District 6
Jane Farthing, President
American Contract Bridge League
Mid-Atlantic Bridge Conference
District 7
Zero Tolerance, D6 policy
Oct/NovArticle by Steve RobinsonFeb/Mar
A Game of Odds (Dec/Jan 2006/7)

I asked my expert panel the following questions. “You have KJxx of clubs in dummy opposite A109xx. ‘Eight ever, nine never’ means that you play the Ace and then the King. In what situations would you deviate and take a finesse?   If the contract is 3NT?    What if the contract is 6 or 7? Are there any noncompetitive situations where you would finesse?”


Bridge is a game of odds. Suppose you lay down the Ace of clubs and when you lead to the next trick, the opponent follows. Without any additional information, there’s a 52% chance that the queen will drop. What are the clues, that makes finessing the better play?  One clue is the opening lead. Players lead trumps from two small but never from Qxx. Therefore if the problem suit is trump, and the opening leader has not made a safe lead, the opening leader is a little more likely to have the queen.


Kit Woolsey: The amount by which playing the AK is superior to playing one of the honors and taking a finesse, is so tiny that it can be ignored. Any side information, or even just a feeling, is easily sufficient to tilt the odds in favor of the finesse. If it is a suit contract and a trump lead wasn’t made, that means that the opening leader is more likely to have Qxx in clubs than xx, because he might have led a club from xx but never would from Qxx. Therefore, the percentage play is to cash the ace of clubs and then finesse the opening leader for the queen.


Carol Simon: Certainly in 7 it will be right to take the finesse - the only problem is which way? Did opening leader fail to lead a trump from the queen or from a singleton?

Since there are three non-queen singletons but three queen doubletons, three queen thirds, and one queen fourth, it is probably best to play the opening leader for the queen.


Another clue could be that you got a club or diamond lead against 3NT. When there is a choice between majors and minor, usually opening leaders choose the major. Therefore, if the opening lead is from a four-card minor (which implies no four-card major)  the opening leader is more likely to have Qxx of clubs.


Mike Passell: I would finesse the opening leader in a 1NT - 3NT auction after a two of diamonds opening lead playing him for three clubs. In a competitive auction I would tend to play the overcaller for a singleton club. When we freely bid seven on our own, I would play the opening leader for the queen if he made a side-suit lead from a non-sequence. I know many top players finesse one way or the other when a singleton is in dummy and another in their hand--- Like everyone else I would play any preemptor for shortness. I try to make the correct percentage play unless something in the bidding or opening lead suggests otherwise.

Barry Rigal: LHO leads a four-card suit against a blind 3NT. Now he is more likely to

have Qxx than xx I think. LHO makes a dangerous lead against a suit contract -- now why did he not lead a trump from xx -- play him for the queen. If RHO has one more card in a specific suit than LHO, the odds are even money on the finesse; two more cards and it’s clear to finesse.


The most popular clue: If one player is known to have two more cards in a non-problem suit than his partner, the shorter hand is more likely to have the queen.


Curtis Cheek: Regardless of the contract, I try to make straight odds plays. I can never recall making what I thought was an odds-against play in order to gain a swing; many an odds-against bid, however. The a priori odds are so close that any clue from the bidding or play will usually lead me to finesse. For example, any opposition bidding that leads me to infer a three-card difference in number of unknown cards. However if the bidding goes 1NT-3NT and LHO leads the 2 (playing fourth best),  I finesse him for the Q based on likelihood he doesn't have a four-card major.


Shawn Quinn: If one hand has at least two more cards in a suit then his partner, I believe the percentage is to play for club shortness in that hand.


Point-count is another clue: an opponent needs the queen for his opening bid or he can’t have it because he’s a passed hand and would open if he had it. One has to be careful because some players open ten-counts and others pass thirteen-counts. Since the odds are so close, any clue can give you reason to finesse.


Ralph Katz: I am sorry but you would need to write a book to answer this one. Simple answer is two fold. 1) The bidding warrants the finesse because of high cards. Don't play lefty for the queen if he passed as dealer and already has shown 11 HCP. 2) Distribution, because of the bidding or sometimes you can count the shape of the defender’s hand and know the finesse is either will work or you have it narrowed down to a high percentage.


Henry Bethe: If I have to play the suit at trick two with no opposing bidding and no other considerations, I play the Ace - King. That is rarely the case. Frequently there are avoidance considerations. That is I don’t mind losing a trick to one of the opponents, but scared of the other. Or high-card count considerations: one opponent did not open and has already shown ten HCP (and the opponent is not Al Roth). The contract is not, in my opinion, relevant. Although in 7 the failure to lead a trump may indicate possession of the queen. The scoring system may be relevant. For example in a pair game you are in a club contract and when the dummy comes down you can see that some of the other tables may play a spade contract the other way. Then you want to play the clubs so you get a good score whether the opponents have one or two club losers.


Sometimes the contract you’re in tells you how to play the suit. Suppose you’re in 5 doubled taking a save against 4. The opponents have taken the first five tricks so down 500 or 800 depends upon guessing the trump suit. If you have an outside AK, play for the finesse because if clubs are 2-2, you could have beaten 4. However, suppose you’re in 5 and you see that 3NT is cold for at least ten tricks. You’re not competing against players in 3NT. They will beat you no matter how you play the clubs. The only pairs you are competing against are those who are also in 5. Take the percentage play.


If you’re in 3NT at matchpoints and you have missed your nine-card spade fit, you have to take the same number of tricks in notrump that the field takes in spades. If playing the AK will leave you one trick short, then you should take the finesse - you’re not competing against other pairs in 3NT.


You might finesse to keep a dangerous hand off lead. Suppose you have a side suit of Kx opposite xx. You might finesse to keep your Kx protected.


Eddie Kantar: I would finesse into the non-danger if there is a danger hand. If the bidding indicates that one player has a long suit, which has at least two more cards in the suit than his partner, I'd play the hand that is shorter in that suit for Qxx. If a player needs the queen to justify his or her bids, I'd finesse through that player. If one player is much weaker than the other, I'd finesse into that hand. Even if the finesse loses, I should get a trick back on the return lead. 


Bobby Lipsitz: Obviously, there could be a "dangerous hand" scenario. Also, a preempt might well determine your play – play that hand to be short in the problem suit.


Jon Wittes: In 3NT, I would certainly take a finesse into the safe hand, if scoring four tricks in the suit would guarantee my contract, and I cannot be beaten if a particular opponent should win the trick, but can be beaten if the other opponent has the queen, and wins a trick with it.  I would also finesse if the bidding indicated that one hand is much more likely to be short in the suit than the other hand. The same applies to 6 and 7 with one or two exceptions. If I were in a team match and was considerably behind with very few boards left to play, I might take an educated guess and take a finesse, particularly if the opponents rate to be in the same contract in the other room, and are likely to be playing the suit straight up. I might also finesse in 6 or 7 in a close match or when I am slightly ahead, if I feel I am one trick higher than they rate to be in the other room, and I feel it's right to take the anti-percentage play to try to take the same number of tricks as they are likely to take in the other room. 


Sometimes the state of the match is a determining factor. However, the odds are still 52-48 and you don’t want to go against the percentages if you’re in 6 since your opponents could be in 3NT at the other table. Unless you’re Superman, able to see what’s happening at the other table, you can never be sure that the opponents are in the same contract that you’re in. If you’re Superman you can see where the Queen is. However, if you have a nine-card fit with 26 HCPs between you, there is a very good chance that your opponents will be in the same game contract that you’re in. That might be a hand to take the finesse.


George Jacobs: If I were down by 71 in the fourth quarter of the Spingold, I would finesse in 6!


Joe Kivel: The contract usually doesn't matter. For example, suppose I'm in 3NT, and making 6 depends on locating the Q. I still play for the drop (without any other secondary info). I'm either getting a good score for making 3NT when Q doesn't drop, or a poor score if it does. So I might as well take the percentage play. On the other hand, on the rare occasion when I'm an underdog, I'm tempted to take the finesse. The odds aren't that much different than playing for the drop.  


Mike Becker: There are reasons to go against the surface odds of eight ever, nine never:

If I can get a partial count on the hand, the percentages change. With one opponent known to have more cards than the other in a side suit or two, a finesse could be indicated thru the player with fewer known cards. For example, if LHO is known to have five hearts and RHO has only three, finesse thru the player with three cards in the suit. But I'd need a math wizard to confirm my suspicion that finessing is the right play.


Against the auction 1NT-6NT your opponent leads the queen in a suit you have Axx opposite Kxx. It is very likely that the leader has QJ10 or QJ9.  Hence that player has two or three more known cards than the other. So finesse against the other rather than play for the drop. Sometimes you can take an inference when an opponent failed to bid, and conclude that with or without the missing queen, one opponent or the other would have bid. If you are substantially behind and need a good score, and want to bet the farm on it, go against the odds and take the finesse.


Kerry Sanborn: When the opponents have shown length in a particular suit, distributional hands, or are bidding a lot with few high cards, I would tend to play the suit NOT to be 2-2. Surely there are many hands where I as declarer cannot afford to have one hand or the other get in. It would be right to play for them not to get in then, as long as losing the trick to his partner is safe and provides enough tricks for the contract. There are other hands where you will make the hand when trumps are two-two even if you lose to the doubleton queen, so you can play for the second round finesse to insure against that. Especially in a slam, you can sometimes endplay an opponent when they win a Qx. In truth there are a lot of situations when it is good technique to play for 3-1 in a suit.


If an opponent bids,, he has length in his suit and is more likely to shortness in the problem suit. If an opponent bids over your 1NT-opener he is more likely to have shortness somewhere.


Marinesa Letizia: Basically, I would always play a competitive bidder for shortness and play his partner for the queen. I would finesse if I were looking for a swing in a match. Otherwise, eight ever, nine never.

David Bird: I would play for the drop unless: (a) one of the defenders had overcalled in a different suit, making him more likely to hold a singleton club; (b) One of the defenders had shown up with notable length in a different suit, making him favorite to be shorter than his partner in clubs. (But beware of the fallacy that the player leading to 3NT is longer than his partner in the suit that he has chosen to lead.) (c) One of the usual avoidance situations happens. You finesse to keep the danger hand off lead.

Fred Hamilton: Since it is quite close, percentage wise, I much prefer to guess what to do if I have led to an honor with both hands following, and now lead toward the other honor. In a vacuum it would still be a guess; at the table the opponents often try to influence you. I seldom get that wrong.


Bobby Wolff: I don't follow any strict rules. Only evidence: 1). Probably if contract is 3NT and it is the first hand I play at Match Points I would tend to play AK. However, depending on who my opponent is, if he has led from a weak four-card diamond suit against my 1NT – P - 3NT sequence I would think that he is more likely to have Qxx in clubs; there is more room in his hand for more than a doubleton since he has refrained from leading from a four-card major; perhaps he is 3-3-4-3. 2.


2) Against a suit grand slam there is a small indication that opening leader has Qxx if he does not lead a trump, but a lot of that has to do with who he is and not so much that he didn't lead a trump. Perhaps the most common trump lead would come from xxx, with xx second and a singleton trump down the list. If we play Roman-Keycard-Blackwood (RKCB) and denied the queen you almost never will get a trump lead, but if we didn't ask we are more likely to get a trump lead from nothing with the opening leader thinking we are solid in trumps. One reason I don't like RKCB is that it is easier to lead against.


3.) There is also a case of nervousness sometimes when an opponent has a key queen while defending even a small slam, but rarely 3NT. On the example hand, since you can guard against 4-0 if you lead the right honor first, sometimes depending on the bidding or lack of bidding and vulnerability, a good declarer should know enough to be able to lead the right honor from the right hand first. Lacking any real evidence remember you are the favorite of being right by playing for 2-2  The percentages favor a 2-2 split around 52% of the time (room in a 12-card unknown hand against an 11) .


Mel Colchamiro: Situations to deviate from eight ever nine never:



There are reasons to play for the drop even though there was a preempt. If you need two ruffs and the player who could have the queen will be able to overruff, then you might as well play the Ace – King. If trumps are 3-1 you will lose a trump trick anyway. If finessing or playing for the drop ensures the contract and the other doesn’t, then ensure your contract.

Zeke Jabbour: These decisions are not always made in a vacuum. After first playing low (where possible) toward the appropriate hand to cash one honor, assuming the opponent follows, I would finesse whenever the auction suggests that I should: one opponent has pre-empted; or one opponent has opened the bidding and there are only 12 or 13 points missing and no other ancillary intelligence develops (e.g. on the opening lead); or an opponent has opened a 1NT with 15 or 16 missing points etc. These considerations
alter the percentages on any given hand. I would also take the finesse into a "safe" hand, should one exist, when it guarantees the contract (especially at IMPs). I MIGHT take the finesse based on psychological factors "at the table" such as a hitch or a hesitation, but I'm not sure that's a valid part of your inquiry.


These considerations would apply at both 3NT and 6. If it is possible for there to be a "safe" hand at the six-level. However, whenever losing a trick would mean that the contract would fail, I would take what I deemed to be the best percentage play. It would also, of course, apply at the seven-level, as well. Psychology may also come into play at the seven-level.  If there was no opportunity for a queen-ask and the opponent does not lead a trump—depending partly on the known or perceived level of ability of the opponents, I might choose to finesse. People tend to make the "safe" lead of a trump against a grand slam, so a failure to lead trump may indicate that the leader knows that you're missing the queen. I almost never lead a trump at seven unless I feel confident that the opponents own the queen of trump.  So it is sometimes unsafe to make assumptions.


Bart Bramely: If all other conditions are neutral, I would finesse only if I got significant distributional information. If one defender had three or more known cards than his partner, I would finesse his partner. If he had two more known cards I would be tempted to finesse, but would look for other clues --one known card extra is not enough.

However, "neutral conditions" almost never occur. Form of the game matters (IMPs, matchpoints, BAM). Then there are inferences from the opening lead and defense, from the bidding, from the known style of the opponents, etc. One obvious inference is that suits will split more often when the opponents don't bid, so even if I get a skewed count in some other suit I might still play for the 2-2 split if the opponents didn't bid.


Then there are situations where finessing is the right play on the deal as a whole, like when the opponent will be end-played if it loses, or when you plan to take two finesses through the same opponent, expecting one of them to win. The contract that I'm in doesn't usually matter. Sometimes at matchpoints you have to take a different play if you're in a contract that is unlikely to be reached at other tables. In a grand slam there is a mild inference that opening leader has the trump queen if he doesn't lead one, but I've gone down using that inference, so beware.


David Berkowitz: In any matchpoint event I would play for the drop. However, at IMPs it would depend on the state of the match. Board one if I felt we were 40% or less I would finesse. If I were losing in a match by more than 20 IMPs or so I would also hook.


Larry Cohen: The contract could be relevant, but almost all cases where I would take a finesse would be based on either the auction or the play-to-date. Surely when one opponent has preempted or shown up with length elsewhere, it will often be right to play him for shortness in clubs.


Grant Baze: My understanding is that it takes a two-card discrepancy between the known cards of the opponent’s hands before the finesse is mathematically correct. This means I would never finesse in noncompetitive auctions and only in competitive auctions in which I thought I knew two cards more in one opponent's hand than in the other opponent's hand (Rarely I am convinced that an opponent could not possibly have bid without a singleton to justify the bid). Exceptions are peculiar to the hand itself. The hand might be such that you take a finesse as a safety play, or because if the finesse loses the opponent is immediately end-played, or some such reason. 


Exceptions are almost never peculiar to the state of the match.  First, "state of the match" swings are extremely rare, because you are not entitled to take away your teammates good game, thereby turning a win into a loss because you took a swing. Second, you would have to be exactly accurate in your evaluation of the state of the match. Third, you would have to be absolutely sure the opponents are in the same contract; otherwise the swing is the contract, not the play. It is probably good advice to plan on taking one or two "state of the match" swings in your life; choose those swings with extreme care.


Steve Bloom: One of my bridge mentors would say, “You can bang out the ace and king, and succeed around half the time, you could finesse West, and win around half the time, or you can play East for the queen, and succeed half of the time. They are all so close, that if you have the slightest inkling one-way or the other, play for it.” This is good advice. The percentages are so close, you may as well go with your gut. Obviously, there are plenty of tactical and hand considerations that influence the play, but I am assuming you are asking about probability factors. Many are obvious, but I will try to point out a few subtler factors.


1) In 7, I am a big fan of treating the trump jack as equivalent to the queen in Key Card auctions. When your opponents bid to 7 announcing all the key cards and the trump queen, you usually lead a trump. When West doesn't lead a trump, there are two reasons: a) he has the queen, or b), he has no trumps. You can handle both cases by starting with a low trump from hand, planning to stick in the jack! That is a suit combination your mother never told you about. (West might not lead a singleton trump if he thought partner could have, say, Jxxx.  No matter, you were always going down if West had a singleton trump, and didn't lead it. You never play the partner of the opening leader for Qxx in these situations.)


2) In 3NT, barring any hand-avoidance issues, most players try to use the opening lead information to gage the distribution. Thus if West appears to have led from a four- or five- card spade suit, with East holding a doubleton, players will play East for the Q.  Those are wrong!  In case a), where West led from a four-card suit, the only real distributional inference is that West had no five-card suit to attack. Playing his partner for the Q means playing West for 4-4-4-1 shape, not likely. In fact, when the lead suggests that West has no five-card suit, then West is much more likely to hold three clubs than East, and the percentage play is to finesse the club through West. 


In case b), you have really gained no information, and all the plays are essentially equivalent.  I usually just follow the nine-never rule. Opening lead information is biased information - the opening leader chose to lead a suit, usually his longest suit. The fact that West holds more cards in that suit than East is pretty irrelevant. A good guideline is this: After, say, 1NT - 3NT, if the opening lead if from a four-card suit, that player is more likely to hold a missing card.  Five-card suit leads are neutral, while longer suits do give you information about the rest of the hand.

When in doubt, play for the drop. “Nine never, eight ever”. However, you’re only getting 52-48 percent odds. Since the difference is small, use any clue to give you reason to finesse. A known long suit held by an opponent changes the odds to make finessing the percentage play.  The fact that you don’t get a trump lead could be a clue.

Don Berman, Web Master.