Symmetric Relays 2.0 ++++++++++++++++++++ One aspect of Symmetric Relays 1.0 was the different response structure for majors and minors. Suit-setting in both-major sequences was also different. Whatever the advantages (recovering a step at the expense of 3NT being sometimes artificial - remember?) that undoubtedly contributed to a critical mass of complexity. Unfortunately there will almost certainly be artificiality even in a Mulberry-like treatment. As Peter observed, whilst a natural jump to 5C as a non-forcing *try* is attractive it makes something of a mockery of investing effort in relays only to be worse placed than those using 4NT as a key-card-ask for a minor. Should these asks be Deadwood rather than straight KCA? As we will see, it's not easy being 'natural'. Showing shape =========== There is not much choice after 1S because all shapes - even 5-5s - bid spades first. So my scheme looks like this: 1S - 2C 2D = (hearts or diamonds) OR special-A 2H = (diamonds or hearts) 2S = special-B 2NT = lower shortage (club length implicit) 3C = 5-4-2-2 OR 6-5-1-1 OR 7-4-1-1 3D = 5-5-2-1 OR 5-5-3-0 3H = 5-4-3-1 3S = 6-4-2-1 3NT = 6-4-3-0 4C = 6-5-2-0 4D = 7-4-2-0 4H = 7-5-1-0 4S = 6-6-1-0 We can choose special-A and -B to contain single-suiters, 5-4-4-0 and 5-3-3-2 (and 4CM!). All specials bid 2S, immediately or via 2D. We can choose 2D to be natural or transfer (to exploit responder's hidden hand). We need to have clubs implied by immediate 2NT and higher responses because PG and I include a non-forcing club hand. We could let this go as our current 1M - 3C = NF, 6+ clubs with 3M seems to be curiously rare. When opener has hearts we could do the same with obvious substitutions: 1H - 2C 2D = (spades or diamonds) OR special-A 2H = (diamonds or spades) 2S = special-B etc. Interlude --------- Interestingly the analogous WA sequence would, nudged up a level (recall they relay with 1NT), look like this: 1H - 2C 2D = (clubs or diamonds) OR special-A 2H = (diamonds or clubs) 2S = spades 2NT = single-suited no shortage 3C = single-suited short club 3D = single-suited short diamond etc. For a heart-minor two-suiter, the spade scheme is used. But when opener has spades, he cannot have 5-5... So though starting higher, those responses look like, 1H - 2C 2S - 2NT 3C = club [lower] shortage 3D = 5-4-2-2 OR 6-5-1-1 OR 7-4-1-1 3H = 5-4-3-1 etc. Which is kind of clever. They spend their extra step gained by the absent major 5-5s after 1H opening to expand the single-suiters early and know what opener's suits by immediate rebid (potentially allowing a natural break in relaying as they might have any long suit etc.). But this is surely a digression - you both will say you'd like a consistent set of responses, that our immediate 2NT and above shouldn't vary with opener's major. What we should do... --------------------- Is spend the extra step (hearts-then-spades-5-5) on something else. My suggestion would be either, a) Maximum (or minimum) 5-4-3-1 as this is the most common shape b) Rare 6-5-1-1 / 7-4-1-1 leaving 3C as only 5-4-2-2, releasing the 3-level to set suits or check on stops rather than clarify against rare types. To be clear, what I'm suggesting is, after the a/b choice above, to have exactly the same responses and suit-setting after 1H and 1S. Again we would choose whether opener rebids in spades by transfer or not. Can anyone think up a mnemonic for the shapes? 54 / 55 / 54 / 64, voids start at 3NT with 64 / 65 / 74 / 75 / 66 / Suit-setting ========= General ------- Unlike Welland-Auken who do not employ two-level responses to show diamonds or the other major (favouring instead a 1NT start to relays), we do not need to agree fits in opener's 3-card major, let alone shorter holdings. potentially however, 3-card diamonds could constitute a fit. So it would seem simpler to only exclude 2-or-fewer non-clubs. However, we should give some thought to clubs. Either we always break out of relays with big club suits or we always include clubs in possible 'fits' - one idea, if they are not real fits, placed at the end (where they won't prejudice more likely fits and be a wake-up call if they occur). Mulberry gives three ways to express suits; weakest via an end-signal, medium by an immediate 4-level bid and strongest by a control assuming Deadwood ask. I've gone for Deadwood A Mulberry-like solution ---------------------- This could work well when 3S is available use it as a puppet to 3NT (which responder promises not to pass). Over 3NT responder uses the 4-level as key-card (Deadwood?) in the named suit. An immediate 4D is an end-signal (puppet to 4H, pass next) and the remainder of the 4-level is natural but encouraging (and NF if game). Say 3H = 5=3=1=4 3S = "Bid 3NT", then 4C/H/S = Deadwood in clubs/hearts/spades (4D undefined) 3NT = to play 4C = Slam try in clubs 4D = end-signal 4H = Slam try in hearts 4S = Slam try in spades That's all very nice because diamonds, which we need for an end-signal, is the unwanted suit. When the shortage is in a major we can use 4-short-major as a surrogate for diamonds. Say 3H = 1=5=4=3: 3S = "Bid 3NT", then 4C/D/H = Deadwood in clubs/diamonds/hearts (4S undefined) 3NT = to play 4C = Slam try in clubs 4D = end-signal 4H = Slam try in hearts 4S = Slam try in diamonds This will still work if opener is short in clubs and all suits are potentially playable but we'll have to go as high as 4NT for the natural try in diamonds. That's not so bad because it give opener a 'last chance' 5C. (Perhaps from a technical point of view it would be attractive to swap 4C and 4NT - "when clubs are short and other 3 suits are 3+, 4NT = clubs" - but I don't see either of you agreeing to that...) Say 3H = 3=5=4=1: 3S = "Bid 3NT", then 4-level = Deadwood in named suit 3NT = to play 4C = Slam try in clubs 4D = end-signal 4H = Slam try in hearts 4S = Slam try in spades 4NT = Slam try in diamonds I think we can manage when 3S is available - or, perhaps when we 'know enough' already. But things get trickier when we have to start at 4C (that is, opener's description of shape has ended with 3S or 3NT). Say 3S = 5=2=1=5 3NT = to play 4C = "Bid 4D", then 4H/S = Deadwood in clubs/spades 4D = end-signal 4H = Slam try in clubs 4S = Slam try in spades Which is OK. Say 3S = 5=1=5=2 3NT = to play 4C = "Bid 4D", then 4H/S/NT = Deadwood in diamonds/spades/clubs 4D = end-signal 4H = Slam try in diamonds 4S = Slam try in spades 4NT = slam try in clubs Because 4D isn't available in either the via-4C sequences or the immediate 4-level (unlike via-3S) the meaning of the via-4L and immediate-4L are the same. When 4C isn't available we have to give up on something. I suggest it be catering for clubs. Note the retention of a NF 4-long-major and 5C. Say 4C = 6=0=5=2 4D = end-signal 4H = Deadwood in spades 4S = Slam try in spades - NF 4NT = Deadwood in diamonds 5C = Slam try in clubs 5D = Slam try in diamonds Too high for an end-signal -------------------------- Whatever we play, it will almost surely involve a 4D end-signal. But once that is taken away we have to have some agreements. In this Mulberry structure we have to give up on the 3-way strength separation. Space is short; let's not cater for clubs when it's not one of opener's suits other than make 5C natural. Whilst 5C can be passed, with good cards for clubs, opener can raise (or correct with solid suit?) as responder might have baled to the major. 4-long-major = Natural, to play 5-long-minor = Natural, to play 5C = Natural, to play 4-impossible-major = Deadwood for long major 4NT = Deadwood for long minor When opener has both majors we have to be a little careful lest the ask takes us too high. 4NT = Deadwood in hearts 5D = Deadwood in spades Say 4D = 7=0=4=2 4H = Deadwood in spades 4S = To play 4NT = Deadwood in diamonds 5C = To play 5D = To play At higher levels still (4H/S) perhaps we should simply play everything as 'natural' but, if there is an impossible-major use that for longer-suit key-card, 4NT for second-suit. If there's no impossible-major then 4NT is double-barrelled key-card for the long suits Say 4H = 5=7=0=1 Pass = possible 4S = To play 4NT = Key-card ask in majors (double Deadwood?) 5C = To play 5D = Last train 5M = NF try or perhaps 4H = 1=7=5=0 4S = Deadwood for hearts 4NT = Deadwood in diamonds 5C = To play 5D = to play 5H = To play WA solution ----------- As mentioned earlier, WA have artificial suit-setting: "4D is an end signal [..] All other bids are optional KCB, the principal of longest suit/lowest game applies. Exception: 4H is never KCB for hearts and 4S never KCB for spades, instead we swap with a lower bid." This is my reading of their methods based on examples from 'Challenge the Champs' (June/July 2015): Say 3S = 1=5=5=2 There are 2 possible trumps suits, hearts and diamonds and (always) clubs (despite it being short [<3] in opener). Suits are set ignoring only 3NT and 4D (end-signal), with cheapest game first, 'real' fits (that is 3+ length) first. 3NT = Nat NF 4C = Deadwood in hearts 4D = End signal 4H = Deadwood in diamonds 4S = Deadwood in clubs Say 3H = 3=5=4=1 Hearts, spades and diamonds count because of 3-card or better, clubs tacked on because they're clubs - all four suits are possible: 3S = Deadwood in hearts 3NT = Nat NF 4C = Deadwood in spades 4D = End signal 4H = Deadwood in diamonds 4S = Deadwood in clubs Say 3H = 5=4=2=2 (1S->2D->3C->3H) There are three suits, diamonds not being a fit. 3S = Deadwood in hearts 3NT = Nat NF 4C = Deadwood in spades 4D = End signal 4H = Deadwood in clubs 4S = Natural! Note that because hearts is a LOWER GAME than spades, if it is a possible fit, then it comes first. Say 4C = 6=0=5=2 4D = end-signal 4H = Deadwood in spades 4S = Deadwood in diamonds 4NT = Deadwood in clubs 5C = Natural 5D = Natural Ostensibly this is a clearer method but in many ways it resembles Symmetric Relays 1.0 suit-setting. Questions ======= - Can we cope with the distribution showing? - Modified-Mulberry or WA suit-setting?