Free Weekly Instant Tournament - October 14 - Board 8

Board 8
Neither vulnerable

♠ A 6  ♥ 10 9 5  ♦ K 6 5  ♣ A Q J 9 8

LHO passes, partner opens with two spades, and RHO passes.

Playing opposite myself, I would pass. While game is possible, it's remote enough that I wouldn't want to risk going minus at the three-level. With three trumps, going minus at the three-level isn't so bad, since you have the protection of the Law: If you go minus, chances are the opponents can make something. With only two trumps, that is less apt to be true, so it pays to be conservative.

Playing opposite robots, however, passing is less clear. Robots often pre-empt with hands I would open at the one-level, so the chance of game is greater. I have missed games passing robot weak two-bids with worse hands than this.

Ron Klinger, in The Modern Losing Trick Count, recommends that, when holding a doubleton in support of partner's weak two-bid, you should invite game with three and half or four cover cards. My aces and king provide three cover cards, and the club queen provides a half. So this hand is worth an invitation by that standard. Since my record opposite robots is poor in these situations, I'll follow Ron's advice.

I bid two notrump, and partner bids three clubs, showing a maximum with a club feature. If spades are solid, we have eleven tricks. The question is whether the opponents can cash four first.

Let's hope not. I bid four spades and everyone passes.

	NORTH Phillip ♠ A 6 ♥ 10 9 5 ♦ K 6 5 ♣ A Q J 9 8
	SOUTH Robot ♠ K Q 10 5 3 2 ♥ 8 6 ♦ 10 4 ♣ K 10 5

West	North	East	South
Robot	Phillip	Robot	Robot
		Pass	2 ♠
Pass	2 NT	Pass	3 ♣
Pass	4 ♠	(All pass)

Do you want to be in four spades with these cards? Against best defense, game is worse than a finesse, since we might have a trump loser. But in practice it's better than that, because the opponents might lead a black suit or might fail to solve the cash-out problem. It's not clear whether that consideration boosts the odds to better than fifty percent. But I don't mind being in game, so I have no quarrel with Klinger's rule.

Unfortunately, the opening lead is the diamond queen, so the ace is offside. And, since my auction revealed my club king to the defense, they know they must cash four red cards immediately. My only chance to make this contract is to convince the opponents I have a singleton in one red suit so they will try to cash three tricks in the other one. 

What's the best way to do that? Since the opponents know they are facing a cash-out problem, all their signals should be count. If I duck this trick, East should play a count card. If that card happens to be higher than the four and is intended as low, I can obscure the message by dropping the ten. But most of the time his card will be readable. Perhaps it's better to cover the queen, forcing East to take his ace and preventing him from signaling.

The defense should be able to solve their problem even after I cover, but it will be more difficult and the opponents might not be on the same wavelength. Of course, robots aren't on any wavelength. They don't signal at all nor do they draw inferences from the play, so this whole discussion is moot. But, just for practice, I might as well pretend I'm playing against real bridge players.

I  play the king and East takes the ace. Should I drop the ten or the four? The ten tells East I have a singleton or doubleton diamond, since I would have ducked the queen with ten third. The four leaves open the possibility that I have one, two, or three diamonds. The more possible layouts I give East to defend against, the harder it will be for him to find a defense to cater to all of them, so the four is the correct play.

I play the four, and East shifts to the heart ace. So far, so good. I play the six; West, the deuce. East now shifts to the diamond deuce. It's all over. West, seeing my ten, isn't going to try to cash a third diamond, since I can hardly have ten third. And indeed he doesn't. He takes the jack and cashes the heart king, as East follows with the seven.

West continues with the heart jack. East follows with the four, and I ruff. My only chance for a good board is to hope that other declarers will be minus in a partscore. Does it makes sense to finesse the spade, hoping East has jack fourth? Could I conceivably get back to average that way?

That might make sense if I thought the field was in three spades. But it's hard to see three spades being a popular contract. Either you make a game try and get to game or you don't. So the contract will be either two spades or four at most tables. I can neither beat nor tie anyone in two spades, so I'm competing only against other pairs in four. That means I should simply take my percentage play. I play a spade to the ace and back to the king. They split. Down one.

	NORTH Phillip ♠ A 6 ♥ 10 9 5 ♦ K 6 5 ♣ A Q J 9 8
WEST Robot ♠ J 9 4 ♥ K J 2 ♦ Q J 9 8 3 ♣ 6 3		EAST Robot ♠ 8 7 ♥ A Q 7 4 3 ♦ A 7 2 ♣ 7 4 2
	SOUTH Robot ♠ K Q 10 5 3 2 ♥ 8 6 ♦ 10 4 ♣ K 10 5

East's heart ace at trick two was a poor play. If I had three hearts and a stiff diamond, West would have to unblock his heart king to beat me. A better play would be to shift to the heart three (showing an odd number of hearts). West would now know what to do.

This works if West has two or three hearts, but what if he has four? How will he know whether West's odd number of hearts is five (in which case he must cash two diamonds) or three (in which case he must cash two hearts)? He will know because East shouldn't lead low with three hearts. He should cash an honor, allowing West to play a count card. Then he will know himself what tricks to cash. An interesting wrinkle: From ace-queen third, East should lead the queen. If West has king sixth, he knows only one heart is cashing, so he can overtake and play diamonds.

The defense must cater to three critical cases: declarer's holding a singleton heart, a singleton diamond, or two red doubletons. Since signals are binary, you can't cater to three cases by signaling alone. (That's why it was important for declarer to play the diamond four at trick one.) But you can often cater to three cases by combining signaling with logic.

I get 4% for this result, which seems a bit harsh for bidding a decent game. Only two other players reached game, and they both made it. One raised two spades to four and got a club lead. Yes, my auction was revealing. But if I was afraid to invite for fear of tipping off the defense, I think I would prefer passing to blasting a game.

The other player bid two notrump, then three notrump over three clubs and also got a club lead. Given the robots' preference for passive leads against notrump, there is some merit to this decision. I did briefly consider it but decided it was too much of a gamble.