Wednesday, February 27, 2013

The hitchhiker's guide to Bridge: part 1

The instruction DON'T PANIC was written in large letters on the front page of the Hitchhiker's Guide to the Galaxy, according to Douglas Adams in the novel of that name (1979). It is also a good thing to remember at times when playing bridge.

There are probably several different aspects of the game where the exhortation is relevant. One is defending a contract with a good side suit. There is a tendency to panic unless we have a good count on declarer's hand. Another common situation is defending a doubled part score, especially when partner is the chief hawk and our hand is dove-like.

The situation that is the subject of this blog occurs during the auction. It's a common mistake of beginners in bridge to keep running from doubled contracts. Just the other day, I doubled a contract at the two-level (we were slated to earn 500) but the opponents panicked and ended up in the exact same suit, but one level higher. We enjoyed our 800.

Here's a situation where the BBO robots demonstrated that they too are subject to this type of panic attack. The hand occurred in a robot tournament, costing me a decent showing. To add insult to injury, the tournaments now require the human to play all the hands so I was the one playing the ridiculous contract. Assuming that 4 gets doubled it would have been worth approximately 12 IMPs to us (-100 against a vulnerable game). 4♠ doubled should also have been -100 although the two times it was played resulted in -500. That would have been worth 4.8 IMPs. The actual contract of 5X (-800 although I could have escaped for -500) was worth -4.1 IMPs.

It was all so unnecessary – and a direct result of panicking. 4 hadn't even been doubled yet. You would think that my robot partner would be happy to have three cards in each of my announced suits. But no, he bids his own suit. When that gets doubled (surprise), he pulls to the strain that he previously eschewed. That gets doubled and he pulls yet again.

Are there any lessons to be learned (by the robot) here? The first lesson is that when partner has promised a two-suiter and voluntarily bids one of the suits "to play", then that suit is at least as good as the other suit and will usually be better. How do we know it's "to play?" In this case, it's clear from the auction that the opponents hold most of the cards and that the robot is broke. A raise would be unthinkable and a preference is only desired with significantly better hearts than diamonds.

The second lesson is that three card support (here in each of partner's suits!) is very good to have. Don't try to bid your own suit, unless it's at least eight cards long!

The third lesson is that once the opponents start doubling, they are likely to double everything in sight. The fact that they doubled 5, doesn't in any way suggest that 5 would play better. And that's going to be doubled too.

Don't get me wrong. I enjoy playing these robot tournaments. The robots are patient, they are fast. They never criticize my inept play. Occasionally they go off the rails, as above, but mostly I love them for being competent, if not brilliant, and dependable partners.


  1. I think I see another example of why I win so few club games ...

    I would not have valued the East hand as close to a 15-17 1NT opening. And if I were South, I would have passed over a strong 1NT. In the end, I am sure that opposite a bunch of me's, 3NT would have been declared and, assuming the North me knows enough to discard all of his cards in one red suit (so that South me has a count on the hand of East), I would expect the defense to be simple: High diamond winning, switch to a black suit, sit back and watch declarer run nine black suit tricks and then win the last three tricks to go along with Trick 1. How can one win a mp event with a boring +-600?

    1. Some interesting thoughts here, Jeff. First, I'd point out that -600 (if you were N/S) would have been a good score (most were -630), at least at matchpoints (this was IMPs).

      But your other comments about opening 1NT and/or intervening made me go back and look at the board. It was played 14 times. In all 14 cases, East opened 1NT.

      In 8 of those auctions, South intervened. The system being Cappelletti here, there was one double (penalty which didn't work out well), a couple of 2C (single-suit) and the rest bid 2H (H+min). Of those 8, only two of us bid a second time (with identical results).

      So, my thoughts: While I am a big fan of upgrading/downgrading suit bids based on texture, I very rarely do it for balanced notrump hands. It's just impossible to intelligently judge which honors will be useful opposite partner's hand. I might upgrade a 14-count to 15 with a decent or better five-card suit and generally good intermediates.

      Now for the intervention. I can't imagine not intervening at these colors. As to whether it was sane to bid 4D I will leave up to readers. The one "formula" that I know for the initial intervention is Mel Colchamiro's rule of 8. Add the lengths of your two suits and subtract the number of losers in your hand: in this case 9 - 6 = 3. Provided that comes to at least 2 and you have at least 6 hcp (as here), you should intervene.

      Personally, I never use that formula at the table because it doesn't take into account vulnerability, texture, other factors but all roads lead to intervention here.

    2. Kaplan and Rubens Evaluator scores the East hand as 13.25. Sounds about right to me.

      Less comfortable about my judgment about the one round intervention (4D is not worth addressing). But with a strong hand and no singleton (meaning lots of defense), why do I need to intervene now? Not familiar with the rule you mention but as it seems to treat 1=4=5=3 the same as 2=4=5=2 and as it seems to treat KQJx/KQJxx the same as AQJx/AKJxx, I am not anxious to learn the rule either.

  2. They never criticize my inept play.

    I had a robot mumble something under his breath once.