Two passes. RHO opens with 1♠. I don't like passing with an opening bid, but this hand is hardly appropriate for a vulnerable two-level overcall, and I can't double without diamond support.

I pass, LHO bids 2♠, partner passes, and RHO bids 3♠, which they play as invitational. I pass again, and 3♠ ends the auction.

A heart lead looks right. But which one? The ♥2 is the systemic lead. But if partner has the ♥Q, I'd rather lead the ♥K. It might drive an entry to partner's hand, and it will avoid leaving me with the only heart guard in case there are squeezes afoot. If partner has the ♥A and the ♥Q is doubleton, my choice won't matter. If partner has the ♥A and the ♥Q is third, either lead could be right. (It's right to lead low if declarer has ♥Qxx. But it could save a trick to lead the ♥K if dummy has it.) Finally, if partner has neither honor, my choice probably won't matter, since I've made a poor lead anyway. On balance, the ♥K looks better, so that's what I choose.

There is a useful technique for analyzing your defensive prospects that I haven't discussed yet in this blog: matching dummy's cover cards with declarer's losers. On this deal, I can credit South with six losers for his game try. We have no unexpected tricks (ruffs, stray jacks, etc.) So if dummy covers two of declarer's losers, we can't beat 3♠. The ♣A covers one loser. If declarer has ♥Axx, the doubleton heart will cover another (now that I've failed to lead a trump). So, for starters, I must assume declarer does not have ♥Axx. Similarly, if declarer has the ♣Q, the ♣J behind my king, will cover a second loser. So I must assume partner has the ♣Q.

I could reach these same conclusions by constructing possible 3♠ bids for declarer and discarding those where he is cold. But it is much easier and faster this way. With this technique, I know the minute dummy hits that, if we are to have any chance to beat this contract, partner must have the ♣Q and declarer must have fewer than three hearts (or we must have the tempi to stop a ruff).

Declarer plays low from dummy, partner overtakes my king with the ace, and declarer follows with the ♥9. Oops. Leading the ♥K from this hand is risky if partner is inclined to overtake it. I overtook Lowenthal's king lead only once. ("Even if I had king-queen," he explained patiently after the session, "why would I lead the king if I needed you on play at trick two? I'd lead low--or perhaps the queen.") Anyway, I now have a new assumption to add to my list. I must assume partner has the ♥Q. If he doesn't, we've compressed two of declarer's losers into one, which is surely all he needs to make this.

At this point, I can place all the major high cards. If partner has the ♥AQ and the ♣Q, as I must assume he does, then declarer needs the rest--the ♠KQ and the ♦AK--for his 3♠ bid. To get declarer's loser count down to six, I must credit him with a six-card spade suit or a four-card minor. With a four-card minor, he probably would have bid it in preference to 3♠, so it's likely he has six spades.

Partner switches to the ♦5--♦A--♦3--♦2. If declarer had AK doubleton of diamonds, partner would have led the ♦J. So declarer must have ♦AKx, and we must establish our diamond trick and cash it before declarer can pitch it on dummy's clubs. Partner's low diamond suggests he has the jack, so it's probably safe for me to lead diamonds myself when I get in. Partner's carding isn't always reliable, however, so I would rather not lead diamonds myself if I don't have to.

Declarer leads the ♠3. That's good news. If he's drawing trumps, then he doesn't have three hearts. I play the ♠2, declarer plays the ♠7, and partner follows with the ♠6. Declarer plays the ♠8 from dummy, partner pitches the ♥3, and declarer plays the ♠4.

So declarer has six spades. If I could trust partner's heart card, then I would know declarer is 6-2-3-2. Partner isn't big on giving present count, however, so declarer might also be 6-1-3-3. We need three more tricks. They must be either a diamond, a heart, and a club or a diamond and two clubs.

Do I need to shift to the ♦Q, hoping partner has the jack? If declarer is 6-1-3-3, we need two club tricks to beat this. So I don't need to play diamonds yet. We have plenty of time to set up our diamond and cash it. If he is 6-2-3-2, all I have to do to stop the pitch is take the ♠10 off the table (in case declarer has 109 doubleton of clubs). Again, there is no hurry to play diamonds. So I might as well try to cash a heart. I'd just as soon declarer didn't know partner had the ♥Q, however. If he did, he might place me with the ♦Q. So I can't afford to play a low heart.

I play the ♥J, intending to switch to a spade if it holds. It doesn't. Partner plays the ♥4, and declarer ruffs with the ♠J. Declarer leads the ♠9 to dummy's ♠10 as partner pitches the ♥7. Now the ♦4--♦8--♦J--♦Q. I exit with a heart. We still have two club tricks coming. Down one.


The diamond finesse was an error. Declarer should have played a diamond to the king and a third diamond if the queen didn't drop, hoping to set up dummy's long diamond for a club pitch. The ♦Q dropping doubleton or diamonds splitting 3-3 is likelier than the finesse.

Perhaps I misdefended in giving him that option. If I had seen declarer's hand, I would have shifted to the ♣K when I was in with the ♠A. If declarer ducks, I would play another club. Declarer can make his contract by ducking again. But if he thinks I have king doubleton, he will rise with the ace, draw my trump, and take a diamond finesse. 3-3 diamonds is no longer an option, since dummy has no entries.

Is there any way for me to find that defense? Not unless partner cards more helpfully. If he switches to the ♦10 rather than the ♦5, I will know declarer has AKJ. If he then drops the ♥7 (present count) on the second round of trumps, I will be double-dummy. I hate to admit it. But I'm not entirely sure I would find the right defense even then. I would have to appreciate that, if I leave the ♣A in dummy, declarer should play ♦AK rather than take the finesse. I didn't appreciate that as I was defending. But maybe that's only because I had too many things to worry about. Maybe if I knew for sure declarer had the ♦J and if I knew for sure the heart wasn't cashing, it would be easier to focus on declarer's options. Or maybe not. We'll never know. Against this particular declarer, shifting to the ♣K is the wrong defense anyway. If he's going to misplay it, why should I risk giving him the contract if he has the ♣10?

Meanwhile, score one for the offshape take-out double. We're cold for 4♥. I know some readers who are pleased to see that, but I'm not changing my mind about offshape doubles yet. I think partner is closer to doubling 2♠ than I am to doubling 1♠.

Fortunately, our opponents did not reach game either. They also defended 3♠ and beat it a trick. I'm a little surprised at that. Not only is ace, king, and a low diamond declarer's best play in diamonds in most scenarios, it is also fairly easy for the defense to mark West with the ♦Q. If the defense starts with a low heart to East's ace, for example, East is marked with the ♥AQ and, by assumption, a club honor (else declarer has two club tricks). Once he shows up with a singleton spade, the ♦Q is a huge favorite to be offside.

Table 1: +100
Table 2: -100

Score on Board 4: 0 IMPs
Total: +3 IMPs