Sunday, September 26, 2010

Match 2 - Board 22

Board 22
Opponents vulnerable

♠ A 9 7 5 7 4 3 A Q 3 2 ♣ A 4

RHO passes. Three and a half honor tricks is hefty for a weak notrump. But the hand is thin. I think I'm about the ten of spades away from upgrading this hand to a strong notrump. I open one notrump (12-14). Everyone passes. This auction didn't work out too well last time. West leads the ten of hearts.


NORTH
♠ K 8 4 3
J 8 2
10 5
♣ K Q J 9






SOUTH
♠ A 9 7 5
7 4 3
A Q 3 2
♣ A 4



WestNorthEastSouth
Pass1 NT
(All pass)


I play low from dummy. (No need to force East to unblock.) East plays the five, and I play the four. East would have played a higher spot if he had one, so the five is presumably his only spot card. It appears he has two honors third or possibly ace-king-queen fourth. West cashes the heart ace. East drops the queen, and West continues with the nine of hearts to East's king. East shifts to the six of diamonds.

If the king and jack of diamonds are in the same hand, it doesn't matter what I do on this trick. The relevant cases are: (A) East has the king; West has the jack. And (B) West has the king; East has the jack. In case (A), I might make three by finessing the queen. I would still need to find spades three-two and be able to duck a spade to East. In case (B), I can make two by playing low, but I would hold myself to one if I finesse the queen. It appears, then, that it's better to play low. Playing the queen, even when it works, may not gain anything. If spades don't break or if West has the long spade, I'm taking the same number of tricks as I would have had I played low. On the other hand, (A) and (B) are not equally likely. If East has the jack, he might have led it, hoping to find his partner with ace-queen-nine.

I don't know if these two considerations cancel each other out or not, but it must be pretty close. Let's try a different approach. What is my goal? How many tricks do I need to take to get a good board? Some pairs may be in four spades, but I don't care about those pairs. I've already either beaten them or lost to them. My decision can't change that. Other pairs may be in a spade partscore making four on a non-heart lead. I don't care about those pairs either, since I have no opportunity to score 180. I care only about pairs in a spade partscore who do receive a heart lead. For practical purposes, I might as well assume that every pair in the room falls into that category.

In that case, if the diamond king is onside and spades break, everyone will be plus 140. The difference between plus 120 and plus 150 will be a full board. If the cards lie less favorably, then some pairs (those who stop in two spades) will be plus 110; others (those who reach three) will be minus 50. The difference between plus 90 and plus 120 will thus be less than a full board. In short, there are more matchpoints at stake when the diamond king is onside than there are when the diamond king is offside. So, even if I judged finessing the queen was 50-50, my expected gain from finessing is more than my expected loss. Since, in fact, I judge it more likely that East has the king than that he has the jack, it must be right to play the queen.

I play the queen, and West wins with the king. You have probably noticed by now that I if I go to great lengths in this blog to demonstrate that a decision is correct, there is a fair chance that it isn't going to work. West cashes his long heart. My only chance to take the rest now is an unlikely spade-diamond squeeze. I pitch a spade from dummy and the deuce of diamonds from my hand. East pitches the decue of spades. That looks promising. There is no two- or three-card holding where a spade pitch makes any sense. So East should have four spades, which will make plus 90 a fine result.

West shifts to club. I run the clubs, trying for the squeeze. There is no squeeze, because my spades are already good. East's spade pitch was from ten third. In all the random deals East generated to choose his play, he never gave his partner queen-jack doubleton of spades. He probably did give his partner queen doubleton of spades. But, since he gives declarer credit for playing double-dummy in his analysis, he didn't see the value in holding spades when his partner has queen doubleton. This is a flaw in the current state-of-the-art in computer bridge that I have no idea how one can program to correct for.

The full deal:


NORTH
♠ K 8 4 3
J 8 2
10 5
♣ K Q J 9


WEST
♠ Q J
A 10 9 6
K J 8 7
♣ 6 3 2


EAST
♠ 10 6 2
K Q 5
9 6 4
♣ 10 8 7 5


SOUTH
♠ A 9 7 5
7 4 3
A Q 3 2
♣ A 4



As it turns out, the overtrick is immaterial. Every other pair in the room is in four spades. Half of them make it; half of them are down two. So this result is dead average.

It's hard to see why everyone is in four spades. If you open with a strong notrump you will get there--and from the right side, since West is unlikely to lead a heart with ace-ten-nine fourth. But I would think that one diamond--one spade--two spades--pass would be the normal auction. The South hand is not good enough for a raise to three spades, and the North hand is not good enough to move over two spades.

Score on Board 22: +120 (6 MP)
Total: 170 MP (64.4 %)

Current rank: 1st

Sunday, September 19, 2010

Match 2 - Board 21

Board 21
Our side vulnerable

♠ K 6 4 K 5 2 J 3 2 ♣ 7 6 5 3

Partner opens one notrump (12 - 14) in first seat, and RHO passes. I pass, as does LHO. How bad can this be? It seems we've reached a normal contract and given the opponents no information. RHO leads the ace of diamonds.


NORTH
♠ K 6 4
K 5 2
J 3 2
♣ 7 6 5 3






SOUTH
♠ J 10 8 3
A J 10 7
9 4
♣ A Q 8



West North East South
1 NT
(All pass)


I'm not sure what the lead of the ace means conventionally. But, since East plays the five on this trick, it appears West has ace-king-queen. I "discourage" with the four. West continues with the king of diamonds; East plays the seven. West continues with the diamond queen, and East plays the ten. If West began with five diamonds, I can't afford to lose any more tricks other than the spade ace. It appears my seven tricks will consist of two clubs, four hearts, and one spade, so I'm going to pitch one club and two spades from my hand. Does it matter which card I pitch first?

East is going to have to find two discards, and I want to make it as hard as possible for him. Clubs will look like a dangerous suit to pitch from. It could cost a trick to pitch from jack third (I could have ace-king fourth) or from king third (I could have ace-queen fourth). Since I have no choice but to pitch a club from dummy on the next trick, pitching a club from my hand now renders either of those dangerous holdings unlikely and simplifies East's discarding problem. Accordingly, I pitch the three of spades.

West continues with the eight of diamonds, I pitch dummy's three of clubs, and East pitches the three of hearts. Unless he has queen fifth, the heart queen is probably going to be on my left. Although that is hardly a sure thing. It's possible any discard looks dangerous to East. And it's also possible he's being sneaky.

If West does have the heart queen, I hope it's not queen-nine or queen-eight fourth. Otherwise I'm going to take only three heart tricks. I may need an extra spade trick if that's the case, so perhaps I should keep three hearts and three spades, hoping to find queen-nine doubleton of spades onside. I could pitch a club now and, on the next trick, decide which major to hold. But it's hard to imagine what would happen to persuade me to pitch a heart. Rather than leave an option open that I can't imagine taking, I'd rather continue with the illusion that clubs is a source of tricks for me. I pitch the eight of spades.

West leads his last diamond. I pitch another club from dummy, and East pitches the heart eight. I'm certainly playing West for the heart queen now. That means I must play East for the spade ace, else West would have doubled one notrump. With the ace of spades alone, East might have pitched a spade to keep his partner from shifting to a spade from the queen. Failure to do so suggests that East thinks it's possible spades are running. Given I'm looking at the spade jack, that means East should have both the ace and the queen. I hope I'm wrong about that. Since Jack isn't a very sophisticated signaler, there is a chance I am.

I pitch the club eight from my hand. West shift to the jack of clubs, East plays the king, and I win with the ace. Does it matter whether I take the spade finesse or the heart finesse first? Not if the heart finesse is winning. But if it's losing, taking the heart finesse first will at least establish a trick. Can I afford to cash the heart ace in case East is pulling a fast one with queen third? No, I can't. I need the heart ace as an entry to my hand. I lead the jack of hearts--four--deuce--nine. Actually, that was a mistake. I should have led the heart ten, because it's harder for West to cover the ten. I can think of layouts--wildly unlikely layouts--where it's necessary to retain the king of hearts in dummy.

Now the moment of truth. I float the jack of spades. As expected, East has both spade honors, so I finish down one.


NORTH
♠ K 6 4
K 5 2
J 3 2
♣ 7 6 5 3


WEST
♠ 7 5
Q 6 4
A K Q 8 6
♣ J 10 4


EAST
♠ A Q 9 2
9 8 3
10 7 5
♣ K 9 2


SOUTH
♠ J 10 8 3
A J 10 7
9 4
♣ A Q 8



This wasn't a normal contract, since a one club opening gives East a chance to overcall. Minus 100 is worth only one matchpoint. Most pairs are minus 90 against two diamonds. Two pairs managed to beat two diamonds, presumably by cashing their five tricks, then leading the thirteenth heart for a trump promotion. Our opponenets have had several bad results from weak notrumps. Now it's our turn, though it would be a different story if we weren't vulnerable.

Score on Board 21: -100 (1 MP)
Total: 164 (65.1%)

Current rank: 1st

Sunday, September 12, 2010

Match 2 - Board 20

Board 20
Both sides vulnerable

♠ 5 4 K Q J 7 5 3 2 ♣ Q J 7 2

Partner opens two notrump in second seat. I raise to three, and RHO leads the four of hearts (attitude).


NORTH
♠ 5 4
K Q J
7 5 3 2
♣ Q J 7 2






SOUTH
♠ A K Q J
8 2
A K 10 6
♣ K 5 3



WestNorthEastSouth
Pass2 NT
Pass3 NT(All pass)


East plays the three. That should show count, so hearts should be five-three. From West's point of view, the three might be low from three or high from three-deuce doubleton. I play the eight to leave both possibilities open. (You can accomplish the same thing with less effort simply by signaling as East does. He gives count, so you give count. Most of the time, that's the most effective way of scrambling East's message.)

It looks right to work on clubs. If clubs are four-two, I have nine tricks. If the opponents continue hearts, that makes ten. If they don't, I can probably develop a tenth trick in diamonds. Might I have a squeeze if East guards both clubs and diamonds? Suppose I lead a club the the king, then a club back to queen and ace. If East plays a heart and West takes his ace, I have a squeeze. But that's a pretty careless defense. West can stop the squeeze by ducking the heart so the count isn't corrected. Or East can stop it by returning a club instead of a heart or by ducking the club queen. So, if clubs are four-two, I'm probably taking ten tricks.

What if clubs are three-three? Now the defense can't afford to allow me to score a second heart trick. Say I play a club to the king; they duck. I play a club to the queen; they duck again. Dummy is now dead, and I must lose at least one diamond trick in addition to the two aces.

Actually, that would be poor play on my part. Since I must play on diamonds if the club queen holds, I should prepare for that possibility by cashing the diamond ace before playing the second club. Now, if I lead a club to dummy's queen and it holds, I play a diamond toward my hand. If East began with queen-jack fourth, he splits. I win, can cash spades, and exit with a heart. The defense must concede a trick to dummy's heart or to my diamond ten. Again, ten tricks.

Barring queen-jack fourth of diamonds on my left, it appears I'm destined to take ten tricks. Three notrump is not hard to reach, so the entire room should be plus 630 on this board. How am I ever going to get above average?

What if queen-jack third of diamonds is onside? Then I can take eleven tricks double-dummy. But how can I afford to play for that? If I lead a diamond to the ten at trick two and it loses (as it will roughly 75% of the time), the opponents can continue hearts and hold me to nine tricks (or even eight if diamonds don't break). Perhaps I can afford to finesse the diamond if I get one club trick in first. Suppose I play a spade to my hand, then play a low club toward dummy. If West has the ace, he must duck. I play the queen. If East has ace third, he must duck. Now I play a diamond to the ten. As long as diamonds are three-two, I'm in good shape. 75% of the time, I take ten tricks like everyone else (albeit a different ten tricks); the rest of the time I take eleven.

Can East prevent this by splitting his diamond honors? If he does, I win and play the club king. They must duck, else I have an entry to repeat the diamond finesse. Now I cash spades and play a heart. They must either play diamonds for me or give me an entry to dummy so I can play them myself. (Yes, I know. East can give me a problem by playing a diamond honor from honor doubleton or third, but let's see him do it.)

This line does entail some risk. I might hold myself to three. But matchpoints is a game for optimists. You don't win matchpoint events settling for average scores, so I'm going to go for it. I play a spade--eight--king--deuce, then a low club. West hops with the ace and plays ace and a heart. That wasn't exactly what I was playing for, but I'll take it. Making five.


NORTH
♠ 5 4
K Q J
7 5 3 2
♣ Q J 7 2


WEST
♠ 10 9 2
A 9 6 5 4
Q 9 8
♣ A 10


EAST
♠ 8 7 6 3
10 7 3
J 4
♣ 9 8 6 4


SOUTH
♠ A K Q J
8 2
A K 10 6
♣ K 5 3


It's hard to see how this can be the right defense. The only layout I can think of where it's right to hop with the club ace is when I have king doubleton of clubs. But why would I play the hand that way if I did? I wonder if my play at trick one had anything to do with West's error. I replay the hand and drop the heart deuce. This time West ducks the club. I replay it again and again drop the eight. Now West hops. If West knows for sure I don't have three hearts, he gets the problem right. If he doesn't know how many hearts I have, he thinks he has a guess. He thinks I could be either two-three or three-two in the round suits, and, for some reason, he prefers to play me for the latter.

Since Jack draws no inferences from declarer's line of play, he has no way of solving this problem. As we've seen in the past, playing the card that keeps Jack in the dark can reap huge dividends. It reaps these dividends less often against humans, since humans have other clues they can rely on. But playing the right card is important against humans, too. The more layouts your opponents have to worry about, the harder they have to work. And the harder they have to work, the more likely they are to make a mistake. It might not even happen on this deal. They might make a mistake later simply because you've worn them down.

One clown played five diamonds, down one. Everyone else was in three notrump making four. Plus 660, as we might expect, is a top.

Score on Board 20: +660 (12 MP)
Total: 163 MP (67.9 %)

Current rank: 1st

Sunday, September 5, 2010

Match 2 - Board 19

Board 19
Opponents vulnerable

♠ 9 8 5 Q 10 7 5 K ♣ A 10 8 7 5

I pass in first seat. One diamond--pass--one spade to me. I double. Jack disapproves of this double, claiming it shows a better hand. Nonsense. A singleton, nine cards in the unbid suits, and eight losers? I'd be happier if the king of diamonds were the deuce, but you can't have everything.

LHO bids two diamonds. The opponents are playing support doubles and redoubles, so this bid should deny three spades. Personally, I think that a non-jump rebid of your suit should deny three-card support for partner even without competition. Jack does not seem to agree, however. (Perhaps that's a reason not to double. If your opponents play more sensible methods with competition than without, perhaps you should stay out of their auction.) RHO bids three hearts. I'm not sure why. Perhaps two hearts would have been non-forcing. LHO bids three notrump, and RHO bids four spades, ending the auction.

RHO is probably 6-4-2-1. With the king of diamonds onside, I'm in danger of losing the club ace if I don't cash it. But is that the best lead? If partner's diamonds are good enough to keep the suit from running, it might be right to lead trumps to protect my heart tricks. This would be a tougher decision if dummy could have three spades. In that case, it might be right to lead a trump at trick one, retaining the club ace as an entry. Declarer would win the trump lead and ruff one heart, but would be unable to return to his hand to ruff a second heart. When I got in with the club ace, I could remove dummy's last trump. But, since dummy holds at most two trumps, I probably don't need to lead trumps at trick one. I can always cash the club ace, then switch to a trump at trick two, holding declarer to one heart ruff. Failure to lead a trump at trick one will cost only if the club ace gets ruffed or if we have a heart entry, which would enable us to prevent even a single ruff.

I lead the club ace and see the following dummy:


NORTH
♠ K 10
8 3
A Q J 10 6 2
♣ Q 9 2


WEST
♠ 9 8 5
Q 10 7 5
K
♣ A 10 8 7 5




West North East South
Pass 1 Pass 1 ♠
Double 2 Pass 3
Pass 3 NT Pass 4 ♠
(All pass)


I don't care for the three notrump bid. I would have bid three spades. One of the reasons to play that two diamonds denies three spades is so that you can give a preference with a doubleton. King-ten doubleton of spades looks like a more important feature to me than queen-nine third of clubs, especially when partner might have club shortness.

Partner plays the club three, and declarer plays the king. I have two attractive defenses. I can lead a heart, hoping partner has the heart ace, or I can lead a diamond, hoping partner has the trump ace. Does the latter defense even work? Declarer can win the diamond ace, then pitch his remaining diamond on the club queen to stop the ruff. But he is now left with two heart losers and no way to dispose of them. If he ruffs them in dummy, he can't stop me from scoring a trump trick.

Which ace is partner more likely to have? His discouraging club suggests tolerance for a heart shift. At least it would if I could trust him. With the spade ace and nothing in hearts, partner should encourage at trick one to say that the "obvious shift" is the wrong defense.

If you'll allow me to hop on the soap box for a moment, here's a good example of how suit preference at trick one can get confusing. With queen third of clubs in dummy, I think most people would play partner's card as attitude: high to suggest a club continuation and low to suggest a heart shift. But what if dummy had the club king? In that case, some would play East's card as suit preference. Attitude makes no sense, they say, when no more club tricks are available. So partner would play a high club with the heart ace and a low club (asking for a diamond shift) with the spade ace. But who says a club continuation makes no sense just because you have no more club tricks? To give just one example: what if partner has the diamond king instead of you? If declarer has a singleton diamond and a doubleton club, a club continuation at trick two, killing the entry to the diamonds, might be necessary. How does partner suggest that defense if his card is suit preference? Should his card be attitude if continuing the suit to kill an entry is an option but suit preference if it isn't? So he plays low to get a heart shift in the first case and high to get a heart shift in the second? Why make things so hard? Why can't a low club always ask for the "obvious" heart shift, while a high club says do something else. Partner may not always be sure what that "something else" is. But most of the time he can work it out based on other clues.

Jack's way of signaling attitude, unfortunately, isn't very helpful. His signal pertains to the suit led, not to the "obvious shift" suit. So all I know from Jack's low club is that he doesn't have a high club honor. To solve my problem, I must rely on the a priori odds. Partner has three sevenths of the outstanding hearts and a quarter of the outstanding spades, so he is more likely to have the heart ace than the spade ace. Accordingly, I shift to the five of hearts.

Partner plays the king, and declarer wins with the ace. I guess my shift didn't matter. If partner has the heart king, he is unlikely to have the spade ace. Declarer plays a spade to the ten, which holds. At that point, I can simply concede:


NORTH
♠ K 10
8 3
A Q J 10 6 2
♣ Q 9 2


WEST
♠ 9 8 5
Q 10 7 5
K
♣ A 10 8 7 5


EAST
♠ 3 2
K 9 6
9 8 7 4
♣ J 6 4 3


SOUTH
♠ A Q J 7 6 4
A J 4 2
5 3
♣ K



The opponents came awfully close to missing a good slam. Make South's club king the diamond king, and slam is virtually cold. I doubt South would have bid any differently.

Let's change South's club king to the diamond king and see how the auction should go. It would be easier if South could bid two hearts over two diamonds. But we have to accept South's three heart bid as a systemic given. As we've already said, North should bid three spades, not three notrump, over three hearts. At that point, it seems fairly routine for South to bid four diamonds. But that's Humpty Dumpty syndrome. Since South is looking at a hand where he would like four diamonds to show a diamond card and slam interest, it's easy for him to forget that's not what the bid means. South knows the partnership has found a trump suit, but North doesn't. So four diamonds simply continues the search for a playable strain.

♠ J x x x x A K Q x K x x ♣ x

is a possbile hand.

That's not to say you can't bid four diamonds to probe for slam. But it does mean partner isn't going to react the way you would like him to. He will not be cooperating with your slam probe; he will be selecting what looks to him like the best trump suit. If you think that information will help you in your decision, go ahead and bid four diamonds.

Personally, I don't think partner's reaction to four diamonds will be of much help. In fact, I have hard time imagining any sequence where you can get meaningful cooperation from partner. You have a one-loser trump suit, a fitting card in partner's suit, and sufficient controls in the other two suits. Any slam try you devise will make it sound as though you're missing one of those features. As a practical matter, I suspect it's best simply to bid Blackwood and drive to slam opposite two key cards. All you need in addition to two key cards is the diamond queen or the heart king. That doesn't seem like too much to hope for.

We get ten matchpoints for minus 680. Since three notrump making six was a popular spot, the opening lead was crucial. We would have received two matchpoints had I not led the club ace. We are back in the lead.

Score on Board 19: -680 (10 MP)
Total: 151 (66.2%)
Current rank: 1st place