Zar points, useful or waste of energy

[Chamaco]

I think calling Justin arrogant for this thread is just unfair -- it was Zar who first got personal.

Where did Zar go personal ?

When he said "if there were two of you it would be perfect?"

 

If this is the crucial point, I think it is no offense:

it is just a normal statement that good judgment is a personal thing and that it works optimally with pards with which we have a complete "bridge feeling".

 

The same works for me or you: if only we could play with a copy of ourselves, all we be well.

 

In any case, this remark was quite decent (and even if it was hironical - I am sorry to say - would not justify anyone to call ZAR a**hole).

IMO Zar was trying to communicate that "good judgment" is harder to quantify than a given metrics, therefore - I myself would argue - harder to teach to weaker players.

 

And in my opinion it's not useless to teach mechanical metrics to advancing players:

just a well as we learned Milton Works point count and learned thereafter to appreciate its limits, we did the same with LOTT, with LTC and so forth.

They are tools, and as a whole they just shed a new light on the possibility to develop a full-rounded "good-judgment", which of course is the overrall goal but not so easy to get to.

 

 

---

Incidentally, I am pretty sure that Justin's judgment works much better than Milton Work count, LTC, LOTT, ZAR,Tysen and "Chamaco" ;) points.

I do not think ZAR points are useful for players of Justin's caliber, but rather as a crutch for more mundane players.

[Echognome]

Agree with Arend above.

 

I think it's fine that Zar came up with a new evaluation method. At least it offers another choice that we may or may not choose to use.

 

I also think that Zar is trying to promote his method. It is not just on BBF that I have seen his posts. However, it is only on BBF that I have seen a serious debate on the merits of his hand evaluation. My opinion is that he has not answered questions very well and has alluded to, if not directly made, personal attacks. I do not find it surprising at all that someone would feel insulted. Perhaps it is a language issue, but I'm not going to rely on that crux.

 

I also believe that Justin just stated what a lot of people felt, but was not as polite as he could have been. Maybe that is being young and brash. I usually consider such a post and cancel it because it I do not believe it is worth the argument. But don't feel alone Justin, for I too have had a post edited here for my sarcastic tongue.

[Guest Jlall]

My original post to ZAR was inappropriate and against the rules of BBF, so I will not try to justify the actual words I said. I wrote it in anger, perhaps you will call this "young and brash" but I call it stupid. Anyways...

 

I do stand by the point I was trying to make in it. Cherdano and Echognome have reiterated it so I will not go into it again.

 

Chamaco: I agree nothing warrants being called an ***hole. As for your other opinions, you are entitled to them. I freely give my opinions and am open and honest on these forums, and sometimes not as tactful as I could be. I believe this leads to better discussions (as long as they do not get personal, as now) and more can be learned by me as well as other people involved if that is the tone of the discussion. I by no means think I know it all about bridge, and I post here more to learn than post my own thoughts. That is the great thing about message forums in general. Sometimes I will argue my point vigorously if I believe in what I'm saying. Some may percieve this as me being a know-it-all, but that is not how it is intended. I am just passionate about bridge and some things involving bridge. I am sorry you feel that I am nice to stronger players and rude and verbally violent to weaker players. I strongly disagree with this sentiment, but again you are entitled to your views.

 

Anyways, I was out of line and escalated things to a place where they didn't need to go. My apologies.

[inquiry]

Zar is trying to

 

1) Promote his method

2) Answer his critics

3) Explain why he thinks his methods are useful

 

He has done a good job on #3, I appreciate the problem he has convincing skeptics. His approach to #1 and #2 could be better, because he appears to have tried a lot to answer #2, but the ground keeps shifting under him with changing topic of attack, that his responses have beome less "responsive" and more reactionary. This actually works against his goal #1.

 

Having said that, ZAR has never claimed that experts need his methods. To the contrary, he has always stated that experts have judgement based upon tens of thousands of hands/years of experience. So the Judgement issue raised and the challenge by world champion justin to take on ZAR in a money match is a red herring, and not of value to the discussion.

 

Second, no matter what school of evalaution you belong too, you know that there are other factors that affect the true worth of a hand (values other than distribution and controls). Zar method, like any other you want to subscribe too, is just a rough evalaution. How do you evaluate a few extra nines? How much better is KJ9875 than KJ5432, etc.

 

One problem people (particulary Richard and Tysen) had with ZAR's approach was it wasn't "scientific enough" (read that as no statistics). Zar has now addressed that issue in a way, he believes, should also help determine the most appropriate method for crediting certain additional distributional factors (smaller standard deviation of number of tricks taken). It is disappointing the champions of the statistical approach have not weighed in on the approach ZAR took due to their attackes in general on his approach. Tysen has been silent although he and ZAR reached at least one similar conclusion, an eight card second fit is not worth all that much. Richard made a comment on the justin/Zar squabble. Only Mike777 commented about the statistical approach demanded by so many here and his reply must have been quite discouragning to ZAR who took this approach to answer critics here, where he said "What happens if we accept that mean-variance optimization isn't sufficient? "

 

The gaunlet has been tossed on the ground. ZAR has shown by a metric demanded by his critics that ZAR (well some flavors of ZAR) gives the "Best" approaxiamtion by his standards. There is some errors here.. the assumptions that all contracts are vul, but that has nothing to do with trick taking potential. But it is time for some of the people demaniding this type of appoach to be taken to step up and say ZAR's conclusions are right are wrong, and why. The data, the statistics, and the conclusions are there for all to see. Tysen, think TSP is better? Show us the numbers (yes it was disappoint ZAR didn't include yours, but you have the ability to run the same test using the same data).

[Chamaco]

Chamaco: I agree nothing warrants being called an ***hole. As for your other opinions, you are entitled to them. I freely give my opinions and am open and honest on these forums, and sometimes not as tactful as I could be.

Justin,

the ability to analyze one's own mistake is the mark of great bridge players and of good men too.

I think being able to say "I was wrong" (while still standing the legitimate different opinions) is a great quality, I really like this of you.

 

I make mistakes much more often than you so I have more practice in saying that !! :unsure:

 

From my part, I can only be happy to hear that some of the impressions I had were not founded :)

 

As far as being sincere and open, I agree with your position, but keep in mind that the written medium is often tricky: the same thing, said with a smile and having a drink with friends, will not sound the same in the cold internet words.

 

The latins used to say "Verba volant, scripta manent", meaning "Spoken words are volatile, but written text remains", so written words are much heavier and better handled with some caution.

 

It is wise to remember this in any field of life ;)

(I wish I would remember to apply myself this advice whenever I am challenged :D )

[Elianna]

"ONNIPOTENTI" (sorry, this is an italian word, I do not know the translation)

You may want "omnipotent", which basically means "all-knowing". :)

Actually, "omnipotent" means (literally) "all-powerful" ; "able to do anything".

 

[for the record, "omniscient" means "all-knowing"]

Yep, you're right! More evidence that English was not my first language. :)

[Al_U_Card]

The same works for me or you: if only we could play with a copy of ourselves, all we be well.

Hmmmmn.....clone myself to get the perfect partner.....perfectly awful perhaps! :) Can't imagine blaming myself for a bad result....but I would HAVE TO! :)

Far, far better to learn to walk in someone else's shoes by experiencing what you would expose them to. The golden rule and a healthy dose of "think before you speak" avoids more problems than it would cause.

[tysen2k]

Zar has now addressed that issue in a way, he believes, should also help determine the most appropriate method for crediting certain additional distributional factors (smaller standard deviation of number of tricks taken). It is disappointing the champions of the statistical approach have not weighed in on the approach ZAR took due to their attackes in general on his approach.  Tysen has been silent ...

 

I've been silent since I've been on vacation. B)

 

Zar's use of standard deviation of number of tricks is the exact method I used a year ago when this thread started.

[Zar]

>

Zar's use of standard deviation of number of tricks is the exact method I used a year ago when this thread started.

<

 

Zar’s use of standard deviation has nothing to do with your measures. For the first time STD is used for OPTIMIZATION of evaluation parameters (rather than just measuring the STD) with SPECIFIC RESULTS on Honor Points assignment, Fits Evaluation, Double-fit Evaluation, Honor Combination Evaluation etc, and then TESTED BACK in the Match, thus pushing the best performer ZP Ruffing from the 0.93 STD to 0.90, then 0.89, then all-the-way-down to 0.82!!! I hope that’s clear enough.

 

>

the 2you are by far the most arrogant: very nice and kind to the player you consider strong, and arrogant and sometimes verbally violent with posters that disagree with you without having any recognized achievements.

<

 

I have no clue how you think I determine that Jlall or Hrothgar (I don’t know neither her nor him) are weak and Xyz is strong ... I couldn’t be able to make any difference – “on Internet nobody knows that you are Dog”, remember?

 

Besides, I share Mike Rosenberg’s opinion on Bridge – “In this game nobody’s any good – winners only make fewer mistakes, that’s all”.

 

>

Of course I might be wrong, as I do not know you,...

<

 

That sounds more like it ...

 

>

Zar points calculates distributional strength using the formula 2a + b – d. This formula is based solely on hand pattern. It does not change based on the number of honors held in the hand.

<

 

Don’t see how stating the obvious contributes to progress – of course the formula for the Distributive Part does NOT change with the Honors – that’s why it is CALLED Distributive Part ...

 

>

because he makes statements, and doesn't explain the reasoning,

<

 

That’s what they call “judgement”, dear :-) Judgement with reasoning is an oxymoron :-) “It’s my judgement, damnit – why should it be based on something other than my “experience” and “gut-feeling”, and why should I explain it when I cannot even articulate it” :-)

 

>

I still thank him for coming with Zar points, they certainly helped improve my judgement

<

 

This meaning of “judgement” is new to me ... :-) These guys use “Judgement” for “shooting from the heap” :-)

 

>

Where did Zar go personal ? When he said "if there were two of you it would be perfect?" If this is the crucial point, I think it is no offense

<

 

I certainly did NOT have anyone personally in mind, that’s for sure. This was a general statement to make a point. Even now I don’t know Jlall and have never seen her playing anywhere in any shape or form, bridge or nonbridge alike – plus even if I new her, I’d never involve someone’s personal skills in a general discussion anyway.

 

>

IMO Zar was trying to communicate that "good judgment" is harder to quantify than a given metrics, therefore - I myself would argue - harder to teach to weaker players.

<

 

AND harder to COMMUNICATE, if you have missed that point.

 

>

I do not think ZAR points are useful for players of Justin's caliber...

<

 

If you mean Justin Hackett, I doubt that too. But he is just one of two twins from the top of the British League rather than being representatives for the intermediate and advanced players, right? I have stated many times (including in the books) that players like the Hacketts, Zia, Rosenberg etc. don’t need ANY method to help them out, Zar Points included.

 

I just came back from a trip to Europe and next week will finish the optimizations of the High-card points and the comparison of 4321 vs. 6421. I’ll certainly let you know when you can download the entire research.

 

Cheers:

 

ZAR

[1eyedjack]

To Zar:

 

In your last post you have I think quoted Hrothgar out of context:

Zar points calculates distributional strength using the formula 2a + b - d

This formula is based solely on hand pattern.  It does not change based on the number of honors held in the hand.

You agreed with that statement, but it was just a preamble to the next paragraph that you did NOT quote and do not appear to have addressed:

Justin is arguing that distributional strength is inversely related to honor strength.

Extremely weak hands benefit from distribution much more than strong hands.

To understand the context of the first quote, it was only mentioned in order to stress that the formula takes no account of the effect speficied by Justin in the second quote. If you are saying that Justin's effect IS accounted for elsewhere in the Zar evaluation then please say how this is done, in order to avoid further wasted time on the matter.

 

I don't know of any formulaic hand evaluation system that DOES take account of this effect, mind, but as long as we are trying to reduce evaluation to mathematics, perhaps it should not be ignored?

[hrothgar]

>Zar points calculates distributional strength using the formula

>2a + b – d. This formula is based solely on hand pattern. It does >not change based on the number of honors held in the hand.

 

Don’t see how stating the obvious contributes to progress – of course the formula for the Distributive Part does NOT change with the Honors – that’s why it is CALLED Distributive Part ...

Listen up ... The reason that I restated the "obvious" was an attempt to simplify matters for your benefit.

 

It might be helpful to review the complete conversation.

 

1. Justin made a comment about hand evaluation

2. You responded (and I quote)

 

"We are talking evaluation methods, you are saying that it’s raining outside ...

I don’t get the picture completely, may an ESL issue"

 

3. I posted a response describing my interpretation of Justin's orginal comment

4. You reply with yet another of your snide little comments, while once again failing to address the main point...

[Guest Jlall]

BTW, I am Justin...and I am a he :( Chamaco's post was aimed at me, not Zar.

[Zar]

>

Distributional strength is inversely related to honor strength. Extremely weak hands benefit from distribution much more than strong hands.

<

 

We are back in the area of stating the obvious ... Of course weak hands will benefit more from distribution, simply because the MAX honor power in short suits is VERY limited:

 

- 0 HCP in a suit which is void;

- 4 HCP in a suit which is singleton;

- 7 HCP in a suit which is doubleton;

- 9 HCP in a suit which is tripleton.

 

Thus the MAX HCP strength is the LEAST limited in the 4333 distribution: 10+9+9+9 = 37 HCP as we all very well know. As we go “wilder” in the distribution patterns, this MAX value diminishes accordingly:

 

- 4432 is limited by 10+10+9+7 = 36 HCP;

- 5332 is limited by 10+9+9+7 = 35 HCP;

- 4441 is limited by 10+10+10+4 = 34 HCP;

- etc.

- 7600 is limited by 10+10+0+0 = 20 HCP;

- etc.

- 13000 is limited by 10+0+0+0 = 10 HCP;

 

and 10 is the “minimal maximum” HCP strength possible. So the MAX-HCP-Strength varies from 10 to 37 HCP for the 39 patterns. NOTE also, that the MORE the HCP the LESS the possibilities to “cover the ground”:

 

– the 37 HCP are ONLY possible with 4333 distribution;

– the 36 HCP are ONLY possible with 4333 OR 4432 distribution;

– etc.

 

Obviously, there are HCP-max values which are covered by MORE than 1 distribution simply because there are 39 patterns and only 28 MAX values from 10 to 37, so the statement follows directly from the Dirichlet Principle.

 

Is this reflected in Zar Points? Of course! With a 37 HCP hand you can NOT get more than 8 Distribution Points, with 36 HCP you can NOT get more than 10 Distribution Points, etc. And in order to reach certain point-boundary like 26 ZP, 31 ZP, etc. with weak distribution you have to have MORE HCP to let you reach the boundary:

 

- with 4333 you need 26-8=18 CTRL+HCP to open!!! That’s about 14 HCP when discount the controls;

- with 4432 you need 26-10=16 CTRL+HCP to open. That’s about 12 HCP when discount the controls;

- etc.

 

Note that there is also the REVERSE dependency:

- with 13-0-0-0 you need 26-26 = 0 HCP + CTRL to open, but you WILL actually have “at least” 10 HCP + 3 CTRL in the hand (all honors in the 13-card suit);

- with 12-1-0-0 you need 26-25 = 1 HCP + CTRL to open, but you WILL actually have “at least” 6 HCP + 1 CTRL in the hand (only the A missing in the 13-card suit);

- etc.

 

Hope that helps – please let me know if you have something else in mind.

 

Cheers:

 

ZAR

[1eyedjack]

Zar

 

I am willing to be convinced that I am wrong. The point that you appear to be interpreting as made by others, and accepting as valid (and correct me by all means) is that as a proportion of the total playing strength of a hand, distributional factors will evidently have a greater proportionate influence on hands lacking high card values than on those having substantial high card values.

 

I do not believe that there is any dispute about that, but I think that you are mistaken in believing that this is the point of contention.

 

The problem that I have is that on hands containing strong high-card values, distributional aspects have less proportionate relevance to the total trick-taking potential.

 

To take a contrived and extreme example:

[hv=w=sakqjhdakqjcxxxxx&e=sxxxxxhakqjdcakqj]266|100|Example 1[/hv][hv=w=sakqjhdakqjcxxxxx&e=sxxxxxhakqjdcakqj]266|100|Example 1[/hv]

 

Both examples can take 13 tricks in comfort. In each case the high card values of each hand is identical. The distributional values however differ between each example. Zar evaluation would suggest that example 1 has a greater expected trick take than example 2.

 

I have been a bit lazy in contriving an example in which there are more than 13 tricks to burn, - an artificial device in order to illustrate a point. If you do not accept that this effect is of relevance at lower levels I am sure that I can produce a more realistic example.

[hrothgar]

Hope that helps – please let me know if you have something else in mind.

I think that there is still a difference of opening

 

The core of your Zar Point formula is based on a linear function:

 

Zar Points = High Card Points (HP) + Distribution Points (DP)

 

Justin is claiming that there is dependency between High Card Points and Distribution. More specifically, he states that there is an inverse relationship between distributional strength and high card points.

 

Assume for the moment that you have a hand with 5521 shape. Your Zar point method says that this hand shape is worth 2a + b - d = 14 "points". This distribution is worth 14 points regardless of whether the hand holds 10 HP or 15 HP.

 

If a hand holds 15 HP and 5521 shape, it evaluates as 29 Zar points

If a hand holds 10 HP and 5521 shape, it evaluates as 24 Zar points

 

Justin claims (but offers no proof) that the linear relationship does not hold true. I'm gonna pull some numbers out of my butt here, but hypothetically, he might say that the a 5521 shape is worth 14 points if it holds 10 HP, but only only 12 points if it holds 15 HP.

 

If a hand holds 15 HP and 5521 shape, it evaluates as 27 Justin points

If a hand holds 10 HP and 5521 shape, it evaluates as 24 Justin points

 

You seem to suggest that you can compensate for using a linear function by appropriate selection of boundary conditions.

 

>Is this reflected in Zar Points? Of course! With a 37 HCP hand you can NOT

>get more than 8 Distribution Points, with 36 HCP you can NOT get more than

>10 Distribution Points, etc. And in order to reach certain point-boundary like

>26 ZP, 31 ZP, etc. with weak distribution you have to have MORE HCP to let

>you reach the boundary:

 

I suspect that your going to run into problems with this approach:

 

Assume that you have two hands, each worth 24 Zar points.

 

One hand holds 10 HP and 14 DP...

The second holds 14 HCP and 10 DP...

 

The same boundary condition applies to both hands. If it is calibrated accurately for one, its going to be off for the other...

[Zar]

>

I am willing to be convinced that I am wrong.

<

 

You are a rare bird in this “expert” section of the BBO :-)

 

>

The problem that I have is that on hands containing strong high-card values, distributional aspects have less proportionate relevance to the total trick-taking potential.

<

 

YES.

 

Let me repeat, just in case: YES, YES, YES. :-)

 

OF COURSE it will have DIFFERENT proportional value.

 

In a hand with 10 HCP and 7600 the amount of GOREN points is 10 + 6 = 16 and the distributional part 6/16 = 38%.

 

In a hand with 10 HCP and 4441 the GOREN points are 10 + 2 = 12 and the distributional part is 2/11 = 17%.

 

In a hand with 10 HCP and 4333 the GOREN points are 10 + 0 = 10 and the distributional part is 0/10 = 0%.

 

Do you make a difference between 38, 17, and 0?

 

>

To take a contrived and extreme example:

<

 

Each hand in the “semi-balanced” hand (the second one) has 26 from HCP and CTRL plus 10 distributional points for the 4432, totaling 36 ZP each. This means the pair has 72 Zar Points, with 67 needed for 13 tricks. Thus, hand #1 has 17 tricks while hand #2 has “only” 14 tricks.

 

Are you convinced? :-)

 

>

I think that there is still a difference of opening

<

 

There is ALWAYS a difference of opening :-)

 

>

If a hand holds 15 HP and 5521 shape, it evaluates as 27 Justin points

<

 

I honestly have no clue what Justin points are (or Tysen or Hrothgar points for that matter) and I cannot give an opinion on something I have not studied in advance.

 

>

Assume that you have two hands, each worth 24 Zar points.

 

One hand holds 10 HP and 14 DP...

The second holds 14 HCP and 10 DP...

 

The same boundary condition applies to both hands. If it is calibrated accurately for one, its going to be off for the other...

<

 

“Calibrated accurately” ... Are you suggesting that:

 

Honor Points

-------------------- = CONSTANT ???

Distribution Points

 

Let me know if this is what your statement actually manifests. Can you please have a look at my example with GOREN above in this posting? I guess this example covers your concern.

 

Cheers:

 

ZAR

[waychan]

ZarPoint is a good method.

but I only use it in " is there be 5m? "

I like 4321, since I can finger opps honors, Zar done?

[hrothgar]

Assume that you have two hands, each worth 24 Zar points.

 

One hand holds 10 HP and 14 DP...

The second holds 14 HCP and 10 DP...

 

The same boundary condition applies to both hands. If it is calibrated accurately for one, its going to be off for the other...

 

 

“Calibrated accurately” ... Are you suggesting that:

 

Honor Points

-------------------- = CONSTANT ???

Distribution Points

 

Let me know if this is what your statement actually manifests. Can you please have a look at my example with GOREN above in this posting? I guess this example covers your concern.

 

Cheers:

 

ZAR

I am most certainly not saying that HP / DP is a constant...

 

I am, however, noting that you use a fixed scale to determine whether two hands produce game.

 

52 Zars for Game at level 4

57 Zars for level 5

62 Zars for level 6

 

This same scale applies regardless of the ratio of HP to DP in the two hands.

[inquiry]

Justin claims (but offers no proof) that the linear relationship does not hold true. I'm gonna pull some numbers out of my butt here, but hypothetically, he might say that the a 5521 shape is worth 14 points if it holds 10 HP, but only only 12 points if it holds 15 HP.

 

If a hand holds 15 HP and 5521 shape, it evaluates as 27 Justin points

If a hand holds 10 HP and 5521 shape, it evaluates as 29 Justin points

I think, Richard, you do Justin and his point of view a great dis-service

 

[hv=s=skqxxxhkqxxxdxcxx]133|100|5521, 10 hcp, "richard claimed justin 29 point (yes.. denfined points were to illustrate his point not real calculation) . [/hv]

 

[hv=s=skqxxxhkqxxxdxcxx]133|100|5521, 10 hcp, "richard claimed justin 29 point (yes.. denfined points were to illustrate his point not real calculation) . [/hv]

 

I would hope no one has to offer a proof that the second hand with 15 hcp is much stronger than the first with 10.

 

Justin's point was actually something quite different. It was the premise that "weak hands" gain more form shape than strong ones. For lack of a better way to say it, two hands with only 8 hcp. one with 5521 would be better than a balanced one. But two hands with 20 hcp, 5521 is still better than balanced, but the "extra" values for the distribution is not as much as the extra values in the weak hand. This is probably not stated as clearly as it could be, but that was his point I am sure, not that weak hands with distribution is worth more than strong hands with the same distribution, that is just silly.

[hrothgar]

I honestly have no clue what Justin points are (or Tysen or Hrothgar points for that matter) and I cannot give an opinion on something I have not studied in advance.

I am not asking you to give an opinion regarding whether Justin points are accurate or not.

 

I made a simple post trying to help you understand Justin's original post...

 

Perhaps I should have been more clear that the numbers that I provided were there to illustrate the point.

[Guest Jlall]

My website on Justin Points (JP) will be up soon.

[hrothgar]

Justin claims (but offers no proof) that the linear relationship does not hold true.  I'm gonna pull some numbers out of my butt here, but hypothetically, he might say that the a 5521 shape is worth 14 points if it holds 10 HP, but only only 12 points if it holds 15 HP.

 

If a hand holds 15 HP and 5521 shape, it evaluates as 27 Justin points

If a hand holds 10 HP and 5521 shape, it evaluates as 29 Justin points

I think, Richard, you do Justin and his point of view a great dis-service

Actually, I think that my example shows that I can't add worth a damn

 

Let me try again

 

Justin claims (but offers no proof) that the linear relationship does not hold true. I'm gonna pull some numbers out of my butt here, but hypothetically, he might say that the a 5521 shape is worth 14 points if it holds 10 HP, but only only 12 points if it holds 15 HP.

 

If a hand holds 15 HP and 5521 shape, it evaluates as 27 Justin points

If a hand holds 10 HP and 5521 shape, it evaluates as 24 Justin points

 

I have indicated the relevant change in red...

I also went back and corrected this in the original post

 

I guess that the point which I was making is sufficently obscure that an arithmetic error crippled your ability to understand

[inquiry]

I don't think ZAR would disagree with the assessment, as stated by justin, that distibution is more valuable to weaker hands than stronger hands. Clearly this is true (I doubt however that it is linear). I think this is one of the reasons that he finds that at very high levels the extra tricks require MORE ZAR points than at lower tricks (I am not going back to read it now, but the going from 11 to 12 tricks required more ZAR points and going from 12 to 13 more still.. the five points per level is an estimate...also, lower levels required fewer per level, duh).

 

The relationship between hcp and distribution is probably determinable by the same sort of methods tysen and ZAR use, but it is all an approximation anyway. The way ZAR describes this relations is to not be stupid when counting ZAR points, to check for aces and controls before blasting to slam. OF course if you are just competing, don't worry about such niceties... use your distirbution to your advantage.

 

As far as where you made your mistake in your math, who knew... you made up numbers and posted them with a statement that was clearly wacky... if justin preached that, his friend arron could create a webpage just about him.

[tysen2k]

Okay, I looked into this a little more. This table compares the trick taking value of average hands for each shape to yarborough hands for each shape.

 

Pattern [space]Average [space]Yarb. [space]Worth
4-3-3-3 [space] 0.00 [space] [space]0.00 [space] [space]-
4-4-3-2 [space] 0.29 [space] [space]0.32 [space] 110%
5-3-3-2 [space] 0.34 [space] [space]0.45 [space] 132%
5-4-2-2 [space] 0.61 [space] [space]0.75 [space] 123%
6-3-2-2 [space] 0.71 [space] [space]0.97 [space] 137%
4-4-4-1 [space] 0.82 [space] [space]0.82 [space] 100%
5-4-3-1 [space] 0.89 [space] [space]0.97 [space] 109%
6-3-3-1 [space] 0.98 [space] [space]1.18 [space] 120%
7-2-2-2 [space] 1.11 [space] [space]1.59 [space] 143%
6-4-2-1 [space] 1.22 [space] [space]1.45 [space] 119%
5-5-2-1 [space] 1.23 [space] [space]1.42 [space] 115%
7-3-2-1 [space] 1.34 [space] [space]1.76 [space] 131%
5-4-4-0 [space] 1.58 [space] [space]1.51 [space] [space]96%
6-4-3-0 [space] 1.71 [space] [space]1.76 [space] 103%
5-5-3-0 [space] 1.71 [space] [space]1.72 [space] 101%
6-5-1-1 [space] 1.81 [space] [space]2.09 [space] 115%
7-4-1-1 [space] 1.87 [space] [space]2.26 [space] 121%
6-5-2-0 [space] 2.08 [space] [space]2.24 [space] 108%
7-4-2-0 [space] 2.09 [space] [space]2.36 [space] 113%

 

If this table is unclear, it means that for an average hand, a 5422 shape takes 0.61 tricks more than a 4333 hand, but a 5422 yarborough takes 0.75 tricks more than a 4333 yarborough.

 

It looks like on average distribution is worth about 1.2x as much for a yarborough than for an average hand.

 

Useful? Probably minimally for initial evaluation. However it turns into a much bigger factor when we talk about adjustments for support when we have more information from the bidding. For example from my RGB article:

 

xxxxx

-

xxxxx

xxx

 

This hand is worth about 4.5 Goren points initially and worth about 10 points if partner opens 1♠, gaining 5.5 in adjustments

 

However, this hand

 

xxxxx

-

AQxxx

Axx

 

is worth 14.5 points initially but 17 after partner opens 1♠, gaining only 2.5 points for the superfit with a void.

 

Tysen

[han]

Cool Tysen. Could you post a link to the RBG thread?