Re: Does it cheat?

Frankly, I don't understand while lot of players focus their attention if
dice are cheated instead if they play well or if a software is correct.
I mean, who can assure or prove that GNU or Snowie, etc. are GODS who never
made a mistake? Why at 2ply evaluation I have a response, at 3ply another
one, and in a rollout another one again? Which is the correct response? And
especially...WHY? Why if you are a "so good" player you can't explain me at
least blunders?
Why with a match equity table i can be in a double/pass and with another one
a double/take? Which is the right MET to use?
Who programmed softwares and on which theoric basis? neural net? ok..and if
neural net is composed by millions of matches played not so good or with
some mistakes from both sides what happens? It should happen that a winning
wrong move/cube decision should be converted in a right one...nice!!
Are we sure about their judgment in equity terms of lucky/unlucky rolls?
And their usual loose style of play does it really pay? Strange, because if
a bad moves can pay between 0.080 to 0.160 but without leaving to opp to
make a 0.700 or more joker, it doesnt seems so bad playing it; maybe because
you can save to swap between a main to a consolation paying a little bit in
advance..in fact, how many international tournaments have softwares won?
In my opinion most of analyzed top players' errors in live matches are due
to an excessive (and probably justified) search of solidness. So I think
they prefer to avoid to leave possible jokers to opps more than play a
"brilliant" move; ...brilliant just if opps doesnt roll the joker, otherwise
immediately hear people whisper: "he could avoid to leave the only winning
shot"
Further, I've seen lot of "hard"situations in which it's difficult to choose
the right move, and also the software doesnt'know what to choose, or choose
it random, estimating EQUAL variation of equity on a list of 8 or more
moves, and this happens not when it has win or lost at 100% or is going to a
sure cube action. So it seems that sometimes also the softwares doesnt know
what to do.

And which is the reason of a software should cheat? It can't win no money no
glory, etc. Just for showing to be 0,02% more good than other softwares?
And worst, why FIBS should cheat? to allow some player to win ZERO money and
other players lose ZERO money?

This kind of question seem to me to be true problems to think about, and not
if a software or a site gives for humans or for itself cheated dice.

Greetings to everybody,
GMK


"Raccoon" <racgammon@xxxxxxxxx> ha scritto nel messaggio
news:1184075890.963110.247140@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
If you believed _some_ people, you might think they _all_ cheat. Some
blasts from the past:

=========================================================================
Does Gnubg cheat?
Date: 18 January 2004

"I prefer to get the dice from random.org because if I let gnubg roll
it's own it somehow miraculously gets exactly what it needs every
time. Never fails. I don't think gnubg ever has bad luck when it
rolls it's own dice."

Date: 31 January 2004

"gnubg is not a bad bot even though it's cube handling is weak. Only
one thing you have to remember, use manual dice and stay well away
from the RNG"

Date: 24 March 2006

"GNU is soooooooo predictable. Playing for 8 points in a 7 point match
with the score at GNU 3 User 2, take a look at user roll
26................the odds of that are? And the odds of the subsequent
rolls are?"

=========================================================================
Does TD-Gammon Cheat?
Date: 29 Jan 1997

"Although I have only played one game (in Grandmaster mode) with the
new program, I would like to relate the following incident to you,
(...) I rolled 6,2 and the program followed with 2,2. I think the odds
of that happening are almost 60-1, certainly not in the realm of the
impossible, but even after only one game, it is a little suspicious."

=========================================================================
Does David's Backgammon Cheat?
Date: 17 June 1998

"I would appreciate anyones thoughts on Davids Backgammon from
Wingammon. I have had several discussions with the Author who states
that the dice are random, but I find that hard to believe. It will
keep me on the back for 4 to 5 throws with 4 points open. I will have
5 points blocked and he will enter in 1 maybe 2 throws. Consistantly
every blot of mine is taken while his blots are rarely taken. (...)"
=========================================================================
Does Gammontool cheat?
Date: 24 September 1991

"> One of my friends claims that [gammontool] cheats ..."
=========================================================================
Does Hyper-Gammon cheat?
Date: 29 June 1996

"(...) I have this game called Hyper-Gammon.... Its a CD game, and the
author purports to have hooked up his PC for two years to solve every
single possible position involved with three backgammon pieces. 32
million positions he claims his computer solved, and there is this
massive 172 megabyte data file on the CD, yet the damn thing still
cheats!! For some odd reason, the only doubles it has ever rolled were
double-6's."
=========================================================================
Does Jellyfish cheat?
Date: 6 June 1997

"While I do not consider myself an AUTHORITY on backgammon, I do know
enough to finally have had enough with this program. I put up with
some really strange and off the wall wins. Sometimes only occuring
after hitting some really very hard combinations. Now, I now that it
does happen in real life, but on a consistent basis!? I got really
curious after somebody had mentioned on this newsgroup on how
jellyfish behaves after doubling or being doubled, because I noticed
the same thing. Over a three day period I played probably over a
hundred games on the hardest level and doubling every single time.
JF3 won every single one of them. Enough said."

Does Jellyfish cheat?
Date: 12 February 2003

"Am I the person who knows that Jellyfish's dice rolls favor the game?
I love the way Jellyfish looks and plays, but I get so frustrated at
the incredable rolls it gives itself. I'm a very good backgammon
player and would appreciate knowing if anyone else feels the same way
I do."

Does Jellyfish cheat?
Date: 10 March 2003

"I'm not a good player. Pretty much a newbie. But if I play Jellyfish
Lite on level 5 and let it roll the dice for itself, I can win maybe 1
in 20 or 30 times. If I tell it to use a file and use a roll generator
I can win maybe 1 out of 4 or better. The difference is dramatic.
Jellyfish will give itself whatever it needs if allowed to roll the
dice itself.

=========================================================================
Does MVP Backgammon cheat?
Date: 28 October 1996

"I bought a copy of a shareware game called MVP Backgammon and I'm
wondering if anyone has had any experience with it. It seems to me
that the rolls are not very random."

"I bought the same program, and I certainly agree. I know the
tendency is to subjectively blame the dice, especially with computer
programs, but I'm convinced that MVP is definitely biased in its
rolls. (...) I'm ready to give up on this program."
=========================================================================
Does Motif cheat?
Date: 12 October 1996
Subject: Re: motif

"I once read this newsgroup in search of an on-line backgammon game to
play and read someone's posting about a rigged game that always seemed
to roll the right roll, defying all odds, the game being the "motif"
listing. And of course assuming that the posting person was not as
good as I, I accepted the challenge and went there to play. He was
right. It is rigged BIG TIME. Am through with it."



.