Wednesday, March 28, 2007

Culbertson's Valuation by Visualization

Ely Culbertson was one of the great geniuses of contract bridge. Ely was born with dual citizenship in Russia to an American father. His family lost everything in the Russian revolution so he came to this country in early adulthood with the only asset he had, an ability to earn a living playing cards. He married a bridge teacher from New York, and with his wife, Josephine, became a major force in auction bridge. When Commodore Vanderbilt invented contract bridge in 1929, Ely quickly grasped its commercial potential, grabbed the game by the throat and took it from the card room at the Cavendish Club to every kitchen in America. Soon he was publishing a magazine (The Bridge World), writing a bridge column, writing the seminal work on Contract Bridge in 1930 (The Blue Book of Contract Bridge), hosting a weekly radio cast featuring a bridge challenge match, forming and operating about 600 bridge teaching studios around United States and partnered with Josephine, playing high stakes bridge marathons against other pretenders to his throne. To give you an idea of his impact on the game, in 1939, when United States was still in the depths of the 30’s depression he made more than $1,000,000 from bridge related revenues.

He was a player’s player with marvelous instincts for the game. Jeff Rubens, the current editor of the Bridge World, said that Ely was known as a brilliant theoretician, yet he literally had dozens of rules relating to all phases of the game which he relied upon to make decisions about his own play. One of these rules was “Valuation by Visualization”.

To paraphrase this rule, it says that in making decisions about whether to make a game try or slam try, you should do so only if “a perfect minimum hand for partner will make the contract a lay down.” You first decide the minimum strength partner can have for his bid, and within that limitation you can logically place specific cards in partner's hand. It must now be a laydown and not be finesse dependent. If you read Frank Stewart’s Bridge column, you often see him employ this technique to justify a bid by saying “partner could hold….” and then giving him a minimum hand.

There is a story about a little known bridge blog author playing in his 2nd National tournament in Atlanta with even a lesser known bridge director. The game was a knockout and IMP scoring was in effect. Toward the end of the 4th set of boards the author picked up:

Void, Kx, Axx, KQxxxxxx.

Across the table he hears director open 1 heart. The match is close and author really didn’t want this on his shoulders, but he bid 2 clubs (not 2/1 game force). Author’s LHO now overcalls 2 spades! Director, for his opening bid, held:

Jxx, AQxxx, xx, Axx (with 2 1/2 quick tricks it meets minimum opening standards for this pair).

Director goes into the tank a minute and finally lays down the 3 club card. (This director administers the rules but doesn’t always follow them). After a pass by the other opponent, author takes a pause in the action. First, to silently thank LHO for his useless overcall in spades, and second to put on his superman suit. Trying to visualize director’s hand, either he psyched and then raised, or he has significant values in the heart suit at least 5 deep and the Ace of clubs. Author, determined to bury partner if he is mucking around at this level, asks for key cards. Director bids 5 hearts (2 without the Q). Now, assigning director a minimum number of points for his opening bid (11) author starts to place key cards in his hand. Author gives him the AQ of hearts and the Ace of Clubs. Based on the bidding the Ace of Spades seems very likely to be with LHO for his 2 level vulnerable overcall. The opposing pair at the other table will surely get to 6 clubs, and it now or never to put this match out of reach, or alternatively look for a new team.

The rest is history recorded with the ACBL. Author quietly slid down the 7 club card, and only after LHO led the Ace of spades did he realize that he had to play it. Fortunately, the hand meets Ely Culbertson’s other criteria, it is a lay down when director tables his hand. It was not until after all the smoke cleared and the “high fiving” ended that director pointed out that author had just bid a Grand Slam at a National Knockout with only 23 combined high card points and "off" one Ace.

If it were not such a good example of valuation by visualization, I would probably tell this story sooner or later anyway. That’s what you get for owning the blog!

Monday, March 26, 2007

In a Minor Key

In a recent talk to the Ocala Duplicate Bridge Club, Jon Shuster of Gainesville presented a selected analysis of 717,102 computer generated bridge hands. One part of his inquiry involved the usefulness and frequency of occurrence of the preemptive jump raise when using the Inverted Minor (“IM”) convention.

For the IM analysis he limited his inquiry to only the preemptive raise to be made with less than 10 hcps, and broke that down into a minimum raise of 1-6 hcps and a good raise of 7-9 hcps. Some criteria need to be mentioned to fully appreciate the IM analysis:

(a) With the minimum preemptive jump raise (1-6), Jon permitted the preemptive raise to be made holding a 4 card major.
(b) With the minimum preemptive jump (1-6) raise he stipulated 5 card support in both clubs and diamonds, but with the good preemptive jump raise(7-9), he permitted the raise to be made with 5 clubs or 4 diamonds (but not on 4333 hands).
(c) Most importantly he imposed a minor suit opening protocol of always opening 1 diamond unless the hand held 2 diamonds and 3 clubs (4423).

It is not necessary that you agree with Jon’s assumptions. Without assumptions analysis cannot be made. Other assumptions may vary the conclusions.

In one part of this analysis he tested the effect of permitting the good preemptive jump raise (6-9) FROM 1 club to 3 clubs with only 4 card club support to determine how this affected the frequency of finding opener with a 3 card suit. His results showed that the frequency of finding opener with a 3 card suit when raising to 3 clubs with only a 4 card club support is anywhere from 2-3 times more likely (depending on the seat in which opener opened). Central to this conclusion is Jon’s assumed protocol in the opening of minor suits.

If you changed the assumption to open “better minor”, I am confident that the relative difference of three card suits would be smaller between clubs and diamonds since you always open the better of the two suits. That is not a pitch for “better minor” nor is it a recommendation that you make a good inverted raise with 4 clubs. Assuming you apply Jon’s criteria, he found that when you make a good IM jump raise with 4 card diamond support, you expect to find opener with a 3 card diamond suit only about 8% of the time. It is easy to see that your support protocol has to be tied into your minor suit opening protocol to make a comfortable decision about the level of diamond support required for a preemptive jump raise.

The more important revelation from Jon’s study of IM preemptive raises deals with their frequency of occurrence. Let’s assume you play duplicate bridge twice a week or 104 times a year. If you use only minimum preemptive minor suit raises (1-6), you will see one about every 4 months. If you use only good preemptive minor suit raises (7-9), you will see one about 11 times a year or once every 9.5 sessions. If you use both minimum and good preemptive minor jump raises, you will see them 15.5 times a year, once every 6.7 sessions.

What are the practical aspects of this. Unless you play several times a week, you have to question whether using inverted minors is worth the effort. Every convention takes up "gray matter," so if you have reached your “Peter Principle”, this might be a good one to consider eliminating. Conventional bids also raise the possibility of missing an alert or missing the bid entirely, which could lead to disaster,

Finally, there is much more to the inverted minor suit raise convention that just the first response. If you are not familiar with the recommended continuations, you very likely will end up with a poor result. My view of this was substantiated by a comment made by Frank Stewart in his bridge column on March 25, 2007. After a discussion of an inverted minor hand, Stewart concluded:

“I don’t care for inverted minors, not because I think the idea itself is theoretically unsound but because the development of the auction will be confused unless a pair spends considerable time in discussion.”

I’ll give you a clue, I think the only correct places to pay these hands is either 3 of a minor or 3 no trump since you don’t have 8 major suit cards. Even with a single raise (10+) you can bet that either 3 of the minor or 3NT will be a top board, but unless you find opener with stoppers in both majors, getting it right can be an elusive proposition.

On the other hand, with long minor suit support and less than 9 points, a part score in no trump is often the best contract. It will be easier to find that contract if you do not take up your own bidding space with IM. The better advice is to simplify your convention card and leave those problems to someone who needs them.

Sunday, March 18, 2007

Losing Trick Count (Part 2)

In Part 1 on Losing Trick Count we learned how to make a LTC count of your hand and also to estimate the LTC of parter's hand. In this second and final post on the subject, we will see how to use the LTC information to make vital game going decisions.

Rule: Once you know your LTC and have estimated partner's LTC, simply add them together and subtact the total LTC from 24. The remainder is the number of tricks you are likely to take on this hand if you have an 8+ card fit.

(1) Opener’s Hand: AKxxx Kxx Jx QJx. (14 hcps/7LTC). Opening Bid is 1 spade. Here are some responses and analysis.

(a) Partner responds 2 Spades. Jxx Qx Qxxx Kxxx (8 hcps/9 LTC). Formula 24- (7+9) = 8. We are predicted to make 8 tricks in Spades. Opener should Pass.

(b) Partner responds 3 Spades Qxxx QJx Axx Kxx (limit Raise). (12 hcps/8 LTC) Formula 24-(7+8) =9. We make 9 tricks in spades. Opener should Pass. Here you would probably bid 4 spades on points alone, but LTC denies that. Note that despite the fit the hand is too square.

(c) 2NT (Jacoby) Qxxx Ax Axxxx Ax (14+ points/7 LTC) Formula 24-(7+7) = 10. We can make at least 10 tricks in spades. With no shortness and a minimum, opener bids 4 spades.

(2) The bidding goes 1s/2c/2d/?. As responder you hold Qxx Qx Axx AKxxx 15 hcps/7 LTC and spade fit. Opening hand vs. 15 hcps. Formula 24- (7+7) = 10. We are favorites to take 10 tricks in spades. Bid 4 spades.

(3) The bidding goes 1c/1s/3s/? Opener is showing 4 card spade support and about 15+ hcps. You estimate it to be a 6 LTC hand . You hold QJxx QJx Axxx xx (10 hcps/8LTC). Quit worrying about the hcps, you have only 8 LTC, so if partner had 6 LTC you ought to make 10 tricks in spades. 24-(6+8) = 10. On the other hand if you held Kxxx Kxx Kxx xxx (9 hcps/9 LTC, pass 3 spades, it looks like you can take only 9 tricks. 24- (6+9) = 9.

(4) Opposite a minimum raise, it is often difficult knowing whether to pass, invite to game or bid game. A good guideline is to bid game directly with 5 LTC, invite partner with 6 LTC and pass with 7 LTC. To accept an invite, partner needs to be on top of his raise and holding a 8 LTC hand. Something like Kxxx Kx QJxx xxx.

The Weak 2 Bid: Setting aside all of the theoretical discussions about disciplined and undisciplined weak 2 bids, if partner opens with a weak 2 bid, it is safe to assume that the hand is 7-8 LTC. If it is 8 LTC is would look something like KQxxxx xx xxx xx. If partner has 7 LTC the weak 2 bid hand could look something like KQxxxx xx QJx xx. If you have a 6 LTC hand, it is too good to be opened with a weak 2 bid, you are liable to pre-empt your own partner. If you have a 9 LTC hand, you better have an understanding partner.

Three level Pre-empts: Do not make the mistake of opening at the 3 level every time you hold a 7 card suit. If you do, some of your hands may be too weak for a 3 level pre-empt (8 LTC hands) and some may be too strong (5 LTC hands). At the 3 level, vulnerability is very important. Until you gain confidence in your own judgments, a good rule to follow is to open a seven+ carder at the 3 level with 7 LTC if non-vulnerable and with 6 LTC if vulnerable. As you move up to the 4 and 5 level with your pre-empts, subtract 1 LTC from each of the 3 level requirements for each level you go up.

Responding to Pre-emptive Bids: Knowing when and how much to raise a pre-emptive bid is one of the hardest judgments in bridge. Start with an assumption about partner’s LTC, and then count how many of his losers you can cover with winning honors in your hand. Assume, for example, that partner bid 2 spades on KQxxxx xx xx xx (7 LTC). To produce game you will need to cover 4 of his losing tricks to bring the total losing tricks down to 3. Counting high card points is not very reliable as Queens and Jacks don’t count for much in this analysis. Your “cover cards” will roughly equate to your Quick Tricks. Thus, you can raise to 4 spades with Jxx AQx AQx Axxx (4 Quick tricks/7 LTC) and have a reasonable play. If you have a singleton, you can also count that as a cover trick if you have at least 3 supporting trump. Thus, Jxx, AQJx, KQxxx x (6 LTC) will give you a play for game.

LTC is not for novices who still need to focus on basics. Using hcps is the only practical way to learn the game, but once you have mastered hcps (advanced novice), you need to learn when to forget hpcs (distributional hands) and use the LTC technique. Pat Peterson teaches LTC in her advanced bidding courses. Again, the book is Modern Losing Trick Count by Ron Linger. Good bridge is a lot of work, but also a lot of fun. If you could buy a good bridge game, who would be interested in playing?

Losing Trick Count (Part 1)

Losing Trick Count (“LTC”) is a valuation tool for determining the trick taking potential of a bridge hand. The concept is widely used by better players. It is most effective on hands that have good distribution. Once the partnership has found an 8+ card fit, LTC can be used to establish the trick taking potential of those 2 facing hands. We do not use LTC to evaluate the potential of two balanced no trump style hands. On these hands we continue to rely primarily on point count.

Determining the LTC of your hand is as simple as “points” once you get the hang of it. Here are the LTC Guidelines:
1. If you are missing any of the 3 top honors in a suit, count each missing honor as a losing trick.
2. The maximum number of losing tricks that you can have in any one suit is three, so each suit card after the top three honors is completely ignored.
3. You can never have more losing tricks in a suit than you hold cards in that suit.
Here is some fine tuning you can apply once you feel comfortable with the basic LTC rules.
1. An unsupported Queen (Qxx) counts as 2 ½ losing tricks, but if it is accompanied by the Jack (QJx), it is only 2 losing tricks.
2. AJx is obviously better than Axx, so we count it as only 1 ½ losers.
3. If you have more queens than aces add ½ losing trick for each extra queen. If you have more aces than queens, subtract ½ losing trick for each extra Ace.

Here are a few practice examples:
1. AK64 KQ93 J3 432 (1 losing trick in spades, 1 in hearts, 2 in diamonds and three in clubs). Total LTC= 7 Hcps 13. Note the application of the principles that we look only at the first 3 cards (see spades and hearts) and that we can’t have more losers than there are cards in the suit (diamonds).
2. AK64 KQ93 A9 432 Total LTC=6 Hcps 16
3. AK64 KQ93 A9 K32 Total LTC=5 Hcps 19

Note: There is an inverse relationship between points and LTC. As points go up, LTC goes down.

Here are three more examples:
1. AK642 KQ94 J3 42 (like hand 1 above except it has a five card spade suit and a doubleton club. Total LTC=6 Hcps 13
2. AK642 KQ932 J3 x. Two 5 card suits. Total LTC=5 Hcps 13
3. AK6542 KQ932 J3 void 6-5 come alive! Total LTC=4 Hcps 13

Note: As the hands gets more unbalanced, the LTC goes down even when the hcps remain the same. Do you see why LTC better reflects the true playability of distributional hands?

The next step is to estimate the LTC of partner's hand based on his bidding action. Here are a few guidelines.
(a) Minimum opener 8 LTC Sound opener 7 LTC
(b) Weak 2 Bid (6-10) Disciplined 7-8 LTC Undisclined 9+ LTC
(c) 3 Level Preempt 6-7 LTC (Vul-NonVul)
(d) 4 level Preempt 5-6 LTC (Vul-NonVul)
(e)Simple Overcall 6-8 LTC
(f) Take Out Double 6-7 LTC
(g) Double and Bid 5-6 LTC
(h) Simple Raise of partner's opening bid 9 LTC
(i) Limit Raise 8 LTC
(j) 2/1 bid, Jacoby 2NT, Splinter 7 LTC or less
(k) 2 Club Opener 3-4 LTC
(l) Jump Rebid of Opener's suit 5-6 LTC
(m) Jump shifts/Reverse 5 LTC
(n) Jump to Game by opener in Suit bid by Responder at 1 level. 4 LTC or fewer

This is not a memory test.Start with opening hands, simple raises and limit raises and proceed from there. In Part 2 on LTC, we will see how to use our LTC skills to make key game going decisions.For futher reference see Ron Klinger's book Modern Losing Trick Count (1986).

Sunday, March 11, 2007

Smolen, Schmolen!

Your lucky day! Partner opens 1NT (15-17). Here are three different hands held by responder that we are asked to bid: (i) QJxxx, Kxxx, x, xxx (ii) KQxxx, J10xx, Jxx, x and (iii) KJxxx, Qxxx, Kxx, x . Notice that each hand is 5-4 in spades and hearts. If opener has 3 spades or 4 hearts you would like to play these hands in a major suit. If opener’s distribution is 2-3-4-5, you don’t want to walk by a playable NT contract. There are several ways to bid these hands, some more sophisticated than others, but as usual, sophistication is of minimal value and the important thing is to have an understanding with partner.

In Standard American you would start with a Stayman ask and opener usually bids 2 diamonds. Well, there goes our heart contract, but now opener is even a bigger favorite to have 3 spades. With hand (i) I next bid 2 spades and hope partner remembers that for us that is “to play.” Even if we have a 7 card spade suit, this hand looks to play better in 2 spades than 1NT. Why didn’t I simply make a transfer to spades? Because the moment I do, partner will show up with 4 hearts!! Note, once I treat this bid as “to play,” it no longer can show an invitational hand. The reason I treat it “to play” is that the same bidding sequence could be “Garbage Stayman”, and I don’t want opener bidding again. You make choices in bridge, and this is my choice.

With hand (ii) I might like to have my 2 spade invitational bid back, but due to my earlier choice, I now have to make a decision about playing in a game contract or playing 2 spades. The hand is too good for 2 spades, so I now bid 3 spades. This bid shows 5 spades and 4 hearts and is a game force. If partner has 3 spades I want him to bid 4 spades, and if he only has 2 spades, to bid 3NT. Hand (iii) is an easy game force and I bid it the same way as hand (ii), but I am now proud to table the dummy. Note that this bidding treatment gets the spade contract wrong sided. While having the opening lead come up to the strong hand has always been thought to be worth ½ trick, there may be some evidence that this advantage is overrated.

If you want to get the 5-3 fit right sided, there is another way to approach the bidding. You can start by transferring opener to 2 spades and then bid 3 hearts. Notice that with 5 spades and 4 hearts we are already at the 3 level. This better be a game force since partner is not going to pass 3 hearts. It is a little more comfortable with 5 hearts and four spades, but that does not cure the problem. Note that if opener has 4 hearts, we have again wrong sided the contract, but this time in the 4-4 suit distribution. Personally, I think the transfer and bid sequence is better left to describe a 5-5 in the majors or a 5-5 major-minor hand.

In an effort to cure the wrong sided problem a convention called “Smolen” was developed by Michael Smolen. You again start with Stayman looking for the 4 card major in partner’s hand. If he bids 2 diamonds, then you bid your 4 card major at the 3 level. This shows opener that you started with 5 spades and 4 hearts (or 4 hearts and 5 spades, as the case may be). Partner with 3 card support for spades bids 4 spades and without three card support bid 3 NT. Smolen can only be used with game force hands, there is no stopping point, but you can use the sequence described in paragraph 2 if you want "to play” the hand at the 2 level.

So, we got Smolen off the ground and all of a sudden over opener’s 3 NT bid we hear responder bid 4 hearts. This doesn’t mean that responder still wants to play in hearts; it is a transfer to spades. Why would responder still want to play in spades? Well, he didn’t tell opener that he had 6 spades because there was no “need to know.” Have you got a better way to show a 6-4 major suit hand? The reason partner didn’t make a Texas Transfer is that he wanted to fully describe this 6-4 hand. Two hands with a double fit are very powerful. My preference is to make a Texas Transfer with 6-4 if I have no interest in slam. If you use Smolen with a 6-4 hand, it usually suggests that responder has mild slam interest.

A comment from Alfred Sheinwold: "Bridge is essentially a social game, but unfortunately it attracts a substantial number of antisocial people."

Saturday, March 3, 2007

Advances to Overcalls (Part 3)

Parts 1 and 2 on advances to overcalls gave you the basics to make a simple raise, to make a preemptive raise, to make a limit raise or to show hands of opening strength in support of partner’s overcall. For those who want to add a little fine tuning to their bidding, there are various conventions to use when you are bidding in competition and wish to furnish overcaller with specific information about your holding in his overcalled suit.

Rosenkranz et. al.
One such convention is called the Rosenkranz Redouble named for George Rosenkranz, a renowned steroid chemist and bridge player from Mexico. When partner makes an overcall and responder makes a negative double, a redouble by advancer shows three card support for the overcall including at least one of the top three honors in the suit. A simple raise also shows 3 card support but denies holding one of the top three honors. Let’s face it, when they open the bidding, most likely we are going to be defenders and we want to get off to a good start. If partner holds AQJxx in the overcalled suit, he wants to know if it is safe to lead the suit. In this case the redouble says I hold one of the top honors in your suit so lead accordingly. Some use Reverse Rosenkranz where the meanings of Redouble and the simple raise are reversed. Choosing between the two formats is a matter of whether you would like to raise the bidding to the 2 level with 3 small cards or 3 to a top honor.

Some also play Rosenkranz Doubles so that the sequence 1c/1h/1s/x shows 3 card support with a high honor and a raise shows 3 card support without the honor. If the sequence is 1c/1h/2c/x, this double would normally be a responsive double showing spades and diamonds. If you want to make this a Rosenkranz Double showing support for hearts, then it cannot be a responsive double. I have played Rosenkranz Doubles and Redoubles in combination and feel that they gain more than they lose. It is something you have to talk through with your partner.

Munson Redouble
This convention was developed by World Champion Kitty Munson Cooper who was introduced in an earlier post. The Munson Redouble shows either Ax or Kx in partner’s suit, or singleton A or K. Any bid other than redouble denies the Ace or King doubleton or singleton. In Munson Redoubles overcaller leads a spot card to advancer which is suit preference. If it is apparent that advancer can get a third round ruff in the suit (overcaller under-leads the Ace or King) then he returns the overcalled suit. Note that a third round ruff would require a 3-3 break which is against the odds, but a ruff even with an over-ruff may still set up a trump promotion for partner. If advancer has a singleton or a natural trump trick, he would take the suit preference signal.

Getting Started
My recommendation is to start with Rosenkranz Redoubles and fine tune your personal preference from there. It is easy to recognize, what else would you be doing by dropping a redouble on top of opponents negative double? Partner will be happy to know that you have 3 card support, and even happier if you have one of the top three honors.