ZAR points

[JoAnneM]

Has there been a discussion on ZAR points? I don't use them but they were introduced on another forum I belong to and one of our resident "experts" has immediately trashed them by saying that no expert player has ever heard of them or played them.

 

I just think stuff like this needs a fair hearing.

[hrothgar]

The following thread provides a reasonable starting point:

 

http://forums.bridgebase.com/index.php?showtopic=2542&hl=

[the hog]

I am curious JoAnne. What is the other forum?

 

Ah Ok, found it. Swan Bridge.

 

The poster who posted the op has posted on this site occasionally. He, (I won't mention names, as he uses the same nic), is very opinionated and really does not know what he is talking about. He stopped posting on this site as many people disagreed with his views. To say that no expert player has heard of them is really lol so hard I fell on my a**e.

 

I know some very fine players who use Zar points and I know some very fine players who only use their judgement in deciding what a hand is worth. They are just another tool for hand evaluation; in my humble opinion no better AND CERTAINLY NO WORSE than other methods.

[david_c]

I'm not a fan. I think Zar points give the illusion of accuracy without actually being accurate. As Tysen says in the link Richard gave, "Zar's method of distribution counting just doesn't reflect real trick-taking ability accurately."

 

It's all very well wanting to take distributional values into account. But if you're going to give a complicated formula, you'd better make sure that this complexity is actually being used to good effect. Given how complicated Zar's method is, it really ought to come fairly close to how experts actually evaluate their hands. But it doesn't.

 

For example - what do you think the difference is between a 4333 and a 4432 shape? Experts seem to rate 4333 as a minor downgrade for the purposes of opening the bidding; personally I'd put a value of about -0.3 HCP on 4333s. Zar points evaluate 4333 as being 2 ZP worse, which corresponds to about 1.5 HCP; that's the same as the value of a queen in this method. No, if you're going to make a system which takes shortage into account, it really ought to give it its proper value, not massively over-value it.

 

Similarly, I don't think it makes sense to try and express distribution very accurately, while ignoring other important factors (such as tens) altogether. Basically you have a choice: either you can use your judgement to evaluate hands, or you can delegate to some complex point-count method. It makes no sense to have a complex point-count method which is so flawed that you have to use your judgement to correct for its flaws.

 

To be fair there are many good ideas in the Zar method. 6-4-2-1 is in many ways a better high-card evaluation method than 4-3-2-1. But this is nothing new, and it seems other people have implemented things like this much better than Zar.

[Free]

There is a search function... Just type in ZAR, select the period, and you'll definitely find some posts of Inquiry because he's a great fan as far as I know.

[NickRW]

There are several discussions of Zar points in the archives - just use the search function.

 

The hcp+controls (or 6/4/2/1 for honours) is much better than Milton Work points specifically for suit contracts. While it is not rubbish for NT, 4/3/2/1 is a little closer to the truth for my money and is probably the best if you don't want to get into counting in halves, quarters and fifths.

 

The distributional element of Zar points (principally the 2a+b-d formula) is roughly right, but is not the best. In particular it postulates a linear relationship between suit lengths and trick taking ability (and Petkov goes to great lengths to prove that it is linear). It is not, however, a linear relationship. Counting shortages 1/3/5 is closer (not a linear formula). Tysen, who was a poster on some of these back threads about hand evaluation, had more to say on that.

 

An additional liability of using Zar points is that, if you pick up a hand like this:

 

♠ Axxxx

♥ -

♦ xx

♣ Kxxxxx

 

then the hand evaluation method says you should open - but what? If you systemically agree with partner to open these 1 of something, you fall foul of WBF rules (that I think all NBOs follow) about opening at the 1 level on less than 8hcp. And, in any event, regardless of the rules, for most systems in current use partner will be disappointed in your hand when he tries to make 3NT with much of his stuff in the red suits.

 

There are ways and means round the above problems - I use a watered down version of Zar points - but I do it with a) a means of opening most 2 suiters preemptively and b] a means of opening strong 2 type hands something other than 1 of something - so partner doesn't get too excited a one level opener.

 

The short way of putting the above is that if you use Zar points - or something similar - you need to think long and hard about its impact on your system - it isn't just about hand evaluation. In particular I don't think it sits well with 2/1 in its current form - SAYC can probably just about be bent to fit if you're prepared for some different 2 level openers - and Acol is more comfortable with it - though not without implications even there (this last sentence being more in the realm of opinion that the rest of what I've written!)

 

Nick

[inquiry]

There is a search function... Just type in ZAR, select the period, and you'll definitely find some posts of Inquiry because he's a great fan as far as I know.

That search will not work, for the simple reason that the software used for BBO forums does not search on 3 letter words...

 

You might try the following...

 

1) Search for "Zar points"

2) You might try a google search (which will sort of work with three letter words) and use this as a search term "ZAR site:forums.bridgebase.com"

[gwnn]

If you systemically agree with partner to open these 1 of something, you fall foul of WBF rules (that I think all NBOs follow) about opening at the 1 level on less than 8hcp.

That's false, it only says "a King below an average hand" and if you're 6520 you can probably argue that it's a non-average hand, particularly considering that you have one ace and one king.

[ONEferBRID]

I am curious JoAnne. What is the other forum?

 

Ah Ok, found it. Swan Bridge.

 

The poster who posted the op has posted on this site occasionally. He, (I won't mention names, as he uses the same nic), is very opinionated and really does not know what he is talking about. He stopped posting on this site as many people disagreed with his views. To say that no expert player has heard of them is really lol so hard I fell on my a**e.

 

I know some very fine players who use Zar points and I know some very fine players who only use their judgement in deciding what a hand is worth. They are just another tool for hand evaluation; in my humble opinion no better AND CERTAINLY NO WORSE than other methods.

In defense of PriorKnowledge (eventhough he needs no defense by me ),

he IS very opiionated because he KNOWS what he is talking about.

 

If people here disagreed with his views, I suggest they review their own "opinionated" views.

 

He didn't say no expert ever heard of ZAR points. He said :

 

"While at the Nationals, during breaks in the Spingold, I asked a half dozen top players, including Jeff Meckstroth and Mike Cappelletti, about ZAR points. None of them had heard of it. One asked me what it was, and as I was describing it, he interrupted me and likened ZAR points to the Rule-of-Twenty, by quoting another famous expert (I forget who), 'If I was teaching a parrot to play bridge, I would teach them the Rule-of-Twenty.' The implication is that bridge judgment overrules any static and complex evaluation system."

 

He went on to say:

"My rule on new bidding or evaluation systems are similar to my rule about diets to lose weight. Until an authority has used ZAR points for a while and succeeded with it, I will not bother with it. Until that time, you should treat ZAR points as an interesting academic exercise with no useful practical application."

[NickRW]

If you systemically agree with partner to open these 1 of something, you fall foul of WBF rules (that I think all NBOs follow) about opening at the 1 level on less than 8hcp.

That's false, it only says "a King below an average hand" and if you're 6520 you can probably argue that it's a non-average hand, particularly considering that you have one ace and one king.

Well, the point is actually arguable.

 

For practical purposes, I am bound by the EBU's interpretation which certainly does have an 8hcp lower limit - and I think this country is not alone in that.

 

As far as the WBF rule is concerned - it certainly does say a king below average - but it defines average as 10hcp in a balanced hand - and gives no definition of what is average in one that is unbalanced. So - yes - you could argue with a WBF TD about it if you want to - I dare say they have some appeal cases for precedents that I don't know about.

 

Nick

[NickRW]

In defense of PriorKnowledge (eventhough he needs no defense by me ),

he IS very opiionated because he KNOWS what he is talking about.

 

If people here disagreed with his views, I suggest they review their own "opinionated" views.

 

He didn't say no expert ever heard of ZAR points.   He said :

 

"While at the Nationals, during breaks in the Spingold, I asked a half dozen top players, including Jeff Meckstroth and Mike Cappelletti, about ZAR points. None of them had heard of it....

I don't dispute whether Meckstroth has or has not heard of it - but Petkov does have this in one of his documents:

 

A lot of people provided editing and theoretical feedback to the Zar Points Bidding Backbone. To name a few in no particular order: Ken Lindsay (Hawaii), John Plaut (Chile), John Gallucci (USA), Beltan Tonuk (Turkey), Vladimir Atanassov (Israel), Andrew Billson (UK), Jacob Davenport (USA), Cam Trenor (USA), Nick Warren (UK), Dave Demers (Canada), Kalle Prorok (Sweden), Piotr Radzikowski (Pola nd), Harry Freeman (UK), Rumen Mantchev (USA), Kees Brill (Denmark), Frank Luithle (Germany), Pavell Boev (USA), Raymond Reynolds (USA), Steve Marks (USA), Ben Dickens (USA), Boris Richter (Germany), Herbert Wilton (USA), Martel Claire (France), John McLeod (UK), Marco Pancotty (Italy), Larry Cohen (USA), Erik Kokish (Canada), Grant Baze (USA), etc.

 

Some of the people on that list, like me, are not known - but I don't think Cohen and Kokish, for example, are unknown or not experts.

 

As for the bit about not bothering with it until some expert has - for many people that is sound enough advice - but we'll never go forward if people don't experiment. It should be noted that many experts have written a great deal of books based around the Milton Work count - and run courses for new people based on it - I won't say that they "have a vested interest" in the status quo because that would be unfair - but there is a huge amount of inertia concerning the accepted methods.

 

FWIW, I have and do use a version of Zar's methods - they do work quite well if you think about the system implications.

 

Nick

[nigel_k]

An additional liability of using Zar points is that, if you pick up a hand like this:

 

♠ Axxxx

♥ -

♦ xx

♣ Kxxxxx

 

then the hand evaluation method says you should open - but what?

The above example illustrates a bigger problem. For distributional hands, the value depends on whether you have a fit and what trumps are. You don't know that when opening the bidding so it makes sense to initially count your distribution as less than its expected value. It's much easier to add than subtract later on.

 

The 'average' trick taking power across all hands with a given distribution is more or less meaningless when bidding an actual hand. The actual value will be more if you have a fit and less if you have a misfit. So find out which then evaluate how much extra your shape is worth.

 

Also, I don't agree with 6-4-2-1. It's true that 4-3-2-1 undervalues aces for suit contracts but 6-4-2-1 is almost equally wrong in the other direction for suit contracts and is just plain inferior for notrump contracts.

 

The most accurate for practical purposes is 11-7-4-2-1. This adds to 25 so you can easily multiply by four and divide by ten to get an equivalent number on the 4-3-2-1 scale, e.g. with KJxx QTxx KQx Jx:

 

2x7 + 2x4 + 2x2 + 1 = 27

27x4/10 = 10.8 HCP

 

I use this at the table all the time now. Obviously you still need to adjust for short honours, honour combinations etc but I don't use any rule for that.

[the hog]

In defense of PriorKnowledge (eventhough he needs no defense by me ),

he IS very opiionated because he KNOWS what he is talking about.

 

If people here disagreed with his views, I suggest they review their own "opinionated" views.

 

He didn't say no expert ever heard of ZAR points.   He said :

 

"While at the Nationals, during breaks in the Spingold, I asked a half dozen top players, including Jeff Meckstroth and Mike Cappelletti, about ZAR points. None of them had heard of it....

I don't dispute whether Meckstroth has or has not heard of it - but Petkov does have this in one of his documents:

 

A lot of people provided editing and theoretical feedback to the Zar Points Bidding Backbone. To name a few in no particular order: Ken Lindsay (Hawaii), John Plaut (Chile), John Gallucci (USA), Beltan Tonuk (Turkey), Vladimir Atanassov (Israel), Andrew Billson (UK), Jacob Davenport (USA), Cam Trenor (USA), Nick Warren (UK), Dave Demers (Canada), Kalle Prorok (Sweden), Piotr Radzikowski (Pola nd), Harry Freeman (UK), Rumen Mantchev (USA), Kees Brill (Denmark), Frank Luithle (Germany), Pavell Boev (USA), Raymond Reynolds (USA), Steve Marks (USA), Ben Dickens (USA), Boris Richter (Germany), Herbert Wilton (USA), Martel Claire (France), John McLeod (UK), Marco Pancotty (Italy), Larry Cohen (USA), Erik Kokish (Canada), Grant Baze (USA), etc.

 

Some of the people on that list, like me, are not known - but I don't think Cohen and Kokish, for example, are unknown or not experts.

 

As for the bit about not bothering with it until some expert has - for many people that is sound enough advice - but we'll never go forward if people don't experiment. It should be noted that many experts have written a great deal of books based around the Milton Work count - and run courses for new people based on it - I won't say that they "have a vested interest" in the status quo because that would be unfair - but there is a huge amount of inertia concerning the accepted methods.

 

FWIW, I have and do use a version of Zar's methods - they do work quite well if you think about the system implications.

 

Nick

So is Rumen. He is a top Bugarian player.

By the way, I did not mention the poster by name; you did him tht disservice. All I can say, onefer, is that his nic is apt in his case.