A9xxx opposite KTxx
This is a holding all good bridge players are familiar with and at least think about taking the safety play, but it is something that many somewhat new players may not realize. A priori, the percentage play to take the most tricks in the suit is to play the AK first but only very slightly. This holding most often leads to an application of the rule of restricted choice. When you play one of the high honors and the Q or J drops, it is right to finesse for the other honor if possible. It's also good technique to start by leading low toward one of the honors rather than just lead the big honor first. That way, when second hand is void, you can still pick up the suit for one loser. This play leads to taking all 5tricks 46% of the time and on average yields 4.42 tricks. But 5% of the time, you'll have to lose two tricks (a 4-0 split is 10% but you can pick up one of the 4-0's by leading low and seeing the void in time to adjust and hold it to 1 loser). So, clearly this is the best way to play the suit in matchpoints unless you are in a good contract that you think the field will not get to.
In imps, however, a safety play may be called for. If you need to guarantee that you get 4 tricks in the suit, lead low and finesse either the T or 9 when second hand follows. Still apply the rule of restricted choice when second hand plays an honor. This will yield 4 tricks every time and might get you 5 tricks if second hand is lazy and doesn't split the honors from QJx. So, this is not a safety play to be taken at matchpoints but may have it's place at imps. Here is an example from a hand played today on BBO.
ATx
K
ATxx
KQxxx
KQJ
Axxx
K9xxx
x
In 3NT, on a heart lead, you have 5 major suit tricks and probably need 4 diamond tricks to make the contract. Otherwise the defense will get enough heart tricks to set you (unless the person with the diamonds has the ace of clubs and only 3 hearts, in which case you can afford to lose 2 diamonds after ducking the 2nd round of hearts but not the 3rd). Yes, many if not most people will make it to 3NT but some might get carried away and go down in 6D so giving up a chance to make 4 is probably worth taking a safety play in diamonds. Win trick 1 with the K in dummy and play a diamond to the 9 at trick 2, and you will be rewarded when west shows out.
As a side note, I see many players, even some rather experienced players, who insist on playing to for the drop instead of a finesse holding 10 cards in the suit missing the K. Yes, there is a 13% chance that you will find a singleton K offside but that is the only case that play will win over a finesse while a finesse wins over the drop when there is Kx or Kxx onside. A 2-1 split is 78% and the chance that the 1 is the K is 1/3 of that or 26%. A 3-0 split occurs 22% of the time and playing for the drop will never work so 26+0 = 26%. A finesse is 50%. 50% or 26% - tough choice...
Showing posts with label safety play. Show all posts
Showing posts with label safety play. Show all posts
Monday, September 20, 2010
Thursday, June 10, 2010
When to take a Safety Play?
Although I do have a preference for matchpoints and consider my overall style better suited for that type of game, safety plays are among my favorite kind of declarer plays. In today's hand, there are a couple of subtle options. South is in 4♥ after an uncontested auction and west leads the ♦Q.
You might be able to make it on some sort of strip and endplay but you don't know enough about the distribution and that's not really a possibility to consider. The only real way to make 5 is to draw trumps, play ace of spades and take a spade finesse. This line will make 5 when spades are 3-3 with the Q onside, an 18% chance. It will make only 3 when the suit is 4-2 and the Q is off (24%) or when RHO has 5 or 6 spades with the Q (7%). Combined, this is a 31% chance you'll go down with this line, and you'll make 4 51% of the time: 3-3 w/ Q off (18%), 4-2 w/ Q on (24%), 5 or 6 spades w LHO or stiff Q with RHO (9%). That means on average, you'll make 3.87.
Let us consider another line of play, a line that is only available on a non-club lead, for a club lead would cut out an entry to dummy. What are the percentages of playing the K and A of spades first and then leading low towards the jack? This will never get 4 spade tricks because if the Q falls, the suit cannot split 3-3, so it essentially gives up the chance of an overtrick, but does it fact make the contract more likely to succeed. This line will make 4 77% of the time: any 3-3 split (36%), any other time the Q is onside (32%), Q or QX offside (9%). You will go down 23% of the time: RHO with at least 4 spades including the Q. On average this makes 3.77 tricks, slightly worse then the previous line.
In other words, playing the AK first wins an extra trick (and makes the contract when the other way goes down) when east has Qx of spades (8%), and the finesse wins an overtrick when west has Qxx (16%).
So, how do you decide what line of play to take? In a matchpoint game? In imps? In matchpoints, you should consider how likely the field is to be in the same contract. In this case 4♥ seems to be an easy and normal contract to reach so you probably should try for the maximum number of tricks. In an average game, the field will take the straight up finesse and the matchpoints to be won by taking the safety play when it wins should be about equal to the matchpoints lost when it loses, relative to the straight finesse. If you decide that a significant portion of the field will be in a partscore or in 3NT making only 3, you would be even more concerned with making the contract for just making would already be a good score.
In imps, it's a little different. You don't care so much about the 1 imp you could win by making an overtrick but you care about not losing the 10 or 12 imps (Depending on vulnerability) when you go down and they make at the other table. If the probability of making an overtrick is 10 or 12 times as big as the potential gain from a safety play, then it's an even chance, but this time the ratio is 2:1. By taking the safety play as opposed to the normal finesse at the other table, you'll push the board 76% of the time, win 1 imp 16% of the time, and lose 10 or 12 imps 8% of the time, so in this form of scoring, it's clear to take the safety play.
| North | |
| ♠ | KJxx |
| ♥ | Jxx |
| ♦ | xxx |
| ♣ | Axx |
 |
|
| South | |
| ♠ | Axx |
| ♥ | AKQxx |
| ♦ | Ax |
| ♣ | xxx |
Let us consider another line of play, a line that is only available on a non-club lead, for a club lead would cut out an entry to dummy. What are the percentages of playing the K and A of spades first and then leading low towards the jack? This will never get 4 spade tricks because if the Q falls, the suit cannot split 3-3, so it essentially gives up the chance of an overtrick, but does it fact make the contract more likely to succeed. This line will make 4 77% of the time: any 3-3 split (36%), any other time the Q is onside (32%), Q or QX offside (9%). You will go down 23% of the time: RHO with at least 4 spades including the Q. On average this makes 3.77 tricks, slightly worse then the previous line.
In other words, playing the AK first wins an extra trick (and makes the contract when the other way goes down) when east has Qx of spades (8%), and the finesse wins an overtrick when west has Qxx (16%).
So, how do you decide what line of play to take? In a matchpoint game? In imps? In matchpoints, you should consider how likely the field is to be in the same contract. In this case 4♥ seems to be an easy and normal contract to reach so you probably should try for the maximum number of tricks. In an average game, the field will take the straight up finesse and the matchpoints to be won by taking the safety play when it wins should be about equal to the matchpoints lost when it loses, relative to the straight finesse. If you decide that a significant portion of the field will be in a partscore or in 3NT making only 3, you would be even more concerned with making the contract for just making would already be a good score.
In imps, it's a little different. You don't care so much about the 1 imp you could win by making an overtrick but you care about not losing the 10 or 12 imps (Depending on vulnerability) when you go down and they make at the other table. If the probability of making an overtrick is 10 or 12 times as big as the potential gain from a safety play, then it's an even chance, but this time the ratio is 2:1. By taking the safety play as opposed to the normal finesse at the other table, you'll push the board 76% of the time, win 1 imp 16% of the time, and lose 10 or 12 imps 8% of the time, so in this form of scoring, it's clear to take the safety play.
Subscribe to:
Posts (Atom)