| Home | Contents | May 25 | Appendix 1 | 05-30-2006 09:50:54 |
Show your thanks and lower your stress level by keeping a gratitude journal. At the end of each bridge game, check three boards that you are grateful for. Research shows you will be healthier and happier if you do, Prevention magazine reports. (Paraphrased from May 30 newspaper.)
Opening
I want something to do, so you specify either point count or controls. 2 
is always the weakest
bid. Then
is Kokish, and forces 2 
with the exception that you can show a long
minor. A 2NT continuation forces game with a notrump-like hand. Stayman and Jacoby apply.
Other bids show hearts and the named suit, and 3NT shows hearts and notrump. should be 22-23. 24 + HCP is a 2 bid.
Warning: Often, signaling is a distraction. It is easy to write signaling rules. More often than not a signal is only a suggestion to partner. Signaling is not the end in itself.
The important element of defense is to visualize the unseen hands. If you are only going to do one thing on a hand, it should be visualize rather than figuring the correct signaling card. It is difficult to write a set of general visualization rules, and to create good visualization problems.
In a single game around the time I wrote this material, I counted eight deals where the opponents misdefended, and one deal where we misdefended. All of the misdefenses were solo errors by one of the partners. Defensive signals had nothing to do with any of the errors. Perhaps one or two of the errors were excusable.
Your leads and signals contain information. We should think about what is the most useful information, and use that approach. The three types of messages coded into signals are attitude, suit preference, and count. Both attitude and suit preference can be ephemeral. Your attitude can be positive, negative, or neutral. If your lead shows or denies an honor, then your lead contains an attitude message. On some deals you do not have a suit preference, and on others both suits are okay. Count is straightforward, but it is rare that count in a side suit led by declarer is useful.
The standard signaling rules are, I think, attitude when you lead the suit, count when they lead the suit, and suit preference sometimes replacing attitude or count. Normal count is high-low with an even number in a side suit, and low-high with an even number of trumps. The count card is current count. Count in a partnership can be rare to frequent to mandatory.
Granovetter's Obvious Shift Principle codes more into an attitude signal. The encouraging signal can mean
The Granovetter's rules for showing count, page 10 of his book, are:
Cases 1, 2, and 3 apply to the opening lead. Case 4 applies when they lead the suit. Case 5 is vague. It can lead to the after the deal question: "How was I to know that you knew that I knew that high cards were known?" Did I miss an "I knew"?
The order that you, as defender, play cards in a suit, depends on the contract, and whether you are leading the suit, following to partner's lead, or following to declarer's lead. Let me paraphrase John Lowenthal: You should load the correct way to play all card combinations as defender into your database. You then fish them out of the database and load them into the foreground window as applicable.
Leads:
Signals versus notrump:
Signals versus suit:
The lists are approximately in order. Attitude signals are given before suit preference signals.
Computer-generated lead tables. Reliance on a double-dummy solver has merit, but is not definitive. It is certainly better than showing a single layout, and announcing +50 for the author's lead and -110 for all other leads.
Gib, the double-dummy solver, used here, determines the number of tricks won after different opening leads followed by perfect play. Gib "perfect" is not real world "perfect." Expert "perfect" would be the percentage play allowing for different distributions and location of honors. According to Gib, expert "perfect" makes mistakes in both offense and defense. Decisions by real defenders and real declarers made early in the play are less likely to be Gib "perfect" than those made late in the play.
The most accurate lead tables would be created for each possible auction, they would use many more deals in building the tables, and they would even consider the defenders vulnerability. What follows is the summary of a generalized lead table built for a 1NT opening, no overcalls, and a responder who does not indicate a long suit other than trumps.
In most of my analyses, it does not matter on whether you are playing matchpoints or IMPs. You need more deals to positively assert that a given lead is better at IMPs. At IMPs, a single 10 or 12 point swing determines whether you win or lose.
You lead top of nothing and that picks off partner's queen. In the Gib world, the only cost is timing. In the real world it costs a fraction of a trick. It is less than half a trick because declarer may not touch that suit. Gib scores top of nothing too high.
Gib does not like trump leads. I add this to Gib's objection to a passive trump lead. When you or partner gain the lead, it is not too late for the crucial switch. It is in one of two suits. You have missed the discovery play of the opening lead in the non-constructive suit.
Gib rates ace leads, unsupported or from ace-king, too high. You have to do the right thing at trick two, and some of the time, that is not obvious.
Gib likes to lead from touching honors. I have been following this advice and feel good about this recommendation. However, the next time you are on lead, you have to decide between the other honor, low in the suit, or another suit.
Gib doesn't like fourth highest. However, fourth highest as well as top-of-nothing are discovery plays useful in later defensive decisions, and can delay the crucial defense decision to late in the deal.
Gib recommends short suits against suit contracts. Gib is happy to lead a singleton king. That is an exposed card in Gib. You probably do not want to lead a singleton queen either.
Gib's least favorite suit length is three cards. My argument to explain this is long-winded.
Then, of course, there is the problem of bias in the deals chosen for the experiment. If I don't set the problem correctly, or you apply the results to a different problem, it could be garbage in - garbage out. Except in building the lead table, I have payed close attention to the bidding. However, people do not bid the way I think they do, and some even bid the way I think they should rather than the way I think they do.
Last week's play topic was "Versus a Weak Jump Overcall". I do not use weak jump overcalls. You made a weak jump overcall to provide experience in defending against them.
We all make several mistakes every time we play, and we have to learn from them. My example: (What's yours?)
1NT* Pass 2 Pass 2 Pass Pass Dble Pass ?
* 12-15
Partner made a takeout double and takeout doubles are to be taken out. Pass collected 100.
Pulling the double to 3 would have been 140, or maybe 730.
Continuations over a preempt. (Robinson) New minor forcing. All other bids not forcing. Hasn't come up yet.
Lightner Doubles and Grand Slam Force. Do we know what these low frequency bids mean?
North-South Vul.
1 2NT 3 Pass 4 5 6 Pass Pass (Dble) Pass Pass Pass
* 16 plus points and artificial.
What do you, as West, lead? At the table, East did not double. The double-dummy solver says
that you make 6 without a heart lead. The East noise had exposed her hand, and partner did
make 6 .
When they interfere with Blackwood: Use Double — Zero-Pass-One because as DZP1, it is a mnemonic not misinterperable.
Law of Total Tricks concept is good. All the toys that come with it are stupid.
Double of their artificial bids when we are not in the auction should be lead directing. This includes Stayman of weak notrumpers.
Fast arrival after a 2/1 response. I think you don't say: "We are going to bid beautifully", and then not bid beautifully.
There are other things we do that are low frequency. You find them in the big club notes.
| Home | Contents | May 25 | Appendix 1 |
