[blml] Decimal HCP ranges.

[Jean-Pierre Rocafort]

Nigel a écrit :

> Tim's suggestion is non-intuitive in some ways.
> [John]
> agreed!
> 
> [nige2]
> Suppose we want to describe a range of 1 HCP --- say 12 HCP.
> My examples translate a 1 point range starting at the worst 17 HCP 
> hand as *12.0-13.0* decimal.
> 
> It seems that John would translate the same range as *11.5-12.5* decimal.
> 
yes

> The two declarations are similar but I started each interval on an 
> integer whereas John started roughly half an interval below.
> 
> Here I regard 12.0 as the *worst* 12 HCP hand; whereas for John 12.0 
> is a *typical* 12 HCP. Hence for John the worst 12 count is 11.5.
> 
> That difference seems trivial and cosmetic -- a matter of taste.

not exactly.
we start with a rough evaluation function: integer milton work HCP and 
nothing else. next we want to refine the function in order to take into 
account other factors than A, K, Q and J for balanced hands. the first 
move if we want to keep a practical link between both functions is that 
the corrections used for the new one are "without bias", that is to say 
that when we consider the complete set of 12 HCP balanced hands, some 
will be upgraded, some will be downgraded and the mean value of the 
hands of the set, using the refined function, will still be 12.
  thus, 12.0 is a typical 12 HCP hand. bad 12 HCP are downgraded to less 
than 12.0; if they still look more like 12 than like 11, they will be 
valued between 11.5 and 12. if they look more like 11, they will come 
between 11.0 and 11.5 or even lower (9.5 ?) if they are that awful.
  so, the most "logical" way to describe the range of "12 hcp hands" is, 
as i see it, 11.5-12.5
  the reason of this difference of understanting maybe that bridge 
players are overoptimistic: they easily upgrade and never downgrade. 
when they pretend their 1NT range to be 15-17, the truth is that they 
embrace some 13 (10%?), many 14 (25%?), all of 15 and 16, only one part 
of 17 (70%?) and no 18, so that the decimal range should be something 
like 14.1-17.2 in place of the presumed 14.5-17.5


> 
> John and Tim, however, seem to be singing from different hymn books 
> because, judging from Tim's earlier post (above) he would declare the 
> same 1 HCP range as *9.5-14* decimal.
> 
> Or have I misunderstood John and Tim again?
> 
> Anticipating quibbles...
> 
> (1) Is suppose a 1 HCP range starting at 12.0 is strictly 12.0 to 12.9 
>    repeating i.e the end is open. 12.0-13.0 is OK to 1 decimal place.

in fact, you should say real numbers instead of decimal which is a 
misguided notion. the principle is to use a continuous scale. if 13 is 
the boundary, the upper interval starts at 13.000...1 and the inferior 
interval end at 12.999...
> 
> (2) The slight advantage of starting each decimal HCP range on an 
> integer is that you avoid negative HCP.
i dont't see this as a problem. considering the set of 0 hcp hands, 
there are good ones that are worth more than 0, normal that are worth 0 
and bad ones that are worse than 0.
> 
> For example. it seems that the range for a 1D negative reply to a 
> strong club is 0.0 to 8.0 in my notation but
> -0.5 to 7.5 in John's ?
- infinite to 7.5: no need to give an inferior limit: there are no hands 
too good for a 1D reply, for what i know about strong clubs.

jpr

> Presumably it is
> -2.5 to 9.0 in Tim's ?
> 

-- 
_______________________________________________
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
_______________________________________________