[Bug-gnubg] Importance of pipcount

[kvandoel]

On Sun, 23 May 2004, Albert Silver wrote:

> > The small  effect of cube errors  on overall performance  allso gives
> an
> > interesting  perspective on  the so-called  "pip-counting"  methods.
> It
> > follows that for almost everyone,  learning to accurately pip-count is
> a
> > complete waste of time.

> That's a very mistaken assumption IMO. You're assuming that the
> pip-count is only of use, or mostly of use, for cube decisions, and
> while it is certainly an unquestionable factor, it also *strongly*
> affects many checker play decisions.

I had thought of this and my handwaving argument against it is:

a) For cube decisions one OCCASIONALLY benefits from knowing the precise
pipcount.
b) For chequer  play decisions one OCCASIONALLY benefits from
knowing the precise pipcount.

>From a) only pipcount would be even less imporant than cube decisions as
it  is a  subset  of  those. b)  Compensates  this a  bit  in the  other
directions  making  it  overall  still  less  important  than  the  cube
decisions.

It's easy to  come up with positions where  it's important; the question
is how often  do they arise in practice? My  personal experience is that
it is  so infrequent that  the effect on  playing strength will  be very
small. But this is just based on my own experience.

If you have  a different experience I will have to  change my opinion on
this.

I actually toyed with the idea of testing it out by hacking GNUBG to get
a wrong pipcount  input to its neural net and letting  it play against a
non-disabled version  of itself and  measure the effect. The  thought of
subjecting  myself to  another smear  campaign by  Douglas Zare  made me
change my mind if I remember correctly.

Kees