Squeezing The Dummy

Wednesday, January 18, 2006

Cooperative Slam Bidding

Bridge is a partnership game.

You hear people rationalizing bids all the time with something like "partner could have had..." Well, usually partner is better informed about what he has than you. Unless he has transferred captaincy and you have to make some kind of decision, that is not a good justification for making a (mis)bid. When in a cooperative auction it can be useful to think about what you have shown and what you have in context of that.

Nothing should illustrate this more than cuebidding sequences to try for a slam. Unfortunately, many times after one player has bid their hand they then bid make the decision to drive to slam themselves instead of trusting partner to do the right thing. This mistake probably costs most players thousands of imps in the long run.

Take the uncontested auction auction 1-2-2-3-4-4-4 in a 2/1 system that does not use serious 3N. Many people would describe 4 as a signoff. That is false, partner is still in the picture and can bid on. 4 is limiting your hand and merely says that you have made 1 slam try and that was all you are worth. You probably have something like a good 14 or 15. Partner can take control with keycard or pass. He also may cuebid again if he is not sure whether or not slam will make (yes knowing how many keycards partner has is not always the answer). The point is a very basic one in bridge; don't bid your hand twice. Still, psychologically people feel like over 4 it's do or die and they must make a decision. They will see their 14, think of some hands that partner can have where slam is cold, and try keycard. What they forget is that most of the hands where slam is cold partner is not going to pass 4.

Labels: ,

Tuesday, January 17, 2006

Right or Wrong?

I had a very interesting play problem come up in the NAOP. It was the final day and we were playing against one of the best pairs in the room. We had a great game going and I thought we were probably qualifying at that point but there were still about 8 rounds left to play. I was in 3N with these two hands:

KJT93
82
J97
QT3

AQ2
AQJT
QT2
A42

The auction was 2N p 3N. This is a typical "insane" matchpoint auction for us. LHO leads the ace of diamonds (asking for attitude). RHO plays the 6. Now LHO returns the 8 of diamonds, I play the jack from dummy and rho plays the 4. What next?

We have 5 spades, 1 heart, 1 club and 1 diamond. The obvious thing to do would be to finesse the heart and then try finessing it again. But what was the 8 of diamonds? It was a suit preference card indicating a heart card. I honestly believe the least falsecarded signal is suit preference, even though my LHO is a very strong player I took it to probably be a true card. I was also confident that if he had 5 diamonds he certainly had a king. Good players will almost always lead small from AKxxx if they have no side entry. If LHO has 5 diamonds and the heart king he will be strip squeezed on the run of the spades.

I thought for a long time and decided to run spades. LHO started with 2 spades and pitched a low club (encouraging) followed by 2 low hearts. RHO pitched 2 low clubs (encouraging). From my hand I pitched a heart and a club. This was the end position:

---
82
9
QT3

---
AQJ
Q
A4

If LHO came down to the guarded heart king and started with 5 diamonds, that only left room for one club. In that position my winning line is to cash the club ace and get out a diamond (see why I didn't pitch a diamond from my hand?). Given that I believe the diamond 8 this is the most likely layout. However, I had a lot of extraneous things to consider.


  • Did LHO stiff his HK and hold on to 2 clubs? This was possible to do in tempo since I took so long to play. I didn't think it was likely because of another pattern people seem to have. When they figure out early on in the play they need to stiff their king, they do it immediately. I'm pretty sure if LHO wanted to go that route he would pitch heart, heart, club in that order.
  • Did LHO falsecard the D8? If he had the club king and the AK of diamonds, looking at 7 points in dummy and 20-21 in my hand he knows partner has very little. A falsecard is very possible, especially since even if his partner does have an entry like the queen of spades, LHO has a slow entry not a quick one so it's not that important. However, again few people falsecard suit preference and it's important to remember not to give your opps too much credit.
  • Did LHO start with 4 diamonds all along? With AK85 he might have chosen to lead a high diamond then had nothing better to do but continue. In this scenario he might have chosen a different lead (a major is preferable on this auction) and he would need the 5 to know that playing the 8 is safe.
  • If LHO does have xx Kxxx AKxxx xx like I think, what will happen to 4 spades? That is the field contract. LHO will lead 3 rounds of diamonds and his partner will ruff and lead back a heart. The clubs then go away on the hearts, so I'd be down one. So even if I go down 1 in 3N I will get an average.

At the table I decided the first 3 inferences were strong but not sure, maybe 80 % each. However, when combined that makes LHO having the hand I think he has to only be a 50/50 shot. If the heart hook is on or diamonds are 4-3, I'm going to get a good result for playing in 3N if I just finesse now. I decided the risk wasn't worth it and I hooked a heart and went down 1. Sure enough we got an average board, and making would have been a top. LHO did have 2452 with the HK.

Looking back on it I'm not sure if I did the right thing. Very often I will play a hand to have a guess in the end position and then back my judgment about the layout 100 %. If I was playing my A game and was feeling very confident I'm sure instead of thinking my inferences were 80 % I would have thought they were 100 % and gone for it. I felt bad for not backing up my judgment in this case because it is one of my strengths in declarer play. However, we did get an average and we did qualify for the nationals. Had I gone for it and gone down for a zero, we wouldn't have qualified. From that point of view I did the right thing.

I still don't know, I'll let you guys decide.

Labels:

Monday, January 16, 2006

NAOP Report

The North American Open Pairs (NAOP) is a unique event. You begin by qualifying at the club then go on to unit, district, and finally nationals. I really love the district finals because there are no weak pairs in the second day and few in the first. District 16 is notoriously tough so the scores are usually very close and the quality of bridge is high. The top 3 pairs get to go to the nationals and the top 2 pairs get their hotel and airfare paid. Usually the money is very nice but ironically the next nationals are in my hometown of Dallas so there will be no airfare or hotel fees.

The district finals were held over the weekend in San Antonio. For the first time I got my dad to go to the club and qualify and was excited about playing with him. One of my dreams is to win a national with my father, it would make it that much sweeter.

The first session of the first day started out miserably. In our first 2 boards we missed two saves and after about 5 rounds it felt like we were having a 20 % game. It's important in these situations to remember that there is a lot of boards left and just to play normally. Towards the middle of the session things finally started going our way and we were getting hot. We ended up with a 55 % game, not horrible but not great either. Considering our start, though, I would have been happy with average.

The second session was uneventful; we just made too many costly mistakes. We had opportunities and let them pass us by. We ended up with an average game. This meant we qualified in 8th position out of 14. The two pairs that were leading had big carryovers, and everyone else would have very little. There was a 2 board max carryover and we got 1/3rd of a board. This put us at a disadvantage but the main thing was we made it to day two.

In day two it is time to take a few more risks. Instead of trying to be in the top half, you're trying to be in the top 3 out of 14 in a good field. Without much carryover you need two good sessions to achieve this. Our first session started out amazingly well. We were getting gifts left and right. Towards the middle bad things started happening, and we didn't seem to be playing as well as we could. We were getting zeroes and they were killing our score. On the last round we had a misunderstanding and bid a grand on a finesse which was....on...whew. Rather be lucky than good. We ended with a 55 % game but a 70 % was literally in reach. I looked at our scores and noticed we had 5 zeroes in 26 boards. That is just way too many. There was good news and bad news with the rest of the people's scores. Jim Griffin/Ken Schutze and Georgianna Gates/Gus Plate both had huge games. They were about even and 2 boards ahead of the rest of the field. Both pairs are very good and very experienced so I didn't expect to be able to catch them. After that there were about 4 pairs very close in score fighting it out for third with everyone else being outside contenders. We were one of those 4 pairs, so if we had a good night session we would probably make it.

The mentality in the final session was just to play tough and not give anything up. Our style is very active so we're used to more bottoms than most but we knew we had been playing softly and doing things we shouldn't have. In the fourth session we finally brought our A games. Our luck was not as good as before but we had some cards and made the best of them. We ended up having a 57 % game which was surprising, it felt much better. In a good field sometimes good games are less good than they feel and bad games also less bad than they feel because there is less variance in skill and actions taken. I wasn't sure if 57 % was going to cut it, but unbelievably both of the pairs that were leading had games in the low 40's. I actually thought we might have won. We ended up second, losing by 1.4 matchpoints to Shawn and Joe Quinn.

There were a few interesting hands. One hand I had J9x AJ9 Q98xx Ax.

RHO opened 1, I passed, LHO bid 1N forcing and partner bid 3. RHO Xed for business and LHO pulled to 4. First, what do you lead? Dummy obviously has a stiff or void in clubs, likely the latter, and we have the red suits locked up. LHO will only have 3 trumps since she started with a forcing NT. I think a trump lead stands out to cut down on the club ruffs. I led one and dummy hit with:

Axx QT8xxx JTxx ---. We were right about the short clubs and 3 card spade support, but we were wrong about having the red suits locked up! Those hearts looked scary. Declarer played small from dummy, partner put in the queen and declarer won the king. Now he tried the king of hearts. At this point I stopped to consider the hand. Declarer is likely 5224, 5134 or 5125. It's possible declarer has 3 hearts as well but he might have played it differently. If declarer has stiff king of hearts it's probably not right to duck, and with Kx he may not hook later so I won immediately, partner showing even count. I now led the ace of clubs to tap the dummy. It's crucial to do this in order to avoid declarer running hearts and having an entry to dummy. Declarer now cashed the spade ace and ran the jack of diamonds to my queen. I now got out the 9 of hearts to kill the final link to dummy. Declarer correctly hooked and had the king of clubs so he was down just 1.

Another hand my partner made a good bid. He held AKQT KJxxx Axx x

I opened 1 (11-15 5+ hearts) and he bid 3 showing a game forcing heart raise. I now jumped to 4 showing 5-5 and decent suits. He tried keycard and I showed 2 with the queen. He then tried 5N and I jumped to 7. What is my hand? What do you bid?

You should know my exact high cards. I must have the AQ of hearts, A of clubs (2 with) and the KQ of diamonds (7). I'm 5-5 so there are 13 top tricks in hearts. In NT there are only 12 tricks. However if diamonds don't behave you can always fall back on the spade hook. If I don't have the ten of diamonds picking up diamonds will be 68 % and the spade hook is 50 % so 7N will be 84%. If I do have the diamond ten it's over 90%. This was from the first day and the goal is just to qualify. He correctly gave credit to the field and bid 7N which made (diamonds didn't break but the spade jack was on). We got about 80 % of the matchpoints and bidding 7 would have been slightly below average. I think the risk was worth it.

There was a very interesting 3N hand against Griffin/Schutze, but I'll leave that for tomorrow's post.

Labels:

Friday, January 13, 2006

Mental Simulation

Most auctions start with each partner describing their hand to the other until one of them is able to take control. Taking control sometimes involves asking specific questions (such as blackwood) and sometimes just means placing the contract. When it comes time to place the contract I find the most effective thing to do is mental hand simulation. Things like combined high card point and losing trick count are now irrelevant as you should have enough information to figure out a range of hands for partner. From that you can determine what is most likely to be the best contract.

Obviously there are time limitations and we are not computers so we can only construct a limited number of hands. To try and get the best range, I like to think of 5 hand types. Horrible minimums, perfect minimums, horrible maximums, perfect maximums and average hands. I like to try and make about 3 average hands and 1 of each of the others. This method that I use was mainly based off of something someone wrote in the Master Solvers Club. I'm fairly sure it was Jeff Rubens, but I could be wrong. Doing this you get a very good idea of where you stand.

Let's try a few hands.

You hold AQ3 AKJ2 KQ62 Q3 and open 2N 20-21. Partner bids 4N inviting a slam and denying a 4-card major. You do not have any method to show 4-4 in the minors, so that shape is possible. At this point instead of thinking "I am maximum so I will bid slam" you should start constructing hands. To do this, we must consider what partner will have. He likely has about 11 points, possibly 12 with no 5 card suit or a very good 10. His possible shapes are 4333, 5332 with a 5 card minor, or 4432 with 4-4 in the minors. With a 6 card minor or 9 minor suit cards you have the methods to bid something besides 4N. In your simulations you want to include all possible shapes.

Horrible minimum: KJx Qxx Ax Jxxxx. Here we are off the AK of clubs and have no play.

Perfect minimum: Kxx xx Axxxx Kxx. Here 6 is gin but 6N will be on a heart finesse.

Horrible maximum: KJx xxx Axx KJxx. With 2 clubs, 3 spades, 3 diamonds and 2 hearts there are 10 top tricks. If the heart hook works you need a red suit to split or have numerous squeeze chances.

Perfect maximum: xx Qxx AJxx KJxx. Here slam is very good, if diamonds split you are cold otherwise you can try a spade hook in 6 diamonds. In 6N you need spade hook or a black suit squeeze.

Average hand: Jxx Qxx Axxx AJx. In 6N we have 3 diamonds, 4 hearts, 1 club and 1 spade off the top. If the diamonds split we just need both black kings to not be offside. If diamonds don't split we likely need a spade/diamond squeeze. 6 is worse.

Conclusion: I did these simulations hands just as I would at the table. I got every shape and HCP range in, and tried to spread the honors out as evenly as possible. On reflection one error I made was always giving partner the diamond ace. Had I done 3 average hands I probably would have given pard one with the spade king and club AK or something. This seems to be a very close case. It seems to come down to whether or not partner has a diamond fit. If I had 5 diamonds available to show 4 diamonds and be forcing, I would choose it. If partner signed off in 5N I would pass, otherwise we would be going to slam. If not, I would probably bid 5N pick a slam trying to get to 6D and otherwise play 6N. Your simulations may come up with a different answer! Usually there won't be discrepancies, but in a case this close there might be.

Let's try one more. You open 1 and partner bids 3 showing a 4 card limit raise. Your hand is 2 AK532 QJ82 Q42. Partner will have about 10-12 support points and has a wide range of possible shapes.

Horrible minimum: KJx QJxx xx Kxxx. Game has no play here and even 3 is not cold.

Perfect minimum: xxx Qxxx KTx AJx. Game is just cold here barring something very bad happening.

The horrible mins and horrible maximums will be the same, all depending on how much wastage there is in spades.

Average hand: QJxx Qxxx Kx Kxx. Here is a normal 11 count with evenly distributed honors. With a loser in each side suit, we have a lot of play. To get rid of the minors we might pitch one on the spades get 2-2 hearts, get a favorable lead, or find the T9 of diamonds third.

Average hand: AQxx Qxxx x JTxx. We must think about unbalanced hands. If partner has a shortness it's likely in diamonds. We have 2 clubs and a diamond to lose and a lot of work to do. They may get a club ruff or lead trumps. They don't always defend perfectly though, and even when a ruff is available they may not find it.

Conclusion: Much depends on partner's spade holding. With the average hands we saw game was close but not cold. I would bid this game at imps and hope for the best, but at matchpoints I would pass. If it is tough to make 4 we'll get a good matchpoint score for 170, but it may just be a win 1 in imps.

Did that feel like hard work? Well it is, but it also will give you much more accurate results than thinking about whether you are minimum or maximum in your high card range. Remember, practice makes perfect and it does get easier.

Labels: ,