[Siegmund]

Bear in mind he published in 1987, and analyzing even moderate numbers of hands in automated fashion wasn't really a viable approach until the mid-90s. He would have HAD to either address a simplified tractable form of the problem (suit combinations), cherrypicked hands from many years of published tournaments with table results, or made some other kind of assumption we'd find unacceptable now.

[stegenborg]

Hi

Thomas Andrews have a lot of material about various point counts here:bridge.thomasoandrews.com/valuations/

I think he finds this to be quite good:
A: 4.2
K: 2.8
Q: 1.8
J: 1.0
T: 0.4

Another good alternative is probably this:
A: 4.4
K: 2.8
Q: 1.6
J: 0.8
T: 0.4

I have actually seen the last one on the convention card of a bunch of players in Oslo, but as 11-7-4-2-1.

Most of us are probably best of using 4321 plus a little judgement, including appreciating aces and tens.

Regards
Stegenborg

[Echognome]

gwnn, on May 30 2010, 04:41 AM, said:

1.03
0.91
1.13
0.98
0.95

For the additional utility (I know there's a neat word for this in economics...) of each honour. I feel pretty surely that this should be a monotonous series, not something that has been arbitrarily perturbed from a bunch of ones.

Neat Economics Word = "marginal"
Neat Mathematics Word = "monotonic"

Main Entry: marginal utility
Function: noun
Date: 1890
: the amount of additional utility provided by an additional unit of an economic good or service

Main Entry: mo·not·o·nous
Pronunciation: \mə-ˈnä-tə-nəs, -ˈnät-nəs\
Function: adjective
Etymology: Greek monotonos, from mon- + tonos tone
Date: 1776
1 : uttered or sounded in one unvarying tone : marked by a sameness of pitch and intensity
2 : tediously uniform or unvarying

Main Entry: mono·ton·ic
Pronunciation: \ˌmä-nə-ˈtä-nik\
Function: adjective
Date: 1797
1 : characterized by the use of or uttered in a monotone
2 : having the property either of never increasing or of never decreasing as the values of the independent variable or the subscripts of the terms increase

[jacksond]

The hand QJ10, QJ10, QJ10, QJxx
much maligned in another thread as hardly worth an opening bid never mind a strong NT in Banzai. It is 23 points in Banzai

Well suppose you find yourself playying in 3NT with

Axx, Kxx, Kxxx, xxx opposite the above hand.

Obviously not a contract of great beauty but it has some genuine chances and certainly if the defense loses a tempo on opening lead there are plenty of potential tricks available.

By contrast if partner has opened A strong NT on
Kxx, Axx, Axx, Axxx. A real strong NT pre Banzai then I think you have nearly zero chance of making 3NT with the A K K hand

certainly i prefer the hand with the queens, Jacks and Tens if I get to 3nt with Axx, Kxx, Kxxx, xxx

incidently i thinlk we all know 4333 hands are a negative and holdings such as KQ bare or QJ10 are small negatives so in any point count method one is entitled to upgrade or downgrade a little from the basic point count.

Lets see some example hands please rather than opinion where you think the 54321 method doesnt work and we can give our views

[Vampyr]

Poor old Milton Work.

[wickedbid1]

 jacksond, on 2011-June-12, 06:05, said:

The hand QJ10, QJ10, QJ10, QJxx
much maligned in another thread as hardly worth an opening bid never mind a strong NT in Banzai. It is 23 points in Banzai

Well suppose you find yourself playying in 3NT with

Axx, Kxx, Kxxx, xxx opposite the above hand.

Obviously not a contract of great beauty but it has some genuine chances and certainly if the defense loses a tempo on opening lead there are plenty of potential tricks available.

By contrast if partner has opened A strong NT on
Kxx, Axx, Axx, Axxx. A real strong NT pre Banzai then I think you have nearly zero chance of making 3NT with the A K K hand

certainly i prefer the hand with the queens, Jacks and Tens if I get to 3nt with Axx, Kxx, Kxxx, xxx

incidently i thinlk we all know 4333 hands are a negative and holdings such as KQ bare or QJ10 are small negatives so in any point count method one is entitled to upgrade or downgrade a little from the basic point count.

Lets see some example hands please rather than opinion where you think the 54321 method doesnt work and we can give our views

This is the sort of analysis they do in the Klinger book, and i think it is very fair.

The basic strong argument of their case is that trad count overvalues a & k & undervalues 10 for hand combos destined to reach nt contracts.

Thin nt games need texture, not Empty A & K combos, essentially.

On the other hand, in suit contracts js & tens in yr hand outside the trump suit (or a secondary fit, sometimes) r usually superfluous, because once the suit is played a couple of times, trumps get in ther from one side or the other.

It might be fair to say that for bidding purposes Banzai count assumes nt contracts & loser trick count assumes suit contracts, and Milton-Work count compromises between the two. It is interesting, tho, isn't it, that these new count methods demonstrate a much larger gap between evaluating nt situations and suit situations than many of us have realized until now?

One problem I hav noticed with Banzai count in practice is it is very difficult to "backpedal" partner out of optimistic suit contracts... based on assuming u hav fewer "texture" points & more quick tricks than u actually do at times. It may be that banzai evaluation should be left to 2nd or third stages of the auction... or else, I think it might be good to start with a SUBTRACTION for the flat empty suit holdings (cuz reevaluating UP later is usually a lot easier than down -- at least with my pards )

[helene_t]

I did a logistic regression analysis (event: 3NT makes. If it makes from one side and not from the other I included both results weighted by 1/2) on the GIB DD database, excluding hands where an 8+ major suit fit was present. The coefficients were almost identical to MW (Goren?) points, except that a queen got some 1.8 points instead of 2. I speculate that the queen might get something closer to 2 points single dummy, as it is more serious to be missing the queen when you won't always know which way to finesse it.

Replacing MW by the new method would only earn you 2 imps per 1000 boards or something like that.

It would be interesting to look at some potentially relevant subset such as for example hands where 1NT-3NT is a plausible auction. Also, people devalue queens and jacks that are in short suits so relative to the regression analysis (which doesn't take the length of the suit into account) they should adjust queens and jacks upwards when they are not in short suits.

So it is possible that something giving more credit to lower honours would be better for 1NT-3NT auctions. However, it is ridiculous to write a book about it without doing some serious statistical research. It is not rocket science. Apart from the subset selection, the modelling is straight forward, and there are plenty of data available. Buying a copy of BridgeBrowser would constitute a small fraction of the costs of publishing a book. Arguing such a case using made-up examples and case stories instead of data analysis is sooooo 18th century, if not even more outdated.

[Free]

I believe this isn't the first time banzai points are mentioned on the forums. Definitely a few years ago, before downvotes were invented...