds, and the ace of clubs. North-South writes 110 in the protocol, while East-West can make 4h. The example is from Hughes, Building a Bidding System, However, there are two drawbacks to opening aggressively. Often, the opponents are the stronger side and get the final contract. In these cases, the less you have told declarer about the distribution, the better. And if partner is the strong hand, you do not want to preempt him. Open High with Weak Hands Thus, the theory is that you should open high with weak hands, or not open at all. This is the idea behind weak two openings and higher-level preempts. Weak Openings Should be Non-Specific I think standard theory is wrong about what a weak opening bid should look like. It is usually recommended that a weak two opening should fulfill demands like: Narrow point interval, say 6-10 HCP. Six-card suit. At least one of A, K or Q. No four-card major on the side. The motivation for this is to make the weak two opening useful for constructive bidding, so that partner does not suffer, in case he is the strong hand at the table. In my opinion, this reasoning is all wrong. The only reason for opening a weak two is that we hope that the opponents have the highest contract. What we actually want is to open as high as we dare, and show as little useful information as possible. Here are my criteria for a weak two opening: Side suits are allowed. There is no reason that the opponents should know that a side suit breaks well. The same goes for higher preempts. Assume that you pick up, non-vulnerable: S Q-J-5-3 h 4 d Q-J c This is a good preempt. Open 3d. You have few enough points that our side will not have a playable contract in spades. The same principle says that the weaker an opening bid is, the less information it should give away. West dealer, North-South vulnerable s K h d c K-J s Q-J s A h J h Q-8 d K d c 6 c Q s h A-K d A-Q-J-10-9 c A-4-3 West North East South 1s pass 3s 4d pass 5d pass pass pass In the Spring Nationals 1992, Paul Soloway opened 1s as West. The lead was the six of hearts. Declarer, David Feldman, played spades to East s ace, and concluded that West must have either the king of diamonds or queen of clubs to open. He laid down the ace of diamonds and made the contract. The weaker the better. Any five-card suit will do, and maybe even a reasonable four. In my opinion, the main error is that East- West used super-weak opening bids in the majors on the one level. Note the difference if West had opened 2s showing 0-9 points 5 (10)

6 and any five cards in spades. Let us assume that the rest of the bidding would have been the same. Declarer has no particular reason to assume the king of diamonds or queen of clubs to be with West. Many systems open one of a five-card major with as little as eight points. In the Swedish Magic Diamond, you open one in a major on a four-card suit and 8-11 points. In the Polish strong pass system Regres, you open one of a major on 3-4 cards and 8-12 points. All these systems aim to be aggressive, but in the five-card major case, you have told the opponents that one of the suits does not belong to them. In Magic Diamond this is less clear, and in Regres, the opened suit may be the opponents best trump (!). Regres is better than Magic Diamond is better than super weak five-card majors. Transfer Preempts and Multi Now assume that we open a preempt, and that partner holds the strong hand. According to our proposed principles, shouldn t we describe our hand, since it is the weaker one, and use transfers? There are several conventions based on this idea, for example transfer preempts and 2d Multi. Again, I think that the reasoning is wrong. The entire idea behind a preemptive bid is that the opponents have the highest contract. If we want to bid something constructively, we should open low. If the opponents have no defense against transfer preempts, no damage is done by the transfer. They will be as preempted anyway. But assume they use a defense like this: X weak takeout or 18+ cuebid strong takeout The cuebid is a bid in the shown suit. Suddenly the opponents have the chance to distinguish between some important hand types. Most important is the weak takeout. You often cannot risk a bid after a natural preempt, but now you can use a probing double. Versus 2d Multi you can use the following defense: X balanced or 18+ 2h/s natural 2N balanced Against a natural weak 2s you can hardly venture a double with balanced. The 2s opening will also preempt hearts, which is not the case with 2d Multi. Assume that you hold the following hand, and your right-hand opponent opens. S A-10-2 h K-J-5-3 d Q c Q-6 If the opening bid is 2s, it is dangerous to enter the bidding. A-10 of spades may be two defensive tricks, while three diamonds or hearts can play out very badly. But if the opponents open 2d Multi, the above hand is ideal for a probing double. You risk almost nothing, and if two diamonds actually is the best contract for the opponents, they have no easy way of finding that out. Preemptive Bids Should Be Natural The conclusion is that preemptive bids should be natural. It does not matter that the weaker hand grabs the trump suit. We are not supposed to play anyway. If the opponents have a good defense against Multi and transfer preempts, we give them unnecessary opportunities to enter the bidding with medium hands. We can also note another thing. If you want to use transfer preempts or Multi, your hands should tend to be constructive. If you prefer really weak, undisciplined preempts (Bergen-style), they should be natural. 6 (10)

7 A current fashion in Sweden is to combine Multi with semi-constructive bids of 2h/s showing and a six-card suit. Here, it may be better to switch the meanings, so that 2d contains the hands, while the direct 2h/s are weak. It will be more difficult to handle the constructive bidding after 2d, but much more difficult for wellprepared opponents to handle the natural weak twos. Another modern idea is to let 2d Multi be either weak majors or strong minors. In this case, 2d may be with strong diamonds. This may be a way of battling the weak double. A pass from responder can suggest to play 2d doubled. Constructive or Preemptive? Which is the more important objective, to bid constructively to the optimal contract, or to try and prevent the opponents from finding their optimal contract? There are two obvious truths. You must be able to reach good contracts, and the practical answer to the question is usually decided by personal style. However, over 70 % of all boards played in modern bridge involve bids from both sides. According to hard statistics, competitive bidding is more important. In my opinion, the conclusion is obvious. You must think about preempting your opponents, and you can count on undisturbed bidding relatively seldom. Being aggressive will pay off, and staying low is less advantageous than it may seem. In today s modern bridge, you need to bid as much as you dare! Intermediate Opening Hands So far we have formulated some principles for strong hands, where probability says that our side is likely to have the highest contract, and for weak hands, where the opponents most probably should declare. Let us turn to the intermediate opening hands, around points. hand, we cannot yet guess what side will have the highest contract. Thus, it makes sense to start frequently and low, but also to be a bit unspecific, until we know who holds the balance of the points. In this interval, I think four-card majors are more efficient than five-card majors, since they will discover more trump fits, especially if responder raises on three-card support. On the other hand, five-card majors will gain on some 5-4 fits that can be raised to the three level immediately. Five-Card Majors When using five-card major openings, you immediately grab a potential trump suit, and you clearly tell the opponents that your suit is not for them. Thus, it is good to apply this to strong openings hands, from hcp and upwards. It is a bad idea to open fivecard majors on very low point ranges. Three-Card Minors If you open five-card majors, you need some way of handling weak, balanced hands, and the most popular method is three-card minors. With a strong hand, you open low, but you are more vulnerable to opponent intervention. South dealer, East-West vulnerable s 7-5 h A-Q-5 d Q-J c s A-10-4 s K-9-3 h K-J-8-3 h d A-K-6 d 7-5 c A-3-2 c K-J-8-6 s Q-J h 9-7 d c Q-7-5 It is not so difficult to see what we should do with such hands. Just by looking at our own 7 (10)

8 West North East South pass 1c 2d X pass 3N pass pass pass Jill Meyers opened 1c and Sabine Auken overcalled 2d. Randi Montin doubled to show strength and one or two majors. But which one? Meyers took her chances with 3N, which went down after a diamond lead. This example from Venice Cup 2001 in Paris shows how vulnerable three-card minors can be. If you have a weak hand, a three-card opening may steal the opponents trump, and if they get the final contract, you have not given away very much information. Thus, you can very well lower the opening strength demands with three-card minors. Three-Card Openings and Canapé In Standard American and Precision, you can sometimes steal the opponents best trump suit when opening one in a minor (one diamond in Precision). This is fine, but would it not be even better to steal the opponents major suit? The old idea of opening your next longest suit, canapé, is nowadays quite unusual, but it may deserve a better destiny than oblivion. If an opening bid shows 3-4 cards in the suit, partner can usually decide whether it is a candidate trump suit or not, but the opponents cannot deduce very much. The bid has a different information value for the different sides. A three-card suit has a higher chance of stealing the opponents trump, the weaker the opening bid is. In Cottontail, we open a 54xx hand with the four-card suit, if the four-card suit is a major, and with a strong four-card diamond and a weak five-card major. With a strong five-card major and a weak diamond, we open the major and rebid it. We also open a 5332 with a three-card suit, which is higher than the five-card suit, to block out the opponents suit. West dealer, no-one vulnerable s E h kn-6 d K-10-5 c D s D-10-4 s kn-6 h E h K-7-4 d kn-9 d D c K-5-4 c E s K h D-5-2 d E c kn-6 West North East South 1s pass 1N pass 2h pass pass pass The typical contract is North-South making two spades. East-West can make two hearts, but spades are higher. The canapé opening makes it more difficult for North-South to find their spade suit. Drawbacks of Canapé There is, of course, a drawback with opening the next longest suit. Sometimes we will miss a good fit that everyone else finds directly after their opening bid. On the other hand, we find some 4-4 and 3-5 major fits faster than the rest of the field. One popular treatment is to always open the major first, whether it is a 5-4 or 4-5. You will find more fits, but the drawback is that you will need another round to tell about your actual distribution. Unspecific Overcalls In Jannersten and Wohlin, Winning Pairs Technique, 1975, (in Swedish), the authors suggest one level overcalls on three-card suits, (especially spades), to find a fit, to 8 (10)

9 recommend a lead, and sometimes to steal the opponents trump. S K-J-8 h 6-2 d c A-K-J-6 When the right hand opponent opens with 1d, Jannersten and Wohlin recommend a 1s overcall. An advantage with such an overcall is that it is unspecific. When the opponents bid on, you have only given away three cards in spades. That is not so helpful for declarer to know. Distributional Hands I have no new ideas about distributional hands. Zia Mahmood describes so-called Pakistani Preempts in his book Bridge My Way, The idea is to use multi preempts to show a number of extra features, but the advantage may be offset by allowing the opponents the weak, probing double. Most likely, bidding as high as you dare and hoping that the opponents will guess wrongly, is the optimal method. Mean Opening For each system below, I have calculated the mean opening, where pass is 0.0, 1c is 1.0, 1d is 2.0, etc. This can be done by multiplying the level of each opening bid with its frequency and adding up the values for all openings in the system. I wrote a program to do this in the nineties. Standard American has a mean opening of 1.48, that is, right between 1c and 1d. The mean opening gives an idea of how aggressive the system is. Of course, it says nothing about other aspects, such as how precise the system is, or how much information is gives away. Some New Bidding Systems If we try to apply the above principles, what would a bidding system look like? First, let us look at a natural system with a constructive flair, (mean opening 2.85): 1c 14+ clubs or balanced 1d 14+ four-card suit 1h/s 14+ five-card suits 1N balanced 2c/d/h/s 8-13 five-card suits The one level openings are unusually strong, and all responses are based on transfers. Apart from the latter, this system looks a bit like Fantunes, the system of Fulvio Fantoni and Claudio Nunes. The system is much more aggressive than Standard American, because of the five-card weak two openings. By the way, I have assumed that all systems use 2N as a weak opening with 5-5 in the minors, and natural preempts. An alternative is to go for weaker openings, but then the one-level openings should be less specific, so let us use three-card openings and canapé, (mean opening 3.13): 1c/d/h/s 10+ three-card suits 1N balanced 2c/d/h/s 0-9 five-card suit The one level openings show at least three cards, and are combined with canapé. Responder does not use transfers, because he is often stronger than opener. Note that we open above 1h on average. This system is rather like the Roman Club of Blue Team fame. Maybe we could call it Aggressive Roman? My guess is that the aggressive version is better, simply because it is going to gain in many competitive situations. Competitive will beat constructive. Well, how would a strong club system look? First the constructive version, (mean opening 3.0): 1c 17+ all distributions 1d preparatory 1h/s five-card suits 1N balanced 2c six-card suit 2d/h/s 6-11 five-card suits 9 (10)

10 The positive responses to 1c should use transfers, and responder should use transfers after 1c 1d. Transfer responses should be used after the 1d/h/s and 2c openings too. But we can also design a weaker version, which becomes more aggressive, (mean opening 3.55): 1c 15+ all distributions 1d/h/s three-card suits 1N balanced 2c five-card suit 2d/h/s 0-9 five-card suits Transfer responses should be used after 1c, but not after the other opening bids. The one level openings are combined with canapé except when clubs are involved. The 1N opening should be strong when vulnerable, for example This is actually the Cottontail Club, the system I play myself. This approach can be made ACBL legal in the following way, (mean opening 3.54): 1c 15+ all distributions 1d three-card suits 1h/s four-card suits 1N balanced 2c five-card suit 2d/h/s 0-9 five-card suits I think that strong club systems are superior to natural systems in the long run, because the limited opening bids are easier to use in both uncontested and contested auctions. Finally, let us look at a strong pass system, (mean opening 2.66): pass 13+ all distributions 1c 8-12 hand with shortness 1d 0-7 all distributions 1h/s card suit 1N 8-12 a five-card major 2c/d 8-12 five-card suits 2h/s 0-9 five-card suits Since the limited opening bids are as weak as 8-12, it is not good if they show long suits on the one level. Thus, we will only approve the unspecific version. In fact, this system is very much like the Polish system Regres. Note, though, that the mean opening of Regres is lower than that of Aggressive Roman and Cottontail Club. The Delta system described earlier is precise but passive, with a mean opening of only It is actually used by responder in Regres, after opener s initial pass. It has one clear drawback, namely that most opening bids are weak (8-12) and show a singleton or void. Thus, it often gives away too much information when the opponents declare. Conclusions The current trend in bidding systems is, if anything, to move away from the artificial experiments of the seventies and eighties, and back to more traditional methods. Fivecard majors seem to gain popularity slowly but inevitably. Several ideas I propose in this article, like canapé and bidding three-card suits, may seem old-fashioned or quaint. They are certainly not becoming more popular. My hope is only to spawn some thinking outside of the current trends. The reader may have noticed that this was not a presentation of one optimally designed system, but at least five. Personally, I believe that either Cottontail or Regres are the best systems. The artificiality of Regres makes it difficult to play with current rules. Therefore the aggressive version of strong club probably is best in practice. The ideas presented here are neither mine nor new. Rather, I owe a lot to many bridge theoreticians and writers, and especially to Slawinski and Ruminski. I only hope I have contributed some easy arguments for and against different design principles for bidding systems. I would also like to thank my partner Anu Uus for experiences gained in playing the Cottontail Club. C d h s 10 (10)