[jtfanclub]

Quote

Couple of things about your interesting analogy.

1) I could go up to Tiger Woods and say "Tige, old son, your putting stroke is all wrong.  Sure, I never score less than 120 myself, but I really know all about putting."  What would the reponse be?  (And rightly so).

But that's Tiger's field. You can't be a good golfer without having a good stroke. Not only does making golf clubs not make you a better golfer, most expert golfers have never designed a golf club.

Quote

2) OK, Golf club design.  Perhaps a keen hacker who does not know what good golfers look for in their clubs comes up with a club that is very good for XYZ.  But the top golfers say "But I don't need to do XYZ, I need to do ABC".  Not getting feedback  from an expert golfer has contributed to the poor design.

But that's the Catch-22. Absolutely, you need feedback from expert players on things like this. But what if experts refuse to test systems not made by experts?

Obviously, there have been exceptions. I don't think CC Wei was a 1-A player, and his system did wonders. But that was because he found himself coach of a team that got to play against the top players. It would be nice if such serendipity wasn't necessary.

In this case, this seems simple to me. Inquiry, for example, could find 100 previous cases of hands where the auction tended to go 1♠-2♣-2♦-2♠ when playing 2/1. Do bidding only from there- no play, and then compare where they got to different contracts. How long would it take experts to bid 100 non-competitive auctions where the first four bids were known? I'll bet it's less time than Fred has already spent on this thread.

Even if Ken's system did turn out better, that wouldn't prove it wasn't too complicated, or an unecessary brain drain. But if it did the same or worse, that would be a strong sign that the system was not superior.

I'm just using Ken's case as an example. I haven't read the book, I'm not much of a system designer myself, I don't have a horse in this race. But it would be nice to see bridge improve beyond the fig leaves of 'gadgets'.

[fred]

Stephen Tu, on Jan 11 2008, 05:36 PM, said:

If you feel this is on par w/ 4 vs 5 cM, then why such a strong opinion about it? Just say "answer is unknown, I don't think it's worth the time to figure it out, thus suggest to play what all the experts play", rather than also including all the insinuations (unnecessary & unwarranted IMO) about inferiority/absurdity?

You just don't get it. Here is the answer to your question. Please read carefully:

Because I thought that people might care about what a successful player thinks, based on his judgment and experience.

Whatever "insinuations" you think I might have made, they only started after a player with considerably less experience and very likely considerably worse judgment than me, posted that his judgment and experience suggested otherwise.

When I pointed out that this person was probably in a minority of 1 and that greatly increased the likelyhood that he was wrong, all hell broke loose.

I am sorry but I really have nothing more to say. Either I have not expressed myself clearly (in which case I apologize), you are not paying attenion, you can't stop arguing even when you know you are wrong, or we are just never going to see eye-to-eye on this matter.

Regardless I have only so many keystrokes in me each day and I am not going to waste any more of those on this discussion.

Fred Gitelman
Bridge Base Inc.
www.bridgebase.com

[SoTired]

Fascinating thread!

One point that nobody has mentioned yet: Judgement is based on experience. Experience is based on the tools and methods used. The vast pool of experts that Fred is referring to who use "pattern" are used to that method. Their judgement is based on that method. They are comfortable with it. They get good results from it. If forced to use Ken's methods, even if it was technically superior, they would probably do poorly for a long time before their judgement caught up to their comfort level.

Personal analogy: I take notes with a pen and pad of paper. My daughter takes notes with her laptop. I can argue why my method is best and so can she. But the fact remains, we do it that way because that is what works best for us by experience. That is our comfort level. If her laptop broke and she was forced to use pen and paper, she would probably do a terrible job. Same with me if my pen ran out of ink and I was forced to take notes with a laptop.

[kenrexford]

mike777, on Jan 11 2008, 02:07 AM, said:

"As an aside, here is an interesting auction.. what does 4♣s here mean (no partnership discussion, just what do you think it should mean)?

1♠ = 2♣
2♥ = 2♠
4♣ .... "

splinter


Partner just told me this is what he bids with

AKxxx..AKxx...xxx...x
I guess he prefers to pattern out via a splinter as opposed to bidding the 3 card suit.

Same for me, but ever=so=slightly better hearts expected (three of the top four honors). Everything else same.

[kenrexford]

I'm expecting Fred to not read this, per his claims. However, just to toss it out and be done.

You have no apparent recognition of how arrogant and condescending you were. You have no idea of whether I have talent or not. You have not read my book, nor do you know what I propose therein, obviously. Having no knowledge, you cannot possibly have an informed opinion, and yet, you reject the idea outright, publicly advise people regarding why (credentials), all while illustrating that a simple auction for you is undiscussed, by the way. You imply, with statistics like "minority of one" and similar comments that no one of talent endorses the idea, implying thereby that you have consulted credentialed talent who has read the book and considered the ideas and that you have a list of such people rejecting the idea after this due consideration. You cannot or will not provide one simple example of anything suggesting how a problem exists. Instead, you cite the "never heard of him" argument.

There are notable "famous" folks who have reviewed the book and the ideas favorably. You can check around if you want. Others have liked some but not all of the suggestions. I expected that. What I prefer is for someone to review the book by reading it rather than reviewing me, especially if you are also under-informed about that latter topic.

[glen]

I will reply in two posts – this post will discuss the lone wolf versus the establishment concern (the crackpot versus the rest of the civilized world etc.). First let me quote from Michael Lewis's MoneyBall

Lewis Michael. Moneyball: The Art of Winning an Unfair Game. 2003. said:

Seven years into his literary career, in the 1984 Baseball Abstract, [Bill] James formally gave up any hope that baseball insiders would be reasonable.  "When I started writing I thought if I proved X was a stupid thing to do that people would stop doing X" he said.  "I was wrong." …
A full decade after James stopped writing his Abstracts, there were still two fresh opportunities for a team willing to take them to heart.  One was simply to take the knowledge developed by James and other analysts outside the game, and implement it inside the game.  The other was to develop and extend that knowledge.
... The whole point of James was don't be an ape!  Think for yourself along rational lines.  Hypothesize, test against the evidence, never accept that a question has been answered as well as it ever will be.   Don't believe a thing is true just because some famous baseball player says that it is true.

Actually what Bill James faced was not just one famous baseball player, but the combined school of thought about baseball developed over many decades.

Here we have a similar situation, the vast expert consensus versus a few “random non-famous” types suggesting possible improvements.

I don’t think anybody should feel offended by Fred’s postings here, as he clearly speaks the truth, even though it might not be nice to hear for some. Lone wolves just have to accept they are not running with the pack. In particular:

>> Fred is not rejecting the idea on the basis of its merits (or lack of them) – he is dismissing it for an advancing player solely because it is not commonly played in the expert community – this is quite typical with the introduction of new approaches and lone wolves need to accept that it happens all the time - Bill James faced this, and so did those, for another example, who believed in aircraft carriers instead of battleships.

>> Fred is not “condescending” when he points out the lone wolves have no proven track record in which to authenticate their ability to discuss these matters. This is simply a reality check. In bridge we have seen quite poor players attempt to pass themselves as world class professionals. I’ve personally seen, nobody in our present company of course, a number of somewhat clueless people write bridge books presented as expert advice.

>> Kind and sympathic reviews of an idea (or a set of ideas), in book and article format, are not sufficient to elevate an idea into best-of-class or better-than-the-experts status. It just means that the book or article presented thought provoking ideas in a decently written manner. Endorsement of the ideas arrives when the top players use the ideas.

I think it is fair for Fred to have made his comments to the person looking for advice. I also believe those who have taken offence from his comments are out-of-line. While it is disappointing for inventors that Fred has decreased the number of guinea pigs available to try out new ideas, he is giving advancing players good advice to stay on the well-trodden path instead of trampling through the untamed forest.

Take an advancing partnership that takes up cuebids as in this thread, and assume that the partnership, at some point, finds themselves unable to reach a good slam on a hand. Afterwards, they consult with a local expert, and guess what they find out – not how to cuebid the hand better, but they are playing the “wrong” methods. If they had used pattern showing methods, and were not able to reach a slam, the expert could provide valuable input into how they could have judged the mesh of values on the hand and suggested a potential sequence to get to the slam. Thus playing the expert consensus methods confers an advantage beyond the method itself.

I believe inventors need to leave the taking up of their ideas to the early adopters, who get enjoyment at being the first to try out the very latest. It is a poor service to the bridge community for inventors to present their ideas as mainstream ready when they are just a few years off the drawing board. Instead the inventors need to understand what the current practices are, accept (and relish) their status as lone wolves, and live with the put downs of their ideas simply because they are new. It’s all part and parcel of being a “mad scientist” leading the way.

[glen]

I have the following three problems with the immediate cue approach:

1) It seems relatively poor for choice of game, which is a very underrated success factor. It seems one is locked into the major suit game or a higher contract, and one can’t play 3NT, which can be a better place to play.

2) It seems poor in the choice of slam strain, as it can’t clearly distinguish the degree of secondary fits.

3) It discloses a lot of information on the way to game. In the 80s some local mad scientists were using this approach against us, and had a cuebid bonanza on the way to 4 of their major. We were presented with a roadmap from the opening lead on. Non-disclosure on the way to game is another underrated success factor.

However it is not like I’m thrilled with the patterning out approach. To use analogies in this thread, for me picking between immediate cue or patterning out is like deciding whether to hit a 6 or 7 iron on the par 5 tee, or like picking the very best sword to use against the approaching tank.

In Hardy 2/1 (blue book) he had the 2M rebid by responder show 12-15 or so, no singleton, and thus places opener in charge of investigation or not. In the evolved style, where 2M covers a wider range of hands, the number one thing the partnership needs to sort out first is do they actually need to explore for slam. Thus some bid must be allocated to opener to show not much extra, and it needs to be something that does not disclose much as opener will play contract in the major. This bid needs to be low enough that responder, with extras, in HCP and/or shape, can show slam interest. However the bidding approach available for responder also needs to be able to offer choice of game.

Methods that disclose too much when the partnership doesn’t have the assets for slam are flawed imo, and methods that don’t offer choice of game possibility will be long term underperformers.

[WrecksVee]

As to pattern showing or cue bidding after 1♠-2♣-2♥-2♠ I decided to see what I could find in bridge literature.

Bridge World Standard did not address this as far as my quick scan could determine.

Kokish and Kraft's "Modern American Bidding" falls in the pattern camp. But the suggested method is conventional. Opener rebids either third suit at the three level to show shortness. Rebids of 2NT, 3♥,3♠ and 4♠ were reserved for =5=4=2=2 with specific definitions.

Steve Robinson's "Washington Standard" 2nd ed. uses cue bidding.

Anyone else have any books that have a view on this?

[glen]

WrecksVee, on Jan 12 2008, 06:30 AM, said:

Steve Robinson's "Washington Standard" 2nd ed. uses cue bidding.

good spot. I see Robinson uses 2M-3M by opener to show a minimum, while immediate cuebidding shows extras.

[kenrexford]

As always, folks have different preferences as to methods, based upon personal assessments of risk-benefit analysis. This got me to thinking about two variations that seem reasonable and may be worth consideration, for the cuebidding camp.

One is a simple change to the impact of a 2NT (poor trumps) cue. One might agree that any cuebidding sequence that includes a 2NT call can end at 3NT. Thus, for instance, consider 1♠-P-2♣-P-2♥-P-2♠. If Opener cues anything but 2NT, the contract would be forced to 4♠ (if game only). However, under the variation, a 2NT call could yield a final contract of 3NT.

A second variation, along similar lines, might be to invert the meanings of a Picture Jump to 3NT and a Picture Splinter in partner's second suit. Thus, after 1♠-P-2♣-P-2♥-P-2♠, Opener might jump to 3NT to show a stiff club, good trumps, great hearts, and no diamond control, a call that Responder might actually want to pass. That would make a 4♣ call by Opener instead show good trumps, no heart control, a stiff diamond, and HHx in clubs, a holding where 3NT is very unlikely to be right.

A third possibility I have considered is also to invert the meanings of 3NT and 2NT, where 2NT is "serious" and bypassing 2NT "non-serious," or perhaps frivolous and non-frivolous, with 3NT being the "poor trumps" cue. The upside is for those who very highly value non-disclosure when a quantitative bash analysis suggests that slam is remote. The downside is that it negatively impacts the nuances available for delayed picture jumping. But, that itself might be worked out sufficiently.

BTW -- Fred and I have made peace. Egos do battle, but gentlemen reach accords.

[kenrexford]

The variational scheme has me thinking of a methodology that may appeal to some who have strong feelings about other schemes. One could easily use a blended scheme. Here is how it would work.

2NT would be serious. One could play that 2♠ over heart agreement is serious and 2NT a non-serious spade message.

That would make 3NT a trump cue. The bidding of three of the agreed major would show good trumps, but that would now mean two of the top three. 3NT would show poor contextual trumps, just like 2NT does in the normal scheme. That would allow the partnership to stop at 3NT with horrifying trumps on occasion. Plus, the entire trump-strength message would be equally transmitted, just later and in a slightly different manner. But, explained.

Now, in the blended scheme, one could agree that 2NT, as serious, or 2♠ if using that, initiates cuebidding. One might be "serious" if one wanted to use a cuebidding-first approach because of tactical concerns.

The "blended scheme" would involve non-2NT auctions, or 2NT bypasses if you will. In that event, the partnership could use pattern bidding. Explaining yourself as "non-serious" might also be a tactical decision based upon a desire to pattern out on that specific hand. Of course, one would need to establish methods for patterning out, and this could be natural, semi-natural, or completely artificial. For instance, perhaps 3♣ shows a 5431 hand, 3D asking. Maybe 3D shows a 5-5 hand, 3♥ asking for the shortness. Whatever.

The idea of the blended scheme is complicated, but it may be of interest and does seem to have some merits.

[fred]

In fairness to Ken I should say that, when I made my original suggestion to the player who started this thread that he bid naturally in this auction and my instincts suggested that "natural" (a more accurate word for describing my preferred style to "pattern") was vastly superior to "cuebidding":

The way Ken thinks of "cuebidding" is not exactly what I was referring to.

For almost all bridge players "cuebidding" is a much less well-defined and much less well-structured concept than the sort of "cuebidding" that Ken advocates. No doubt Ken's form of "cuebidding" is considerably more effective than the form of "cuebidding" that I would expect an "average player" to use (or be able to use) in an auction like this one.

Ken has obviously given a great deal of thought to these matters. While my instincts still suggest that it is better to play "natural" here, it would not surprise me in the least if *something* was better. If I had to guess I would say that neither "natural" nor "Rexford-style-cuebidding" is "best" - probably "best" is that 2S starts a pure relay sequence, but responder has ways to break the relay if his hand is more suitable for a cuebid, a natural bid, or some specific asking bid the answer to which would not be conveniently forthcoming from the relay. If that is true then I think that "pattern" is probably the first thing that systems designers would try to resolve when defining the relays.

I would also guess that "normal cuebidding" is (by far) the worst of all of these alternatives. That was the main point I was trying to make in my original post in this thread.

Fred Gitelman
Bridge Base Inc.
www.bridgebase.com

[y66]

Good thread. Will save this one and discuss these points with fellow intermediate pard:

* problems opener faces after 1S 2C : 2H 2S;
* possible alternative solutions (current mainstream expert practice as well as new ideas); and
* tradeoffs between alternative solutions.

Haven't seen much discussion along these lines in *any* books on 2/1 and nothing that approaches the level of thoughtfulness (and intensity) in this thread.

re: OfficeGlen's Moneyball and Bill James analogy, isn't it incredible to see what Theo Epstein and the Red Sox have done with James' ideas? And I'm a Yankees' fan!

[bhall]

officeglen, on Jan 11 2008, 10:07 PM, said:

I have the following three problems with the immediate cue approach:

1) It seems relatively poor for choice of game, which is a very underrated success factor.  It seems one is locked into the major suit game or a higher contract, and one can’t play 3NT, which can be a better place to play. 

2) It seems poor in the choice of slam strain, as it can’t clearly distinguish the degree of secondary fits.

3) It discloses a lot of information on the way to game.  In the 80s some local mad scientists were using this approach against us, and had a cuebid bonanza on the way to 4 of their major.  We were presented with a roadmap from the opening lead on.  Non-disclosure on the way to game is another underrated success factor.

However it is not like I’m thrilled with the patterning out approach.  To use analogies in this thread, for me picking between immediate cue or patterning out is like deciding whether to hit a 6 or 7 iron on the par 5 tee, or like picking the very best sword to use against the approaching tank.

[snip]

Methods that disclose too much when the partnership doesn’t have the assets for slam are flawed imo, and methods that don’t offer choice of game possibility will be long term underperformers.

Officeglen has the right of it. And mathematically speaking, there are over a thousand sequences available between 2S and 4S to distinguish 1) both hands minimum; 2) opener minimum, responder extras; 3) opener extras, responder minimum; and 4) both hands extras.

It wouldn't be rocket science to arrange for case 1 to concentrate on choice of games, while the others declare early that slam is the object and ask/show according to simple schemes. You can take Ken's work, and his recent comments, as proof that the schemes need not involve some convoluted, impossible-to-remember coding. However, I'll bet that such a scheme would take many years to penetrate the expert community,primarily because there would be a significant memory burden imposed in a situation where the gains would be rather infrequent.

Many years ago, I showed (in BW) how to play 6-way game tries over 1M-2M. Today, nobody plays them, not even I.

[dburn]

bhall, on Jan 12 2008, 02:40 PM, said:

And mathematically speaking, there are over a thousand sequences available between 2S and 4S

Over how many?

Mathematically speaking, the number of uncontested sequences starting with bid 1 and ending with bid n (bids from 1♣ to 7NT are numbered from 1 to 35) is [Fibonacci number n+1] (the Fibonacci series, for non-mathematicians such as myself, is 1,1,2,3,5,8,13,21... where each term is formed by adding together the two preceding terms).

For example, there are eight auctions starting with 1♣ and ending in 1NT ("bid 5", and the sixth Fibonacci number is 8):

1♣-1♦-1♥-1♠-1NT
1♣-1♦-1♥-1NT
1♣-1♦-1♠-1NT
1♣-1♦-1NT
1♣-1♥-1♠-1NT
1♣-1♥-1NT
1♣-1♠-1NT
1♣-1NT

The number of sequences starting with 2NT ("bid 1" when the last bid was 2♠) and ending in 4♠ is thus the same as the number of sequences starting with 1♣ and ending in 2NT ("bid 10"). The 11th Fibonacci number is 89, which is the number of sequences available between 2♠ and 4♠. I have a truly marvellous proof that 89 < 1000, which unfortunately this text box is too small to contain.

Besides, a couple of pages ago I showed all of you how you could play both cue-bidding and pattern-showing in these auctions, and you didn't take any notice. Obviously you are all ignoring my views because I am not a talented player and I don't live in Washington. It's an outrage.

[dburn]

Ignore the foregoing - it was complete rubbish. The Fibonacci numbers have only to do with how many relay sequences you can have given that you start somewhere and responder to the relays isn't allowed to bid beyond somewhere else, while the relayer always makes the cheapest available bid. The word "relay" should therefore have appeared between "uncontested" and "sequences" in the original post; that it didn't is what I believe mathematicians call a "function" of the number of glasses of wine I'd drunk before posting.

[bhall]

dburn, on Jan 13 2008, 08:10 PM, said:

bhall, on Jan 12 2008, 02:40 PM, said:

And mathematically speaking, there are over a thousand sequences available between 2S and 4S

Over how many?

Mathematically speaking, the number of uncontested sequences starting with bid 1 and ending with bid n (bids from 1♣ to 7NT are numbered from 1 to 35) is [Fibonacci number n+1] (the Fibonacci series, for non-mathematicians such as myself, is 1,1,2,3,5,8,13,21... where each term is formed by adding together the two preceding terms).

For example, there are eight auctions starting with 1♣ and ending in 1NT ("bid 5", and the sixth Fibonacci number is 8):

1♣-1♦-1♥-1♠-1NT
1♣-1♦-1♥-1NT
1♣-1♦-1♠-1NT
1♣-1♦-1NT
1♣-1♥-1♠-1NT
1♣-1♥-1NT
1♣-1♠-1NT
1♣-1NT

The number of sequences starting with 2NT ("bid 1" when the last bid was 2♠) and ending in 4♠ is thus the same as the number of sequences starting with 1♣ and ending in 2NT ("bid 10"). The 11th Fibonacci number is 89, which is the number of sequences available between 2♠ and 4♠. I have a truly marvellous proof that 89 < 1000, which unfortunately this text box is too small to contain.

Besides, a couple of pages ago I showed all of you how you could play both cue-bidding and pattern-showing in these auctions, and you didn't take any notice. Obviously you are all ignoring my views because I am not a talented player and I don't live in Washington. It's an outrage.

Ah, but it's not just the sequences that end with 4♠ we need to consider. Every bid between 2♠ and 4♠ (inclusive) may either occur, or not. Some will not include 4♠ because 3N is passed out, others because 4♠ is bypassed in slam exploration. There are, in fact, 2 to the power 10, or 1024, such sequences. Plus their continuations. Including the one in which none of the 10 bids appears, because opener jumps to 4N or higher. And the one in which all of the bids appear (whatever that may mean).

I am not suggesting that every one of the sequences can be given a unique, and useful, meaning. I did it with the 32 sequences between 2M and 3M, but filling this larger space would be a monumental effort in efficient information exchange.

[dburn]

bhall, on Jan 13 2008, 10:27 PM, said:

And the one in which all of the bids appear (whatever that may mean).

Funny you should mention that. Only today I was commentating on an international match, and East-West solemnly bid:

2♣-2♦-2♥-2♠-2NT-3♣-3♦-3♥-3♠-3NT.

Down one, with 4♠ cold. It seems that whatever bids may mean, sometimes they don't mean enough.

[paulg]

dburn, on Jan 14 2008, 04:37 AM, said:

bhall, on Jan 13 2008, 10:27 PM, said:

And the one in which all of the bids appear (whatever that may mean).

Funny you should mention that. Only today I was commentating on an international match, and East-West solemnly bid:

2♣-2♦-2♥-2♠-2NT-3♣-3♦-3♥-3♠-3NT.

Down one, with 4♠ cold. It seems that whatever bids may mean, sometimes they don't mean enough.

Kokish strikes and means that the weak hand plays the contract when they hold spades.

In the CBAI-Scotland match, neither pair was so affected but, bizarrely, had Puppet Stayman auctions where responder bid the 4-card major that he/she held thus ensuring that the weak hand, again, played the contract. At least they were in spades though after 2♣-2♥-2NT-3♣-3♦-3♠!!-4♠