Saturday, November 27, 2010

Finessing to Eliminate a Guess

A common card combination we see almost every time we play bridge is something like A432 opposite QT65. You need to play this for 3 tricks. We all learn early that the correct play is Ace and then small toward the Q-10. If your opponent follows with a small card on the second lead, then you are left with a guess as to the location of the King so your chances are 50-50 of getting your 3 tricks.

Louis Watson’s Play of the Hand at Bridge (1933) remains the classic reference for playing the bridge hands correctly. In a sub-part dealing with advanced finesses, Watson displays a hand that is somewhat similar to the one discussed in the above:

QT98 (Dummy/North)

A742 (Declarer/South)


Again the object is to take 3 tricks. Assume no bidding or prior play that suggests the location of specific cards. If you apply what you know from the first example you would lay down the Ace and then play low to the queen-ten, again subjecting yourself to a guess on the location of the King. What Watson has done is add some middle cards his hand to create finessing positions in each hand instead of just one. Q-T in the dummy which we can use to finesse the jack and A-7 in the hand over the King once the Jack is played. If you start with the Ace you will give up one of these critical finessing positions so don’t lay down the Ace. If you remember just that, you will dramatically increase your chances for 3 tricks. But this is Watson’s hand and he wants you to start by leading the Queen from the North hand. Here is a summary of his explanation:

Scenario I: First, give East Kx and give West Jx. You can distribute the 5th card to either defender. West has the King and it makes no difference whether he covers or not. If he covers you play the Ace and you have the three tricks you need by simply smoking out the Jack. If West doesn’t cover the Queen wins, and since there are only 3 cards left in the suit after trick one, and you cannot lose more than 1 trick. Scenario II: Now switch the location of the King and Jack. If West has the King and captures the Queen, you are down to 3 cards missing the Jack. The next time South gets the lead he runs the 10 through East finessing the Jack. Scenario III. Give KJx to East. East can cover or not, but he will still only gets one trick in the suit. Scenario IV. Give West KJx. In this case we cannot prevent West from getting two tricks. We have three winning scenarios and one losing scenario. By leading the Queen first you have eliminated the guess on the finesse and increased your chance of making 3 tricks in the suit from 50% to 75%.

Hugh Kelsey in his book Bridge Odds for Practical Players (1980) has a little different take on the same combination. It is the same idea, but he says that if you start with the 10 from the North hand rather than the Queen, you increase your chances by another 2% to 77%. This is because you will win 3 tricks not only when the array breaks 3-2, but also when when East has KJxx as well. As usual, he is right.

Mike Lawrence recently gave this age old problem a different approach on his excellent web site Bridge Clues http://www.bridgeclues.com/. His co-host is Anne Lund, a bridge expert, director and teacher from California. If you have not visited the site, I urge you to do so. There are both bidding and play problems at 3 different levels and they change daily. Mike suggests rather than leading the Queen or ten from the dummy, that you start from the South hand and lead a small card toward the Queen.

We have now changed the orientation a little because declarer will win three tricks any time both honors are in the same hand and also when East holds the King and West holds the Jack. It will lose if East holds the Jack and West holds the King. But let’s examine that winning scenario a little closer.

You are West and hold Kxx. When I lead small toward the QT98 what are you going to play? Well I hope you didn’t go up with the King. If you ducked my plan is to put in the 10 which your partner will win with the Jack. I will now lose two tricks since you still get your King. If you popped the King, I am ducking and your partner’s Jack gets smothered on the next trick. You turned my 75% chance into a 100% chance. You screwed up the only combination that wins for you.

Let’s make the cheese a little more binding! You hold Kx and partner holds Jxx. I lead small to the Queen. You are one helluva bridge player if you ducked smoothly and held onto your King. If you go up with the King same result, wine into vinegar. If you duck it is correct for me to put in the ten which loses to pard’s jack. Now I finesse pard for the King the next time I get in (also the percentage play) and you smile and win with your stiff King. Fixed Again!!

A couple of observations. If you play small to the queen you are no worse off since against perfect defense you can win three tricks only 75% of the time. It is human nature to take tricks when offered and preying on natural instincts is often winning bridge. If you go into the tank with the Kx declarer might figure it out and drop your King doubleton on the second lead even though the a priori odds on the play are less than 3%. So don’t turn a 3% play into a 100% play either. Did I mention it is a tough play???

This may be a bit hard to follow in prose, so get out 13 cards and set them up in any of the hypothetical arrays that I have suggested. If I am wrong, feel free to write either Watson or Kelsey in Bridge Heaven. Please don’t bother Mike Lawrence. If you have figured out some 4-1 distributions or 5-0 distributions that produce nightmares, that is not what this blog post is about. The residue will split 2-3 2/3rds of the time and I am going with the odds. Did you catch that memory hook?

Sunday, November 21, 2010

Surviving with Queens and Jacks in No Trump

Queens and Jacks, even if they are supported by other non-honor cards, are not very valuable in suit contracts, since tricks that can be won only after 2 or 3 leads of the suit often disappear. In no trump contracts, these middle cards take on more value as they can become stoppers in the suit and produce a trick. The following discussion assumes entries are not a problem.

1. Assume you are declaring a no trump contract. You hold QJx in a suit and dummy has xxx.
(a) What is the probability this holding will provide a stopper in the suit and produce a trick if opponents lead the suit?
(b) Does it make and difference if opponents do not lead this suit and force you to lead it?
(c) If your contract was dependent on this combination producing a trick would you still bid it?

2. Assume again you are declaring in no trump. You hold Qxx and when the opening lead comes down you see Jxx in the dummy.
(a) Is this combination any different that the first combination in result?
(b) If so, is it more or less vulnerable if opponents lead the suit?
(c) If your contract was dependent on this combination producing a trick would you still bid it?

3. Does the fact that declarer hold 6 cards in the suit have any practical significance in either case?

Here is the discussion of each issue:

1. The holding of QJx in any hand will produce a guaranteed stopper except where the A and K are both sitting behind the QJx. Thus, if held by declarer (always South), it is a winner when East holds AK (25%), or when the A and K are split between the two defenders (50% since they can split 2 ways) but loses to AK West (25%). Thus it is a guaranteed stopper and will produce a trick 75% of the time. With this holding it does not make any difference if opponents lead the suit or if declarer is forced to lead the suit, the probabilities remain the same. With a 75% probability I want to be in the contract every time. It’s like asking if you would like a 75% game. If West holds the AK, hopefully your disappointment will be shared by many others.

2. In the case where Qxx is in one hand and Jxx in the other, the situation changes somewhat. Note that if opponents lead this suit, it will produce a trick 100% of the time as long as declarer ducks on the hand to first play to the trick. With this holding you are happy to have the suit played on opening lead. That is the good news. The bad news is that if declarer if forced to play this suit it becomes a 50% proposition and will not produce a winner when the A and K are divided between the opponent’s two hands. If you merge the two probabilities you still have a 75% chance of making a trick, so I still want to be in the contract.

3. Note in each case you had 6 cards in the suit. One of the concerns is opponent’s holding in the suit will split 5-2 and they may be able to run 5 tricks before you get your track shoes on. The defenders will hold a 5+ card holding only 1/3 of the time and only 1/6 of the time will opening leader have that holding. This provides some additional element of safety. You probably noticed that QJx opposite xx will also produce a stopper and a trick 75% of the time. But now you only have 5 cards in the suit and the odds of one of the opponents holding 5+ cards in the suit has doubled.

If the suit is not led on the opening lead, this information can be helpful in planning your further play. Do you take the opportunity/risk to set up a trick in another suit for an overtrick or should you take your tricks and run. With QJx only one holding of four can hurt you. With Qxx opposite Jxx, two holdings of four can doom your contract. Do you run the risk for an overtrick? If the scoring is IMPs, whether a game or part score contract, I would take the guaranteed plus score and not take any risk for an overtrick. In matchpoints I would try for the overtrick with either a 75% or 50% chance of success. The odds feel right to me.

What if there is an overcall and partner makes a Western Cue bid? Do you treat QJx as a stopper? What about Qxx? We ask questions, you decide!!

If you are defending and sense that declarer may have one of these holdings it is seldom right to lead the suit unless you see your tricks are going away. When you declare, you may have sensed that it is often advantageous to put opponents in the lead and let them solve your problem. Now you know why.

Wednesday, November 17, 2010

Game Tries-- Are you Game or Do You Just Smell That Way?

No offense intended in the title. My first disclosure is that I made a game try yesterday afternoon accepted by my partner and we both ended up like Peking Duck on a spit. Here is my hand so you can savor it. AQJ9xx, xx, AK, xxx. The auction was 1s/2s. Your bid Oswald! Here are the understandings of the partnership: 2nd suit game tries (a/k/a help suit tries) and in a competitive auction a re-raise asks about the quality of trump support.

Many experts (including Ron Klinger and Mel Colchamiro) and countless bridge teachers have advised their readers and students to follow the rule of 5-6-7. Bid game with 5 Losing Trick Count (LTC), make a game try with 6 LTC and pass with 7+LTC. In fairness to Mel, he later enunciated a different and safer rule for 5-4-2-2 hands but that is not germane to my hand. My hand is a 6 LTC hand, so I started looking for a way to invite.

I didn’t have a second suit (you can’t use 2nd suit help suit on a 3 card suit) and I wasn’t particularly interested in partners trump quality since I was 50-50 to pick up the King even if partner didn’t have it. This gave me pause for thought, but I apparently need a longer pause and more thought. I decided that given the limitations of our agreements it might be reasonable to use the re-raise as an omnibus game try asking partner to simply evaluate the quality of his raise. Standing ready to absolve partner of any blame if she did not figure this out, I bid 3 spades, partner bid 4 and we learned to make perfect circles on our personal scorecards.

What observations will help avoid this huge embarrassment the next time?

1. When I couldn’t find the right bid I should have passed. My hand also demonstrates that there are times when “Help Suit” will not help and trump quality is not the primary concern. We (and I think most regular partnerships) needed a more flexible agreement on game tries.

2. My hand met all the prominent tests for a game try, 6 LTC, 3 ½ Quick tricks, 5 controls and a good 6 card suit. Actually, the “don’t blame me refrain” is somewhat weak. For one thing, half my points and 3 of my controls were in a 2 card suit. You want your controls and touching honors in suits in which you have length so you can make tricks from little cards. Second, even with a 6 card suit, I had the worst possible distribution: 6-3-2-2. Third, if I visualized the hand that partner would need to have to make game, it really boils down to controls, Ace, Kings and singles and voids. Queens and Jacks are worthless. Although a short suit game try (asking partner where he had shortness or 2nd round control) might have been better that was not our agreement. Even if our bidding had more precision, my game try was too aggressive for matchpoints.

3. Partner never saw an invite that she didn’t want to accept and it didn’t help that it was the last set of boards. We suffered the ultimate indignity of going down 2 in 4 spades when the field was making 2 spades. At least we played it well! This brings me to another point. You need an understanding with your partner about who does the stretching to reach games. In matchpoints (in spite of my bid) I prefer sound game tries by opener so that responder can stretch a bit on hands that smell like game. This strategy will keep us at 2 of the major on high risk trials. The bottom of the barrel is to be in 3 of a major down one because nobody else thought your hand was worth a try. If the scoring is IMPS, I think the understanding should be just the other way around. Games are a big premium (particularly vulnerable) and you can’t get there if you don’t try. David Berkowitz recently reiterated a comment that Jim Jacoby make to him about IMP game tries: “Don’t make them vulnerable, just bid the game.”

4. Another way to add more certainty to game tries is to use "semi- constructive" major suit raises. With this understanding partner does not raise the major unless he has a really good 7+ to 10 hcps. With lesser hands and 3 card support he first bids 1NT (forcing) and then supports the major at the 2 level. This helps prevent useless game tries since you know in advance that partner does not have a strong raise. The argument against this is that 2 of the major is more preemptive than 1NT and forcing opponents to enter at the 3 level has its virtues in competitive auctions. Choices, Choices!

Steve Robinson, a well known bridge expert, conducts a bidding survey among his expert friends every other month. It is published on the District Six web site http://www.districtsix.org. Recently he asked 20+ bridge professionals about game tries and as you can imagine he got some support for every kind of game try known to the bridge world. There were two consistent themes.

Very few experts like 2nd suit game tries for the very simple reason that they give the opponents too much information about the opening lead and defense. Eric Compton said “Disclosing your hand at IMP’s is losing bridge.” Larry Cohen said “I don’t like to tell.” Marty Bergen said “the very last thing declarer should do is tell the enemy which suit he is weak in.”
I have to say that the re-raise system we used did not give away any information t our opponents or to each other for that matter!!

What has evolved at advanced levels are two way and three way game tries where one of the options is to ask responder to further describe his hand. Many of these trials ask responder in what suit he would accept a game try. Responder bids his positive responses up the line and the major suit at the 3 level as a rejection of the trial. This discloses nothing about opener’s hand and only gives information about the dummy. Not much of a gift!

One such system would have the next bidding level after the raise ask responder for a suit in which he would support a trial. (Ex. 1h/2h/2s or 1s/2s/2NT). In the heart example 2NT is a surrogate bid for spades so there is no overshoot. If 1s/2s/2NT, then if responder bids 3 diamonds it says he accepts in diamonds but would not accept a trial in clubs.

If this were a two-way system, then any bid other than the next level by opener asks a different question. In many systems that is a short suit trial, again revealing nothing about opener’s hand. So 1h/2h/3c might ask about 2nd round control of clubs. Alternatively, 3 clubs could be a 2nd suit game try if that’s your agreement.

This is an advanced topic and for some of my readers may have to be put on the shelf for later. But there is something that all of us can take from this post. If you see 1s/2/s and then some other bid by opener, ask what the bid means. Players are notoriously bad about alerting their game tries. If the game try is a second suit game try (which it will be in most club games) start thinking how you can use this information and responder’s further bidding to direct your opening lead and defense.

Thursday, November 4, 2010

Mike Lawrence Visits the Blog -- OOPS!!

In a recent blog I noted that I was reading Mike Lawrence’s book The Complete Book on Overcalls, a 2009 overhaul after 30 years in the marketplace. In the blog post I noted that Mike stated “Good matchpoints is almost never good bridge. You have to learn to play badly.” That was simply his way of emphasizing that matchpoint competitive bridge has changed and the conservative advice from “back in the day” is often not effective in today's matchpoint battlefield.

In my light, lively and rollicking writing style I stated “Gee Mike, I have been saying that for some time, you reading my blog?”

It is really humor and not a cheap shot, just an absurdity aimed in my own direction. The funny thing is that I got an e-mail from Mike Lawrence yesterday saying that he had visited my blog for a “look.” He said “I did not go through more than a couple of items but it looks like you have done a lot.” Fortunately he took no umbrage at my weak humor, so “all's well that ends well.”

For those who may not know Mike’s accomplishments, he may be the one of the most celebrated bridge competitors ever. He is a member of the Bridge Hall of Fame, a member of the original Dallas Aces, and has won 3 World titles and 18 National titles. When they have credentials like that you can engage in humor (if you only pick on nice guys).

This may be a good point to reiterate my advice on competitive bidding. Yes it has changed, yes it has gotten much more aggressive, yes it lets some of us stragglers back in the game. Here is the kicker: It is not an “automatic game leveler.” Initially you will benefit from a more aggressive stance, but that advantage will disappear if you do not back it up with an effective risk reward system that tells you “when hold them, when to fold them and when to get out of Dodge.” Rely on intuition and windage? That will work about as good as it did last time. It may be a time to join me in a little study and get new standards that will refine your competitive risk reward analysis.

Mike did say that he has new software coming out on November 15 which analyzes 333 frustrating moments in the bidding. Apparently it is not just cutting edge stuff and relies, in large measure on common sense. Less cutting edge and more common sense sounds appealing. The cost $33.95 with postage. You are on your own, since I have not seen it prior to its release.