Monday, August 8, 2011

Unusual over unusual

Unusual over unusual is one of those very dangerous conventions which has so many alternatives and each exponent of the "convention" assumes that everyone else is using the same alternative.  Other examples of ambiguous conventions are four-way transfers, Jacoby 2NT, 5NT, fourth suit forcing.  I'm sure we can all think of a few.  Perhaps the most common problem of this nature is in so-called "Standard American" where after the auction starts 1♠ p 2♣ there are various opinions regarding which continuations are forcing -- something that very few casual partnerships actually discuss.

But back to "Un vs. Un".  When RHO makes a two-suited call, typically an "unusual notrump" or a Michaels cuebid, we have two suits in which to cuebid, one to raise, one to make a possibly natural call and of course double and our own notrump bid.  That's six possibilities, not including jumps.  Double is penalty oriented and while it has its own challenges, we won't discuss it here.  Let's assume for the moment that raises and notrump bids are non-forcing.  Indeed, the raises are all "to play", while it makes sense to consider a 2NT bid (if such is possible) as invitational because it's extremely unlikely that playing 2NT is going to be the par result in one of these competitive auctions.  That leaves three suit bids for which we may assign meanings.  Clearly both cuebids are forcing, almost by definition, while the new (fourth) suit bid could be considered forcing or non-forcing.  I'm going to assume for now that we will agree that the fourth suit bid is non-forcing, although I am aware of systems where it is forcing.  However, I believe such systems are unnecessarily complex for all but the most practiced partnerships.

The non-forcing new suit bid should be something like the poorly-named "negative free bid" after a single-suited overcall: sufficient points to compete and a decent six-bagger or a very good five-bagger.  So we are left with an appropriate meaning for the two cuebids.  It makes sense to use one for a "limit" raise (or better) in partner's suit and one for a forcing bid in the fourth suit.  But which is which?

Probably the most popular method is to use the lower ranking cue-bid to suggest the lower-ranking of "our" two suits: either the limit raise plus, or the fourth suit, depending on which suit partner actually opened with.  This is all very easy to remember but it is wasteful of our now extremely limited bidding space when partner actually opened the lower-ranking of our two suits.  That's because to make the forcing bid in the fourth suit, we must make the higher-ranking cuebid, which therefore consigns the lower-ranking cuebid to oblivion.  Here's why that's bad (this was taught to me by a couple of good theoreticians: Dick Wagman and Bruce Downing).

Let's take an example hand: ♠A3 AQT72 Q5 ♣J652.  Let's say you open 1 and LHO calls 2NT.  Playing hi/hi-lo/lo, partner cuebids 3 to show that he has a forcing hand with spades.  What are you going to bid now?  You can't bid 3♠ because that would suggest three decent cards in support, you can't rebid your hearts because they really aren't good enough and you're reluctant to bid 3NT with such tenuous stoppers in the minors.  But you can't pass either.  You'll have to decide which of these actions is the least evil.

This is where the "good" way of playing Un vs. Un comes in: we assign the higher cuebid to the message that describes responder's hand most completely: the limit/plus raise.  We assign the lower cuebid to the message that puts the final contract in most doubt: the forcing new suit response.

Let's go back to the hand above with the new auction: 1 (2NT) 3♣ [showing spades].  With your hand you can bid an artificial 3 to show that you have no other reasonable call.  Partner can now clarify: 3♠ with rebiddable spades and game-forcing values; 3NT to show five spades and minor-suit stoppers, 3 with three-card heart support (or perhaps two to a high-honor) and game-forcing values.  If instead the sequence had gone 1 (2NT) 3 (pass) 3 (pass) 4, that would show four trumps and probably some slam interest since with only interest in reaching game, responder could have bid 4 to start with (fast arrival, preemptive and/or willing to double their minor suit sacrifice).

Isn't that better?  And just as simple?


  1. By my recollection, Marshall Miles makes similar arguments in his book on Competitive Bidding in the 21st Century. In your example, determining the proper strain when responder has spades is more difficult than determining the proper strain when responder has support for opener's hearts. Accordingly, said Miles, the lower cue bid -- allowing more room to ascertain strain -- should be used on the responder hands with a new suit.

    So ... "yes" to your closing questions: using lower Q for "other" suit and higher Q for opener's suit, is better and all methods that use consistent interpretations are equally as easy to remember.

    Just one disagreement with your post. Jumping to 4 of partner's major over the 2NT interference should be, I think, preemptive while any hand with limit raise or better needs to start with the higher cue bid. Thus I would interpret the auction that begins with the higher cue bid and then bids game not as slam going (a slam invitational hand would have to choose to control bid a second time, and have "five level safety" opposite an opener that has shown a minimum) but rather as perhaps a hand with no more than three trumps and minimum opening bid values.

  2. Thanks for your comment, Jeff. Your "disagreement" with my post seems to agree with it 100% assuming that "Some slam interest" and "slam going" are two different kinds of support. Perhaps something like ♠KT52 ♥J943 ♦AK86 ♣7 versus ♠KT52 ♥KJ93 ♦AK86 ♣7 (this latter hand might cue-bid 3♠ over 3♥ if that suits partnership style). The kind of hand that might go straight to 4♥ without going through the cue-bid might be ♠JT52 ♥KJ943 ♦AT86 ♣-. Finally, a hand that will probably cue-bid and then launch into key-card: ♠KQ52 ♥KJ93 ♦AK86 ♣7. Such fine gradations might not be possible if the opponents keep on bidding of course. I've constructed these four different hands rather quickly so my apologies if your judgment and mine aren't quite on the same page - or if one of my hands has 12 or 14 cards.

  3. Robin, here is a possible hand I would cite for the support-showing cue bid and then raise of 3H to 4H: KJxx, Kxx, AJxx, xx, no fourth trump is required; all that is required is three trumps and a hand strong enough to want to be in game even though partner/opener, by rebidding only 3H, showed a hand that would not accept a game invitation. Opener is not invited to launch a slam try over the second round 4H call by responder because all that responder is saying, by raising 3H to 4H, is that he thinks you can make a game. Four card support is possible, but should not be counted upon by opener.

    Personally,I would jump to 4C (a splinter) on both JTxx, KJ9xx, ATxx, -- and on KTxx, J9xx, AKxx, x, each seven loser hands with some quick tricks. To me, a 4H jump is a hand with less defense than even the first of the two hands above.

    What to do with a stronger hand such as your example of KTxx, KJ9x, AKxx, x is a problem ... the type of problem that makes opponents want to continue preempting! Splintering and then bidding on over a signoff is not something I like to do, but might consider in this case.

  4. I need to talk to you about credit for this idea in a book I"m writing. Please contact me.