Follow Up Question for Muratk
William Womack
There are two points that you and other dim brained morons
like you can't seem to get no matter how many times you
are explined:
1- The "gambling tournament circle" you are talking about
is quite apparently a small number of "inbred/incestious
group" who is incapable of conceiving that there could
exist a bigger and more real world outside of their little
primitive and/or imaginary world... You dumb shits are
probably looking at what I am writing here like some
isolated Amazonian tribe looks at an airplane playing
past over them.... Wrrroooommmm..... :)))
2- No matter how confused you are about the differences,
beating a bot claimed to be world-class is not the same
as beating the human world-class champions...!
Think...
Think...!
MK
</quote>
I am trying to following your reasoning but I have some questions to
try to clarify your position.
You have stated that you think Bots are bad for backgammon. As best as
I can follow you you believe this because the bots style of play is
inferior, and therefore those that play like the bots are playing and
inferior style of backgammon compared to the bots. Is that a correct
summary of your position?
It is generally recognized that there are some weaknesses in the bots
and that a skilled player can often times beat the bots by steering
toward these types of positions. Whether this can be done often enough
for a human to consistently beat the bots is open for debate because
it is not always possible to steer games into these positions.
However, if I am following your claims you claim to be able to beat
the bots using what you would consider your normal playing style
rather than a specific anti-bot strategy. Is that correct?
I gather you comment about the "gambling tournament circle" is because
you believe that the players in this circle have accepted the bot's
way of play (in the vast majority of cases anyway) as correct and tend
to play as much as possible like the bots. Therefore, a style of play
such as yours is going to be rejected by these people because the bots
evaluate it as inferior, even though you would claim that it is
superior and that if others would just recognize that bot's style is
really inferior than they would see there was something better. Is
that pretty much your position?
I also believe, trying to gather from what you have posted, that you
think that the current crop of wc players tend to do well when there
error rates are low (according to the bots) because they are playing
others that try to play like the bots as well are thus are both
playing an inferior style of backgammon. Is that a fair assessment?
muratk
> I am trying to following your reasoning but I have some
> questions to try to clarify your position.
I would be glad to clarify. I am glad that you are taking
what I wrote seriously.
> You have stated that you think Bots are bad for backgammon.
> As best as I can follow you you believe this because the
> bots style of play is inferior, and therefore those that
> play like the bots are playing and inferior style of
> backgammon compared to the bots. Is that a correct
> summary of your position?
Yes, except the phrase "compared to the bots" and the word
"inferior".
The first one may be a typing error because the beginning
of your paragraph is in the right direction.
I don't think you meant to compare the inferior style of
the bots to the style of the bots.
And about the word "inferior", personally, I had used words
like "metallic", "sterile", etc...
The "bot style" may be superior or inferior, depending on
whose style you are comparing it to...
But if you are referring to me specifically, yes, you may
say that apparently the "bot style" is "inferior" to "my
style"...
> It is generally recognized that there are some weaknesses
> in the bots and that a skilled player can often times beat
> the bots by steering toward these types of positions.
> Whether this can be done often enough for a human to
> consistently beat the bots is open for debate because
> it is not always possible to steer games into these
> positions. However, if I am following your claims you
> claim to be able to beat the bots using what you would
> consider your normal playing style rather than a specific
> anti-bot strategy. Is that correct?
Yes. Correct.
Furthermore, I am knowingly less consistent in my play
that an average rated player may be.
I have never made an effort to memorize the best opening
moves or best second round moves, etc.
To the contrary, if similar positions arise too many
times in a row, I will be more likely to make a totally
"awful rated" move just to break the monotony and give
myself an extra challenge. (I am more likely to make
such moves especially if I have a comfortable lead).
> I gather you comment about the "gambling tournament
> circle" is because you believe that the players in
> this circle have accepted the bot's way of play (in
> the vast majority of cases anyway) as correct and
> tend to play as much as possible like the bots.
Yes, I still remember comments about "Jelly Grendel"...
> Therefore, a style of play such as yours is going to
> be rejected by these people because the bots
> evaluate it as inferior, even though you would claim
> that it is superior and that if others would just
> recognize that bot's style is really inferior than
> they would see there was something better. Is that
> pretty much your position?
Yes.
Furthermore, unless I am a bg genius to be discovered,
I resent those so-called "world class" player for not
seeing even as mush as I can see as an amateur...!
That's when I really get frustrated and start calling
them ass-kissing partners in crime and such...
It blows my mind that none of those people can see
what I am seeing... :((
Also, as a side comment, my argument is a valid one
in general, in any similar situation but I don't want
to distract from the focus of this discussion.
> I also believe, trying to gather from what you have
> posted, that you think that the current crop of wc
> players tend to do well when there error rates are
> low (according to the bots) because they are playing
> others that try to play like the bots as well are
> thus are both playing an inferior style of backgammon.
> Is that a fair assessment?
Yes.
Since I discovered how to play "money sessions"
against the most promoted bots, I played 100 games
each against Snowie4, XGR+, GnuBG and Jellyfish32.
In case it would matter at all, I won 210 against
190 games, but only 36 against 44 gammons (perhaps
with the consolation that I won the only backgammon
in those 400 games against Snowie4).
But those were "money sessions", so focusing on the
points won, I came out +311 against those four bots
(+126 against Snowie, +76 against XGR+, +51 against
GnuBg and +58 against Jellyfish).
I have been looking for patterns in my own games
for years but I only could come up with a few very
obvious ones.
Now that I am getting better at analyzing sessions
and exporting them in pretty HTML formats, I have
all those 400 matches ready to be posted on the
Internet, if anybody else would be interested in
looking at them to recognize patterns, etc...
Since I'm not very likely to start playing in
tournaments, maybe we all just will have to wait
until somebody else who plays like me eventually
wins an important tournament and draws attention
to the "other reality"...
MK
William Womack
following should happen:
a) You show a positive result against the strong player who will be
playing as much like a bot as possible. In this case a bot analysis of
the matches should show you with a much higher error rate than the
strong player even though you ended up winning.
Or
b) The strong player is able to adapt to your style and will begin to
play moves that are are "un-bot like." In this case I think you
contention has been that the strong player would probably show a
positive result against you, although that is not all that important
to discussion. In this case, however, a bot analysis of the match
would should the strong player playing at a much higher error rate
than normal because they are making moves that the bot considers poor
in order to adapt to your play.
Am I missing something here?
muratk
> So if you were to play "live" against strong players
> one of the following should happen:
>
> a) You show a positive result against the strong
> player who will be playing as much like a bot as
> possible. In this case a bot analysis of the matches
> should show you with a much higher error rate than
> the strong player even though you ended up winning.
Yes, in essence I agree but "as much like a bot as
possible" is a slippery proposition.
I'm sure you are aware that the results would depend
on "how much like a bot" my opponent plays...
> Or
>
> b) The strong player is able to adapt to your style
> and will begin to play moves that are are "un-bot
> like." In this case I think you contention has been
> that the strong player would probably show a positive
> result against you, although that is not all that
> important to discussion.
Well, actually, it could be important to the overall
subject. I will agree that I said I would not be as
brave to bet money against world-class human players
but I never conceded that "they would probably show a
positive result against me" either.
Since I never had the opportunity to play against any
world-class human players, I/we just don't know what
would happen if I did.
You are raising a very good question though. What if
I did well (anywhere from not losing as badly as
expected to actually winning) against world-class
human player...?
Maybe the bots are doing just fine and I/we should
take a broader look at this...??
> In this case, however, a bot analysis of the match
> would should the strong player playing at a much
> higher error rate than normal because they are
> making moves that the bot considers poor in order
> to adapt to your play.
Again, in essence, I agree. You are raising excellent
questions.
Basically, you are proposing that "adjusting to one's
opponent's play would necessarily involve making poor
moves on purpose".
Since the bots are not capable of doing this, they
would still rate them as inferior play, (whether such
deliberate moves are made by a human against the bot
itself or by a human against another human).
But, wait! The beauty of your question in the fact
that there is a difference between making such moves
against a bot which can't adjust back at your poor
moves versus making them against a human player who
adjust right back at you...!
Let me illustrate/elaborate. Player-A starts playing
like the bot and Player-B adjust his play to take
advantage of it. The bot rates Player-A world-class
and rates Player-B beginner.
Player-B is succeeding with his strategy. Player-A
realizes it and also adjusts his play which ends up
unlike the bots as well. Now the bot rates them both
as advanced.
Player-B realizes that he has been deciphered and
further adjusts his play which "coincides" with the
way the bot would play. Thus, the bot promotes him
to supernatural, and demotes Player-A to awful...
Having said these, I guess I am realizing even more
that a bg bot that beat the best human players is a
long, long ways away...
It's very likely that the "world-class backgammon
circles" are full of run of the mill ("bot-like")
players of average intelligence who just can't do
any better, and that people who could have better
things to do in life than gamble for less than
minimum wage per hour spent...
> Am I missing something here?
Were you looking for a contradiction? Everything
you said make sense to me and I don't see any
contradictions.
I doubt that we will get any so-called world-class
gamblers interested in this subject, (which could
be enormously useful for the sake and future of bg),
but I find this discussion very stimulating and
enjoyable. Please continue...
MK