[mike777]

luis, on Nov 7 2005, 12:16 PM, said:

I don't think you can say that judgment is mechanic.
With the same hand after the same bidding I might take a different decision based on the state of the match, scoring, pd, opponents, what happened in the day so far, etc etc.

ok all of that sounds like a set of rules to me but thanks. Agree a very long set of rules but still rules.

1) You decide what factors are important.
2) Based on these factors and combination of factors you make a decision.


example...factor 1 and 5 and 8 and I feel hot I bid game but same factors and I feel cold I pass.

[Robert]

hello pigpenz

The LAW is certainly sometimes wrong, however, it does serve as possible bidding guideline. Larry Cohen sometimes appears to violate the LAW when he bids in Bridge World auctions. Maybe he 'knows' something that I do not and maybe he is making some unknown(to me) correction?

The LAW has a lot of corrections 'if' it functions properly. Your 5332 opposite a 5332 shape is just one of those corrections. It the hands match up exactly with the doubleton opposite each other, the number of losers will be 'way above average.'

You devalue a holding of 5332 opposite a 4(5) card holding. WHen I refer to opening a 5332 major hand, I often refer to it as a 'death trap' hand because of the number of losers. Any other 5 card opening shape has at least one less loser.

If you hold a fairly normal 5431 shape opposite either 5332 hand 'your' LAW tricks will be much closer to normal. Two 5431 hand with the shortness 'not matching' will take a lot of tricks. You can often estimate how the hands 'fit' and make corrections in the LAW.

It is often overlooked that the LAW is based on the HCP being fairly well divided between the two pairs. If my pair has 40HCP your 4-4 card fit will not take anywhere near 8 tricks. Even if I hold near game values, the total tricks should be 'adjusted' upwards in view of my added HCPs.

I use a 'Stayman in Doubt' method in my 1NT bidding to 'avoid' playing a 4-4 major fit with two 4333 hands facing each other. A 4333 shape is devalued by most serious players for suit play.

One advantage in 'knowing' the number of trumps is that many experts normally 'show' the trump queen when they 'know' that they hold a combined ten+ trumps. AKxxxx xxxx and Axxxx Kxxxx will not often care if they are missing the Queen(and Jack) of trumps.

This does not always work. I remember that I once held AK 8 times and had xx in dummy. The first trump lead found RHO showing out. I never claimed to be a good card holder.

Regards,
Robert

[pigpenz]

yes Robert
Ginsberg and some French person did a statisticaly analysis of The Law of Total Tricks, its not a fool proof method. The anlaysis does show how close it does apply.
I think the bridge world one did an anlaysis from bermuda bowls.

As far as bidding in Bridge World...the challenge the champs contest is a different beast altogether Its to test bidding systems and pairs ability to handle difficult hands with their systems opposite the other pairs system.

I believe Granovetter and Rubin had the best record in the early 80's or late 70's where they won challenge the champs for who knows how long with their relays system.

[han]

luis, on Nov 7 2005, 11:57 AM, said:

The LAW is a good tool, as any tool it is not perfect for every problem you have.
But it is indeed a good tool, adjustments and general judgment are still needed, there's no way to blindly take a decision just based on the LAW.

"I fought the LAW" is in my view a poor attempt to sell a book attacking a popular topic. The stats are completely biased, I can write a book titled "I fought the finesse" presenting 1000 hands where kings are singleton offside and what does it prove? Absolutely nothing.
Instead of "fighting" the law a book about guidelines to understand how to better use the law, applying adjustments and generic statistics would be much better.

Please keep these posts coming Luis, they make my day.

[Kalvan14]

It looks like there is a kind of consensus: LOTT is a somehow blunt, but helpful tool for beginners and advancing players.
I am not the guy for sampling and statistical analysis, but I have the feeling that LOTT is at its best evaluating semi-balanced hands. It has a negative bias when evaluating high-level, very unbalanced hands. It makes sense.
Lawrence's book does not reveal a new Gospel. It suggests adjustments to the Law, and formulas for correcting. Fine and good, but possibly not very helpful: this book is again for a mass market, beginners and advancers. Otoh, while the LOTT was marketed in an easy, practical way, the SS adjustments require more work.
If this becomes popular (doubt it ), it will take 10 minutes to bid any hand

In my view, it's like the MWP and Goren's adjustments for shortness were not much different, or off the mark. Whatever.

Expert player make their own adjustments, based on experience, card sense or table presence; and they listen to other people bidding.
Very often, with hands that look very similar to the beginner, their bidding is quite different. And most of the hands are bid and played on automatic pilot, or thereso.

[Chamaco]

Kalvan14, on Nov 7 2005, 11:03 PM, said:

Lawrence's book does not reveal a new Gospel. It suggests adjustments to the Law, and formulas for correcting.

This is not quite true.

Lawrence's book criticizes the Law bringing up unfair examples, e.g. examples with duplication of shape/values, and other minuses that were already addressed by Cohen as "needing adjustments".

He goes on using statistics in an unfair way to say how man times the law does not work.

Finally, he suggests using a method of numericaly assessing the trick-tacking potential, that is NOT a real, practical, alternative to the Law, because it has different purposes (the Law tries to assess whether a jump raise will be a good sac opposite the alternative contract by opps, FTL just tries to assess whether we'll go down, regardless of which alternative contract opps have).

So, it's fair to say that the Law of Total Trick is not a Law.
It is also fair to say that often it fails and that it should not be used as a panacea.

But the way it was done specifically in the book sounds more like a battle of religion and/or a marketing affair.


Critics are always food for thought, but bringing down an opposing theory by using unfair examples and a misuse of statistics is bad.

[whereagles]

luis, on Nov 7 2005, 04:57 PM, said:

"I fought the LAW" is in my view a poor attempt to sell a book attacking a popular topic. The stats are completely biased, I can write a book titled "I fought the finesse" presenting 1000 hands where kings are singleton offside and what does it prove? Absolutely nothing.
Instead of "fighting" the law a book about guidelines to understand how to better use the law, applying adjustments and generic statistics would be much better.

Comparing a well-documented (though poorly written) book like "Fought the Law" with singleton-kings-offside "Fought the finesse" is a rather unfair comment, Luis. Come on, you know better than that. Every book obviously has some bias, but Lawrence's book is FAR from "completely biased".

I agree Lawrence's SST/WP evaluation method sucks in terms of practicality and that the law of total tricks rules supreme in that respect. But to dismiss months of serious work by Lawrence and Wirgen, labelling it as "populist attack on a mainstream topic", is going too far.

[Chamaco]

whereagles, on Nov 8 2005, 10:49 AM, said:

Comparing a well-documented (though poorly written) book like "Fought the Law" with singleton-kings-offside "Fought the finesse" is a rather unfair comment, Luis.

The book is not quite well documented:
there are MENTIONED the summary of hundreds of hands, but the precise conditions of the analysis and simulations are not illustrated.
A "well-documented" summary usually should allow the readers to verify whether he can agree/disagree with the boundary conditions, as well as to double check if there are mistakes.
Basically, "well-documented" means delivering access to the raw data and the methods, to enable someone else to verify whether te math analysis might output a different hypothesis than the one presented by the authors.

Also, claiming that LOTT fails when it deviates by ONE TOTAL TRICK is very silly: ONE TOTAL TRICK means more or less half a level, that is, the margin that is left to hand evaluation and adjustments.

If we indeed consider the times when the LOTT deviats by at most +/- 1 TOTAL TRICK, the % of success is quite significant (if I recall correctly, in the 75-80% range).

Finally, the most interesting parts of such books are concrete hands and examples, but the chosen examples are carefully selected for worse case conditions (4333, shape duplications and /or duplication of values).

But such factors are used to adjust in ANY method: Cohen suggested it in his 2nd book, and in any case they should be used even in more traditional evaluation methods (e.g. Lawrence's book on Hand evaluation).


Hence, the documentation and the argumentation is very partial, and above all absolutely UNSCIENTIFIC, despite the way the data are presented might lead a naive reader to believe that the analysis is objective.

The misuse of statistics can demonstrate almost any nonsense, if one can play with numbers and select the right case histories.

[Kalvan14]

I was not trying to defend I fought the Law.
Actually, when I read the book, i found it more than a bit annoying. It was truly a kind of war of religion, which made me think there was something personal between Bergen and Lawrence.
This said, the idea of finding a kind of unified field theory for bridge is quite funny
I would anticipate that this is not possible. Luckily so: were it possible, I would stop playing bridge.

[CarlRitner]

Kalvan14, on Nov 8 2005, 07:17 AM, said:

It was truly a kind of war of religion, which made me think there was something personal between Bergen and Lawrence.

That's twice now I've seen Marty Bergen instead of Larry Cohen, the author of the LOTT books. These two used to be partners but Marty doesn't really have any involvement in this discussion.

It's a great discussion, by the way; things like this help players at all levels to understand where the other folks are at in terms of hand evaluation techniques, and opinions of the different approaches.

The evaluation techniques expressed in I Fought The LAW are quite valid, if you have the mental acumen and capacity to do this at the table during the auction; apparently Lawrence and Wirgen feel there are a sufficient percentage of players capable of this to warrant the publication of the book.

Until the ACBL permits players to keep a computer running, as an aide during the auction, I'm afraid these techniques are beyond my level

Cheers,

Carl Ritner
ACBL Library Used Bridge Book and Magazines

[Chamaco]

Carl, on Nov 8 2005, 03:18 PM, said:

The evaluation techniques expressed in I Fought The LAW are quite valid, if you have the mental acumen and capacity to do this at the table during the auction; apparently Lawrence and Wirgen feel there are a sufficient percentage of players capable of this to warrant the publication of the book.

Their suggested method sure seems a good one, but my point was that it is not an alternative to the LOTT, because it has different purposes: the LOTT goal is to find the PAR of the hand, even if you go down (because opps have a beter score), the FTL method just tries to compute how many tricks your side can make, regardless of the opps potential.

That's a competely different story, and that's why it is silly to compare the 2 methods.

=====

I'll repeat here that I am not criticizing their suggested method, but I strongly dislike the inappropriate way they attacked the theory of the LOTT: they used biased examples and inappropriate statistics.

[barmar]

Chamaco, on Nov 8 2005, 04:42 AM, said:

So, it's fair to say that the Law of Total Trick is not a Law.
It is also fair to say that often it fails and that it should not be used as a panacea.

The LOTT is a law in the same sense as the Law of Averages. When randomness is involved, that's about the best you can generally get.

[whereagles]

Chamaco, on Nov 8 2005, 10:58 AM, said:

Basically, "well-documented" means delivering access to the raw data and the methods, to enable someone else to verify whether te math analysis might output a different hypothesis than the one presented by the authors.

Hence, the documentation and the argumentation is very partial, and above all absolutely UNSCIENTIFIC, despite the way the data are presented might lead a naive reader to believe that the analysis is objective.

That is true, but you forget something: Lawrence's book is for the general audience. It is NOT a scientific article. You have to reach a comprimise between precision and understanding of the point you want to make. What good does it make to fill dozens of pages with curve-fittings and simulation data? People just want to know the conclusions of the study, and that is what Lawrence gives them.

As for the examples shown, even a 100% honest analyst, which Larry Cohen is NOT - he clearly 'fits' the hands to prove his point, cannot escape a bias. The point, however, isn't to provide results of random hands, but to give you food for thought.

[Chamaco]

whereagles, on Nov 8 2005, 04:25 PM, said:

Quote

That is true, but you forget something: Lawrence's book is for the general audience. It is NOT a scientific article. You have to reach a comprimise between precision and understanding of the point you want to make. What good does it make to fill dozens of pages with curve-fittings and simulation data? People just want to know the conclusions of the study, and that is what Lawrence gives them.

That is true, but the book is presented under the aura of "making an objective statistical analysis".
It should be pretty clear that it is not so.

As far as being a book for a general audience, I have nothing against that, as long as this is quite clear, and that the critics are not biased with carefully chosen examples

Quote

As for the examples shown, even a 100% honest analyst, which Larry Cohen is NOT

Sure, I agree on this.

Quote

- he clearly 'fits' the hands to prove his point, cannot escape a bias.

I disagree.
If he wants to bring these examples, it's ok, but then he should mention that Cohen himself would downgrade hands with mirrred shapes or duplication of values or 4333.
You cannot criticize a hand evaluation methods based on its misapplication, this goes beyond good faith.

Quote

The point, however, isn't to provide results of random hands, but to give you food for thought.

I agree and I like "food for thought", but I do think that this should be done in a fair way.
There are many examples where the law fails genuinely and these examples are very interesting, but using systematically examples where it fails because applied in the wrong context does not seem to me the proper way to "feed" any thoughts

[pigpenz]

I have not read Lawrences book but I am sure both of them use hands to fit their purposes. The only way to look at either is to just sit down and look at hand records after handrecords and come to your own conclusions.

[Chamaco]

pigpenz, on Nov 8 2005, 04:43 PM, said:

I have not read Lawrences book but I am sure both of them use hands to fit their purposes.  The only way to look at either is to just sit down and look at hand records after handrecords and come to your own conclusions.

That's fair, but when I decide to do that, I would not buy a book.

Buying a book that titles "I fought the law", I expect a *somewhat* but not *totally* biased analysis: I know pretty darn well that the Law fails in given hands (I paid my prices for that), and indeed I wanted a good read about that that would add more to the subject, besides the adjustments mentioned by Cohen in "Following the law".
But, buying a book with such a title I expected an analysis more detached and objective.

As it is, the analysis is pure desctructive: it appears that the LOTT is close to useless, which is not true either.
This is rather annoying: if, like you correctly say, the only reliable approach is "do your homework" (which I do) then I'll stop buying bridge books, but that would be sad, meaning that bridge authors are not capable to analyze some topics with a decent deal of objectivity.

The part of the book that has some values is the one on their method to compute trick-taking potential, although the results are often very close to the application of a straightforward and oversimplified method such as LTC.
Besides, it's not really what one expects to get when he buys a book with such a title...

======

What makes me even more sour is the fact that for me this is a "deja vu" from the chess literature.
Pre 1995 chess literature had a good standard, because every book chapter had real contents, selected by human persons with specific cases histories that were fitting with a given topic.

Then, starting the mid 90s, the big databases became available, and the quality of many books dropped. Lots of books became just big database dumps, because dumping and analyzing by PC lots of data became quite fast, quite faster than writing a REAL book, and putting real effort in analyzing single, complicated, and deep positions.

This "Fought the Law" book gave to me the same impression, in a way.

[whereagles]

Chamaco, on Nov 8 2005, 04:51 PM, said:

Then, starting the mid 90s, the big databases became available, and the quality of many books dropped. Lots of books became just big database dumps, because dumping and analyzing by PC lots of data became quite fast, quite faster than writing a REAL book, and putting real effort in analyzing single, complicated, and deep positions.

Once the technology is there, people use will it. I don't consider it as good or bad... it's just a sign of the times.

By the way, about biasing: I once read a chess book by Alekhine and his writting was VERY dogmatic. As you see, even the greatest players of all time fall into the trap of biasing

[Chamaco]

whereagles, on Nov 8 2005, 05:27 PM, said:

By the way, about biasing: I once read a chess book by Alekhine and his writting was VERY dogmatic. As you see, even the greatest players of all time fall into the trap of biasing

Alekhine played and wrote in the years 1910-30s.
Besides, he was universally considered one of the most objective commentators of the time, one that introduced the modern way of commentary (the same followed by Bronstein in one of the best books of all time, the commentary of the Zurich 1953 tournament).
Many think that especially his ability in annotating, leaning from errors, and selfcriticisms is what led him to defeat Capablanca in the world title match.

Capablanca or Nimzowitsch articles/books (same period) appear to the modern reader even worse in terms of dogmatism, and we are talking of really great players.


It is only natural that all of these books seem overly simplified to the reader of the 2000, in between we had many generations of other authors that have elaborated many nuances from the time these giants were playing :-)

If any of them was living, playing and writing today, he's be writing ina different way and most of its articles would go over our head, trust me

[Chamaco]

Quote

Once the technology is there, people use will it. I don't consider it as good or bad... it's just a sign of the times.

I do not hate technology: if someone uses a new technology to improve teaching or the level of documenatton and play, then it's great.

For instance, the tools Fred introduced (Bridgemaster, BBO, and now Full Disclosure) are indeed revolutionary, but they do not drop the quality of the products, they indeed enhance them.

Instead, when you do use large AMOUNT of data, but lose depth and originality in the coverage, that's where we lose quality.

Technology is not evil: if hi-tech is used for a better product, I am all for that.

[Canuckstan]

Well, since we veered off into chess, maybe this is the place to ask about the Hocus Pocus books, which compare chess and bridge. I didn't buy it but skimmed one of two volumes in the book store. I just couldn't get my head around what exactly the book was trying to say so saved my money. But it was close!