[NickRW]

Free, on Mar 14 2008, 12:43 PM, said:

Point count is just easy to explain to beginners (not like the rest of this game).  And that's the end of it.

So why make point count more difficult when you know it won't improve your bridge?

I don't agree with this sentiment. For those who have "been around", perhaps, hand valuation is largely a matter of simply looking at a hand. For a very few novices, who are generally good at card games, that may perhaps also apply. However, for the majority who are new to the game and for quite a percentage of those that aren't, improved methods of hand valuation can pay significant dividends.

I am not a bridge teacher - but I have taught my teenage children. At first I taught them 4321 methods because that is what is in all the books. However - they don't read the books - so that wasn't a very good reason - at least not in their case. One day they arrived in what might seem to be a perfectly reasonable 4 spade contract - 24 HCPs between them, 5/4 fit, ruffing possibilities in both hands. Unfortunately the missing 16 HCPs were all 4 aces and the ruffing possibilities were not voids. One down - entirely attibutable to the 4321 system not counting aces highly enough - using the 6421 system they would have correctly stopped in 3.

After I taught them 6421 and something vaguely sensible for distribution suddely they started finding slams that it took me years to learn to bid - and successful 6 level saves against slams too. When I was a teenager learning to play myself we marvelled at stories of such things in the press - my kids were doing it after only a few months of learning.

So, yes, for the experienced, new methods of hand valuation probably are just a waste of brain cells - but you guys on this forum seem to largely fall in the more experienced category - you forget what it used to be like before you learnt judgement.

Nick

[ASkolnick]

The 6421 system is a perfectly reasonable system and is included in a different method of counting called ZARS. I have gone over many hand records and it seems to usually land you in the right contract based on the counting provided you have a fit. If not, sometimes distribution plays too vital a role.

Count:
(HCP+Controls) This gives you your 6421 count.
+
(2 longest suits)
+
(Diff long-short) Means with a fit you have roughing values.

Instead of using 13 as your magic number, 26 becomes the magic number, with 52 being game and each level higher 5 points.

I am not claiming its better since all I am doing is quantifying the fact people saying "Look, I have a good hand and its 5-5", but sometimes people like seeing the value on paper.

[jtfanclub]

lowerline, on Mar 17 2008, 05:34 AM, said:

There is a difference in the value of an honour between trump and NT contracts. In trump contracts, the value of honours is more like:
A = 4.5
K = 3
Q = 1.5
J = 0.75
T = 0.25
(A = KQ = 3Q; K = 2Q; Q = 2J; J = 3T)

X4/3

A=6
K=4
Q=2
J=1
T=1/3

Not coincidental that your relative values and ZARs come out to the same, when you use a different total.

I kind of like it, because it totals to 52 points (excluding tens). You have 13 cards, and on the average, 13 points. It also means that on the average you win one trick per 4 points, and a trick has 4 cards. I think if we were starting over, this is the count I'd use.

Unfortunately Goren points are pretty much locked in for beginners, and required for alerting and annoucning points.

[NickRW]

jtfanclub, on May 8 2008, 04:12 PM, said:

Unfortunately Goren points are pretty much locked in for beginners, and required for alerting and annoucning points.

Too much lethargy amongst the bridge authorities I'm afraid. Any given national bridge authority could change it if they actually wanted to improve the game in their area. Many of them seem however to want to cater for what they think the will of the majority is, rather than actually provide leadership.

Nick

[NickRW]

jtfanclub, on May 8 2008, 04:12 PM, said:

Not coincidental that your relative values and ZARs come out to the same, when you use a different total.

I corresponded quite a bit with Zar Petkov at the time he was putting up his web site and developing it. I can confirm, both form theoretical work and practical experience at the table that his system is an improvement over Milton Work, Goren and the like.

It isn't, however, as good as you can get. 2 points for a 5 card suit is too much and is pushing the limit for even a six carder. The formula for distribution (2a+b-d) is on the money *on average*. But the value of "a" does not respond linearly. The "b-d" part is pretty good though.

And, one also has to remember that it is not as good as you can get for NT. My results suggested that 6421 was marginally better than 4321 in this regard, but not by much. 5-3.5-2-1 is more like correct. And distribution is worth only a little at NT. Maybe 4432 is quarter of a point better than 4333 and so on.

(Back to suits) Tysen, who used to post on this board came up with 1 point for length over 4 and 1/3/5 for shortages (this is in conjunction with 6421 of course). That is good, if a little prone to get you too high when you later discover a less than ideal fit. I'd recommend this method for responder, who has a little better idea about fit than opener does.

Nick

[neilkaz]

jtfanclub, on May 8 2008, 11:12 AM, said:

lowerline, on Mar 17 2008, 05:34 AM, said:

There is a difference in the value of an honour between trump and NT contracts. In trump contracts, the value of honours is more like:
A = 4.5
K = 3
Q = 1.5
J = 0.75
T = 0.25
(A = KQ = 3Q; K = 2Q; Q = 2J; J = 3T)

X4/3

A=6
K=4
Q=2
J=1
T=1/3

Not coincidental that your relative values and ZARs come out to the same, when you use a different total.

I kind of like it, because it totals to 52 points (excluding tens). You have 13 cards, and on the average, 13 points. It also means that on the average you win one trick per 4 points, and a trick has 4 cards. I think if we were starting over, this is the count I'd use.

Unfortunately Goren points are pretty much locked in for beginners, and required for alerting and annoucning points.

Off course if you like this method, just stick to the 52 pts and to be really nitpicky you can adjust for tens noting that the average hand is supposed to contain 1 ten. So if you have no tens substact 1/3 of a pt.

I do agree that there are better counts than 4321. I use 4321 and hope that I continue to improve my adjustments as my experience gains.

Zars is good, but when you don't fit a fit, it seems less good.

I expect there will be more small advances in this field to come from computers, but one thing to keep in mind is the thousands of bridge book written using 40 HCP count and that 40..ie 10 per hand is an easy number to work with mentally.

Then again, how hard would it be to use 52 pts, again noting we have 52 cards and 4 pts should equal a trick ?

[neilkaz]

I found this very interesting link

http://members.cox.n...club/chpt02.htm

[ASkolnick]

About ZARS count (I have talked to him as well),

1) yes, you can come up with things more accurate than 6-4-2-1, but you also need to include practicality as well. I am sure 5.85-3.72-1.96-.85 is more accurate, but who is going to remember to the x decimal place.

2) There are negative adjustments in misfit hands (minus 2 lengths)

3) Some of his premises about 5-3 fits are inconsistent on whether suits should be bid versus NT.

4) Short honors need to be deducted and interior spots need value. Also, interior spots are better when working with working honor and not in long suits. Kaplan created a value for this in a simulator, but once again, to what extent are you learning.

But we are just talking about a reasonable solution without using a calculator.

[DrTodd13]

IMHO, BUM RAP is a better starting point than Milton.

A=4.5, K=3, Q=1.5, J=0.75, T=0.25.

Still the same total points in each hand. You quickly learn how to add this up almost as fast as Milton. The same modifiers apply as you would with Milton like deducting for doubleton honors and adding for concentrated values and nice combinations of honors. For those relayers, it also have the nice property of decreasing the frequency with which you have to lie about point count because you don't have the required number of controls.

[NickRW]

DrTodd13, on May 12 2008, 10:06 PM, said:

For those relayers, it also have the nice property of decreasing the frequency with which you have to lie about point count because you don't have the required number of controls.

I'm not a relayer, far from it, but I can echo the general thought there. We've had quite a number of successful auctions such as 1H-5H-6H since using this 3:2:1 ratio for A/K/Q. It sounds terribly unscientific going 1H-5H-6H and not check for aces - but it seems to be much less necessary when you're properly valuing them in the first place

Nick