Not logged inRybka Chess Community Forum
Up Topic Correspondence Chess / Correspondence Chess / Anybody wants to try to beat (almost) unattended Houdini ?
- - By punter11235 (**) [pl] Date 2011-06-30 23:20 Edited 2011-06-30 23:29
Hi,

I have no idea how strong cc players are so I would like to find out :)
The challange I offer is as follows:
-I play white
-I let Houdini run for about 15 minutes in infinite analysis mode on my weak hardware (some old 2 core)
-I hand pick one of the top moves proposed by the engine but never any move weaker than 0.15 eval than the top engine choice (I am 2300 ELO)
-As long as I am in the book I hand pick one of the book choices (the books I will be using are r4 book, Hiartz book and Zeeman 3.5final book)
-I have draw odds :) (but I will play normal game trying for advantage, not simplifying to draw)

I will post moves whenever I have time to run the engine which will be probably averaging 2 a day or something like that.
Anybody wants to try to beat me with those stipulations ? :)
Parent - - By Turbo (***) Date 2011-07-01 01:45
Punter, your (hopefully virus-, doubtfully bug-free) Chessrazor-software should indicate you that playing against an unattended engine is just bum hunting for an experienced correspondence chess player.
And no, using books, hand-selecting moves, letting 15min be 15h, asking Kasparov's advice is not what we really would call "unattended"... :-)

(btw., draw odds with white - how original, do you call a Hold'em flop with nuts only?)
Parent - - By punter11235 (**) [pl] Date 2011-07-01 09:28 Edited 2011-07-01 09:31
It's not my software anymore (for like 2 years now), please don't mention this topic anymore cause easy to hurt me here ;)

I wondered if that would be good terms cause I have no idea how difficult is to beat white with those terms. I want to play white because I am somehow convinced I would be getting owned with about 100% probability playing black.
We could change it to "completely unattended Houdini on weak hardware" if someone would like to play then :)
I have this impression when running Houdini that it's somehow showing "the truth" about the position (even when blunder checking my blitz games) I would love to be proved wrong on this one thus the challange :)
Parent - By parmetd (***) Date 2011-07-01 11:08
I guarantee Houdini is not showing "the truth" and Houdini is an engine largely hardware dependent.
Parent - - By Turbo (***) Date 2011-07-01 12:18
For a real correspondance chess master like "Banned for Life"/Alan or selected others on his forum its very easy to construct (closed) out-of-the-book positions where engines just shuffle pieces in the top 0.15 eval-range, thats no real challenge for them. Just look at a few high-quality WBCCC-games here... and all losers also highly used engine-assistance and often were sure they could draw it (=engines showing those beautiful zero-evals way to long time...) until it was too late... btw., "unattended engine"+books+handselecting+IMadvice+xy = correspondence chess :-) = no odds necessary to prove a point.
Parent - - By ppipper (*****) [es] Date 2011-07-25 17:42
+1000
Parent - By Mark Eldridge (****) [gb] Date 2011-07-25 18:12
+2000
Parent - By Morpheus (****) [pl] Date 2011-07-25 21:28
+1 hehe True! :lol:

Good point!
Parent - By parmetd (***) Date 2011-07-01 05:17
Draws odd AND white? HAHA A good 80% of corr games are draws anyways before giving color and draw odds. You said try to *beat* then declare yourself a winner if they draw?
Up Topic Correspondence Chess / Correspondence Chess / Anybody wants to try to beat (almost) unattended Houdini ?

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill