Monday, April 20, 2009

Double, double (Updated)

Some years ago, I wrote a series of haiku about Lightner doubles:

Double is Lightner
Asking for a heart?  A spade?
A diamond or club?

Double is Lightner
Asking for whatever suit
That partner wants led

Double is Lightner
But, when they rewind, I think:
Was it stripe-tailed ape?

Try a passive trump
The only lead to beat it?
They run thirteen tricks

While Mr. Lightner had a very nice concept, the pitfalls are many.  The following situation came up in a Swiss event yesterday:

At red v. white, you hold AKTxxxxx – xx Kxx in second seat.  The auction proceeds:
6H-P - P-P/X

At one table the partner of this hand passed, at the other he doubled (and all passed).  What would you lead in each case?  Assume you have a good partner but haven’t specifically discussed this kind of situation.  If you would bid differently, feel free to chime in as well.

Update:  In fact partner of the given hand held Qxxx Txx Axx xxx.  The double was based on a suspicion that the opponents were saving (probably not a good double even if it meant that, with only one trick.)   A diamond was led at the table with the double and the contract was unbeatable, with declarer having J AQJxxxxxx  x  Ax and dummy -- K  KQJTxxx QJxxx.  At my table I declared 6H undoubled -- I'm with Drew and Kenny who led a club here, but the opponents led a spade tapping dummy and I had to take the club hook and go down.

Should the double be Lightner?   I think so.  A better hand could have maybe bid 5H as a slam try rather than 5S, after which a double would clearly just mean it's our hand.  But it's not totally clear, so the hand on lead could have guessed that partner didn't have a diamond void and led something else, especially since it's not clear where you get a second trick other than SA if partner is ruffing a diamond.  It's best to have as many partnership agreements as possible, but if on uncertain ground it's ok to look at your hand and make a practical guess.  Best to know what partner means and not be in that position! 

BTW you can make 5S on an endplay, so anyone who saves at red/white beats a club leader this time (although I don't think I would.)

Thursday, April 9, 2009

Bidding problems from District GNT

Many bidding problems are clearly a bit of a guess -- if you can reason well about the percentage guess, the long-run rewards are significant. Here are two problems from a close District GNT semifinal which we won last weekend (final at a later date.)

1. You hold Ax AKQT9x xxx Jx. At neither vul., partner opens 3c in second seat, passed to you. Your style is "modern standard." Your call?

2. You hold xxx x Jxx AJ9xxx. White v. red, partner deals and opens 3S -- the auction continues 3S-4H-4S-5S-P-6H-?? Are you bidding?
(stories follow)


#1: I decided opposite most hands we needed at least a club hook, more often needed a hook and heart break, and sometimes had no play for game. The few times game was good didn't seem to compensate and I passed. Partner held xxx x xx AQxxxxx (how traditional) and just made 3C. At the other table my hand bid 3h and preemptor bid 3nt, which even on a non-diamond lead is under 40%. They won the spade lead and lost a club hook -- down 5, we won 8 imps. Needed some luck, but I think I made the percentage guess.

#2: A difficult guess -- you know the save will be super-cheap, but mightn't partner have a stiff club, so you can beat 6H even when the opponents have their bids? An agonizing decision where you can lose 12-14 imps either by a phantom save or by failing to save. I can't tell you for sure what I would have done, because partner did it for me -- holding KQJxxxx xxx x Tx he bid 6S over 5S, so the problem I posed never arose. Against imperfect defense he got out for 500 and we won 14. (His lefty holding Kx clubs needed to pop king to cut communication at a crucial moment -- he led a low club from hand to discourage this, good technique.)

Thursday, April 2, 2009


Have you noticed that your partners and opponents (not you, of course – you are a very sound bidder) seem to be more likely to have 14 hcp than 17 for a 15-17 notrump? This might not be an illusion, and they might not be wild bidders, either.

Let’s assume that 5-3-3-2, 4-4-3-2 and 4-3-3-3, including 5-card majors, are considered balanced. The 5-3-3-2s represent about 1/3 of the balanced hands. I feel safe in saying a typical expert upgrades the majority of these. Of course, you wouldn’t upgrade QJ Jxx Jxxxx AKQ :-) – but you would upgrade AJT9 KTx Kx KT9x or the like. Let’s assume 1/3 of all balanced hands are upgraded, for now. I don’t think this represents wild overbidding. Subjectively, I’m fairly certain there are people who upgrade more than 1/3 of hands, maybe close to ½.

The important objective statistic is that balanced 14s are about 2.3 times more common than balanced 17s. (Thanks to Richard Pavlicek’s excellent webpage for some of these stats; as he notes, the hcp distribution is different for balanced vs. unbalanced hands.) If your opponent (or partner) upgrades 1/3 of the 14s to 1NT, and 1/3 of the 17s to the 18-19 range, then when he opens a 15-17 1NT you should assume he is about 1.15 times more likely to have 14 than 17! If all the 15s and 16s are treated as 15-17, the overall percentages are that a 15-17 NT will contain

14 hcp: 17% 15: 39% 16:30% 17: 14%

Just hypothetically, if someone upgrades half of all balanced hands, the percentages become:

14 hcp: 24% 15: 37% 16:28% 17: 10%

So 14 would be almost as likely as 16, and much more likely than 17.