DinDIP

Freaks are infrequent but often a problem when they do arise. I've played a symmetric system where it was possible to show every shape, with the shapes ordered on the basis of frequency (so 3D = 5431, 3H = 6421, 3S = 6430, 3N = 7420, 4C = 7510, 4D = 8410 etc). I've also played where there are no ways to show freaks and the 4C+ bids either were not used to show anything or were used to show strong versions of shapes already shown, typically the shape shown by the bid of 3N so the 3N bid was limited and passable. Another alternative is to follow the original symmetric idea which was to make the bid that showed the most frequent shape with an analogous pattern and then make an impossible bid. So, 7510 is analogous to 5431 and 8311 to 6322. Of course, that still doesn't provide a clear answer about how to differentiate between a 5431 that is too strong to pass 3N and a 7510 that is too shapely to pass 3N, although it is easy to make special rules (such as 4C or 4D is always the freak). I've not found any of these satisfactory so looked further and decided that the ones that were most frequent and caused the greatest problems were those with 8+card suits. As a result, my partner and I now play that we can show 8+card single-suited hands, and we can bid some freaks, when the hand is limited, by showing the closest shape and then making an impossible bid. Our rules: Freaks We can show some freaks: • 4♣ when showing a one-suiter shows an 8+card suit (doesn't matter whether teller is limited or unlimited) • A limited two-suiter can make a NAT, descriptive bid if it is otherwise anti-systemic. – An unlimited hand can't do this because such bids are zooms with the shape shown by 3N (see below). For one-suiters 3N 7-3-3-0, 5-8SP 4♣ 8-card suit (4♦ = relay [4♥ = singleton, 4♠+ = void], 4N = RKC) 4♦ 7-3-3-0, 9SP 4♥ 7-3-3-0, 10SP etc where SP are slam points (3, 2, 1, kingleton = 1, stiff Q = 0)

I agree that's the rationale for TFRs but my point is that the first TFR doesn't need to be 1N. The situation after a one-level overcall is different to that after a 2M overcall, where space is at a premium and it is a favourable tradeoff on balance for TFRs to start with 2N. In my experience a NAT, NF 1N is too valuable a bid to give up. This is true if opener's hand is 11-13 BAL. And it's even more true when opener has some 17+ hand. 1N isn't the only option with the problem hand (KT63 A982 9843 K) -- maybe a takeout X is better to show the four-card H suit. But change responder's hand to something like KT6 A98 9843 K32 and 1N is clearcut. If you are forced to pass with hands like this you make it impossible for our side to declare 1N -- a typically advantagous contract at all forms of scoring -- when opener is minimum and BAL. And you put serious pressure on opener whenever he has a strong hand. Should he reopen with xxx KQx AKJx KJx or the like if you pass 1S? I think opener has to be able to assume that responder will have acted over 1S on virtually all hands with sufficient strength to force to game opposite 17+. (Note that this is not true when the overcall is 2M, which is why getting to the two level before the 1C opener has the opportunity to clarify his strength and shape is so important.) David

I presume the rationale for playing 1N+ as TFRs is so there is commonality with auctions after two-level overcalls. But I think that's wrong. A natural, limited 1N is too valuable to give up for any TFR method, and space is not at a premium as it is when they overcall 2M. Partner and I play (after our strong-only club) that 1N is NAT, NF and 2C-2S are TFRs to the next suit or, if a TFR to their suit, a BALish hand without a stopper. There is some loss of space when responder has long clubs but the gains from being able to TFR into the other suits and show a NAT limited hand with a stopper outweigh that loss. I think the gains even greater when playing a two-way 1C opening when 1N is an even more likely contract, and especially as you really want to be able to show a BAL 8-11 with a stopper when opener is strong. David

Funny you should ask: partner and I have just had a protracted discussion about this. He, from a Precision background, was used to playing X = 5-7 any with O's suit response F1. I, from a Moscito-like background, was used to playing X = takeout, 5+HCP, neg free bids at the two level and modified Rubensohl at the three level (with 2N NAT, NF). My other partners had some bad experiences with off shape hands so we had also adopted ELC after doubling 1M or 2m (meaning, e.g., we would X 1S with 3-4-(51)/2-4-(52)/2-4-(61) 5-7 and make the cheapest rebid if partner bid the minor we didn't have). He worried about ELC; I worried about O's NS in response to a X being F1. So, I generated 200 deals with semipos responding hands to see what worked best (recognising that the sample is limited in size). From this analysis I concluded that partner and I were both right: ELC was often dangerous and rarely necessary. In fact, it was mostly right to pass with length in their suit with less than GF values unless you had a 6+card suit or a good 5-card suit you could show at the same level. And playing O's rebid as F in response to a takeout X got us too high too often. So, this is what we ended up with (these rules cover what we do after a 1M overcall as well; over lower interference we continue relaying): • Pass = nothing to say or penalty pass • X (if their bid NAT or showing a suit) = takeout, at least semi-positive (if their bid ART, not showing a suit) = as though they opened with the same bid (so often values but takeout of the higher-ranking suit if 2D multi or similar) • New suit at same level = NAT over D/H/S interference, NF at two level, F at one or three levels = transfer over C interference = transfer over NT interference, with C = both majors (unless NT is TFR to C) • If NS is competitive, transfers start at the next level, either from 2C or 2N. • Transfer into / bid of their known 4 card M = natural AT ONE OR TWO LEVEL, stopper ask at three level • Transfer into / bid of their known m or 5+ card M = stopper ask • JS at three level = NAT, GF, good suit • Notrump bids are NAT, except over their 2M 1N = 5-bad 8, NF 2N = good 8+ (but prefer X or suit bid with flexible hand) 3N = 14-16 if 2N is F 4N = 14-16 if a single jump • 4m = leaping Michaels (H+m if they haven’t shown a M), weak or strong If R doubles at the one- or two-level this is classic takeout, semipos or better, or a flexible GF+ (where X is a better start than TFR and NS/cue) • O's non-jump responses, including in NT, show a min and are NF. Any bid by R, including a raise, is GF: a NS shows a 5-card suit. • O's J response is GF and shows 5+(4 good) cards. (This is NOT a fast arrival situation because R may have a GF flexible hand lacking support.) A NS by R shows 5 cards and a hand that was strong enough to GF. A cue initially asks for a stopper if below 3N. • O's jump NT response is NAT and GF. • With other GF hands O cues and we bid 4+card suits up the line (except that R shows 5M before a lower-ranked 4-card suit). • O's jump to game shows a min hand with a 6+card suit and a shapely hand. • (If available) O's below game double jump (say 1C 1H X P 3S/4m) sets the suit and is slammish: R cues with any GF hand. • If advancer bids/raises, O's free bid shows extras and is GF if at the three+ level; X of raise is responsive (of NS is penalties) and shows extra values or shape, and so is also GF. • If advancer redoubles, O's bid suggests a strong preference (so usually 5+card suit); a P shows no preference. O's jumps are still GF. If R makes a semipos+ TFR then O accepts the TFR with min hands. Bids (including 2N) are F and show extras (strength and/or shape). A Jraise is GF and JS are SPL. The TFR scheme we use is based on Rubens's original Bridge World article, so (over their 2M) 2N is 5-7 with 6C or a GF with long C and at least a half stopper. With a GF and less than a half stopper R bids 3S (high but descriptive). After a TFR to their suit (BALish, no stopper) O's accept denies the ability to bid 3N or a NS and is Baron-like. Over 1M, we feel 1N as a NAT response (5-bad 8, stopper) is too valuable to give up so we start TFRs at 2C. A TFR to their suit again shows BALish, no stopper but now O's acceptance is Lebensohl as a NS is GF. David

A decade or two ago I read an article or letter to a bridge magazine suggesting the rules of bridge be altered to allow for a zero notrump. This was a bid that would rank below one club and require the declaring side to take 6 tricks. The point of the suggestion was not so much that it would be a playable contract (there were rules I now forget about whether and when it could be the final contract) but that it would give bidders an additional step. IIRC, the author suggested that natural bidders could use this to play weak and strong notrumps, while strong clubbers would have a strong opening and a natural one-club opening at the one level. (And relayers and forcing pass players would have all the room they needed and could relay to their heart's content.) My recollection is that it was in The Bridge World but I haven't found it there. Can anyone help? Thanks in advance David

I think the answer depends on the partnership philosophy about handling intereference. If it is to revert to defensive mode (i.e. bidding as though they had opened with that bid/call) then XX should be takeout of diamonds or of both majors, and values with penalty interest if the double shows a strong hand or is undiscussed. If the partnership's approach is (when the auction is low enough -- as it is here) to ignore the interference and continue with our constructive bidding then XX should be an extra-strength relay, which allows 1H to be NAT. As our partnership's agreement is to continue relaying when their intereference is lower than 1H, XX should be a relay. However, from a theoretical viewpoint, there's a lot to be said for the idea that when they make a shape-showing bid and we have only shown strength our bidding should revert to shape-showing as the auction is likely to be competitive. It's possible to combine that with retaining XX as a relay: pass with min BALish hands (and awkward UNBAL hands such as 4441s) and widen the range of the NAT, shape-showing bids to 16-19, even 20; and play a 1N rebid as 18-20. That way O only relays when he is really strong (21+ or equivalent) David

Yes, and 1M-2N might only be good three-card support. Just to clarify, the 3N I was describing was in the auction 1♠-2♦-2♠-3N. A direct 3N over 1M shows a min BAL GF raise with three-card support no stronger than Hxx and stoppers outside. Partner treats bids like these as RKC for S. I'm working on persuading him that such a definition is unnecessary when a forcing raise is available as responder can always start with that and then use RKC. But there are more important things to get him to adopt so that's way down my list at present. Snap, even though (again) a slam INV raise is available (3♠ -- F because game INV hands start with a 3♣ raise that promises a GI with three-card support. This is a great method as we have lots of auctions that go 1M-3♣-4M and the opps have very little information to go on -- at most a negative inference from 4th's failure to double. Contrast that with auctions like 1M-NS-NS-3M[GI] where we have bid three suits to help the opps on lead and defence.) He could but our style is to bid NAT unless there's a very good reason for not doing so. Not obvious here what that is; after all, 1♠-2♣-2♠-3♦ leaves even less room for O on the third round. Partner's sequence must be better. David

Playing with a good -- and aggressive -- partner against quality opps in a serious teams playoff match, we had this auction. Where did we go wrong and what suggestions, if any do you have, given our (imperfect but partner's preferred) methods (old-fashioned SA with some tweaks)? First board of the match we held [hv=pc=n&w=sakq953hqd73ck764&e=st8hkj54daqj5caqj&d=s&v=n&b=15&a=p1sp2dp2sp3cp3sp3nppp]266|200[/hv] By way of explanation: Our partnership agreement is not to stretch on the first board: being aggressive is fine, but not stretching. (That might be semantics for some but the distinction is meaningful for us.) 2♠ was NF and did not promise 6; a NS, 2N or 3♠ would have been GF 3♣ 2N instead here would have been INV but NF; 3N would have been NF but promises S tolerance (doubleton or stiff honour) so O(pener) knows when to remove when holding a six-card suit 3♠ All actions here would be GF: 3♦ would usually be three-card support, could be Hx at a pinch; 3♥ would be asking for a stopper (partner would bid 3♣ with a hand like x KJT AQJxx Qxxx as 3N would promise better S); 4♠ would show 6 or 7 good spades (a picture jump: 3♠ then 4♠ would show weaker S); 4♥ would be a SPL (over this partner's 4♠ would be a cue, not an offer to play); 4♣ would be a hand strong enough to make 5♣ -- weaker hands would bid 3♥ then pass 4♣, while a direct 5♣ suggests minimum with good clubs, often 5 (4♠ over a direct 4♣ is an offer to play, not a cue) P (over 3N) 4♣ here was undiscussed but would show 6S and 4C: in related auctions we have bid like this with shapely hands with minimum high-card values so partner might take this as NF. The partnership agreement is that when choosing between equally suitable or equally flawed calls (such as O's call over 3♣) choose the cheapest. We know the hand would be easier using other methods. But should we have done better given what we were playing? Thanks David

If 2N promises two hearts then it makes sense for 3H to show a good 5-card suit as responder can just bid 4H with a 6-card suit. (And if responder is an UPH -- and so might still have slam interest -- he can bid 3H then 4H or show a control/SPL with a strong 64.) It probably is reasonable to require a doubleton heart for 2N: the auction is GF so O must be interested in a choice of games. Of course that means some other hands that want to offer a choice of games have to relay (unless playing the methods discussed below), say a hand like KJxxx x Axx AKQx where you might want to play 4S if partner has Ax or Qx. There are a number of possibilities. My preferred method is to play 3D asks for stopper(s) in doubletons (denial method, following usual rules for ordering suits) 3H asks for AKQP 3S shows a SPL in the major teller has shown (the shorter suit if teller has shown H+S or C+D), initially choice of games but could be a slam try, so teller should bid above 4C with extras (can be showing AKQP, RKC responses or NAT -- i.e. control-showing -- as you wish). 4C asks for controls (A=2, K=1) Of course, you have to decide whether Qx qualifies as a stopper: in Ultimate Club only Kx and Ax counted. But that adds some complexity -- too much for some. In which case, given Shevek's methods, it seems reasonable to define 3S as asking for a choice of games. It could be asking about the higher (or lower) of the doubletons or the long suit, or just a general direction shot. The exact choice should depend on what 2N (instead of 2S over 2H) means -- no point duplicating options. In which case, maybe it makes sense to reserve 3S for hands like KJxxx x Axx AKQx where O is looking for a choice between teller's doubleton M and 3N. David

Help please! I'm reviewing Peter Winkler's book "Bridge at the Enigma Club" -- great book that you should read if you haven't yet done so -- for The Bridge World. I wanted to clarify whether it is true, as a number of websites state, that encrypted bidding is banned. The ACBL, WBF and ABF (as well as other BF) system regulations explicitly state that encrypted signals are not permitted. But I can't find anything that proscribes encrypted bidding. Have I missed something? Are there jursidictions where encrypted bidding is permitted? Thanks in adavnce David

I presume this is intended for Shevek, the original poster. Having played played forcing pass with him on and off since the early 80s, I'll volunteer my tuppence ha'penny's worth: * A forcing pass provides very useful extra room for relaying, relay breaks and natural bidding. That's mostly a big plus, although there are occasionally range issues that wouldn't arise if playing a strong club (such as having the weaker hand relaying). But the big downside is that pass is more vulnerable to interference because the passer guarantees so little strength. * The 8-12 openings are great for competitive auctions, including ones that we might not otherwise get to enter (or ones that don't become competitive because we preempt them). But they also give away information to the opponents which can help declarer form a better picture of the opposing hands. And the advantages of opening BAL 9-12s with no 4+M is less than for the openings that show a suit. * The 1S fert works surprisingly well. A large part of that ATT is because the opps don't know how to defend, but even good players with well-thought-through methods find it hard to judge correctly, so many of the potential -1400s are avoided. So, at a systemic level, we do anticipate problems like this: we expect to miss games and slams when the opps compete after we pass in the expectation that the gains elsewhere will compensate. But it's still a good idea to see if those outs can be avoided. And, because there is no body of knowledge to draw on that is comparable to that for NAT bidders, it's necessary to ask others whether what was done was normal or if some alternative were better. My view is that neither O nor R can act over 3D without risking too much so this is just one of those boards. David

It's a pity too many posters focus on extraneous issues (is this a sensible system base?) rather than the generic problem. To put it in context, imagine you're playing a 17+ strong club when O holds AKJTx AKxx Axx x EXCEPT that the combined partnership assets can be an A lower, which increases the risk. (If you play a 16+ 1C take away the SJ, if a 15+ 1C make O's spades AQTxx. In each case you have the problem of whether to reopen -- and if so how -- after an amorphous strength-showing opening, a NV 3D preempt by LHO and a NF pass by responder.) I think it's clear to pass after a 13+ opening: responder's average number of diamonds figures to be closer to two than one at the vul (because many opponents will bid 3D at favourable on a six-card suit). And the average count for responder's hand figures to be less than the 9 or so you'd otherwise expect ((40-(15+7))/2) because R would have acted with many 10 counts and virtually all 11+HCP hands -- as well as most 9s and some 8s with shortness. This means that most of the time you're bidding for the partscore with no clear way to find the best fit -- double is more attractive with 4-5-3-1. But with stronger hands -- and a stronger 1C opening -- it becomes less clearcut. If my 1C were 17+ I'd reopen with AKJTx AKxx Axx x as the risks are smaller and the upsides greater (a decent chance of making game opposite a yarborough with five hearts). And I'd reopen with a X as that maximises the chance of finding game. So, it does matter how strong your amorphous opening is, and stronger is better. But, as all system designers know, one needs to take account of the other effects: I think a 15+ 1C is best because it allows me to open most 10s and many 9s that are shapely without making O's range excessive. David

Playing most relay systems this is easy, so how about my preferred NAT system (modified KS): P 1♥ (sound openings in a minor; 5+) 3♣ 3♦ (fit-showing*; cue) 3♠ 4♠ (cue**; Kickback) 4N 5♣ (1; Q ask) 6♣ 7♥ (2+/top3 in C; to play) * Promises 1+/top2 and {2+/top3 or 3/top5) ** Shows better than minimum (else 3♥) David

In KS by the book (see the KS Updated notes at http://www.bridgeworld.com/default.asp?d=editorial_dept&f=edgarkaplan/ksupdated.html) opener cannot have 15+ to rebid 2D after 1D-1M. 2D has a narrow range, much narrower than STD. Kaplan defines it as "Severely limited; 12 to 14 and not 14 if good 6-card suit." (See section B-10.) With 15+ opener bids in a new suit unless strong enough for a GF 3D rebid. 1D-1M-2C is treated like a reverse. (So KS players will rebid 2D with hands that STD players would rebid 2C with.) It's different from STD -- very different in parts -- but it all hangs together way better than most other systems. David

1. what you don't realize is that opps won't compete overagressively if you have a penalty Dbl available. So they will have a 9 card fit pretty much all of the time. This makes it easier in this case for responder to just bid 4♥ since he can expect partner to have 3+♥. 2. there's a reason why this semipositive structure was formed based on 5 card Majors. These responses are made to make competitive auctions easier, not to have something which is different for fun. If you first deny a 5 card M with your SP (1♥), then you can safely bid your 4 card M later on if necessary. It's with reason why Richard and I are really against a structure where you combine ALL semipositives into 1 response. 1. Maybe, but it's still the case that it's harder to find a fit if all you've done is to say something about your hand's strength and nothing about its shape. If O's pass is the equivalent of a takeout X and R bids 4♥ how does O know if R has 4 or 5 or 6H? There's still a lot of guesswork involved and this on hands where we know we have the values for game and often for slam. 2. I agree that showing 5+M semipos hands will often work well. But that comes with a price; on other hands O's decision about whether and how to compete will be affected by knowing whether R has 4 cards in a major. An ambiguous response like 1♥ that lumps all semipos without 5M doesn't provide O with that information. And Marston's structure means 2N and 3-level responses are not available to show other semipos hands with (say) length in the minors. That's less likely to be critical (because the suit is a minor and not a major) but there is still some cost. As always, different structures have plusses and minuses; and when the plusses come up the system looks great. David

Well, that may make sense now but the original and second versions of Moscito didn't use TFR openings. Marston's original acronym was Major Oriented Strong Club In Trouble Often On the OP's substantive question: it is possible to use a Moscito strong club and different openings (even five-card majors). Alternatively, Shevek has played a majors first-principle structure without relays with some success when visiting ACBL-land. Don't have all the details to hand but 2m were NAT denying 4M and 1M promised UNBAL hands so that a 1N response was F and could be used with all GI+ hands, and 2/1 after 1M were NF (as usual). (Can't remember whether whether 1D and 1N were both BAL hands or whether 1N remained BAL without 4M and 1D became BAL with 4M -- both are sensible.) Those who have read Peter Winkler's great new book ("Bridge at the Enigma Club") will recognise the similarities with his suggestions. David

While there is some merit in trying to optimise how individual sequences such as this are handled, I think that is only sensible for a partnership where both players have computer-like recall which does not affect their mental effort on other aspects of the game. For the vast majority of us it's MUCH better IMO to have universal principles that apply to all similar auctions. So, if your normal rules are that you continue to relay/reverse relay over interference of one or two steps then do that here. If your normal rules are that intereference breaks the chain and you revert to NAT bidding do that. There will be some auctions where you waste a step or reduce or even eliminate your chances of penalising the opps. But, IMO, that's a reasonable price to pay for being certain that you and partner will ALWAYS know what's going on. Having said that, if you do want to consider adding another rule to your agreements then you might try what Shevek calls "bubble doubles". The idea is that the double always shows the suit the opps have bid, so increasing the opportunity to penalise them. (It isn't perfect because you don't know whether the double shows 5432 or AKJ9xx in their suit.) So, after 1C (P) 1D (1H) P = relay X = 4+H 1S = 4+S etc After 1C (P) 1D (1S) P = relay X = 4+S 1N = 4+H etc I've played this in two partnerships but never had the penalty X come up and one partner (not Shevek) got two auctions wrong (because the doubles meant we had to add a step to a higher response to show 4441/5440). But that's a very limited sample. David

If you trust the opps then they are likely to have a nine-card fit to bid 3♠, which makes 4♥ by responder rather than X more attractive. David

Sure, but auctions like 1♣* - P - 1♦* - (1♠) strong; art GF; natural overcall 1N - (3♠) - ? are the same reason other people like to play natural positives. Any time your opponents manage to jam the auction to 3♠ where your side has made only 1 descriptive bid is going to work out poorly for you on average. True, but semipositives bring opener into the auction in a way that 1D (0-7) can't do. If 1D here showed a GF of any shape and then we have a 1S overcall, opener is now empowered to make a useful descriptive bid with his minimum hand. If responder had instead made a descriptive (or ambiguous) semipositive, then opener may have been able to do something. Say the auction went 1C P 1N showing 5-7 and 5H (a la Moscito), opener has now an easy raise of hearts. 1. I'm with Rob. This auction is difficult whenever 1♦ just shows strength, whatever the range. Sure O has more options after a 1♦ pos but how do you find a 5-3 H fit with confidence? 2. These hands are easy with some semipos responses but not others. So, after a 1N response that shows 5+H, O has enough (with the promoted SQ) to bid 4♥. But what if the response were a majors-first principle 1♠ showing 4+H UNBAL? If 4th now bids 2S 1♣ (P) 1♠ (2♠) would you bid 3♥ as O when you might have a 4-3 fit? David

Shevek and I have played a strong pass system with each other or in the same team for more than 25 years, when regulations allow and we've thought it a good thing to do. The structure we use is P = 13+ 1m = 7-12, 4+H/S 1H = 9-12 BAL, no 4M 1S = 0-7(8 BAL) 1N= 5+D, <4M 2C = 5+C, <4M with symmetric relay continuations when we are strong enough. Shevek has compiled records of how the different openings have gone (IMPs against the datum) for a number of these events. As one might expect, the 8-12 openings that show a suit are winners; we're happy to break even after opening 1H, even with the range narrowed to 9-12. Pass tends to be a small loser, mainly because competition is so difficult: O's range is very wide. But another reason is that we sometimes end up with thw weaker hand relaying and that's usually not as good as when the stronger hand relays. The 1S fert is almost always a winner. It's hard to tell whether this is because the opponents mostly use poor methods to counter it, or because they lack experience about what to do so their judgement is MUCH worse than in routine auctions, or because it is genuinely hard to bid sensibly after so much room has been taken away by the fert. We see elements of all three in many deals. Our experience is that we do go for occasional numbers but they're rarely where the fert loses. Much more damaging -- or, at least, more memorable -- are the number of auctions where the opponents are forced to guess because they have no room and do so successfully. David

That treatment -- 2C as forcing, like a reverse -- was played after a response of 1H or 1S but not after a 1N respoonse, even though there is a good theoretical argument (as enunciated by Frances) for doing so. In a weak notrump context, I've played a 1N response to 1C as promising C, with O rebidding 3C with weak hands and 2C (forcing) with strong ones. The rationale is the same. David

In the early-mid 2000s Magnus Lindkvist played (with Peter Fredin) a system with a three-way club like this. His old system cards were available online (at ecats bridge) so you could check there. He was playing relays over the other suit openings and used some of those modules after 1C as well. The system card attachments provided quite a lot of information about how they handled intereference and, if you look at hand records from the BB and European Chmapionships they played in. you will get a good feel for the style. David

Thanks for all the helpful suggestions. As it happened, partner decided that he too was interested in trying out the suggestion of Ulf Nilsson (and played by other Swedish experts) -- see The Bridge World Nov 2009 -- that all 5332s be treated as BAL hands (i.e. including 5M332) so we adopted the following structure in 1st and 2nd (in 3rd and 4th we open 1M/1D with 5M/D332): 1C = 5+C UNB or 4=4=1=4 or (optional) 4=1=4=4 or BAL 15-20 (including all 5332s) 1D/1M = UNBAL or semiBAL Simulations showed that, a priori, O had a BAL hand 57% of the time (9.3% of 1C openers were 5M332s) but that his average C length was over 4.1 and 67% of the time O had 4+C. A hand with C length was more likely when O's LHO passed: only 47% of O's hands were BAL; 5M332s were only 6.7% of O's hands; O's average C length was 4.5; and O had 4+C 72% of the time. (These figures are not entirely accurate as they assume that the opps could bid 2C, 3C and 4C as NAT overcalls of the 1C opening.) Thus, in response to 1C, I decided that it was important for R(esponder) to be able to show 4+C support as soon as possible on many hands, i.e. avoiding the ambiguous 1S response favoured by many. Therefore our response structure is: 1D = 4+H, may be 0-4 if 5+H or (4H and 0-3C) 1H = 4+S, may be 0-4 if 5+S or (4S and 0-3C) 1S = 4+D, denies 4M unless GF, may be 0-4 if 6+D or (5D and 0-3C) 1N = 4-5C BALish, denies 4M, (4)5-8(9) 2C = 4+C, denies 4M or longer D, (9)10+ 2D = 4D and 5C, 5-8 OR strong JS in D with (semi)SOL suit or 4+C 2M = strong JS 2N = BAL, denies 4M, 12-15 or 18+ 3C = (5)6+C, 4-8 3D/3M = 5+C, SPL, denies 4M 3N = BAL, usually 3=3=(43), 16-17 After 1C-1R-? we play 1M = 15-17 BAL, not 4M unless 4333 1N = 18-20 BAL, not 4M unless 4333 Over both of these we play 2C as a puppet to 2D (O can break to bid 2M with 4333 or 2H with 3=5=(32)) and 2D-3C as TFRs (2S TFRs to NT so we can rightside NT after a 1M acceptance of the TFR response). After 1C-1S-? 1N is 15-18 and 2N is 19-20 BAL. Whenever O shows BAL many sequences allow him to show 5M332 if R is interested in that information. One of the neat bits is the 1N response: over that 2C is not needed to show a min UNBAL hand (as this just invites the opponents to find their fit knowing we have at least an 8-card fit and two limited hands). With those hands O rebids 3C -- balance now at the three level if you dare! So 2C is used as a puppet for O to show 5M332s or GI hands with long C. It's early days but our first outing (a serious three-day Swiss) showed few system issues. And the 500+deals I've bid in practice (and more with partner) suggest there are some real advantages of the method. There are also downsides: finding 5-3M fits can be tricky after the opponents intervene. And, like other methods that lump lots of BAL hands into the 1C opening, it can be difficult for O to distinguish in competition between BAL and UNBAL hands. David

We play F2NPR: a new suit is forcing to 2N, preference or rebid. Simple, easy to remember, seems to be as good as any other rules (all of which are more complex or less efficient). David

In my experience, it's hard to find rules that are useful, easy to use and sufficiently accurate. But the process of working out why they don't work is very valuable -- with the proviso that the process involves bidding lots of hands to work out why. (Simulations can help but are less valuable than bidding actual deals.) I found when I did this years ago -- trying to work out a better method of hand evaluation -- that I never did discover a new metric that met the necessary criteria (useful, easy to use and sufficiently accurate) but my hand evaluatiion skills improved significantly because of all the deals I examined, trying to ascertain why the new value did (or did not) work on a particular deal. So, keep on bidding those hands! David