Not logged inRybka Chess Community Forum
- - By somepeter [sk] Date 2010-07-27 17:36
Does anybody know how works analyze Monte Carlo? When I press F1, Help, there is not any word about it.
Parent - - By NATIONAL12 (Gold) [gb] Date 2010-07-27 23:32
Monte Carlo is a joke.trust me.i wasted over 100 hours on this on a Skully before i realised it.i just let it run before i realised it was usless.i suppose i had one to many that night.
Parent - - By nebulus (****) [no] Date 2010-07-27 23:45
I'd say it depends on how do you use it and what for.
Parent - - By Felix Kling (Gold) [de] Date 2010-07-28 00:08
what do you say it would be useful for? :)

I'm convinced IDeA or infinite analysis is always superior.
Parent - By nebulus (****) [no] Date 2010-07-28 00:28
Comparing it to IDeA or infinite analysis is like comparing apples and oranges. :smile:
It's just statistics. Nothing more, nothing less. You can compare it to success percentages in the database.
Parent - By Geomusic (*****) Date 2010-07-28 07:51
is there a position that lead to forced mate using MC other ave's couldnt find?
Parent - By Moz (****) Date 2010-07-28 10:10

> what do you say it would be useful for? :)


It's useful as a tie-breaker and a source of inspiration in endgame (or near endgame) positions when the branching factor goes supernova.  By itself, it's practically useless but when I've analyzed a position very deeply with other methods like IA or Idea I have found that Monte Carlo is good at finding outliers.  It's the disparity between MC and the other methods that gives the MC results meaning.

To put it another way: Monte Carlo Analysis is terrible at providing answers, but it's great at dropping hints and offering clues that sometimes lead to new insights and interesting discoveries.
- By somepeter [sk] Date 2010-08-01 17:31
Thank you for helping me.

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill