Frequently Asked Questions about Zar Points

 This page provides the answers to some questions I've been asked over and over again - so I just decided to put them in FAQ-page for people to read whenever they decide to ask me a question :-)

 Are Zar Points too complicated to calculate?

 This statement is kind-of a “common” concern, and honestly my “unbiased” opinion is that there isn’t such a “giant leap of a difference” in complexity between:

 - calculating 3*v + 2*s + d (where v is the number of voids, s is the number of singletons, and d is the number of doubletons in the hand);

 - calculating (a+b) + (a-d) where a, b, c, and d are the suit-lengths in descending order;

Habit, familiarity and convenience are probably some of the factors here, but certainly the “perception” of complexity that you point out is there to some extend.

Can you use Zar Points "secretly" while your partner continues to use Goren/Bergen?

The answer is “yes, you can”. In fact Zar Points have been used in the Bermuda Bowl (by a very few teams of course) and one of the players was using them exactly that way – just to “back” his decision-making process in though situations.

The key points where you can use Zar Points that way are at the opening-bid decision and at the point of "invitational" sequences, be it for Game or Slam. 

Do Zar Points comply with the WBF standards for opening hand?

The WBF defines an opening hand as "a hand which is better than the average hand by a Queen-worth". Thus, a hand with 12 HCP is the minimal opening hand if you happen to do hand evaluation via HCP count - 10 HCP for the average, plus 2 for the Q.

As you can read in the "Never Miss a Game Again" article, the average Zar Points hands has 13 Zar Points from HCP and Controls, plus 11 Zar Points from distribution for a total of 24 Zar Points.

So, a hand of 24 Zar Points plus a Q-worth (2 points) = 26 Zar Points is an opening hand, so Zar Points comply with WBF.

It is interesting to note that all the 3 methods - Goren, Bergen, and Zar - fit EXACTLY the minimal conditions imposed by the World Bridge federation. Since the "average" hand has 5-3-3-2 distribution with 10 HCP, here are the three calculations:

1) Goren gets 10 for the HCP plus 1 for the doubleton = 11 Goren Points for the average hand - a Queen-worth from the opening bar of 13 Goren Points.

2) Bergen gets 10 for the HCP plus 8 for the 2 longest suits = 18 - a Queen-worth from the opening bar of 20 Bergen Points.

3) Zar gets 10 + 3 = 13 for the HCP + Controls, plus (5+3) + (5-2) = 11 for the distribution for a total of 24 Zar Points - a Queen-worth from the opening bar of 26 Zar Points.

Do Zar Points comply with the EBL and ACBL standards for opening hand?

The European and American Bridge Leagues, contrary to the World Bridge Federation, relate the opening bids to HCP and state that an opening hand has to have 8 HCP + "compensating distribution". To quote the EBL, for a hand with 9 HCP you have to have either a "a 6 card, or two 5 card suits". I am not going to argue the equivalence of 1 6-card suit and 2 5-card suits, but the point is that the rule allows for some kind of distribution compensation.

Further, the EBL states that for the opening bids "at the one-level the "Rule of 18" shall serve as the guideline for judging whether it meets normal opening bid expectations". That is, the sum of your HCP and the lengths of your longest 2 suits (a+b) should be equal to at least 18. Zar Points comply with the Rule of 18 also (see the article for proof).

Isn't the value of HCP diminished in Zar Points count?

It is hard to say "diminished" since it is calculated through tons of equations - the accurate way to say it is that the values of HCP is "accurate" (as it is the case with all the other components of Zar Points, for that matter).

Compared to the "normal" weight of HCP, their value is "diminished" in two different ways:

1) The amount of points you get from distribution is BIG compared to the amount of points you "collect" from the HCP (say, a hand with 11 HCP and 5431 distribution will get 13 points for the distribution - more than the HCP itself);

2) You add the Controls, which pushes the value of the HCP even further (say, in the above-mentioned 11-HCP hand you'd get additional 5 points if the 11 HCP are A-A-K).

How does the Zar Points method compare with Losing Tricks Count (LTC)?

LTC doesn't directly address the distribution (or shape) per se. What I have in mind is the following (examples directly taken from the LTC page of bridge-forum.com):

How does the Zar Points method compare with the Charles Goren method?

The article proves in 3 different ways that Zar Points are 3 times better than Goren and Bergen as hand evaluation methods.

As a general bidding performance, Zar Points are proven to deliver 2 times better bidding performance via a "head-to-head" comparison through over 300,000 boards in the zones of part-score, game in Major, game on Minor, slam, and Grand slam. All the records are available from the "Support" menu for download. 

How does the Zar Points method compare with the Marty Bergen Rule of 20?

The article proves in 3 different ways that Zar Points are 3 times better than Goren and Bergen as hand evaluation methods.

As a general bidding performance, Zar Points cannot be compared to Bergen since Bergen does not address the bidding per se, but rather only serves as an indicator for opening the bidding. 

How can you scale the Zar Points to the Goren or Bergen methods?

There is a separate section called "The Conversion" which addresses this point. It shows how to "convert" or "scale" Zar Points (the distributive part of it, of course) to Goren points and then to Bergen points - this was an idea of Jeff Rubens, the Editor of "The Bridge World", which I liked a lot.

It is also shown there, that Zar Points are better than Goren in Goren terms and better than Bergen in Bergen terms - a very powerful tool for comparison indeed.

How does the Zar Points criteria for opening compare with the experts' judgment?

There are a lot of examples of "light" openings by real experts in real tournaments throughout the article.

As I have put it in a letter to one of the experts, "If ANY expert EVER opens a hand that does NOT have 26 Zar Points, this would be an indication that the expert is drunk :-) The other case around - failure to open a hand that does have 26 Zar Points would be an indication that the expert is asleep :-)

Which distribution count is better - short suits (Goren) or long suits (Bergen)?

I don't know which one is better, but I hope you already know which one is the best :-)

Goren does not reflect suit lengths DIRECTLY, but he DOES do so indirectly, whether he intended that or not. I believe I gave a good example with the 5-5 two-suiter. In Goren you get 3 points for that 5-5 lengths (either 2+1 for a singleton and doubleton or direct 3 points for a void).

The Bergen method only emphasizes adding your two longest suits, but again - one can argue that Bergen gives 10 points for a void or a combination of a singleton and doubleton in the above-mentioned hand :-) All these correlations stem from the fact that all the four lengths are "squeezed" in a sum of 13.

How do you value the suits you hold (say clubs vs. spades)?

Zar Points assign one point for having the spades suit, but you can only use that point in borderline decisions - say, you have 25 Zar

Points and hesitate to open ... Now, if you have the spades suit (4 cards at least) you can say that you have 26 Zar Points actually and to open the bidding.

What's the rationale? Not only the convenience to outbid the opponents at the same level, but the convenient re-bid after your partner bids something on level one. As a general rule, you should know in advance what your eventual re-bid would be prior to opening the bidding, That's where holding the spades suit counts.

Does Zar Points count work for NT contracts?

The Zar Points were originally calculated via tons of boards in the suit Game and Slam areas. However, there were many regression tests performed on NT contracts (all DB-records are available for download here under the "Support" menu)- and the results as you may see there were consistent with the results for suit contracts.

You realize, of course, that the reason for that is simple - it's in the ratio between the HP and DP portions of the Zar Points. With NT-oriented hands, the portion of the HP (HCP + Controls) will be much more heavier, compared to the corresponding picture in the suit-oriented hands and contracts.

Has anyone before tried an approach similar to Zar Points?

The best way to answer that is via the discussion with Roger Eymard from France:

>

It looks as a great effort to improve and complete the initial work by Jean-René Vernes, published in 1966 (Emile-Paul editor, Paris), and in which has been firstly demonstrated the Law of Total Tricks, by an approach similar to yours (solving a number of equations to determine the values of a lot of parameters).
<

I have never heard of the gentleman ... I DO remember that we were using The Law back in the 70-es in Bulgaria in the university bridge club, but even back then I didn't know who The Law of Total Tricks was "invented" by ...

Now that you told me that, I made a search on the Web and I was surprised to read an interview, in which he talks about that and the fact that he had discovered it back in 1955 and started discussing it in 1958 ... interesting ... Here is a "cut" of it:

"J'ai découvert la loi des levées totales vers 1955. J'ai commencé à en parler, à partir de 1958, dans une série d'articles, et je l'ai publiée sous sa forme actuelle en 1966 dans " Bridge moderne de la défense " . 

How did Zar Points evolve?

The Zar Points themselves (the DISTRIBUTIVE part of it) were an independent "discovery" which initially had the form of (a + b) + (c -d) and was targeting the difference in playing power with 5-4 two-suiter between

- 5-4-2-2

- 5-4-3-1

- 5-4-4-0

The comments of Eric Kokish for example are regarding this first light version where the current Zar Points were called "Recursive Zar Points" and were presented as an improvement over my initial count of (a + b) + (c - d). After I ran a number of boards (equations) through them, the first version was dropped since it was of the SAME complexity as the first one, and (more importantly) it fit the 6-4-2-1 HCP count (Milton + Controls) closer. The equations themselves DID already include the distribution part of Zar Points in the form (a + b) + (a - d) and involved the coefficients for Honors, Discount honor combinations, Upgrades for 2-suit-concentration of Honors, and the adjustments for additional length (over 8) and honors in the partner suits. The Controls were "separated" later from the 6-4-2-1 count for easier-to-remember purposes (i.e. to preserve the Milton HCP 4-3-2-1 count which people are accustomed to). The inclusion of the Distributive points solved a lot of problems with length calculations and adjustments and the way you combine these lengths and differences in lengths. There were other experiments which "didn't make the grade" - plus the coefficients were rounded and the equations re-run for regression testing, of course. The final version is the one you know. Note also, that virtually all the boards were either game in major or small slam with virtually no duplication, since this was the initial target I had in mind. The yielded values for the levels were about 52 and 62 respectively for game in major and slam, again - in the "no fat" or "no duplication" cases (the corresponding values for the (c-d) version were 48 and 58 and the opening level was 24). That's the story.