Sunday, April 30, 2023

Free Weekly Instant Tournament - Apr 28 - Board 1

Board 1
Neither vulnerable

♠ J 9   K J 9 2   A K 5 4  ♣ K J 8  

Two passes to me. I open with one notrump, partner transfers to two spades then bids three notrump, which I pass. LHO leads the eight of diamonds.


NORTH
Robot
♠ A K 8 6 3
5 4
Q 6 3 2
♣ Q 3






SOUTH
Phillip
♠ J 9
K J 9 2
A K 5 4
♣ K J 8


West North East South
Robot Robot Robot Phillip

Pass Pass 1 NT
Pass 2 Pass 2 ♠
Pass 3 NT (All pass)

The diamond eight can't be lowest from four (assuming West would lead the jack from J1098), so unless it is a singleton, diamonds are splitting. That gives me four diamond tricks, two spades, and two clubs (after knocking out the ace). While it looks natural to attack dummy's long suit for additional tricks, that's not necessarily the best plan. It depends on what my objective is.

If I am content to take one more trick, it might be better to attack hearts. Whatever heart tricks the opponents have, they can probably take whether I lead the suit or not. But if I attack spades, I may set up tricks for the defense that they would be unable to take by themselves.

Say, for example, I win the diamond in my hand and lead the spade jack. It loses to East's queen, and East switches to a heart. I play the jack. West wins with the queen, returns a heart to his partner's ace, and East plays another heart. Now what? The defense has already taken three tricks, and I still have to knock out the club ace. Taking nine tricks before they take five is going to be problematic. I would be in much better shape if I hadn't lost a trick to the spade queen.

Still, this is matchpoints, and it is a contract everyone should be in. So I'm not content to make my contract. If I can manage four spade tricks, I have a shot at ten or eleven tricks. And I should be able to take four spade tricks more than half the time. So I'm going to try to set up the spade suit.

What's the best way to attack spades? One possibility is to lead the spade jack from my hand. If it's covered, I win and lead a spade back to the nine. If the jack loses to the queen on my right, I cash the ace and king, hoping to drop the ten. Ignoring five-one breaks, this takes four tricks whenever the queen is onside or the ten drops in three rounds. So it loses only to queen doubleton or queen-ten fourth on my right.

Another possibility is to lead toward my jack. If East plays low, I insert the nine. If this loses to the ten, I finesse West for the queen. If I can count on East to hop with queen doubleton or queen third, this plan loses only to queen empty fourth on my right. This would not be a good plan against an opponent capable of ducking with queen doubleton or queen third. But the robots assume declarer is double-dummy, so they will never duck with those holdings. Against a human, you would have to make a judgment. But against robots, the latter line is clearly right.

One problem with this line is it requires two dummy entries. If diamonds break, I have them. The diamond six will be a fourth-round entry. Against a human pair, I wouldn't be worried. But the robots do lead singletons against notrump occasionally. If this lead is a singleton, I may want to change my mind about how I play the spade suit. So perhaps I should test the diamonds before attacking spades. If the six of diamonds isn't going to be an entry, I may change my plan.

But how do I test diamonds? Suppose I win the first diamond in my hand, then lead to the queen. It turns out diamonds split, so I proceed with my plan of leading a low spade off dummy. East hops with the queen and plays a third diamond. This kills the last sure entry to my hand. I may have just cut myself off from the third club. I suppose I could unblock dummy's diamond six to give myself a hand entry. But the whole point of this line was to use the six as a dummy entry.

To keep sufficient entries to both hands, I can't afford to play an extra rounds of diamonds. I'll just have to win the first trick with the queen and hope they split. If they don't, perhaps West has the club ace and I get to dummy via the club queen.

I play the diamond queen from dummy. East plays the seven. That's a worrisome card. It's the only diamond lower than West's eight, so it means West doesn't have a doubleton. If we assume he would have led top of a sequence, his only possible holding are a stiff eight or J98. Since J98 is not a particularly attractive lead, a stiff diamond is starting to look likely.

I lead a low spade from dummy. East hops with the queen, and West follows with the seven.

East shifts to the heart ace. I play low, and West follows with the seven. East continues with the heart six. I see no point in hopping. I put in the jack, and West follows with the eight. We've reached this position, with the lead in my hand:


NORTH
Robot
♠ A K 8 6
--
6 3 2
♣ Q 3






SOUTH
Phillip
♠ J
K 9
A K 5
♣ K J 8

East is a passed hand and has shown up with the spade queen and the ace-queen of hearts. So the club ace is on my left, and I don't need diamonds to split any more. I can cash the spade jack and lead a club toward the queen. If West hops, he scores the third trick for the defense, and I'll have the rest: four spades, three diamonds, two hearts, and a club. Ten tricks.

If West doesn't hop, I'll take eleven tricks. If diamonds split, the fourth diamond is my eleventh trick. And if they don't split, I have East caught in a red-suit squeeze, assuming he has the long hearts.

Unfortunately, I can't imagine West's ducking the club ace if I lead low. He pretty much knows that's the last trick for the defense, so he has nothing to gain by ducking.

Is there any chance I can get him to duck the ace if I lead the club king? Suppose diamonds split and the six is an entry. If West doesn't know that, perhaps he will be afraid of letting me get to dummy with the club queen. If I unblock the spade jack then lead the king of clubs, perhaps he will duck to deprive me of the entry.

At this point, I would rather be playing against humans than robots. Robots don't forget which spots have been played. If diamonds split, they will know for certain that the six is an entry, whereas a human might not be sure. Is it possible the defense might think I started with ace-king third of diamonds? This would be a strange way to play the hand if that were the case. But I suppose it doesn't hurt to try. I do have to cash one diamond first, however, just to make sure they do split. If they don't, leading the club king would be a mistake. 

I cash the diamond ace--nine--deuce--ten. Now the spade jack. Everyone follows. Finally, the club king. West, unsurprisingly, doesn't fall for it. He takes his ace. Making four.


NORTH
Robot
♠ A K 8 6 3
5 4
Q 6 3 2
♣ Q 3


WEST
Robot
♠ 7 4
10 8 7
J 9 8
♣ A 10 6 5 2


EAST
Robot
♠ Q 10 5 2
A Q 6 3
10 7
♣ 9 7 4


SOUTH
Phillip
♠ J 9
K J 9 2
A K 5 4
♣ K J 8

Plus 430 is worth 89%. East had queen-ten fourth of spades, so the normal play in the spade suit would not have worked. Most of the other declarers in three notrump either made exactly three or went down. I was rewarded for insulting my opponent. 

 

Update on 1NT Study

I mentioned in an earlier post that I was keeping records of my results with balanced 14-point hands. I had been advised that it was right to open these hands with one notrump in robot individuals, and I was skeptical. Some readers asked me to keep them advised of my findings.

Originally I was going to compare my results with the average of those who opened with one notrump. I've changed my mind about that approach. For one thing, it proved to be a lot of work. In some tournaments, it meant reviewing what happened at 65 other tables. For another, it's not clear that the comparison would even be meaningful. The average of what other players did is probably a poor proxy for what my own result would have been had I opened with one notrump.

What I have decided to do instead is to keep track of my results on all balanced hands with 14 to 17 HCP. My one notrump openings with 15 to 17 HCP will serve as the control. If I do worse on average with the 14-point hands where I open with one of a suit, it suggests I'm going something wrong. There is no guarantee that I would have done better had I opened with one notrump, but there is at least a suspicion.

The study has just begun, but the preliminary results are already interesting. Incidentally, I am excluding hands with a five-card suit, since it's not clear how to evaluate them. Should a 14-point hand with a five-card suit be treated as 14 points or 15 points? Simply throwing those hands out avoids the need to answer that question. So all these hands are "4432" or "4333" patterns:

HCP Opening Observations Avg Result
14 1C/D 13 60.8%
15
1NT 8
77.7%
16 1NT 5 76.6%
17 1NT 11 77.4%

An unusual feature of this study is, if I decide to start opening one notrump with 14-point hands, I don't have to worry that my average will go down with 15- to 17-point hands. In real life, partner might change how he responds if I widen my range. But the robots are oblivious to my habits, so they will continue to bid the same way. My results with 15- to 17-point hands should be unaffected.

Sunday, April 23, 2023

Free Weekly Instant Tournament - April 21 - Board 2

This week I'm returning to reporting on the Weekly Free Instant Tournament on BBO. If you haven't played it yet, you may wish to do so before reading on.

Board one was boring. After a Blackwood auction, you can bid seven notrump and claim before the opening lead is made. So I'm going to skip to

Board 2
Our side vulnerable

♠ K 7 3   Q J 9 8 6   Q 2  ♣ A J 4  

RHO passes. I bid one heart. Partner responds with one spade and I bid one notrump. Partner bids two clubs, artificial and forcing, I bid two spades, and partner raises to four.

LHO leads the five of diamonds.


NORTH
Phillip
♠ K 7 3
Q J 9 8 6
Q 2
♣ A J 4






SOUTH
Robot
♠ A Q 9 6 4
--
K J 8 4
♣ Q 10 6 2


West North East South
Robot Phillip Robot Robot
Pass 1 Pass 1 ♠
Pass 1 NT Pass 2 ♣
Pass 2 ♠ Pass 4 ♠
(All pass)


I'm off the diamond ace and possibly the club king. If those are my only losers, I'm making five. But I might have a trump loser, and I need to find some way to dispose of my fourth-round diamond loser. The fourth round of clubs isn't an issue if the club finesse loses, but if it wins I'd like to find a way of avoiding that loser as well.

The diamond five could be low, or it could be high from five-three doubleton. I can ruff my fourth diamond in dummy, but I'd prefer not to, since there is a variety of ways doing so could create a trump loser. There is no particular reason to unblock by playing dummy's queen. I might as well play low and hope East takes his ace, obviating the need for a ruff. He probably will take the ace, since he doesn't know who has the king.

I play low from dummy. East indeed plays the ace. Which diamond should I play? If the five is high from five-three doubleton, I want to conceal my eight, making it possible that the lead is low from three. If the five is low from three small, I want to conceal my four, making it's possible that the lead is from a doubleton. It probably won't matter, but if you make a practice of keeping the opponents in the dark whenever possible, it pays dividends. In this case, I'm not sure what the lead is from. But the robots like leading from small doubletons, so I'll assume that's what West has and play the four, concealing the eight.

East shifts to the club eight. This is probably from shortness, quite likely from a singleton in the hope that West has a trump entry. If so, I have three club tricks. Is it possible to avoid losing the fourth round of clubs?

One possibility is to play a heart from dummy at some point and hope East hops with the ace. Then I can ruff and, if West has the heart king, maybe I can squeeze him in hearts and clubs. In general, trump squeezes require two dummy entries. To see why, consider this four-card position with the lead in your hand:

(A)

NORTH
Phillip
♠ --
Q J
--
♣ A 4






SOUTH
Robot
♠ 9
--
--
♣ Q 10 6

West can hold king-nine third of clubs and the heart king and there is no way to take the rest. In this position, however, you do have the rest:

(B)

NORTH
Phillip
♠ --
Q J
--
♣ A J






SOUTH
Robot
♠ 9
--
--
♣ Q 6 2

The second club entry makes the difference. If West holds two clubs, you can take the finesse, cash the ace, dropping his king, and ruff back to your hand. If he holds three clubs, you can lead to the jack, ruff out his heart king, and return to dummy with the club ace.

This position (with the club king already played) works as well:

(C)

NORTH
Phillip
♠ --
Q J
--
♣ J 4






SOUTH
Robot
♠ 9
--
--
♣ Q 6 2

With the clubs unblocked, this is now a simple squeeze. All you have to do is cash the last trump.

What this means is I can't afford to play low on this trick. If I do, West will simply duck, and I will wind up in position (A) with no squeeze. If I play the ten or queen, I reach either position (B) or (C), depending on whether West covers or not. In either case, the squeeze will operate--provided I can isolate the heart guard.

Accordingly, I play the ten. West covers with the king and I take the ace. The heart ruff can wait. I should make sure I don't have an issue in the trump suit first. I play a low spade from dummy--five--ace--deuce. Now the spade four--eight--king--ten. Trumps are breaking, so I might as well ruff the heart now. The sooner I do that the better, since the less East knows about the layout, the more likely he is to hop with the ace.

I play the heart six. East plays the ace, and I ruff. The deal is over. Here is the current position:


NORTH
Phillip
♠ 7
Q J 9 8
Q
♣ J 4






SOUTH
Robot
♠ Q 9
--
K J 8
♣ Q 6 2

I draw the last trump, play a diamond to dummy, and return to my hand with a heart ruff.  Oops. That was careless. Any time you say to yourself "the deal is over," you set yourself up to make a mistake. I should have led the heart queen, just in case East began with ace-king and decides to cover. Of course I am always going to ruff it. But why not give East a chance to make a mistake?

Fortunately West has the king. I cash my diamonds, squeezing West, and take the rest. Making six.


NORTH
Phillip
♠ K 7 3
Q J 9 8 6
Q 2
♣ A J 4


WEST
Robot
♠ J 8 2
K 7 4
5 3
♣ K 9 7 5 3


EAST
Robot
♠ 10 5
A 10 5 3 2
A 10 9 7 6
♣ 8


SOUTH
Robot
♠ A Q 9 6 4
--
K J 8 4
♣ Q 10 6 2

It turns out I didn't need East to hop with the heart ace. He began with only two trumps, so if he doesn't hop, I can still make six by leaving the last trump outstanding and ruffing a club in dummy. It doesn't hurt to try this. If East gets a club ruff, the trick comes back, since I can now ruff my club loser.

My only alternative is to play West for both heart honors and continue with my plan of squeezing him. Normally I would assume West couldn't have ace-king of hearts and not have led the suit. But dummy did bid hearts, so I can't make that assumption here. Still, a doubleton trump in East's hand is more likely. So, even though I hadn't decided what I would do if East didn't hop, I suspect I would have opted for the club ruff.

Making six is worth 89%.  Some declarers dropped a trick by putting up the diamond queen at trick one. Every declarer who got past that test failed the second test by playing low on the club shift. Unfortunately, they weren't punished. West foolishly played the king, giving declarer the whole club suit. Had West played low, those declarers probably would have been held to eleven tricks. I doubt they would have found the partial trump elimination. It's not an especially good field, as evidenced by the fact that the boring board one, which should have been flat, was worth 100%.

Sunday, April 16, 2023

Zenith Daylong Reward - April 13 - Board 14

Board 14
Neither vulnerable

Last week, after 382 posts, I broke my commitment to choose deals for this blog at random. Now that there's a precedent, I feel emboldened to do so again. This is how slippery slopes work. But there are so many fascinating aspects to this deal that I had to post it. It features an interesting bidding problem. It features a challenging play problem. It illustrates the biggest flaw in the way bridge robots are designed. And it finishes with a Shyamalanian twist ending.

First, some preliminary remarks. Gabrielle Kamil, who has managed some impressively high scores in robot individuals, told me that in these events you should routinely open one notrump with 14-point balanced hands. I wasn't convinced. But to show how open-minded I am, I decided to keep a log. I sometimes evaluate a 14-point hand as a strong notrump. But if don't on a particular deal, I will record my result along with the average result of everyone who opened one notrump. If, over time, their results are better than mine, I'll change my strategy. 

Unfortunately, if their results are worse than mine, the test will be inconclusive. What I'm really interested in is what would happen if I had opened with notrump. Since there is no way to know, so I'm using the field average as a proxy. In practice, I suspect my results would be better.

I began this project recently, so today's deal is only the second entry in my log. On the first entry, I scored 61% opening one club with

♠ K 9 8 7   A J 5   K 10  ♣ K 9 6 3.  

Those who opened with one notrump averaged 21%. (Partner has a bad hand. If you open with one notrump, you go minus. If you open with one club, you let the opponents into the auction. They reach a good game in their four-four spade fit, but it goes down on the bad trump break.)

This time, I pick up:

♠ A K J 5   A J 10 6   J 7  ♣ 6 5 4.  

This hand looks a lot more like a strong notrump opening than the previous one. It has three and a half-plus honor tricks, so it qualifies by pre-HCP standards. But I still don't like one notrump, because I hold both majors. If partner has less than nine HCP, opening with one club may enable us to find a four-four major-suit fit that those who open with one notrump will miss.

I bid one club, and partner bids two clubs, showing a limit raise or better in clubs.

Now I wish I had opened with one notrump. I want to bid game. But two notrump by me is non-forcing, and three notrump shows 18 or 19 HCP. The standard way to solve this problem is to temporize with two of a major, then bid three notrump. But I'm disinclined to do that. If I bid two hearts and partner bids two notrump, we are probably playing this from the wrong side. I'd like to make sure to get the notrump bid in first.

Maybe I can. Since this is a best-hand tournament, it is unlikely partner will have a bid over three notrump anyway. So I doubt it will hurt to let him believe I have 18 or 19 HCP. In fact, it may help, since it may induce the opponents to misdefend. I once saw Pete Hollands make this bid for similar reasons. Accordingly, I bid three notrump. Partner raises to six notrump.

Well. That didn't work out. Maybe this was a bad idea. Now that I think of it, when Pete did this, his partner was a passed hand. The one-notrump openers will say I got exactly what I deserved. I underbid with one club, then grossly overbid with three notrump, trying to solve the problem I had created for myself. 

There's nothing I can do about it now. I pass, and West leads the spade three.


NORTH
Robot
♠ Q 10 4
8 7
K 10 2
♣ A Q J 10 9






SOUTH
Phillip
♠ A K J 5
A J 10 6
J 7
♣ 6 5 4


West North East South
Robot Robot Robot Phillip


Pass 1 ♣
Pass 2 ♣ Pass 3 NT
Pass 6 NT (All pass)

Six notrump seems a bit aggressive with partner's hand. But partner is closer to his bid than I am to mine, so I can hardly complain. What do I need to make this? I need the club finesse for starters. That brings me up to ten tricks. If the ace and queen of diamonds are onside, maybe I can score two diamond tricks to come to twelve.

The spade three is hard to read. It could be lowest from three or four, fourth best from five, or high from a doubleton.

I play low from dummy. East plays the eight, and I win with the jack. I play the four of clubs--three--ten--eight. So far, so good. I need to repeat the club finesse and take two diamond finesses, so I need three entries to my hand. If spades are three-three, I have two spade entries, but using the heart entry is problematic. If I lead a heart to the ace, I set up heart tricks for the defense.

Perhaps I can squeeze West down to a stiff heart. If I play a spade to my hand, repeat the club finesse, run clubs, and play another spade to my hand, I come down to this position:


NORTH
Robot
♠ --
8 7
K 10 2
♣ --






SOUTH
Phillip
♠ 5
A J
J 7
♣ --

Now the last spade squeezes West. If he comes down to ace-queen tight of diamonds, one diamond play from my hand suffices. If he comes down to three diamonds and one heart, I can play a diamond to dummy and return to my hand with the heart ace and impunity.

Another possibility is simply to cash the club ace, hoping the king drops. If it does, I need only two entries to my hand, so I don't need to play for three-three spades.

Which layout is more likely? Three-three spades or a doubleton club king? Before I decide, there is a third line to consider. I can play a heart to my jack. If West doesn't hold the diamond ace, he will expect me to have it for my three notrump bid. So, after he wins the heart, he will see no reason to break diamonds. He will probably exit passively with a spade. Then I can cash the spades and clubs, coming down to this position:


NORTH
Robot
♠ --
8
K 10
♣ 9






SOUTH
Phillip
♠ --
A 10 6
J
♣ --

If East began with four hearts, the last club squeezes him.

East can thwart this plan by hopping with his heart honor. But I can't imagine anyone's finding that play from honor fourth. This looks like a better plan to me than either of the previous two lines I considered.

I play the heart seven; East, the deuce. Does it matter whether I play the ten or the jack? Probably not. But the ten might make my heart suit look too threatening. The stronger West thinks my heart suit is, the more likely he is to play his partner for the diamond ace in desperation.

I play the heart jack, and West takes the queen. He doesn't find the diamond shift. But, unfortunately, he continues hearts. That's another way to thwart my plan. The heart continuation destroys my entry for the squeeze. I play dummy's eight, East plays the nine, and I win with the jack. The good news is I may not need the squeeze. Dummy's hearts were eight-seven. So if East began with king-nine third, my heart six is good.

Do I have any chance if East began with four hearts? Maybe I can execute a pseudo-squeeze. I can cash the clubs, pitching two diamonds, then cash spades, finishing in my hand. If East thinks I'm 4-3-3-3, he will hold the diamond ace instead of the long heart at trick thirteen. I could cash the heart ace first to see if the king drops, but I see no reason to do that. If it drops now, it will drop later. And cashing the heart ace may make it easier for East to count my hand. If there is only one small heart outstanding, East may be suspicious when West doesn't pitch it.

I play the club five--seven--jack--diamond five. Oops. Clubs are four-one. Now I have a new problem. I need to repeat the club finesse. Since I have to win trick twelve in my hand for the pseudo-squeeze to work, I need two hand entries. I'm back to overtaking the spade and hoping for three-three spades.

Or am I? If the heart king is dropping, I don't need the pseudo-squeeze. I have the rest in top tricks. Maybe there was a reason to cash the heart ace after all. Then I would know for sure whether I had to overtake the spade.

Can I infer how spades split? East has a stiff club, so it's unlikely he has a doubleton spade. That means West's spade three was probably lowest from three or high from three-deuce doubleton. I'll play the spade ten and see what East follows with.

On the spade ten, East plays the deuce. It appears, then, that West's lead from lowest from three. I overtake the ten with the king and repeat the club finesse. Then I run clubs and overtake the spade queen. Luckily, both opponents follow. I have now reached this position:


NORTH
Robot
♠ --
--
K 10 2
♣ --






SOUTH
Phillip
♠ 5
A 6
--
♣ --

On the last spade, my opponent pitches a heart, clutching his diamond ace. My heart ace drops his king, and my heart six takes the last trick. Making six.

So where is the Shyamalanian twist ending I promised you? It turns out it is West, not East, who is pseudo-squeezed. This is the full deal:


NORTH
Robot
♠ Q 10 4
8 7
K 10 2
♣ A Q J 10 9


WEST
Robot
♠ 7 6 3
K Q 5 3
A 6
♣ K 7 3 2


EAST
Robot
♠ 9 8 2
9 4 2
Q 9 8 5 4 3
♣ 8


SOUTH
Phillip
♠ A K J 5
A J 10 6
J 7
♣ 6 5 4

Before you are too critical of West's defense, let me point out that he defended correctly according to his programming. When he was in with the heart queen, he knew I had at most ten tricks. He could cash the diamond ace for down one. But if his partner had the heart ten, he could beat me two by continuing hearts.

Double-dummy, continuing hearts can't cost. At worst, it gives me the heart ten for my eleventh trick. So playing a heart either gains or breaks even versus cashing the diamond ace. But while the play entails no risk double-dummy, it does entail a risk single-dummy. It requires West to read my shape and discard accordingly.

Since plus fifty is probably a 100% board anyway, taking any risk at all of letting me make this can't be worthwhile. But the robot doesn't even see that there is a risk. As far as he's concerned, a heart can't possibly cost. This is the biggest flaw in the robots' design. In analyzing the play, they assume everyone is double-dummy. Usually in making this assumption you give your opponent too much credit. Sometimes, as here, you give yourself too much credit. 

Of course, while West's defense may be understandable given his limitations, I can't say the same for his opening lead. Or his final pass. I don't understand either one.

I wish I could replay the board to see if my choice of the heart jack rather than the ten made a difference. If I played the heart ten, would West think it was impossible for his partner to hold the jack and just cash the diamond ace? He already knew I didn't have the 18 HCP I promised. Who knows what assumption he would make about the heart jack?

Anyway, I now have entry number two in my log. Opening with one club scores 100%; opening with one notrump averages 54%. 

And, by the way, I was wrong about this being a best-hand tournament. While most robot individuals are, the Zenith Daily Reward tournaments are not. It's fortunate I didn't remember that. Had I done so, I never would have bid three notrump.

Sunday, April 9, 2023

Zenith Daylong Reward - April 1 - Board 12

Board 12
Our side vulnerable

I don't usually pick and choose the boards I discuss in Gargoyle Chronicles. I follow a pre-determined procedure to select a board before I play it, effectively randomizing my selection. I have two reasons for this approach:

(1) Michael Rosenberg once told me that all bridge deals have something interesting about them. If you can't see it on a particular deal, you just have to dig deeper. Selecting a deal at random to write about forces me to do that, and I sometimes discover things I wouldn't otherwise see.

(2) If I choose the deals I discuss, it is hard to avoid focusing on deals where I do something good. There is much to be learned from examining errors, and choosing boards at random ensures that a fair portion of the deals I discuss will include errors. 

Nonetheless, after 382 posts following this approach, I can't resist making an exception this week to discuss a deal I particularly enjoyed:

♠ K   K 8 6 3   A K 8 3  ♣ A 7 6 5  

LHO passes, partner opens with one club, and RHO bids two spades, weak. I make a negative double, and partner bids three clubs. Since spades weren't raised, there is good chance partner has enough in spades to provide a stopper opposite my stiff king. I could bid three spades to make sure. But I doubt partner would bid three notrump with, say, jack third, which may suffice. Even three small may suffice, since RHO might duck the opening spade lead to preserve communication. So if I'm content to play three notrump, I should probably just bid it.

Am I content to play three notrump? What do I need to make a club slam?

♠ x x x  A x  Q x x  ♣ K Q x x x 

is enough to make slam cold on normal breaks. Since that is a perfect minimum--actually a jack less than a minimum--I have enough to invite a slam by Culbertson's Rule.

But it's hard to construct an intelligent auction to investigate slam opposite a robot. Part of the problem is we have no way of stopping in four notrump. If I bid past three notrump, I'm pretty much committed to bidding a slam, since I have no desire to play five clubs at matchpoints. I don't have a slam drive, so, as a practical matter, I think I should settle for three notrump, though I'm not happy about it.

I bid three notrump, everyone passes, and LHO leads the spade five.


NORTH
Robot
♠ A 10 8 2
A 5
9 5
♣ K Q 10 8 2






SOUTH
Phillip
♠ K
K 8 6 3
A K 8 3
♣ A 7 6 5


West North East South
Robot Robot Robot Phillip
Pass 1 ♣ 2 ♠ Double
Pass 3 ♣ Pass 3 NT
(All pass)


Six clubs looks pretty good, and some will bid it. But there is no sense in worrying about whether it makes or not. I'm not competing against those pairs any more. My only concern is outplaying the other declarers in three notrump.

I have eleven tricks off the top, assuming clubs come home. Squeeze chances don't look good. RHO stops spades. If West guards one of the red suits, I would have a double squeeze except for the fact that both red-suit threats are in my hand. One of the threats must be behind your opponent for a squeeze to operate.

Maybe I can execute a simple squeeze against East. If he has four cards in one of the red suits, can I squeeze him somehow? I don't see how. I need to duck a trick both to correct the count and to isolate the threat. And if I do that, they can attack my entries. Say, for example, I duck a diamond, playing East for four diamonds. Now the defense can simply play another diamond, and I don't have an entry in either threat suit. Another necessarily condition for a squeeze is that you must have an entry in one of the threat suits.

I might as well get a better picture of the hand by running some clubs. Once I'm dealing with specifics rather than generalities, something might occur to me. I win the spade lead in my hand as East plays the spade six. I play a club to dummy's king. East discards the spade three. So East is probably 6-4-3-0 or 6-3-4-0. I play a club to my ace, and East discards the diamond deuce. He wouldn't be discarding from his four-card red suit, so he must be 6-4-3-0.

Now that I know what he has, what happens if I run the clubs? East must come down to seven cards. He must hold three spades and four hearts, else I can duck out a trick in one of those suits. So he must discard all his diamonds. If he started with two diamond honors, I'm home. I can lead the diamond nine and drive the remaining honor, setting up my eight. In fact, maybe all I need is for him to have started with one diamond honor. If West holds two honors fourth in the four-card end position, I can endplay him.

I run all the clubs. East, perforce, holds three spades and four hearts, pitching both his remaining diamonds, the seven and the jack. I don't need all my hearts anymore, so I pitch the heart three. We are down to this position:


NORTH
Robot
♠ A 10 8
A 5
9 5
♣ --


WEST
Robot
♠ --
x x x
Q 10 6 4
♣ --


EAST
Robot
♠ Q J 9
x x x x
--
♣ --


SOUTH
Phillip
♠ --
K 8 6
A K 8 3
♣ --

Now it is West who is squeezed. I cash the spade ace, pitching another heart. West must hold four diamonds, else I can duck out a diamond trick, so he has to come down to a doubleton heart. I cash the king and ace of hearts and float the nine of diamonds to endplay him. Making six.


NORTH
Robot
♠ A 10 8 2
A 5
9 5
♣ K Q 10 8 2


WEST
Robot
♠ 5 4
J 10 4
Q 10 6 4
♣ J 9 4 3


EAST
Robot
♠ Q J 9 7 6 3
Q 9 7 2
J 7 2
♣ --


SOUTH
Phillip
♠ K
K 8 6 3
A K 8 3
♣ A 7 6 5

This is not a particularly difficult deal to play. You just run the clubs and it plays itself. I like it because it combines so many different types of squeezes. It is a delayed ducked squeeze, with the threat of establishing a trick by ducking in all three threat suits. It is also a double squeeze, involving a triple squeeze on one opponent and a simple squeeze on the other. It has elements of a guard squeeze (where an opponent is squeezed out of an honor that protects his partner against a finesse), although in this case the "finesse" must be accomplished by endplaying that partner. Finally, the concluding simple squeeze is non-material. West is squeezed out of an exit card, not a guard. So what should we call it? A delayed-duck, quasi-guard, non-material double squeeze-endplay?

Plus 690 is worth 62%. Four pairs did reach six clubs. I was curious if anyone had found a sensible way to investigate, but everyone who bid six clubs got there by blasting. So I don't blame myself for not reaching slam. If you can score 62% simply by playing three notrump correctly, I don't think the risk-reward ratio is right for blasting to a speculative slam.

Sunday, April 2, 2023

Free Weekly Instant Tournament - March 31 - Board 8

Board 8
Neither side vulnerable

Phillip:

Today's guest on Gargoyle Chronicles is ChatGPT. While, strictly speaking, ChatGPT is not a bridge player, he has done an incredible amount of research on the game and is happy to share his insights with us as we analyze this week's deal. Welcome, ChatGPT, and thank you for coming.

ChatGPT: 

Thank you for having me on Gargoyle Chronicles! I'm excited to share my insights on bridge and help analyze this week's deal.

Phillip: 

With neither side vulnerable at matchpoints, I pick up

♠ 6 4   K Q J   A Q J 9 5 3  ♣ K 10  

LHO opens with one spade, partner passes, and RHO bids one notrump, forcing. What would you do now?

ChatGPT: 

With your strong and distributional hand, you have several options. One possibility is to bid 2, which is a transfer to hearts, showing at least a five-card heart suit and asking partner to bid 2. If partner has four hearts and a weak hand, they will bid 2; if they have a strong hand or a different four-card suit, they may bid something else. 

Another option is to bid 2♣, which is a waiting bid. This shows a good hand with no immediate fit for partner's suit and asks partner to describe their hand further. They may bid a second suit, which could help you find a fit, or they may repeat their notrump bid with a balanced hand. 

A third option is to jump to 3, which shows a long and strong diamond suit with game-forcing values. This bid tells partner that you have at least five diamonds and are interested in exploring the possibility of a diamond contract. 

Ultimately, the best choice depends on your partnership agreements and your personal style. Each bid has its advantages and disadvantages, and you need to weigh the risks and rewards of each option. However, given your strong and distributional hand, it's likely that you will end up in a game contract of some kind.

Phillip:

Interesting ideas. I thought my choice was between passing or bidding two diamonds. If one notrump were not forcing, I might have passed in the hope of going plus defending one notrump. I'm not sure passing is right even then. But since one notrump couldn't end the auction, I saw no reason not to introduce my suit. So I bid two diamonds.

ChatGPT: 

Bidding 2 is also a reasonable choice with your hand. It shows a long and strong diamond suit and invites partner to bid their next highest suit, which could help you find a possible fit or even explore the possibility of a diamond contract.

Passing with your hand is also not out of the question, especially if you have a strong defensive hand and think you might be able to set the contract. However, with so many potential points and a strong suit of your own, it's understandable why you chose to bid instead.

As always, the best choice depends on your partnership agreements, style, and the specific context of the auction. But bidding 2 with your long and strong diamond suit is definitely a reasonable option.

Phillip:

Thank you. Over two diamonds, LHO bid two hearts, which was passed back around to me. Would you sell out or bid on?

ChatGPT: 

With your hand, bidding on is definitely a reasonable choice. You have a strong hand with a long and strong diamond suit, which suggests that you have a good chance of making a game contract in diamonds. Moreover, your partner has not yet had a chance to show any strength or distribution, so there may be additional potential for finding a fit or other good contracts.

At this point, you could bid 3 to show your continued interest in diamonds and your game-forcing values. This should encourage partner to bid again if they have any additional values or a diamond fit.

Alternatively, you could jump to 4 to show your long and strong diamond suit and your willingness to play in game opposite a minimum hand from partner. This bid should discourage partner from bidding on if they have a minimum hand with no diamond fit.

Ultimately, the best choice depends on your partnership agreements and your personal style. But with your strong hand and long diamond suit, bidding on to game seems like a reasonable choice.

Phillip:

Yes, with eight playing tricks, I wasn't about to sell out. But four diamonds seems a bit aggressive to me. I settled for three diamonds. Everyone passed and LHO led the heart ace. 

Dummy hit with ♠ A 5 2 10 5 3 10 6 2 ♣ 8 7 6 5.

[All hand diagrams are for the readers' benefit. They were not part of the conversation.]


NORTH
Robot
♠ A 5 2
10 5 3
10 6 2
♣ 8 7 6 5






SOUTH
Phillip
♠ 6 4
K Q J
A Q J 9 5 3
♣ K 10


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


ChatGPT:

Bidding 3 was also a reasonable choice with your hand, and it's better to err on the side of caution than to risk going down in a high-level contract.

After the opening lead of the A, you can see that you have only two heart tricks, so you will need to look elsewhere for additional tricks. One option is to try to establish additional tricks in the diamond suit by playing the A and then a low diamond towards your hand. If the diamonds are splitting favorably (either 3-3 or 4-2 with the longer diamonds on your right), you may be able to establish additional tricks in that suit.

Another option is to try to establish tricks in the spade suit, either by setting up the ♠K or by ruffing spades in the dummy. This will depend on the distribution of the spades and how many entries you have to the dummy.

In any case, it's important to count your winners and losers carefully and make a plan for how to take additional tricks. You have a strong hand with a good chance of making the contract, but you will need to play carefully and use your resources wisely.

Phillip:

The ace of hearts is a strange lead, so I know quite a bit about the opponents' hands. West would have led a spade from king-queen or queen-jack, so East probably has the spade queen, East would have shown a spade preference with three spades or with two-two in the majors, so he rates to be two-three, making West six-four. West would probably have preferred a singleton club lead to the heart ace, so that means West is 6-4-1-2. He might be 6-4-0-3. But if he is, I have an unavoidable diamond loser, so I might as well assume he is 6-4-1-2, giving me a chance of picking up the diamond suit.

ChatGPT:

Your analysis of the hand and the opponents' likely distribution is spot on. Given that West is likely 6-4-1-2, you have a good chance of picking up the diamond suit if you play it correctly.

Assuming that West has a singleton club, you can win the K in your hand and then lead a low diamond towards the dummy's ten. If West plays low, you can play the Q from the dummy, which will force West to take the K. You can then win any return and play another diamond, establishing the suit for three tricks.

If West plays the K on the first diamond, you can win with the A and play another low diamond towards the dummy's ten. This time, West will be forced to play low, allowing you to play the Q from the dummy on the next trick to establish the suit.

If West has four diamonds to the king, you will need to rely on a favorable split or a squeeze to make the contract. But given your strong hand and the favorable distribution, you have a good chance of making the contract if you play your cards right. 

Phillip:

East played the heart four at trick one, and I dropped the jack. West continued with the heart seven, East played the six, and I won with the king. 

I had eight tricks and could take either a diamond finesse or a club finesse for a ninth. Unfortunately, I had only one dummy entry, so I couldn't take both. If I took the diamond finesse, I had to lead the ten to repeat it, so I wouldn't have a dummy re-entry in diamonds if the finesse lost. I knew diamonds weren't two-two, so I couldn't have a third-round re-entry. It appeared I had to try one finesse or the other. Rather than guess which finesse to take, however, I saw a way to give myself an extra chance. Do you see it?

ChatGPT: 

Yes, there is a way to give yourself an extra chance to make the contract without having to guess which finesse to take.

Instead of immediately playing a finesse, you can cash the A and then play the Q, discarding a small club from your hand. This has the effect of "unblocking" the diamond suit and creating an extra entry to the dummy. If West has the K, he will be forced to take it, and you can win any return and play the J, hoping to drop West's remaining diamond. If the K does not drop, you can still try the club finesse later, using the extra entry to the dummy to finesse both ways if necessary.

If East has the K, you will need to rely on a favorable split or a squeeze to make the contract, but by unblocking the diamond suit first, you give yourself the best chance of success.

Phillip:

Yes. Cashing the diamond ace, hoping to drop the king, is a possible line. If it fails, you can then take the club finesse. That gives you two chances instead of just one. Unfortunately, if the diamond king doesn't drop, you can be pretty sure the club ace is offside. It's barely possible for West to have an opening bid with neither of those two cards. But most of the time, he will have at least one of them to open. Furthermore, East, with an ace, a king, and the spade queen, might have bid over two hearts. So cashing the diamond ace can't be my best play.

I decided on a different discovery play. I led the club king out of my hand. If it lost on my right, it would now be likely that West had the stiff king of diamonds. If it lost on my left, I would take the diamond finesse. Of course, West might thwart my plan by unblocking the club queen from queen doubleton and scoring a ruff with his diamond king. But, since my alternative is taking a diamond finesse, it doesn't hurt to try this maneuver. Maybe he won't find the unblock. Maybe it's not even possible; he might have queen-jack doubleton.

ChatGPT:

That's an interesting idea as well. By leading the ♣K, you gain additional information about the distribution of the opponents' hands. If East shows out, you can be fairly confident that West has the K, and you can take the diamond finesse with more confidence. If East follows suit, you can try the club finesse, knowing that West is unlikely to hold both the ♣A and the K.

However, if West does have both the ♣A and the K, this line of play could result in two losers instead of one. If West wins the ♣A and returns a spade, you will be forced to lose a spade and a diamond, regardless of whether you take the diamond finesse or the club finesse. But the chance of this happening is relatively small, so your line of play is still a reasonable one.

Phillip:

I led the club king, West took the ace, and East followed with the four. West continued with the nine of hearts, East played the deuce, and I won with the queen. I could cross to dummy to take a diamond finesse. But I decided that could wait. There was no harm in exiting with a club. Sometimes if you put the opponents on play, they do something foolish. I led the club ten, West played the jack, and East followed with the deuce. West continued with the club three to East's queen, and I ruffed with the diamond three.

That was disappointing. West was 6-4-0-3. The diamond king was onside, but I still had a diamond loser. Was there any possibility of picking up the king? It would require gross misdefense, but I did see one chance. Do you?

ChatGPT:

It's always good to be alert for the possibility of a gross misdefense, even though it's unlikely. In this case, the only way you could pick up the K would be if West held both the ♣A and the K, and made a mistake on the club play at trick 10.

Suppose you lead a diamond to the queen at trick 10, and West plays the J. This would be a very strange play if he holds the K, as he could just play the king and end the defense. So, you might infer that West has made a mistake and played the J instead of the ♣J (which he intended to play as a suit-preference signal).

If you now lead another club, West might misinterpret the situation and play low, thinking that East has the K and is trying to set up the suit for you. If that happens, you can ruff the club and lead another diamond, forcing West to win the K and give you the ninth trick.

Of course, this line of play relies on a very unlikely misdefense by West, so it's not a practical option in most cases. But it's still interesting to consider all the possibilities, as you never know when you might encounter a similar situation in the future.

Phillip:

I hadn't thought about suit-preference confusion. This was my idea: 

Suppose I crossed to dummy with the spade ace and led the diamond ten. East would duck. I could repeat the finesse [to reach the following position:]


NORTH
Robot
♠ 5 2
--
6
♣ 8






SOUTH
Phillip
♠ 6
--
A Q 5
♣ --

and exit with a spade. If West won and led the thirteenth heart, I could ruff in dummy and coup East at the end. I couldn't see that happening, especially since East would probably win the spade with his queen, but I had nothing to lose by trying. I did, however, need to make sure to keep my diamond five, so I could underruff the six at trick eleven.

ChatGPT:

Your line of play is also interesting. If you lead the diamond ten from dummy and East ducks, you can continue with another diamond and ruff in your hand. Then you can lead the spade ace and a spade, hoping that West wins and leads a heart, which you can ruff in dummy to set up your last diamond. However, this line requires that East has the diamond king and that West has the thirteenth heart, which is unlikely.

If you do decide to try your line of play, make sure to be careful with the order of your discards. You need to keep your diamond five for the endgame, but you also need to make sure to discard correctly on the hearts to signal to your partner which suit you want to keep. Good luck!

Phillip:

I didn't have much luck actually. I played the spade four--three--ace--jack. Now the diamond ten. East played the four, and I unblocked the nine. Unfortunately, West pitched the heart eight. There went my only chance. He no longer had the option of giving me a ruff-sluff. I repeated the diamond finesse but still had to lose to the diamond king, so I was down one.


NORTH
Robot
♠ A 5 2
10 5 3
10 6 2
♣ 8 7 6 5


WEST
Robot
♠ K 10 9 8 7 3
A 9 8 7
--
♣ A J 3


EAST
Robot
♠ Q J
6 4 2
K 8 7 4
♣ Q 9 4 2


SOUTH
Phillip
♠ 6 4
K Q J
A Q J 9 5 3
♣ K 10

ChatGPT:

It's too bad that it didn't work out, but it was a good effort nonetheless. Sometimes, even the best efforts don't yield the desired results. It's important to remember that bridge is a game of probabilities, and sometimes the odds just aren't in your favor. Don't get discouraged by one result and continue to focus on making the best decisions possible in future hands.

Phillip:

Excellent advice. But it turned out to be a pretty good result. Minus 50 was worth 86%. Some players sold to two hearts. Others passed over one notrump, which had a strange effect. Without the diamond overcall, East had no reason to slow down the auction, so he corrected two hearts to two spades. Now, when South bid a belated three diamonds, West knew to compete to three spades. It's not surprising that bidding late rather than early worked out badly. It often does. But the reason it worked out badly in this case is a little surprising.

ChatGPT:

Yes, bidding late in the auction can often lead to confusion and suboptimal results. In this case, the delayed bid of three diamonds by South allowed West to make an informed decision to compete to three spades, as opposed to the blind decision he would have made had South bid diamonds earlier in the auction. This underscores the importance of making your intentions clear early on in the bidding process and not waiting until later to make your bid. Additionally, it's important to consider the potential impact your bid might have on the subsequent auction, as it can influence your opponents' decisions and potentially lead to unexpected outcomes. 

Phillip:

Thank you for your assistance. I had no idea you knew so much about bridge. It was a pleasant surprise. While some of the things you said didn't make a lot of sense to me, our discussion did make more sense than some I have had on BridgeWinners. 

ChatGPT:

You're welcome! I'm glad that I could be of help to you. As an AI language model, I've been trained on a wide range of texts and topics, including bridge, so I'm always happy to share my knowledge and insights. If you have any other bridge questions or if there's anything else I can help you with, feel free to ask!