Challenge Match - Jazlene vs. Phillip - Board 7

No surprises on Board 6. I'm still trailing by seven imps.

Board 7
Both sides vulnerable

♠ A 7 6 4 3  ♥ K J 6  ♦ 3  ♣ K Q 9 4

I open with one spade in first seat. LHO overcalls with two diamonds. Partner passes, and RHO bids two spades, showing at least a limit raise in support of diamonds.

We discussed doubles of artificial bids back on Board 2, so you know I think double here should be take-out of diamonds. With a singleton diamond and support for the other suits, I should act. And a double of two spades, which partner can pass if he has nothing, is the safest way to get back into the auction. Passing, then making a balancing double of three diamonds is considerably more dangerous. Unfortunately, the robots play this double shows rebiddable spades, so I have to pass.

LHO bids three diamonds--pass--pass back to me. This is why I wanted to double two spades for take-out. It could easily be right to compete. But forcing partner to act at the three-level when both opponents have shown good hands is too dangerous. I have no choice but to sell to three diamonds. Fortunately, Jazlene and I are forced to play the same methods, so I'm not at a disadvantage as I would be in a normal field.

Partner leads the king of spades, and the following dummy appears:

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

West	North	East	South
Robot	Robot	Phillip	Robot
		1 ♠	2 ♦
Pass	2 ♠	Pass	3 ♦
(All pass)

Declarer plays the spade jack from dummy. Sure. Why not? He has the ten in his hand.

My agreement in a regular partnership is that discouraging suggests a heart shift. (When dummy's side suits are of equal length, discouraging suggests the suit with fewer honors. Clubs has two honors; hearts has one.) If I encourage, I'm saying I don't want a heart shift. Partner must use his judgment in deciding whether continuing spades or shifting to a club makes more sense. On this hand, I think he would decide to shift to a club, which is the suit I prefer. So I would play an encouraging seven.

Note there is no need to play suit preference here, as I'm sure some would. If you have well-defined rules for which suit a discouraging signal calls for, attitude works just fine. If you want the obvious shift (defined here as hearts), then you play low. You don't have to stop and ask yourself whether this is an attitude situation (where low asks for a heart shift) or a suit preference situation (where high asks for a heart shift). If you want the non-obvious shift, you encourage. If it's so clear that continuing the led suit makes no sense, partner will make the right shift, since he knows you don't want the obvious one.

There is nothing to gain by playing attitude sometimes and suit preference at other times, so why risk confusion? Even if it's clear this time which signal applies, there will always be borderline cases where you and partner won't agree. So why switch ever? If your card is always attitude, you can't have an accident.

Robots, however, play neither attitude nor suit preference, so it makes no difference which card I play. I choose the three.

Declarer plays the five, and partner continues with the spade eight. I wonder if declarer played the spade jack at trick one as a clever ploy to make West think I hold ace-ten over dummy's queen. Robots don't draw inferences, so partner won't realize I can't possibly hold the ten. Who better than another robot to understand that? I'll have to remember that trick.

Dummy plays the deuce. I win with the ace, and declarer plays the ten.

In bridge, as in many games, there are two general strategies of defense: Containment and Attack. The goal of Containment is to deprive declarer of tricks. The goal of an Attack is to take tricks ourselves before declarer can take his. The fundamental task of defense is deciding which strategy is better on a given deal. The reason for making this distinction is that we analyze differently depending on which strategy we adopt. For Containment, we focus on declarer's tricks. For Attack, we focus on our tricks.

On this deal, it may be easier to count declarer's tricks than ours, so let's start with Containment. What high cards does declarer have? It's hard to see his having less than ace, ace-king in the red suits for his two-level overcall. And if he's missing one of those cards, we might beat him whatever we do. So let's assume he has them.

What's his shape? If he has six diamonds, then he has eight top tricks (six diamonds and two aces), and we can't possibly stop him from eventually taking a spade trick. So we need to assume he has only five diamonds.

Ace-king fifth of diamonds and the heart ace gives him four or five diamond tricks (depending on who has the jack), the heart ace, the club ace, and two spades: eight or nine tricks in all. If we return a spade for partner to ruff, we eliminate one of his spade tricks, reducing him to eight tricks. (The diamond jack now becomes irrelevant. After partner ruffs, declarer will always have five diamond tricks.) Can declarer find a ninth trick somewhere? If he has the heart ten, he can. If not, I don't see where a ninth trick is coming from.

The next step is to double-check our plan by constructing a concrete layout. Sometimes an abstract plan has a flaw that is difficult to see unless you play through it. Let's give declarer

♠ 10 x  ♥ A x x  ♦ A K x x x  ♣ x x x

If we return a spade, declarer will pitch a club as partner ruffs. Partner will then shift to a club. Declarer will win his ace, draw trump ending in dummy, and pitch his last club on the fourth spade. If one of those heart x's is the 10, he can develop a heart trick and make his contract. If not, he's down. Yes, our analysis seems correct. 

We now have a provisional plan. The next step is to try to improve on it. Can we defeat the contract if declarer has the heart ten?

Perhaps an Attack will do better. A successful Attack requires us to establish and cash five tricks. If partner has the club jack and declarer has three clubs, we can shift to the club king and establish two club tricks. That gives us two spades, a heart, and two clubs. Again, we need to test our plan against a concrete layout. Let's add the heart ten to declarer's hand above so that the spade return will fail:

♠ 10 x  ♥ A 10 x  ♦ A K x x x  ♣ x x x

We shift to the club king. Declarer wins and draws trump, ending in dummy. They don't split, so he can't take his spade pitches. He must lose two clubs and a heart. Possibly a diamond as well if partner has the jack.

So Containment (returning a spade) works if partner has the heart ten, and Attack (shifting to the club king) works if partner has the club jack and declarer has three clubs. Since Attack requires two assumptions instead of one, Containment is more likely to be right.

Or so it appears at first glance. Let's confirm we're right about declarer's needing three clubs for the club shift to work. Give declarer:

♠ 10 x  ♥ A 10 x x  ♦ A K x x x  ♣ x x

If we shift to the club king, his best play is to win and cash a spade, pitching his club loser and letting partner ruff. This won't hurt if diamonds split, but it saves a trick (compressing his diamond and club losers) if partner has jack fourth. Partner will now try to cash a club. Declarer will ruff, draw trump ending in dummy, pitch a heart on the last spade, then play a heart toward his ten. We get only four tricks: two spades, a heart, and a ruff.

Yes, a club shift fails if declarer has a doubleton club. What about our assumption that partner needs the club jack? If declarer has jack third of clubs, we can take only one club trick. But maybe that's enough. If partner has jack fourth of diamonds, perhaps we can take two spades, one club, one heart, and a diamond.

Let's check it out. Give declarer

♠ 10 x  ♥ A 10 x  ♦ A K x x x  ♣ J x x

We already know a spade continuation fails on that layout. What happens if we shift to the club king? Declarer wins in dummy and leads a second club. If we hop, declarer unblocks his jack. We continue with a club to dummy's ten to kill the entry. Now what? Declarer cashes a spade, pitching one heart as partner ruffs, then draws trump ending in dummy to pitch his last heart. We get our club trick, but we don't get a heart trick.

Perhaps, instead of hopping with the club queen, we should duck and let declarer win his jack. This deprives him of using the club ten as a dummy entry. If we do that, declarer can continue with a third club. We win and find ourselves in this position:

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

We're endplayed. A heart or a club shift gives declarer a trick, taking care of one of his small hearts. He can then cash a spade, pitching his last heart as partner ruffs. He loses only four tricks: two spades, one club, and a ruff. Our only safe return is a diamond, and that gives declarer two dummy entries. Declarer can now pitch one heart on a spade as partner ruffs, and eventually pitch his last heart, again losing two spades, a club, and a ruff.

Yes, a club return does indeed require finding partner with the jack and finding declarer with three clubs while a spade continuation requires only finding partner with the heart ten. So a spade continuation is the better play.

I return a spade. Declarer pitches the three of clubs, and partner ruffs with the nine of diamonds. Partner shifts to the eight of hearts.

Hearts? Why are we breaking hearts, partner? Oh, I see. He's hoping I have the ace, so I can play another spade and kill that trick as well. Let's hope the shift is from ten-eight-seven. If so, we still have a heart trick coming.

Declarer plays the deuce from dummy, and my jack forces his ace. Declarer draws three rounds of trump, ending in his hand, and leads the nine of hearts--ten--queen--king. I return the six of hearts. Partner overtakes with the seven. Down one.

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

Partner had both the heart ten and the club jack, so either Containment or Attack would have worked.

Jazlene and I don't agree on this one. She initially played as I did but later decided later that a club shift was better. Watch her play on Jaz Plays Bridge, then watch her post mortem.

Challenge Match - Jazlene vs. Phillip - Board 6

Board five was a push, so I'm still down seven imps.

Board 6
Opponents vulnerable

♠ J 7  ♥ K Q J 6 4 3  ♦ K 2  ♣ K Q 10

Six heart tricks, half a trick in diamonds, a trick and a half in clubs. Seven in all, not quite good enough to open with one heart and rebid three hearts.

I bid heart, LHO overcalls with two diamonds, partner passes, and RHO bids two notrump. I have a pretty good hand, but partner couldn't act over two diamonds, and RHO's two notrump suggests hearts may not be breaking. So I pass, as does everyone else.

What should I lead against two notrump? Should I try a heroic low heart in case declarer has ace-nine fourth and partner has the ten? LHO might have removed two notrump if he had a stiff heart, so if declarer does have four hearts, partner probably has a singleton. Hoping it's a stiff ten is a bit much. I'm better off hoping declarer has only three hearts. But at least, if a low heart lead turns out to be right, I can say I thought about it.

I lead the king of hearts and see the following dummy:

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

West	North	East	South
Phillip	Robot	Robot	Robot
			Pass
1 ♥	2 ♦	Pass	2 NT
(All pass)

Declarer plays a curious ten of hearts from dummy. Partner plays the eight; declarer the six.

If diamonds are running, we aren't beating this. Can we beat it if partner has jack third of diamonds? Declarer then has five diamonds and two aces. The tooltip says two notrump show 11 HCP. There are 15 HCP missing, so even if declarer has stretched a point, partner has at most five. That means he can't have both the club ace and the spade king. Declarer must have one of those cards, so he has eight tricks even after losing a diamond.

One possibility is to switch to clubs, playing partner for ace fourth. But if he has that, we can always cash clubs when partner wins his diamond trick.

What if partner has jack fifth of clubs? In that case, I must switch to clubs now to knock out the ace, giving us a heart, four clubs, and a diamond. Is that possible? That gives declarer a 6-3-2-2 shape. I don't think so. My best shot is to continue establishing hearts and hope partner has jack-ten fourth of diamonds, so that the diamond suit isn't a threat. 

I continue with the queen of hearts. Oops. That was a mistake. I'm supposed to lead the jack, since I want a club return when partner wins his diamond. Good thing partner pays no attention to my signals. Partner follows with the deuce of hearts and declarer wins with the ace.

Declarer plays the three of diamonds--deuce--nine--ten. Declarer can't have a stiff diamond. If he did, he would be finessing the queen, not the nine. So dummy's diamonds are now good. 

Partner shifts to the deuce of clubs, and declarer takes the ace. This is the position:

	NORTH Robot ♠ A 4 ♥ -- ♦ A Q 8 5 4 ♣ 7 3
WEST Phillip ♠ J 7 ♥ J 6 4 3 ♦ K ♣ K Q

We're not beating this. So it's all about stopping overtricks now. After declarer runs the diamonds, I'm left with four cards. What should they be? If declarer has king-ten of spades, I must keep both spades, else declarer can cash the ace and finesse against partner's queen. So my last four cards will be

♠ J 7  ♥ J  ♦ --  ♣ K.

Actually that might not work. If declarer holds jack doubleton of clubs in the end position, he can duck out my king. I can't decide what to hold until I see what declarer keeps. If he holds three spades, I have to keep both spades. If he holds two clubs, I must keep both clubs. If he holds a heart, I must keep the heart jack. If I'll have to pay attention. And here I was hoping to catch a nap. Oh, well.

Declarer has played the club ace. I follow with the ten. (In real life, this card would be face down on the table and I would now turn it over. That's how a defender can pause play to give himself time to absorb what's happened without misleading declarer into believing he is thinking about the current trick. Those who designed online play forgot about this option. There should be a "play card face down" button.)

Declarer plays a diamond to the ace and cashes the queen. Partner plays six-jack in diamonds; declarer pitches the four of club; I pitch the three of hearts.

On the next diamond, partner pitches the deuce of spades. The robots almost always discard count cards, so partner is presumably 5-2-3-3 and declarer is 4-3-2-4. Declarer pitches the nine of hearts. If my construction is correct, declarer is down to four spades and three clubs. I pitch the heart four. Since declarer is out of hearts, I can now throw all my hearts if necessary. 

On the next diamond, partner pitches the spade three; declarer, the spade eight. Declarer still has three spades left, so I can't pitch a spade yet. I pitch another heart. On the last diamond, partner pitches the five of spades; declarer, the ten of spades. Declarer is down to two spades and two clubs. I can afford a spade now, so I pitch the spade seven. We are down to this position, with dummy to play:

	NORTH Robot ♠ A 4 ♥ -- ♦ -- ♣ 7 3
WEST Phillip ♠ J ♥ J ♦ -- ♣ K Q		EAST Robot ♠ ? ? ♥ -- ♦ -- ♣ x x
	SOUTH Robot ♠ ? ? ♥ -- ♦ -- ♣ x x

Declarer leads the three of clubs from dummy. What's he doing? Why not cash the spade ace? Is he trying to endplay me? Maybe he thinks I'm now down to

♠ K x  ♥ J  ♦ --  ♣ K.

If declarer is trying to endplay me, he must have the queen of spades. So this is the position:

	NORTH Robot ♠ A 4 ♥ -- ♦ -- ♣ 7 3
WEST Phillip ♠ J ♥ J ♦ -- ♣ K Q		EAST Robot ♠ K x ♥ -- ♦ -- ♣ x x
	SOUTH Robot ♠ Q x ♥ -- ♦ -- ♣ x x

The right defense is for me to win with the king of clubs and cash my heart. Partner must stiff his spade king and hold onto a club. Now if I lead the spade jack, declarer might get greedy, duck it, and go down.

Playing with a partner I trusted, that's exactly how I would defend. Partner can see that stiffing his spade king is our only chance to beat this, so he should go for it. But robots aren't that tough. If I win the club and cash a heart, partner is going to come down to a doubleton king of spades like a wimp.

What if I win the club king and lead the jack of spades immediately, without cashing my heart? Declarer might still duck and go down. This line is a lot less convincing than letting partner stiff his king. But playing with a robot, it's the best I can do.

Partner follows to dummy's club lead with the five, and declarer plays the jack. I win with the king and play the jack of spades.

Declarer doesn't bite. He hops with the ace. Making two.

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

Maybe I was too quick to go into stop-the-overtrick mode. We still had a chance to beat it that I didn't see. My goal should be to represent

♠ K J 7  ♥ K Q J 6 4 3  ♦ K 2  ♣ K 10

and entice declarer into endplaying me. 

How would I discard if that were my hand? I would probably pitch a spade early, so that's what I should do. If I pitch a spade at my first opportunity, a good declarer would be disinclined to think I had stiffed my jack. So in the end position, he would "know" I had king-jack of spades left and might well go for the endplay. And my lucky mistake of giving the wrong suit-preference signal at trick two would reinforce that illusion. (I would, of course, tell declarer I was thinking ahead and had falsecarded with the heart queen on purpose.)

Would declarer really risk his contract for an overtrick at IMPs? He should if he thinks the odds are good enough. Yes, you are risking six imps to gain one. But if you think it's ten to one you've read the position correctly, going for the overtrick is the percentage play. Why throw that 0.36 imps away?

What if you're wrong and go down? Then you have to be philosophical about it. I once lost a Spingold match by two imps. After we compared, my teammate told me that, on the last board, he had gone down in one notrump by playing for a "sure" overtrick. When he showed me what had happened, I couldn't blame him. He had made the right play. On a different day, that overtrick might have swung the match in our favor, and I would be congratulating him on his thoughtful play. So I could hardly take issue now. Besides, it was Edgar Kaplan. I would assume he had done the right thing even without seeing the deal.

Be sure to go to Jaz Plays Bridge and see what Jazlene does on Board 6. Let's hope she doesn't bid a fearless three hearts over two notrump and pick up a couple of imps.

Challenge Match - Jazlene vs. Phillip - Board 5

Jazlene bid slam on board four and picked up 12 imps. I'm trailing by seven imps going into

Board 5
Our side vulnerable

♠ A K 10  ♥ K 8  ♦ A 10 9 8  ♣ K 9 8 7

Two passes to me.

Despite holding only 17 HCP, this hand is too good to open with a strong notrump. My HCP are all in aces and kings, and I hold two tens and no card lower than a seven. K&R point count evaluates the hand as 18.9 points. 

Still, I've never had much success opening one of a minor instead of one notrump with hands that the Work point count undervalues. Yes, sometimes you reach game or slam that you would miss after a one notrump opening. But that's aiming for a narrow target. More often, it seems, you get a poor result by letting the opponents in at the one-level, especially at this vulnerability. Since partner is a passed hand, at least I don't have to worry about missing a slam. So I decide to make the tactical underbid of one notrump and hold my breath.

LHO passes, and partner bids four hearts, a transfer to spades. I can let my breath out now. I bid four spades, and everyone passes. LHO leads the deuce of diamonds.

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

West	North	East	South
Robot	Robot	Robot	Phillip
	Pass	Pass	1 NT
Pass	4 ♥	Pass	4 ♠
(All pass)

If the lead is from jack third or fourth, I'm taking four diamond tricks. But the robots don't like leading from honors, so it's likely the lead is a singleton. I play low from dummy, East plays the jack, and I take my ace.

I can pick up four-zero trumps in either hand if I cash the ace first, so I cash it. West plays the five; East, the six. I continue with the spade king and both opponents follow. West held the jack.

It looks as if I'm losing the diamond king and the club ace to make five. If East has the ace of clubs, is there any way to convince him to duck when I lead the club jack from dummy? Maybe. If East thinks I'm missing the heart king, then dummy's heart is a loser. If he hops with the club ace and I have the club king, then the heart loser goes away. So he might duck, hoping I have king-ten and misguess.

At least a human East might. Robots never play for a misguess; they assume you're double dummy. But even if I'm double dummy, it could be right to duck. Say I have queen-ten of clubs. Then hopping allows me to take a ruffing finesse against West and pitch my heart loser.

This means if I want to make it dangerous for East to hop, I must keep the heart ace in dummy. If I lead a heart to the ace and play a club, he has no reason to duck. He can simply hop with the ace and play a heart.

So leading a heart to dummy can't possibly work. What about a spade? The problem with leading a spade to dummy is I now have no fast entry to my hand. If East is worried about losing a heart trick, he can hop with the club ace and shift to a heart. I can't get to my hand in time to take my putative pitch. 

Leading a heart to dummy doesn't work. Leading a spade to dummy is better, since it requires a tad more care on East's part. But it probably won't work either. A diamond is the only option left. Let's give it a try.

I play the eight of diamonds. West pitches the six of hearts, and I play low from dummy. East takes his diamond king and returns the diamond seven. Yes! Perfect! I overtake in dummy with the queen as West pitches the four of clubs. We've reaching the position I was aiming for:

	NORTH Robot ♠ Q 9 8 7 ♥ A 7 ♦ 5 ♣ J
	SOUTH Phillip ♠ 10 ♥ K 8 ♦ 10 ♣ K 9 8 7

Now when I lead the jack of clubs, East has a legitimate reason to duck. If I have queen-ten of clubs and no heart king, he must duck to stop the second overtrick.

I lead the club jack. East hops with the ace. Oh, well. I claim the balance. Making five.

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

East had ace-queen of clubs, so he had no temptation to duck. Still, I don't understand his diamond return after winning the king. Couldn't I have the same hand with the queen of hearts instead of the king, in which case he must shift to a heart?

The only reason I can think of to play a diamond is he thought I had forgotten to draw the last trump and his partner was ruffing. A priori, playing for the ruff is probably percentage. But I can't have that hand. Until we teach robots to draw inferences, their card play is never going to get above a beginner level.

My one notrump opening did keep the opponents out of the auction, although it probably didn't matter. I'm not sure why West didn't bid anyway. Though perhaps there isn't much point playing Cappelletti, where you announce to partner that you have a suit to bid but you aren't going to tell him which one. I would certainly overcall one notrump with two hearts if it were natural.

While my swindle didn't work, this hand is an excellent example of the necessity of looking at a deal through your opponent's eyes. Sneaking a club through has a better chance to work if you sit in your opponent's seat and imagine a scenario where ducking makes sense. 

This has become my favorite type of play problem. When I was a beginner, my favorite hands were winkles. But any fool can execute a winkle. Even a robot. Seeing a deal through your opponent's eyes is hard. And fun. And often more profitable than it was on this deal.

Now take a look at the deal from Jazlene's eyes at Jaz Plays Bridge. I suspect the board is a push and she's still up seven imps. We'll see.

Challenge Match - Jazlene vs. Phillip - Board 4

My match against Jazlene Ong continues. Board three was another push, so I'm still up the five imps I gained on board one. Now on to

Board 4
Both sides vulnerable

♠ A Q  ♥ A K Q 5 4 2  ♦ A 10  ♣ Q 4 2

LHO and partner pass, and RHO opens with three spades.

My choices are four hearts and three notrump. If they lead a spade and hearts break, I have nine tricks in notrump. If they don't lead a spade, maybe dummy will have an entry in hearts. In any event, nine tricks rates to be easy in notrump. I might not have a tenth trick in hearts.

I know some would double with this hand, considering it too good for either three notrump or four hearts. Whether double makes sense or not depends on how you play this auction:

You	LHO	Partner	RHO
			3 ♠
Double	Pass	4 ♣/♦	Pass
4 ♥

If this auction shows a strong hand with an independent heart suit--a hand too good to overcall with four hearts, then double is the right call. But I don't believe that's what it should show. On the contrary, I believe this auction should deny an independent heart suit. It shows doubt about strain, not doubt about level. Typical hands would be:

♠ x x  ♥ A K x x x  ♦ A x x  ♣ A K x

or

♠ x x  ♥ A J 10 x x  ♦ A x  ♣ A K Q x

Obviously those of us who double and bid four hearts with these hands don't expect partner to pass four hearts with a bad hand and a singleton heart. Those who double with the hand I'm currently holding presumably do.

You can agree to play either way. But if you don't play that double shows doubt about strain, I think you are making a serious mistake. With an independent suit, there is always something you can bid, namely, your suit. While you may have to guess how high to bid it, you're never stuck for a bid. But what exactly are you supposed to do with my example hands if you can't double and follow with four hearts? You have no sensible call with a hand type that's quite common.

I'm not sure how the robots play. But I can't get into too much trouble if I bid three notrump, so that's what I do. LHO bids four spades, and partner doubles.

If partner has a little something, I can probably manage ten tricks in notrump. So at matchpoints, I would consider bidding four notrump. At IMPs that doesn't make much sense. It's unlikely we're beating four spades less than two, so if I pull to four notrump, I'm trying for a measly four-imp pickup. And the downside is huge. If they can run clubs or if hearts don't break and four notrump goes down, I'm losing 12 imps. Those aren't good odds. There is no vig in pulling a double to a game at IMPs when you can collect a game bonus by passing.

Pulling the double to six hearts, however, is another matter. Now at least I have some upside. What do I need for slam to be cold? King of diamonds, king of clubs isn't quite enough. I need the club jack or the queen of diamonds in addition. Even king of diamonds, ace of clubs isn't enough. I still need a slow trick in one of the minors. I also have no guarantee that hearts are solid, although partner's presumed spade singleton increases the odds that they are. Finally, there is the possibility that the defense can start with ace of clubs and a ruff.

In short, there are three ways bidding slam can be wrong: Partner has the wrong cards, hearts don't break, or the defense gets a ruff. While any one of those chances may be small, they add up. And, because of the pre-empt, bad breaks are more likely than they would be a priori. It's close. Add the jack of clubs to my hand and I would certainly bid six hearts. But with this hand I decide to defend.

I pass, LHO passes, and I lead the king of hearts.

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

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

Partner plays the three and declarer follows with the seven. I don't see any reason not to try to cash another heart. But before I do, I must decide what I'm going to do if declarer ruffs and plays a low diamond. If I intend to duck, I need to do so in tempo. I know my robot declarer won't pay any attention to my tempo, but it's good practice to pretend I'm playing against a human. Should I duck or not?

Hopping saves a trick if declarer is 7-1-1-4. But ducking won't cost the contract, and declarer is more likely to have a doubleton diamond than a singleton. So I intend to duck.

I continue with the queen of hearts. Partner plays the jack, and declarer ruffs with the spade three. One of these days, I'll figure out how to formulate a rule that partner's play on this trick is surrogate count--that is, count in diamonds--so I'll know for sure whether to duck or not. It's clearly the most important piece of information for me. But I'm not sure how to make an unambiguous rule for when it applies.

Declarer doesn't play a diamond, however. He leads the king of spades. That suggests declarer does have a singleton diamond. If he had a doubleton, why not try to reach dummy to finesse against the spade queen?

I win with the ace of spades as partner follows with the deuce. I shift to the deuce of clubs. Partner takes the ace, and declarer follows with the six. Partner continues with the club jack, and declarer takes the king. We've reached this position with declarer on lead:

	NORTH Robot ♠ 9 4 ♥ -- ♦ K J 9 8 7 ♣ 7
WEST Phillip ♠ Q ♥ A 5 4 2 ♦ A 10 ♣ Q

Declarer continues with the eight of spades to my queen, and partner pitches the three of diamonds.

That card is count--even from a robot. And it should show five, making declarer 7-1-1-4 with a stiff deuce of diamonds. If partner has four diamonds, he must have a spot higher than the three, and he should play it.

I know, however, that the robots are fond of bloodless count. With four, he would pitch his next-to-lowest diamond. Some unenlightened humans do that, too. It makes no sense. If you're going to signal, signal loudly. Either play your lowest card or play the highest one you can afford. What do you have to gain by playing a card partner might be unable to read?

Since the robots stubbornly refuse to listen to me on this matter, partner could easily have four diamonds. So I'm on my own. If declarer is 7-1-2-3, I should cash my club queen and lead a low diamond, making declarer guess. If he guesses right, he's down three and we collect 800. If he guesses wrong, we collect 1100. If he's 7-1-1-4, however, I must cash the diamond ace. If I underlead, he will hop with the king and get out for 500.

If I underlead, how likely is he to get it right? I would certainly bid three notrump with the diamond queen instead of the ace. And partner's auction would be the same. So I see no reason declarer shouldn't guess wrong. On the other hand, as I noted at trick two, declarer's failure to try to get to dummy for a trump finesse suggests he has a singleton diamond rather than a doubleton.

Another consideration is the risk vs. the reward. The difference between 500 and 800 could be worth more imps than the difference between 800 and 1100. If Jazlene is in game at the other table, trading plus 800 for plus 500 converts a four-imp win into a four-imp loss, for a net loss of eight imps. Converting 800 to 1100 converts a four-imp win to a ten-imp win for a net gain of only six imps. You don't need to know the IMP table off the top of your head to reach this conclusion. It's sufficient to know that, the way the IMP table is constructed, converting a gain into a loss is costly. If that's a possibility, trying to convert a gain into a larger gain is generally not worth the risk.

I cash the ace of diamonds, settling for plus 800.

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

Declarer did have a doubleton diamond. Who knows if he would have guessed right or not had I underled?

We're making six hearts. Partner didn't have much, and it's a fine contract, so perhaps I was too conservative in passing the double. To check, I did a simulation using DealMaster Pro. If I give partner 8-10 HCP and specify spades are three-seven, six hearts makes 74% of the time. I'm not surprised it's a favorite, but I wasn't expecting it to be that much of a favorite. I clearly made an error in passing the double. How long will it be before DealMaster Pro comes out with a version that can be installed in your head? DealMaster Pro 7.0 - for Windows, Mac, or cerebrum.

I also made an error in the play in forgetting that slam was a possibility at the other table. If Jazlene is in six hearts, the odds on underleading the diamond ace change dramatically.

If declarer is 7-1-1-4, then six hearts is going down. Collecting 500 instead of 800 means I win 12 imps instead of 14, so the underlead risks practically nothing. But if declarer is 7-1-2-3, six hearts is making. Now I'm losing 12 imps if I collect 800. But if I underlead and declarer misguesses, I hold my loss to eight imps. In short, I'm risking two imps for a chance to collect six. Clearly the odds favor the underlead.

Again, you don't need to memorize the IMP table to reach this conclusion. It suffices to know that converting a win to a larger win or converting a loss to a larger loss generally makes little difference. Your primary goal at IMPs is to beat the other result or, if you can't, to get as close to the other result as possible. Even without knowing the actual imps at stake, we know that getting closer to 1430 will gain more imps than moving away from 1430 will lose.

So how did I do? Will Jazlene bid slam and obliterate my five-imp lead? Check out Jaz Plays Bridge to find out.

Challenge Match - Jazlene vs. Phillip - Board 3

The second board of my match against Jazlene was a tie, with each of us pushing the opponents up to the three level to no avail. Now on to Board 3, which I hope will offer some avail.

Board 3
Opponents vulnerable

♠ K 9 4  ♥ K J 6  ♦ A J 8 3  ♣ Q 7 5

I open with one diamond, partner bids one heart, I rebid one notrump, and partner raises to three. West leads the ten of spades.

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

West	North	East	South
Robot	Robot	Robot	Phillip
			1 ♦
Pass	1 ♥	Pass	1 NT
Pass	3 NT	(All pass)

I play the jack from dummy, East covers with the queen, and I win with the king.

If the lead was from a doubleton ten, my nine is a stopper. But that would mean East passed over one heart with ace-queen sixth of spades. I suppose that's possible, since he is vul against not. It's also possible that West led the ten from ace-ten third. The ten is the normal lead from honor-ten third to start an unblock. Imagine, for example, I held the same hand without the spade nine. If West leads the ten from ace-ten third and I take the king, the defense can cash four spade tricks. If West leads low, he blocks the suit.

What do I need to do to make this? I have four diamond tricks, two hearts, and one spade. I need two more. I can develop two club tricks by knocking out the ace. If West wins the ace, they can't beat me. If East wins and plays a spade, I have to guess whether to duck, playing West for ace-ten third, or to hop, playing West for ten doubleton.

It's not much of a guess. A timid East might pass over one heart with ace-queen sixth of spades and out. But no one would pass with ace-queen sixth and an ace. Well, maybe Harold Berlin. But he's not around anymore. So if East shows up with the club ace, the spade ace is 100% to be on my left. 

Furthermore, even if East forgot to look at his hand during the bidding, he's probably looking at it now. If he holds ace-queen sixth of spades, for all he knows, his partner has led from ten-nine doubleton and the spades are running if he cashes them from the top. How could he ever underlead his ace and risk West's winning his stiff nine? So I have two reasons to duck a spade return. 

If I were actually worried about "misguessing" spades, I could play a diamond to dummy and a heart to my jack.  If I can take four heart tricks, I won't need to knock out the club ace. Losing the heart finesse to West won't hurt, since if he continues spades, he gives me my ninth trick. If he doesn't, I can just revert to playing on clubs. But I'm not getting the spades wrong. So why throw an imp away for no reason? 

Which club should I lead? Since dummy has the six, it makes no difference whether I lead the five or the seven. I choose the seven out of habit. West hops with the club ace and cashes the ace of spades, presumably hoping his partner has the nine and the suit is running.

I claim the rest. Making five.

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

West had ace-ten-eight fourth? That holding I wasn't expecting. In a real game, I would wonder if partner had flashed his hand and West had caught a glimpse of the jack. In that case, leading the ten might make some sense. It turns out I was wrong when I said it wouldn't hurt to lose a heart finesse to West. Since West holds four spades, I could go down if the finesse lost and East had the club ace.

I want to go back to a statement I made earlier. I said that it made no difference whether I led the club five or the seven, because dummy held the six. What if dummy held KJ43? Would my choice make a difference now?

It might. Suppose, for example, that the six is West's lowest card. If I lead the five and West follows the six, East will know West played low. But if I lead the seven, East won't know whether the six is low or the start of an echo with six-five. When dummy holds the six, however, the five and seven are equals, so it's impossible to introduce an ambiguity. On this particular deal, scrambling a count signal probably won't matter. But it's good to get in the habit of keeping the opponents in the dark when you can.

About fifteen years ago, I wrote a whole blog on the rules for declarer's spot-card plays, showing how to create the maximum ambiguity in a variety of situations. If you go back and read all the posts from the last fifteen years, I'm sure you'll find it. But before you do, head over to Jazlene's YouTube channel to get her comments on Board 3.