[blml] Decimal HCP ranges.

[Jean-Pierre Rocafort]

Nigel a écrit :
> tiny correction to summary
> [Tim West-Meads]
> The two are, IMO, different. 14+ - 17 means one opens about 50% of 14
> counts (but not about 5% of 15s) (14)- 17 means one opens around
> 15-20% of 14s (and not about 15% of 15s) "About 15-17" means maybe 5%
> of 14s and 5% of 18s.
> 
> I also use decimal variants (11.5-14, 14.7-17).  Indeed Chien Fou and
> self had an argument about whether our 3rd/4th seat SNT should be
> 14.7-17 or 14.8 to 17.  We left the CC as 14.7-17 on the grounds that
> I open 14.7 while he opens 14.8 (despite the fact that I play the
> cards at  least 0.4 better than him) due to limited space. The decimal
> notation  has the advantage of being orally reproducible but the key
> thing (IMO)  is giving walrii and novices and indication that one is
> (by their standards) deviant.
> 
> [nige1]
> If I understand Tim's *decimal* notation, I like it is better than the
> *bracket* notation; which is better than the *plus* notation; which in
> turn is better than the *true range* (unqualified); and almost
> anything is better than a partial range  :)
> 
> But unsure I've completely got to grips with Tim's decimal notation. 
> Please would Tim clarify a possible misunderstanding for me by 
> checking the following summary and examples (to *one decimal place*)

there is a trap to be avoided when comparing standard notation and 
decimal notation.
when using integer notation, boundaries of neighbouring intervals are 
distinct: 12-14 and 15-17 intervals cover all ranges from 12 to 17 HCP
when using decimal notation, the strength evaluation function is 
supposed to be continuous and adjacent intervals have common boundaries. 
the translation in decimal notation of the previous ranges would be: 
11.5-14.5 and 14.5-17.5
17.5 means that all hands considered nearer from 17 than from 18 are in 
the interval and hands nearer from 18 than 17 are outside.
for "walrus" who only refer to raw HCP and open 1NT balanced hands of 
15, 16 and 17, the accurate decimal zone is 14.5-17.5 which effectively 
corresponds to a width of 3 HCP.
15.0-17.0 would mean that, relating to factors other than HCP, you 
evaluate half of 17 hcp hands to be inside and half to be too strong.
> 
> SUMMARY: By evaluating other relevant factors according to your agreed
> methods, you can order eligible hands with a given raw HCP from worst 
> to best.  Presumably the worst hands with a given HCP are at the
> bottom of each *one point range* and the best at the top for example
> 12.0 might be QJ2 QJ2 QJ2 QJ32 whereas 12.9 might be ATxxx ATx ATx xx
> 
> EXAMPLE A. You open over 90% of otherwise eligible 17 HCP hands but
> nothing else. Surely your decimal HCP range is not *17.0-17.0* but
> *17.0-17.9* or conceivably *17.1-18.0*?
no: 16.6-17.4
> 
> EXAMPLE B. You open 30% or otherwise eligible 14 HCP hands and 90% of
> otherwise eligible 17 HCP hands. Surely, your decimal HCP range is not
> *14.7-17.0* but *14.7-17.9*?
i would say: 14.2-17.4
> 
> EXAMPLE C. Like me, you open 20% of otherwise eligible 15 HCP hands
> and 10% of otherwise eligible 18 HCP hands. IMO, your decimal HCP 
> range is *15.8-18.1*

the mean value of the strongest hands you open 1NT is broadly under 18 
according to your judgement:
15.3-17.6
...
> 
> EXAMPLE D. You open less that 5% of otherwise eligible 14 HCP hands
> and less than 5% of otherwise eligible 18 HCP hands. Surely your
> decimal HCP range is not *15.0-17.0* but *15.0-18.0*?
> 
> As the last example shows, when a percentage is under 5%, Tim's
> notation allows the occasional partial range to slip in under the back
> door.
> 
> Hitherto, I would have argued that the true range in example [D] is
> *14-18 HCP*; but since no BLMLer agrees with my opinion, I am happy to
> compromise and accept a decimal range of 15.0-18.0 HCP  :)
> 
> Tim's suggestion is non-intuitive in some ways.

agreed!

jpr

> Nevertheless, IMO, it
> deserves to be enshrined in the law-book because it would enforce a
> beautifully simple uniform basis for disclosure.
> 
> Thank you again Tim! We Walrus of the World salute you!
> 


-- 
_______________________________________________
Jean-Pierre Rocafort
METEO-FRANCE
DSI/CM
42 Avenue Gaspard Coriolis
31057 Toulouse CEDEX
Tph: 05 61 07 81 02      (33 5 61 07 81 02)
Fax: 05 61 07 81 09      (33 5 61 07 81 09)
e-mail: jean-pierre.rocafort at meteo.fr

Serveur WWW METEO-France: http://www.meteo.fr
_______________________________________________