Sunday, February 26, 2023

Free Weekly Instant Tournament - February 24 - Board 3

Board 3
Opponents vulnerable

♠ A K 8 7  ♥ A Q 10 5  ♦ K 8 4  ♣ 6 4  

I open with one notrump in first seat. Everyone passes. I like this auction. One notrump is frequently a difficult contract to declare. And in this event, where your system is foisted upon you, everyone should be in this contract. So this deal should offer a good opportunity to outshine the field.

LHO leads the eight of clubs.


NORTH
Robot
♠ Q 9 6
♥ 7 3
♦ Q 10 6
♣ K 10 9 7 2






SOUTH
Phillip
♠ A K 8 7
♥ A Q 10 5
♦ K 8 4
♣ 6 4


West North East South
Robot Robot Robot Phillip



1 NT
(All pass)


The lead is probably top of a doubleton. I have three spade tricks, one heart, and one diamond, bringing me up to five. I need two more to make this. There are prospects for more tricks in all four suits. That's what makes playing one notrump challenging.You frequently have lots of options with little information at the start. Your goal is to find a flexible line that leaves as many options open as possible until you get a better picture of the layout.

There's no point in thinking about this too hard now. The defense is in charge for the moment. Let's see what they do.

I cover the eight with the nine, and East wins with the jack. He shifts to the deuce of diamonds.

That was unexpected. A heart shift looks more natural. Diamonds must look like a promising source of tricks to East. He probably has ace fourth or fifth. He might have the jack as well, especially if he has five. In this layout, switching from ace-jack fifth gives me a second diamond trick I wouldn't have managed by myself. But his partner might have the king, or I might have king doubleton. Also, giving me a second diamond trick might gain in the long run if the key to the defense is running the diamond suit.

I play low, West plays the nine, and I win with dummy's ten. I'm up to six tricks now. My first instinct is to play a heart to the ten. If that drives the king, I've made my contract. If it loses to the jack, I can decide whether to use my entry with the spade queen to repeat the finesse or to try for four spade tricks.

But maybe I can do better. What happens if I play on clubs? If East wins with the queen and plays another diamond, he gives me a dummy entry. I can now drive the club ace, giving me two club tricks. Even if East began with five diamonds, the defense can take only three diamonds and three clubs, so I've made my contract. If he began with four diamonds, I'll make an overtrick.

More likely, East will win the club queen and shift to a heart. Then I'm no worse off than if I had played a heart myself. In fact, I'm better off, since I've cut the opponents' communications. Having stripped West of his club exit, I may be able to endplay him.

The way playing a club could work out badly is if East wins and shifts to a spade. Now I lose my chance to take two heart finesses. But, looking at that dummy, he isn't apt to shift a spade. If he does, I'll worry about what to do at that point. I'll still have a variety of options.

I play a low club from dummy toward my six. East hops with the queen, and West follows with the three. As expected, East shifts to a heart, the four. I finesse the ten. West wins with the jack and shifts to the three of diamonds. 

Here is the current position:


NORTH
Robot
♠ Q 9 6
♥ 7
♦ Q 6
♣ K 10 7






SOUTH
Phillip
♠ A K 8 7
♥ A Q 5
♦ K 8
♣ --

I would like to win this in dummy so I can continue clubs. What's the best way to do that? If I play the queen and East began with ace fourth, he may duck to preserve communication. Then I can continue clubs. If, instead, he takes his ace, I can no longer set up clubs. But, thanks to my club play at trick three, I may now be able to endplay West.

Say, for example, East takes the diamond ace and continues diamonds. I win in my hand and play ace of spades (unblocking the nine) and a spade to the queen, reaching this position:


NORTH
Robot
♠ 6
♥ 7
♦ --
♣ K 10 7






SOUTH
Phillip
♠  K 8
♥ A Q 5
♦  --
♣ --

If West has played low spades, I play a spade to the king. If spades break, the long spade is my seventh trick. If West has long spades, I toss him in to force a heart return, reaping the reward from my club play at trick three.

If West has played an honor, I play a spade and, if East follows, finesse the eight. If I lose to jack-ten-third, West is still endplayed.

More likely, if East chooses to win the diamond ace, he will shift to a heart. I must then decide who has the heart king. If it's East, I must finesse. If it's West, I must hop, cash the diamond king, and execute the same endplay. I can make that decision later. Let's see what happens on this trick.

I play the diamond queen. East ducks. Driving the club should work now. Unless, that is, East started with five diamonds. Then the defense will take three diamonds, three clubs, and a heart for down one. If that's the case, I need to abandon the club suit and go for the endplay.

Is that possible? If East started with five diamonds, he knows the diamond suit is dead. The natural play is to win and continue hearts. It would be quite diabolical to duck, assuming that would entice me to continue clubs. A robot, who assumes declarer is double-dummy, would never do that. And there are few humans who would be capable of it. So I'll stick with my plan.

I play a club. East wins and cashes two diamonds. Making one.


NORTH
Robot
♠ Q 9 6
♥ 7 3
♦ Q 10 6
♣ K 10 9 7 2


WEST
Robot
♠ 10 3
♥ K J 9 8 6 2
♦ J 9 3
♣ 8 3


EAST
Robot
♠ J 5 4 2
♥ 4
♦ A 7 5 2
♣ A Q J 5


SOUTH
Phillip
♠ A K 8 7
♥ A Q 10 5
♦ K 8 4
♣ 6 4

+90 is worth 79%. Almost every declarer played a heart at trick three. Some went down, some made it, depending on how they proceeded after that. A heart is certainly the intuitive play. But I think a little reflection will reveal that attacking clubs and retaining your flexibility is a better approach.

Sunday, February 19, 2023

Free Weekly Instant Tournament - February 17 - Board 2

Board 2
Our side vulnerable

♠ A 6 2  ♥ A 8 7 6  ♦ K 9 3  ♣ A 4 2  

Pass on my right. I open with one notrump. Partner bids four diamonds, a transfer to hearts. I bid four hearts and everyone passes. LHO leads the king of clubs.


NORTH
Robot
♠ Q 9 3
♥ Q 10 9 5 3 2
♦ A
♣ J 10 9






SOUTH
Phillip
♠ A 6 2
♥ A 8 7 6
♦ K 9 3
♣ A 4 2


West North East South
Robot Robot Robot Phillip


Pass 1 NT
Pass 4 ♦ Pass 4 ♥
(All pass)


Thanks to this lead, I have only one club loser. I can pitch one of dummy's spades on my diamond king, leaving me with one spade loser. So to make this, I need to hold my hearts losers to one.

I can guarantee at most one heart loser by playing a low heart from dummy and covering whatever card East plays. But there are two problems with that line. (1) I lose a trick unnecessarily if West has a stiff king. And (2) If West wins with a stiff king or jack, he may be able to give his partner a club ruff. It would be embarrassing to go down in a cold contract by taking a "safety play."

At IMPs, I wouldn't care about (1), but (2) would be a concern. A stiff king or jack in West's hand is a twice as likely as a void. So if I judge the risk of a club ruff is more than 50%, the deep finesse is wrong. 

Since I'm playing matchpoints, I don't need to bother making that judgment. I expect everyone to be in game. So consideration (1) all by itself makes cashing the heart ace marginally better than the deep finesse. Once you consider (2) as well, it's not even close. Cashing the heart ace is the percentage play.

The next thing to consider: Is there is any way to avoid the spade loser? Can I strip the hand and endplay somebody?

Suppose I win with the club ace, play a diamond to dummy and a heart to the ace. Everyone follows low. I cash the diamond king, pitching a spade from dummy, and ruff a diamond. For the endplay to work, I need the hand I toss in to have the spade king. But I also need him to have no small club to exit with.  I have two ways to play for such a layout. I can exit with a club, hoping West began with king-queen doubleton, or I can exit with a trump, hoping East wins the trick and began with a singleton club. Neither seems likely, but it's worth a shot.

I play a low club from dummy. East plays the club seven, and I win with the ace. I play a diamond to dummy--deuce--ace--queen. Now a low heart. East plays the four, I play the ace, and West discards the club three.

So much for the percentage play. Now I have two heart losers, so I need the endplay to make my contract. I cash the diamond king--eight--spade three--diamond five. I ruff a diamond in dummy. West plays the six; East, the ten. This is the position


NORTH
Robot
♠ Q 9
♥ Q 10 9 5
♦ --
♣ J 10






SOUTH
Phillip
♠ A 6 2
♥ 8 7 6
♦ --
♣ 4 2

West, holding king-queen third of clubs, would not have pitched his only safe exit card, so I can't endplay West. My only chance is that East began with a stiff club. I play a trump. East cashes his two trumps and exits with the diamond jack. I ruff in my hand, pitching a spade from dummy, and drive the club queen. Making four.


NORTH
Robot
♠ Q 9 3
♥ Q 10 9 5 3 2
♦ A
♣ J 10 9


WEST
Robot
♠ J 10
♥ --
♦ 8 7 6 4 2
♣ K Q 8 6 5 3


EAST
Robot
♠ K 8 7 5 4
♥ K J 4
♦ Q J 10 5
♣ 7


SOUTH
Phillip
♠ A 6 2
♥ A 8 7 6
♦ K 9 3
♣ A 4 2

Making four hearts is worth 93%. Most of the field is in four hearts going down, which isn't surprising. You have to have the foresight to cash the diamond ace at trick two. If you cash the heart ace first, you no longer have the entries to strip the hand.

I was lucky to find East with a stiff club. But good luck does you no good if you don't set yourself up to take advantage of it. A friend once told me that's a valuable life lesson that she learned from bridge.

Two declarers made four hearts, but neither with the line I took. One declarer took the pseudo-safety play in the trump suit. When that worked, she should have made an overtrick on the endplay. But, perhaps blinded by her elation at getting the trumps right, she failed to see the endplay and made only four.

The other declarer tried an interesting gambit. She ducked the opening club lead. In theory, the duck gives away a club trick. But how can West play you to have done such a thing? Looking at jack-ten-nine of clubs in dummy, he will surely conclude his partner has the ace and will continue the suit, giving the trick back.

What, then, have you gained? If the other cards are favorable, you can now endplay East when he has a doubleton club, which is twice as likely as a singleton. I like this line. I wish I had thought of it.

West did, in fact, continue clubs. When East ruffed, declarer must have had a moment of panic. Unless the ruff was with a natural trump trick, the gambit had backfired. Fortunately it was (at least on a normal handling of the trump suit), so the duck broke even. Since the endplay was intact, declarer made the game. (Actually, East exited with a spade, giving away the contract immediately. But I assume declarer would have executed the endplay had the defense put her to the test.That was obviously her plan, else the duck at trick one made no sense.)

I left unanswered the question of how to play the heart suit at IMPs. The critical question is, if West has a stiff king or jack of hearts, what is the chance of running into a club ruff? If it is better than 50%, the "safety play" isn't safe.

If we assume West has the club queen, there are five ways for him to hold six clubs, ten ways to hold five, ten ways to hold four, five ways to hold three, and one way to hold two. So at a first approximation, the chance of a club ruff is 15 to 16. It's actually a little less than that, since these cases are not equally likely. But there is another factor to consider. King-queen third is not an attractive lead. Even king-queen fourth without the ten or nine isn't especially attractive. If we factor in that consideration, cashing the trump ace becomes a stand-out at IMPs as well.

Sunday, February 12, 2023

Free Weekly Instant Tournament - February 10 - Board 1

Board 1
Neither vulnerable

♠ 3 2  ♥ K J 8 5  ♦ A Q 10  ♣ A 10 7 3  

Two passes to me. I open with one club. LHO passes, partner bids one heart, and RHO overcalls with one spade.

I bid two hearts, showing four-card support. In the days before support doubles and redoubles, I would have bid three hearts. After interference, two hearts showed a minimum three-card raise, three hearts showed a minimum four-card raise, and a cue-bid (or two notrump over a double) showed a hand that would have bid three hearts without competition. The support double is an improvement. It allows you distinguish between three- and four-card raises while keeping the auction low, and, perhaps more importantly, it allows you to show a three-card raise when you hold better than a minimum.

Over two hearts, LHO raises to two spades, partner bids three clubs, and RHO bids three spades. With a maximum in high cards and no spade wastage, I have a clear acceptance of partner's invitation. How should I accept? We are both limited, so we know we don't have values for a slam. But it's important to prepare for a four-spade sacrifice, so I need to make the most descriptive call I can to aid in that decision. With this hand, I have no feature to emphasize. I'm balanced with scattered values. The most descriptive call I can make is a simple four hearts.

I bid four hearts, LHO bids four spades, and partner passes. Is his pass forcing? I'll give my usual answer to this question: Who cares? If partner thinks it's right to defend, he will double. If he doesn't want to stop me from bidding on if I'm so inclined, he will pass. In this auction, he will probably double any time he has two or more spades and pass when he has spade shortness. Since he is limited by his failure to open, if he passes I will have a fair picture of his hand and will be well placed to make the right decision. While I'm unlikely to choose to defend four spades undoubled, there is no reason I shouldn't do so if it looks right.

I think we tend to get hung up on whether a high-level pass is forcing or not when it often doesn't matter. Double shows better defense than offense and pass shows the opposite. What more do you need to know? If it's not clear from the auction that it's our hand and the opponents are saving, then we don't want pass to be forcing. And if it is clear, why do we need an agreement? Partner won't sell undoubled even without one.

Of course if you invert the meaning of pass and double, as some do, then you need a clear and unambiguous understanding of when pass is forcing, so you will know when this inversion applies. To my mind, that's a good reason not to play that way. No matter how clear you think your agreements are, there will always be fringe cases you didn't think of. While I can see an advantage of such an inversion, I'm not convinced the advantage is sufficient to compensate for the risk of a misunderstanding.

In any event, what should I do over four spades? With seven losers, it's hard to see taking eleven tricks opposite a passed hand. With a flat hand and scattered honors, I have a much better hand for defense than for offense. It's clear to double.

Double ends the auction. Now what should I lead? We appear to have all three side suits under control. If declarer has no source of tricks, there is no need for an active defense. My goal on the opening lead is to avoid giving away a trick. A trump lead looks like the best way to accomplish that. So I lead the deuce of spades.


NORTH
Robot
♠ Q J 8 6
♥ A 4
♦ 9 8 6 2
♣ J 6 4


WEST
Phillip
♠ 3 2
♥ K J 8 5
♦ A Q 10
♣ A 10 7 3






West North East South
Phillip Robot Robot Robot


Pass Pass
1 ♣ Pass 1 ♥ 1 ♠
2 ♥ 2 ♠ 3 ♣ 3 ♠
4 ♥ 4 ♠ Pass Pass
Double (All pass)

What do I know about the layout? Declarer probably has six spades for his three-spade bid, so partner appears to have the singleton his auction suggested. It is likely to be a singleton honor, since South did not open with two spades. Declarer has at most three hearts and at most three clubs (since partner would not bid three clubs with fewer than three).

Declarer plays a low spade from dummy and partner wins with the ace. Partner shifts to the diamond four.

Why not a heart? With four small hearts, partner would probably lead one to get me off a potential end play. And with any five-card holding, he would want to establish and cash our heart trick as soon as possible. Not doing so risks declarer's later playing ace and a heart for a throw-in. The fact that partner didn't make the obvious heart shift suggests the shift is dangerous from his point of view.

What holding would make it dangerous? A shift from queen fourth might enable declarer, with jack-ten third, to set up a heart trick for a club pitch. Similarly, a shift from ten fourth might give declarer a trick if he has king-jack-nine. I can't see how a shift from any five-card holding would hurt, so I suspect partner has one of those two holdings. Also, he appears to be unconcerned about dummy's fourth diamond becoming a winner. Perhaps he knows that's not a danger. Perhaps he is 1-4-4-4, making declarer 6-3-2-2.

Declarer plays the diamond seven, and I win with the ten, Declarer surely would have played the diamond king if he had it, so partner has that card. That's seven HCP accounted for. 

Is there any reason not to continue with a passive defense? Can declarer possibly take a pitch somewhere? Suppose declarer holds this hand:

♠ K x x x x x  ♥ x x  ♦ J x x  ♣ K Q

If I don't play a heart now, our heart trick goes on dummy's jack of clubs. But why would partner switch to a diamond on that layout? If my deduction that partner has only four hearts is correct, I don't think any tricks can go away. So I continue with another trump.

Declarer wins in dummy, and partner pitches the five of diamonds. Partner would not pitch from four diamonds, allowing declarer to establish dummy's long diamond. So partner must be 1-4-5-3, making declarer 6-3-1-3. It also means partner has the diamond jack. With that pattern, partner is limited to 10 HCP given his failure to open, so he has at most the heart queen or the club queen.

Declarer plays a diamond from dummy and ruffs it. I unblock the ace to maintain flexibility. If declarer had the king-queen of clubs, he would simply drive the club ace. The fact that he has set about a strip suggests partner's queen is the club queen. Declarer must have queen third of hearts and king third of clubs. Not bad. I've worked out the entire layout. If my inference at trick two is correct, I can even place partner with the heart ten.

Declarer plays the six of hearts--eight--ace--three and ruffs another diamond. He now leads the nine of hearts. I play low, and partner wins with the ten.

If partner leads a third heart to tap dummy, declarer can lead dummy's last diamond and pitch a club, endplaying partner. Partner must tap declarer with a diamond to prevent that.

Partner agrees. He exits with the diamond king. Declarer ruffs and leads the heart queen. I cover and declarer ruffs in dummy. We have reached this position:


NORTH
Robot
♠ J
♥ --
♦ --
♣ J 6 4


WEST
Phillip
♠ --
♥ J
♦ --
♣ A 10 7


EAST
Robot
♠ --
♥ x
♦ --
♣ Q x x


SOUTH
Robot
♠ K
♥ --
♦ --
♣ K x x

There is nothing we can do. Declarer can simply duck a club. Whichever hand wins this trick is endplayed.

Declarer leads a club to the nine. Henry Bethe would let the nine hold and laugh. But I don't have the confidence to make plays like that. ("Oh, you had the king of clubs, partner? Sorry. I must have miscounted.")

I win with the ten and cash the club ace. Down two.


NORTH
Robot
♠ Q J 8 6
♥ A 4
♦ 9 8 6 2
♣ J 6 4


WEST
Phillip
♠ 3 2
♥ K J 8 5
♦ A Q 10
♣ A 10 7 3


EAST
Robot
♠ A
♥ 10 7 3 2
♦ K J 5 4 3
♣ Q 8 2


SOUTH
Robot
♠ K 10 9 7 5 4
♥ Q 9 6
♦ 7
♣ K 9 5

Was four hearts making? If South leads his singleton diamond and partner wins and leads a heart to the jack, he'll make it. If he leads a heart to the king, he'll go down. Nobody played four hearts, so it's hard to say what would happen.

Plus 300 is worth 93%. Four spades doubled was a popular contract, but most defenders either led a heart, giving away a trick immediately, or led the club ace and continued clubs. There is no need for an aggressive defense when you have all the side suits bottled up. 93% is a generous reward for an opening lead that should be routine.

Sunday, February 5, 2023

Free Weekly Instant Tournament - Februrary 3 (Mendelssohn's Birthday) - Board 8

Board 8
Neither vulnerable

♠ K 10 4  ♥ A K 3  ♦ A 6 5 3  ♣ 8 6 5  

Three passes to me. I have only 14 HCP, but I do have three and a half honor tricks, which makes it a borderline one-notrump opening. Still, given the "4333" pattern and the absence of texture, I don't think the hand quite makes it across the border. So I open with one diamond.

Partner responds with one notrump and I pass, ending the auction. RHO leads the ten of hearts.


NORTH
Phillip
♠ K 10 4
♥ A K 3
♦ A 6 5 3
♣ 8 6 5






SOUTH
Robot
♠ A Q 9
♥ J 7 6
♦ 9 7
♣ Q J 10 3 2


West North East South
Robot Phillip Robot Robot


Pass Pass
Pass 1 ♦ Pass 1 NT
(All pass)


I have six cashing tricks. I can possibly develop three more in clubs to make nine, although I will have to give up the lead twice to do so. How many tricks can the opponents take if I give up the lead twice? Without knowing who has the heart queen, how diamonds split, or how the club honors lie, it's impossible to say. But let's make a guess. 

My first decision will be whether to hop with the heart ace or duck. For now, let's say I hop. I play a club. They win and play a diamond. I win and play another club. They win and cash however many diamonds they can. If they can cash at least three, they are holding me to at most eight tricks, so I am pitching winners as they run diamonds. This is important to note, because it means ducking the heart can't gain in this scenario. I already have more tricks than I can use, so scoring the heart jack won't matter. If the opponents can take at least three diamond tricks, my right play at trick one is to rise with the heart ace.

Can they? Not necessarily. In fact, if diamonds are five two and the hand with long diamonds has a stiff club honor (or neither club honor), I can hold them to one diamond trick. But most of the time they will be able to take at least three. So hopping looks like the right play. 

Except I didn't ask the right question. It's not "can they take three diamond tricks?" It's "will they?" They can't see my hand. And a diamond shift may not look all that attractive.

If the defense isn't going to find a diamond shift, I might as well duck at trick one. If I duck and clubs come home, I'll take ten tricks if the lead was from the queen and nine if it wasn't. If I hop, I may end up with only eight tricks. Say West has queen-ten-nine fifth of hearts. I rise and play a club. East wins and plays a heart through my jack. If West has the other club honor, they take three hearts and two clubs--five tricks without ever touching diamonds.

So it appears my best strategy is this: Hop if they are going to find a diamond shift; duck if they aren't. Will they find it? A lot depends on their precise diamond holdings. The shift will be easier to find from a sequence than from a broken suit. But diamonds is dummy's longest suit after all. And dummy's three small clubs may make clubs an attractive shift to East. If East wins the heart queen at trick one, he may lead clubs for me. Rightly or wrongly, I decide to duck. East plays the four, and I win with the jack. 

Hurdle one cleared. But I still have to hope they don't find the diamond shift. If they do, my extra heart trick may be worthless. 

What's the best way to attack the club suit? If East has a stiff club honor, I must get to dummy to lead a club. But how? Getting to dummy with a heart risks letting West establish his heart suit. Getting to dummy with a spade exposes the fact that spades aren't a threat and may make it easy to find the diamond shift. 

At least it would playing against a human. A human might decide I would never play a spade unless I had both missing spade honors. But robots don't think that way. A robot will think ace third of spades is a possible holding for me. So releasing dummy's spade king actually makes it more likely that a robot will see spades as an attractive shift.

A spade to the king, then, serves two purposes: It enables me to make my best play in clubs, and it serves as an "idiot's delight." (Howard Chandross' name for the tactic of gratuitously cashing winners to induce an inexperienced defender to play that suit.)

I play the spade nine--deuce--king--seven. Then the club five--ace--deuce--four. East is delighted to shift to the spade five. I win and drive the club king. Making four.


NORTH
Phillip
♠ K 10 4
♥ A K 3
♦ A 6 5 3
♣ 8 6 5


WEST
Robot
♠ 8 3 2
♥ Q 10 9 5
♦ Q 10 4
♣ K 9 4


EAST
Robot
♠ J 7 6 5
♥ 8 4 2
♦ K J 8 2
♣ A 7


SOUTH
Robot
♠ A Q 9
♥ J 7 6
♦ 9 7
♣ Q J 10 3 2

Plus 180 is worth 79%. About half the field ducked at trick one and half rose with the heart ace. It turns out the idiot's delight was unnecessary. If you duck the heart to your jack and play the club queen, West grabs the trick to lead spades through dummy's king. So everyone who ducked made four.

If you rise with the heart ace and play a club, the defense still doesn't find the diamond shift. But East hops with the club ace and continues hearts, which suffices to hold it to three.

The only declarer to take fewer than nine tricks is the one who opened with one notrump and got to game. Weirdly, the reason opening one notrump failed was because it wrong-sided the contract. It was easy for the defense to find five tricks after a spade lead by East. Or maybe that's not so weird. This doesn't look like a hand that needs to grab notrump. Perhaps opening one notrump under strength with no tenaces is a bad idea.

Sunday, January 29, 2023

Free Weekly Instant Tournament - January 27 - Board 7

Board 6
Opponents vulnerable

♠ A 7 6  ♥ J 10 9 2  ♦ A 8 5  ♣ Q 9 4  

I pass in first seat, LHO opens with one club, partner bids one heart, and RHO doubles.

My choices are two clubs, showing a limit raise, or three clubs, showing a "mixed raise," that is, a good pre-emptive raise. Is this hand worth a  limit raise? I have 11 HCP, but the club queen may be worthless and I'm "4333." I also have nine losers. A typical limit raise has only eight. Still, I do have four good trumps, and my two aces mean loser count will undervalue my hand. So I'm going to value this hand as a limit raise and bid two clubs.

LHO passes, and partner bids two hearts. With four trumps, I will compete to three hearts if pushed. But they don't push. Partner buys it at two hearts, and RHO leads the club three.


NORTH
Phillip
♠ A 7 6
♥ J 10 9 2
♦ A 8 5
♣ Q 9 4






SOUTH
Robot
♠ J 10 4
♥ K Q 8 6 5
♦ K 7 6 4
♣ 10


West North East South
Robot Phillip Robot Robot

Pass 1 ♣ 1 ♥
Double 2 ♣ Pass 2 ♥
(All pass)


I have five losers: one in each suit plus a second loser in spades. Four losers are unavoidable. Is there any way to avoid the second spade loser? If diamonds are three-three, I can pitch one of dummy's spades on a diamond, then ruff a spade. If they aren't, perhaps I can strip the hand and endplay someone. I'll need both spade honors in the same hand for that to work. If someone has both spade honors, it's probably East, since West might have led a spade from king-queen.

What are the opponents' shapes? I know spades are four-three. What about clubs? West might have competed to three clubs with four-card support. On the other hand, East might have competed--or at least have doubled two clubs--with six. One of the opponents seems to have one more club than he should.

I play a low club from dummy. East takes the ace, and I follow with the ten. I haven't seen the robots falsecard at trick one, so West should have the king. What about the jack? If East had the jack, he would normally play it unless he was afraid I had a stiff king.

When might he worry about that? If East expects his partner to compete with four-card club support, then the only time he would worry was if he had six clubs. But even then it's not clear letting me score a stiff king would cost. If I have a useful discard on the club queen, losing the club ace wouldn't matter. All in all, it's highly likely that the club jack is on my left.

At trick two, East shifts to the spade three. I play the jack, West plays the queen, and I take the ace. If East has the spade king, I have avoided my second spade loser.

If East does have the spade king, as seems likely, that was a dangerous shift. It suggests East has ace fifth of clubs and was afraid I could set up my club queen for a spade pitch. So I suspect clubs are four-five and it was West who under-competed.

I lead the heart jack--three--five--ace. Ace? I'm surprised that card is on my left. That gives West the spade queen, the heart ace, and the king-jack of clubs, leaving East with the spade king, queen-jack of diamonds, and the club ace. Would he have opened with that hand? The robots don't typically open with ten high-card points unless they have a fair amount of shape. Maybe I'm wrong about the club jack.

West shifts to the diamond deuce. I'm making one overtrick already. Is there any way to take a second one? Maybe if West is 4-2-4-3, I can squeeze him in the minors. For that to work, I'll need to keep the diamond ace as an entry to the club threat. So I play low from dummy. East plays the jack, and I win with the king. I play a heart to dummy. East pitches the club deuce. 

Here is the current position:


NORTH
Phillip
♠ 7 6
♥ 9 2
♦ A 8
♣ Q 9






SOUTH
Robot
♠ 10 4
♥ K Q 8
♦ 7 6 4
♣ --

West has four spades, three hearts, and at least three clubs, so he can't have four diamonds. There is nothing to do but draw the last trump and concede a spade and a diamond. Making three.


NORTH
Phillip
♠ A 7 6
♥ J 10 9 2
♦ A 8 5
♣ Q 9 4


WEST
Robot
♠ Q 9 8 5
♥ A 7 4
♦ 10 9 2
♣ K 8 6


EAST
Robot
♠ K 3 2
♥ 3
♦ Q J 3
♣ A J 7 5 3 2


SOUTH
Robot
♠ J 10 4
♥ K Q 8 6 5
♦ K 7 6 4
♣ 10

As you may have noticed, there was no squeeze even if West held four diamonds. To come to nine tricks, I need to concede a spade and a diamond. And once I concede two tricks, I can't possibly take ten.

But the point of this blog is to discuss what goes through my mind as I'm playing, and retaining an entry to the club threat did pass through my mind, even though on closer examination it makes no sense. Should my lack of rigorous thinking worry me? I don't believe so. Retaining the right matrix to facilitate a squeeze is the bridge equivalent of putting your rooks on open files. You don't necessarily need to see an immediate benefit. It's just a good idea.

East's spade shift didn't cost and was arguably the right play. I rated to be 3-5-4-1. So with diamonds three-three, he was never going to enjoy a third-round spade trick on a passive defense. If his partner had held the spade ten instead of me, a spade shift would stop the overtrick. Of course, if that were the case, he would be wishing he had pushed us to the three level.

I'm not sure what the best use is for a double of two clubs, but I know from past experience that the robots play it as showing rebiddable clubs. I don't know why East didn't take advantage of that agreement here. If he had, his partner would have competed to three clubs.

This result is worth 82%, which seems generous. Apparently most declarers got a different defense, prompted by a different auction, and didn't see how to take nine tricks. It's true that one generally pitches losers from one's hand, not from dummy. Pitching a loser from dummy to enable a ruff isn't a theme one sees every day, so it's easy to have a blind spot.

Sunday, January 22, 2023

Free Weekly Instant Tournament - January 20 - Board 6

Board 6
Opponents vulnerable

♠ Q J  ♥ K Q 10 2  ♦ A 8 7 5 2  ♣ A J  

RHO passes. I could open with an offshape one notrump. In general, the reason one does that is to avoid a rebid problem. A "5422" pattern with 17 HCP does not usually create a rebid problem, since you have enough strength to reverse. But with four honors in my doubletons, this hand looks more like notrump than like diamonds and hearts. Besides, the robots play unplayable methods after reverses. Responder should be able to rebid his suit without creating a game force. Since the robots don't play that way, I try to avoid reversing with robots.

Accordingly, I open with one notrump. Everyone passes, and West leads the diamond queen.


NORTH
Robot
♠ 10 9 7 3
♥ 6 3
♦ K 10 9
♣ 10 6 3 2






SOUTH
Phillip
♠ Q J
♥ K Q 10 2
♦ A 8 7 5 2
♣ A J


West North East South
Robot Robot Robot Phillip


Pass 1 NT
(All pass)


Since West is missing the ten and nine, this lead is presumably from shortness. It is probably from either a singleton queen or queen-jack doubleton. It is possible, though unlikely, that the lead is from queen-small doubleton or queen-jack third.

If I think West has queen-jack, I should win this trick in my hand. I will then have two dummy entries with which to lead up to my king-queen of hearts. Queen-jack doubleton is slightly more likely than a singleton queen, so that seems like an attractive plan.

But I'm not convinced. Often the way to resolve a guess is to ask what will happen if you guess wrong. In this case, I don't mind so much if I lose to West's stiff jack, since any return he makes will be helpful. If I win this trick in my hand, lead another diamond, and West shows out, the position is awkward. East will have a productive club shift available when he gains the lead. And the blockage in diamonds may cause problems for me.

For these reasons, I decide to play West for a stiff queen. I take the diamond king. West plays the three; I play the deuce. I lead the nine, East plays the four, and I duck. I guessed wrong. West wins with the jack.

West now shifts to the five of clubs. This gives me a club trick, which I hope compensates for the misguess in diamonds. I have four diamonds and two clubs. If I can develop a heart trick in time, I will have seven.

I play low from dummy, East plays the eight, and I win with the jack. It's not clear how the clubs lie. It's possible West shifted from king-queen. It is also possible East has one or both honors and is finessing against dummy's ten.

I can now set up one heart trick to make my contract. Can I make an overtrick? I don't have the tempo to go after a spade trick, but I have ways to try for two heart tricks. One way is to finesse against the jack, either before or after driving the ace. If the finesse loses, the opponents will take two spades, two hearts, a diamond, and at least two clubs, so I will go down.

A safer approach is to lead a diamond to dummy and a heart to the queen. If East has the ace and West, the jack, neither opponent can afford to come down to a singleton heart when I run diamonds. With each opponent holding two hearts, I may be able to cash the club ace and exit with a spade, forcing them to give me a second heart trick.

This offers at best a 25% chance rather than a 50% chance. But it's safer. And the opponents may misdiscard in the end position even if the heart honors don't lie favorably. Besides, it's not clear how important the overtrick is. Some declarers will be in diamonds. Unless they manage to make three, which seems unlikely, the gain of an overtrick may not be worth the risk of going down.

I lead a diamond to the ten. West pitches the five of spades. He's probably echoing from a four-card suit.

I lead the three of hearts from dummy, and East plays the eight. The queen is the right play, since it makes it harder for West to duck the ace. West takes my queen with the ace and continues with the jack of hearts.

That's surely ace-jack doubleton. The robots assume declarer is double-dummy, so West thinks he has no chance to score his jack if I have king-ten left. He might as well lead it and hope to set up heart tricks for his partner. So it appears West is 4-2-2-5.

I'm up to eight tricks. I don't see much chance of a ninth, but who knows? I'll run the diamonds and see what happens.

I win the heart king as East follows with the four. On the diamond ace, West plays the club seven. I discard a spade from dummy, and East discards the spade six.

If I have read the position correctly, West has three spades and three clubs remaining. East has two spades, two hearts, and a singleton club. West has shown up with eight high-card points already. With eight more, he might have doubled one notrump, so he is limited to two of the four black-suit honors. Most likely, the spade honors are split and East's club singleton is an honor. So the position probably looks something like this:


NORTH
Robot
♠ 10 9 7
♥ --
♦ --
♣ 10 6 3


WEST
Robot
♠ H x x
♥ --
♦ --
♣ H x x


EAST
Robot
♠ H x
♥ 9 7 5
♦ --
♣ H


SOUTH
Phillip
♠ Q J
♥ 10 2
♦ 8
♣ A

I cash the last diamond, and West pitches the spade four, completing his echo. If I'm right that the opponents' spades are two-two, dummy's long spade isn't useful any more. Ten third of clubs might prove useful, however, especially if I'm right that East has a singleton club honor. Should I pitch a spade and hold ten third of clubs?

No, the benefit of holding ten third of clubs is an illusion. My goal is to hold the opponents to two tricks. They can certainly take two spade tricks, so I can't afford to lose a club trick as well. I might as well keep three spades in dummy in case I'm wrong that spades are two-two.

I pitch a club from dummy, and East pitches the five of hearts, confirming that hearts were two-five. Is there any way to hold the opponents to two tricks? Not that I can see, but maybe the opponents can think of something. It can't hurt to exit with a spade.

I lead the spade queen. West takes the ace, and East follows with the eight. West exits with the club nine. East plays the queen, and I take the ace. I'm out of ideas. There is nothing left to do but cash out.

I cash the heart ten. Surprisingly, both opponents follow, so my heart deuce is good. West started with ace-jack third of hearts, not ace-jack tight, and East erred with his heart pitch. Making three.


NORTH
Robot
♠ 10 9 7 3
♥ 6 3
♦ K 10 9
♣ 10 6 3 2


WEST
Robot
♠ A 5 4 2
♥ A J 7
♦ Q J
♣ K 9 7 5


EAST
Robot
♠ K 8 6
♥ 9 8 5 4
♦ 6 4 3
♣ Q 8 4


SOUTH
Phillip
♠ Q J
♥ K Q 10 2
♦ A 8 7 5 2
♣ A J

Plus 150 is worth 86%. Those who opened with one diamond typically reached two diamonds after a take-out double by West. Some made two diamonds; some went down.

Tuesday, January 17, 2023

Rules For Scrambling the Opponents' Count Signals

This is a reprint of a post from twelve years ago. The discussion in this week's deal of whether to lead the seven or deuce from your hand with J72 reminded me of it.

In the ensuing discussion, I will use the name 'Fred' to refer to the defender from whom you wish to conceal information. I will use the name 'Ethel' to refer to his partner, that is, the defender whose holding you are trying to obscure. Sometimes you know ahead of time which defender is which; sometimes you don't. Sometimes Fred may be, for example, whichever defender holds the ace.

Rule 1 - If you wish to represent a particular holding for Ethel, signal as if you held that holding.

Most players know this rule when it comes to attitude signals. Declarer should play high to encourage and low to discourage (or the opposite if the opponents signal upside down). Many players do not know, however, that the rule works with count signals also.


NORTH
♣ Q J 10 9


SOUTH
♣ 7 6 2


Fred leads the king of clubs against a heart contract; Ethel plays the four. If Ethel has four-three doubleton, you want Fred to think that it is possible she has three clubs. So you signal as if you had three clubs (which, by sheer coincidence, you do). Since the opponents play standard count, you play low. ("Low" means any card lower than the card Ethel played. If you have choices, you should choose among them at random.)

Suppose, however, you want Fred to continue clubs. Perhaps you know--for whatever reason--that Ethel has three clubs (presumably eight-five-four), and you want Fred to waste a tempo cashing his club ace. You want to represent a doubleton in Ethel's hand, so you signal as if you had a doubleton. Since the opponents play standard count, you play high. Any card higher than Ethel's will do. By concealing the deuce, you leave open the possiblity that Ethel has four-deuce.

Rule 1 is fine if you have a specific objective in mind. But what if you don't? What if you just want to play the card that has the greatest chance of scrambling the opponents' signals? In that case, you need Rules 2 and 3.

Rule 2 - With three or more cards, play your lowest or second lowest card, choosing whichever is closest to middle in rank. If the opponents signal upside down, choose similarly between your highest or second highest spot card. 

For example,


NORTH
♣ K Q J 10


SOUTH
♣ 8 5 2


Which card should you lead toward dummy to scramble Ethel's count signal? Assuming standard signals, your choice is between the deuce and the five. The five, being a middle-ranking card, will work more often. Playing the deuce is right only if Ethel has four-three doubleton, since it leaves open the possibility that she has eight-five-four. Playing the five is right anytime Ethel has three clubs, since it leaves open the possibility she has a doubleton.

Note that if dummy had the nine of clubs instead of the ten, the five would work in even more cases. It would work against against any three-card holding and against any even holding that includes the ten (provided Ethel is unwilling to signal with the ten), since it leaves open the possibility she has ten-eight third.


NORTH
♣ K Q J 10


SOUTH
♣ 9 8 4


Now the choice is between the four and the eight. Again, the middle-ranking card, the four, is the percentage play. The eight will create an ambiguity only if Ethel has seven-six-five. The four will create an ambiguity any time Ethel has an even number.

Rule 3 - With a doubleton, give correct count with one exception: If Ethel is apt to have a card she doesn't want to part with, give false count. 

Why should you usually signal an even number with a doubleton? Because, in general, it is easier to scramble Ethel's odd signal than it is to scramble her even signal. To scramble her odd signal, all you need to do is to conceal one card lower than the card she plays (or higher if they signal upside down). To scramble her even signal, you must conceal two cards higher than the card she plays (or lower if they signal upside down).



NORTH
♣ K Q J 10


SOUTH
♣ 9 2


Signal as the opponents would. Play the nine if the opponents play standard signals; play the deuce if they play upside down. You hope that Ethel has an odd number of clubs, in which case playing the proper card will make it possible that she has a doubleton. It is impossible to create an ambiguity when Ethel has an even number of clubs.


NORTH
♥ K Q 9 8 3


SOUTH
♥ 7 2


In this layout, there are two cards outstanding that Ethel might well choose not to signal with: the jack and the ten. Ethel is apt to have one or the other of them, so the exception applies, and you should give false count.

Why does this situation create an exception? Because there are now many three-card holdings from which Ethel's play is automatically ambiguous:

♥ J 6 5
♥ J 6 4
♥ 10 6 5
♥ 10 6 4

Ethel will play low from any of these holdings, but she will also play low from:

♥ J 5
♥ J 4
♥ 10 5
♥ 10 4

(Actually, it doesn't matter whether she will play low from these doubletons or not. All that matters is that, from Fred's point of view, she might play low.)

Normally, against standard carders, you want to retain the deuce when you hold a doubleton. That way, if Ethel plays low from three, it will be possible she holds a doubleton. But if Ethel holds the jack or ten, you don't need to do that. It is already possible that she holds a doubleton. Her signal from three cards is already ambiguous without your having to do anything.

Therefore, you need to shift your attention to scrambling her signal when she holds an even number. And, as we have seen, the way to do that is signal an odd number. If Ethel plays her highest spot card from:

♥ J 6
♥ J 5
♥ J 4
♥ 10 6
♥ 10 5
♥ 10 4
♥ J 6 5 4
♥ 10 6 5 4

then retaining the seven leaves open the possibility that she has three. (If they signal upside down and she plays the lowest spot card, then retaining the deuce leaves open that possibility.)

One addendum: Declarer's choice of equals rarely matters. So, whatever card the above rules instruct you to play, you are free to play any other card of equal rank. Randomizing among equals will help to keep Fred from  exploiting these rules to divine your holding.

Am I being too cautious in saying "rarely" instead of "never"? No. I can't think of a case where declarer's choice among equals matters for an even-odd ambiguity. But it can certainly matter for a two-or-four ambiguity:




NORTH
♣ K Q J 3


SOUTH
♣ 9 8 2



You lead the king from dummy, and East plays the seven, presumably from seven doubleton. By Rule 1, if you want to represent three cards in East's hand, you should play low; if you want to represent four cards (so that you might have a singleton), you should play high. But, if you play high, you must play the eight. The nine will not work, since East would play the eight from eight-seven fourth. This falls under the heading of "playing the card you're known to hold."

So there you have it. A simple set of guidelines for scrambling the opponents' count signals. What's particularly nice about these rules is that the opportunity to apply them typically comes up several times per session. So, if you didn't know these rules already, your expected per session score has just gone up perhaps a couple of matchpoints.