Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
AI Factory Backgammon founders slams XG!
285 views
Skip to first unread message
Nasti Chestikov
Dec 16, 2021, 11:00:07 AM12/16/21
to
In the comments.
https://backgammonapps.blogspot.com/2021/11/my-first-android-review-backgammon-by.html
Calls its analysis "defective"!
https://backgammonapps.blogspot.com/2021/11/my-first-android-review-backgammon-by.html
Calls its analysis "defective"!
Frank Berger
Dec 16, 2021, 4:12:36 PM12/16/21
to
I think on redit it was also mentioned. I assumed the developer has not the slightest idea about the game and his comment proves that...
Timothy Chow
Dec 16, 2021, 10:37:41 PM12/16/21
to
what it looks like. Rollason's comments make no sense. If I
give him the benefit of the doubt, then maybe he checked what
move AI Factory Backgammon makes in that position and found that
it doesn't play 6/2(2) after all, so he thinks that the reviewer
must have made a mistake entering the moves. But even under that
very charitable reading, the rest of what he says makes no sense.
If the reviewer really did incorrectly enter 6/2(2), then XG's
evaluation of that move as terrible is perfectly reasonable, as
even Jeff Rollason agrees. What did Rollason expect, that XG
would magically know that AI Factory Backgammon *would* play
23/21 even when you ask XG to evaluate some other play? Maybe
XG has an audio feature where it would speak to the user, "No,
I see you're trying to study AI Factory Backgammon, and the move
you're asking me to evaluate isn't what AI Factory Backgammon
actually plays here!"
---
Tim Chow
peps...@gmail.com
Dec 17, 2021, 7:34:23 AM12/17/21
to
On Friday, December 17, 2021 at 3:37:41 AM UTC, Tim Chow wrote:
> On 12/16/2021 11:00 AM, Nasti Chestikov wrote:
> > In the comments.
> >
> > https://backgammonapps.blogspot.com/2021/11/my-first-android-review-backgammon-by.html
> >
> > Calls its analysis "defective"!
> Is your point here to criticize Jeff Rollason? Because that's
> what it looks like.
I don't agree that it "looks like" that at all.
> On 12/16/2021 11:00 AM, Nasti Chestikov wrote:
> > In the comments.
> >
> > https://backgammonapps.blogspot.com/2021/11/my-first-android-review-backgammon-by.html
> >
> > Calls its analysis "defective"!
> Is your point here to criticize Jeff Rollason? Because that's
> what it looks like.
I think that the default assumption is that the poster agrees with the "slam".
Paul
Nasti Chestikov
Dec 17, 2021, 12:04:40 PM12/17/21
to
I suspect he isn't the author of the program and he's been assured by whatever developer he has employed that the dice are honest (they aren't, trust me) and that it's playing to the best of it's ability at level 7 (count them, seven levels of difficulty).
See
https://www.youtube.com/watch?v=wNlnoNchYRI
for poor play
or
https://www.youtube.com/watch?v=EXKDfqJ8Cag
for cheating.
MK
Dec 19, 2021, 4:40:27 AM12/19/21
to
On December 17, 2021 at 5:34:23 AM UTC-7, peps...@gmail.com wrote:
> On December 17, 2021 at 3:37:41 AM UTC, Tim Chow wrote:
>> Is your point here to criticize Jeff Rollason? Because that's
>> what it looks like.
> I don't agree that it "looks like" that at all. I think that the
> default assumption is that the poster agrees with the "slam".
I had the same first impression as you also, but it looks like
he has been posting about the AI Factory bot being too weak
and cheating. I don't remember him saying XG-dung yet. So,
maybe you should try to give people the benefit of the doubt.
MK
> On December 17, 2021 at 3:37:41 AM UTC, Tim Chow wrote:
>> Is your point here to criticize Jeff Rollason? Because that's
>> what it looks like.
> I don't agree that it "looks like" that at all. I think that the
> default assumption is that the poster agrees with the "slam".
he has been posting about the AI Factory bot being too weak
and cheating. I don't remember him saying XG-dung yet. So,
maybe you should try to give people the benefit of the doubt.
MK
MK
Dec 19, 2021, 4:55:21 AM12/19/21
to
On December 17, 2021 at 10:04:40 AM UTC-7, Nasti Chestikov wrote:
> ..... absurd comments from a developer who has marketed
Then, we need to clarify "best among what exist".
If you would say that all of the bots "out there" are no good,
then I would wholeheartedly agree that XG may be at least
one of the best of them...
> I suspect he isn't the author of the program and he's been
> assured by whatever developer he has employed that the
> dice are honest (they aren't, trust me) and that it's playing
> to the best of it's ability at level 7 (count them, seven levels
> of difficulty).
"7 levels" of difficulty reminds me of Jellyfish's "7 levels"...
What you are describing, i.e. not having the mastery of the
code because it was stolen from another bot, is probably
more common than what most would suspect. (See my old
posts about XG being based ob Gnubg code in violation of
its license). The only difference may be that tha AI Factory
people may not be good enough petty crook scumbags as
others, yet...
MK
> ..... absurd comments from a developer who has marketed
> a (very poor) backgammon program and yet chooses to
> diss the best program out there.
First, we need to ask "best in what"?
> diss the best program out there.
Then, we need to clarify "best among what exist".
If you would say that all of the bots "out there" are no good,
then I would wholeheartedly agree that XG may be at least
one of the best of them...
> I suspect he isn't the author of the program and he's been
> assured by whatever developer he has employed that the
> dice are honest (they aren't, trust me) and that it's playing
> to the best of it's ability at level 7 (count them, seven levels
> of difficulty).
What you are describing, i.e. not having the mastery of the
code because it was stolen from another bot, is probably
more common than what most would suspect. (See my old
posts about XG being based ob Gnubg code in violation of
its license). The only difference may be that tha AI Factory
people may not be good enough petty crook scumbags as
others, yet...
MK
0 new messages