Why many think GNUBG is "rigged."
mont...@lycos.com
since it comes up so often, but I have refrained from doing so because
I don't really care - I just use the software to detect weak points,
and don't care if I "win" against it or not. However, it is
irritating when I go back to play a game from a certain position and
keep having GNUBG roll automatically, and it keeps hitting, time after
time:
GNU Backgammon Position ID: 2N4GBiC2M3MACA
Match ID : UQmrAAAAAAAA
+13-14-15-16-17-18------19-20-21-22-23-24-+ O: gnubg
| O O | | O O O X | 0 points
| O O | | O O O |
| | | O |
| | | O |
| | | |
v| |BAR| | 5 point match
| | | |
| | | |
| X | | X |
| O X X | | X X X X | Rolled 62
| O X X | | X X X X O | 0 points
+12-11-10--9--8--7-------6--5--4--3--2--1-+ X: user (Cube: 2)
In this position, after running with the back man, GNUBG hit five
times in a row, before rolling a 5,5, and thus not hitting. There
seems to be a point in many games were GNUBG wants to hit, even though
it is a significant underdog to do so, and you have to roll it again
and again before finally, on the fifth or sixth time, it does not hit,
even if it is an 11 in 36 to hit. If this only happened once in a
while, or if it happened on a 25 out of 36, it would not look quite
unlikely.
As I've said before, don't take the automatic dice seriously, but
instead just go back and play the key the positions without the
jokers, so that you learn more than you would in a joker-filled game.
Obviously, there are plenty of these games even when you roll out
manually, but if you are using GNUBG as I am, that is, as a learning
tool, you want to maximize your experience, rather than becoming
irritated as GNUBG hits you time after time on an 11 in 36 shot.
Michael Petch
on fingers... The troll.
;-)
On 8/9/07 3:41 PM, in article
1186695672....@g12g2000prg.googlegroups.com, "mont...@lycos.com"
<mont...@lycos.com> wrote:
[removes the stupidity]
Stick
Are you kidding ... has to be the best humor that gets injected in my
day sometimes. I've given up on every explaining anything to him
again but still look forward to his posts, always bring a smile ;)
Stick
On Aug 9, 11:41 pm, Michael Petch <mpe...@capp-sysware.com> wrote:
> Its so difficult.... Errrrrrr.... Don't.... wanna say it.... Feed.... Sits
> on fingers... The troll.
>
> ;-)
>
> On 8/9/07 3:41 PM, in article
> 1186695672.961514.56...@g12g2000prg.googlegroups.com, "monty1...@lycos.com"
>
> <monty1...@lycos.com> wrote:
>
> [removes the stupidity]
ri...@notyahoo.com
wrote:
>
>Are you kidding ... has to be the best humor that gets injected in my
>day sometimes. I've given up on every explaining anything to him
>again but still look forward to his posts, always bring a smile ;)
>
>Stick
If you really want a smile then complete ignore Monty's posts. The
more you ignore him, the more outrageous his posts will become in
order to get attention. After a while when no one responds to him, he
will find another group to troll.
Rich
>
>On Aug 9, 11:41 pm, Michael Petch <mpe...@capp-sysware.com> wrote:
>> Its so difficult.... Errrrrrr.... Don't.... wanna say it.... Feed.... Sits
>> on fingers... The troll.
>>
>> ;-)
>>
>> On 8/9/07 3:41 PM, in article
>> 1186695672.961514.56...@g12g2000prg.googlegroups.com, "monty1...@lycos.com"
>>
>> <monty1...@lycos.com> wrote:
>>
>> [removes the stupidity]
>
--------------------------------------------------------------------
=======================================================================
=======================================================================
"I don't have the least interest in verifying my theory because I have
already adopted it as correct to myself, from both a logical and
practical standpoint." - Grunty
mont...@lycos.com
comprehension disorders. Whether or not GNUBG is "rigged" is
irrelevant. My point was that those unfamiliar with it should not
worry about the "rig" factor, because the software is very useful.
I'm concerned that many beginners will play GNUBG and think that they
will never be good at backgammon, and then play poker or something
else. Instead, they should play humans and use GNUBG for analysis
purposes, as I've pointed out in more than one previous post.
David C. Ullrich
Just for fun I set this position up and played a 62. gnubg then
rolled 44. Didn't hit. Curious.
Let's see what happens next. In a money game I'd double - I
don't have any experience with match play, scared of big
cubes, so I roll. gnubg says no, I should double. Ok, I double.
gnubg takes, and I easily win the game for a 4-0 lead.
Hmm. Crawford game. I'm in a little trouble a few times,
but I come in on four-point boards twice, I come in on
a five-point board at a key spot, and now gnubg has four
man behind a full prime and I'm not behind anything.
Oops, I just roiled my only bad number and gnubg hit!
I can understand why people think it's rigged...
Bounced twice on a pathetic three-point board...
gnubg wins the Crawford game.
I start the next game with a 42. gnubg doubles.
Let's see, what is this "match" stuff? I guess
I have a free drop, but since I started with a
42 I take, saving the free drop for later.
Few rolls later I'm in a little trouble. Woops,
I roll 61, the only thing that hits! Hah.
Oops, I'm in trouble again... wow, I roll
a 63, escaping and hitting. The poor guy's
got three men behind a four-prime/board, one
on the bar, I'm out.
What the hell is that? Oh right, a gammon doesn't
cost it anything here. Whoops, I roll 33,
hitting the outside blot and pointing on an
inside blot... Three men closed out, let's just
be a little careful here...
gnubg comes in and concedes the gammon.
Ok, so I won the match 8-1. But that spot up there
where it hit. Really, must be rigged.
>As I've said before, don't take the automatic dice seriously, but
>instead just go back and play the key the positions without the
>jokers, so that you learn more than you would in a joker-filled game.
>Obviously, there are plenty of these games even when you roll out
>manually, but if you are using GNUBG as I am, that is, as a learning
>tool, you want to maximize your experience, rather than becoming
>irritated as GNUBG hits you time after time on an 11 in 36 shot.
Good of you to share your insights.
************************
David C. Ullrich
Gregg Cattanach
That really reminds me of the old 'joke': "So..... have you stopped beating
your wife?"
Or in monty's terms, "Whether or not you are beating your wife or not is
irrelevent. It's only important that you use lots of ointment on her cuts.
Now I'm NOT saying you're beating your wife... No, that's not it at all."
Troll, through and through.
mont...@lycos.com
thinking, Gregg, because then you and your ilk might not look quite so
silly. There are some things that are undeniable facts. One is that
many have posted on this NG about their thoughts on GNUBG's dice
software. Another is that people like yourself (I like the phrase
"rubber room boys," but whatever works for you is fine) tell these
posters that it is an issue of "selective perception," which certainly
could be true. Thus, my original post on this thread is consistent
with these facts, and I was pointing out to those who had similar
experiences that they should disregard what appear to be nearly
impossible sequences of rolls, and just use the software for what it
has to offer, rather than be concerned about winning against it.
Your analogy is ludicrous, since there are no circumstances that would
render it acceptable to beat one's wife. On the other hand, you
"rubber room boys" don't deny that the unlikely dice sequences happen,
but are insisting that it must not be "rigged." While this may be
true for software that is open source, though I don't have the
expertise to investigate it myself, and this is obviously true for
many others as well. Furthermore, one could do manual rolling, but I
don't bother to do this because, as I said, one can just go back and
let GNUBG roll it out again. When I've done this, I've been surprised
at how often a 12/36 hits not just the second time, but the third,
fourth, and fifth times. Now if this is "selective perception," then
those new to the software should realize that while the statistical
probability of this happening is very small, it does happen more than
one might think. And even if it were "rigged," it would not matter,
because it is still indispensable software for those who want to
improve significantly (assuming one does not pay for Snowie or another
one).
With your inability to consider exactly what I was saying, and trying
to accomplish, Gregg, you have played the role of "troll," actually.
What is a beginner to make of all this, for example? That many
backgammon players are nasty and rude? What in the original post
would have been unclear to a beginner who just started to use GNUBG to
help improve his/her game? Do you honestly think that the original
post would have led anyone to think that GNUBG was "rigged?" Why not
just respond with a post pointing out that it's not uncommon to see
these kinds of highly unlikely sequences once in a while, if one plays
against GNUBG a lot?
My guess is that there are issues here that transcend backgammon, but
if you don't realize that at this point in your life, then I doubt you
ever will.
ri...@notyahoo.com
>It's too bad nobody appears to have taught you about critical
>thinking,
I had to get a new keyboard. That will teach me to read this troll's
posts.
Rich
Michael Petch
On 8/11/07 3:33 PM, in article jnasb3dcc810dd660...@4ax.com,
mont...@lycos.com
"credentials." Those who can't argue a point often resort to claiming
that an "expert" has "great credentials," rather than examining the
evidence. You have provided yet another example of someone lacking in
critical thinking skills.
ri...@notyahoo.com
Here is an example of Monty's critical thinking skills:
On 24 Mar 2006 20:05:24 -0800, "montygram" <nazzt...@lycos.com>
wrote:
>The problem with GNU is that too often the game is decided by a big
>double, usually in GNU's favor.
>If you don't believe me, then just do what GNU suggests is your best
>move and you will see. I have played way too many games against it to
>think there is even the possibility that everything is "on the level,"
Monty relies on his selective perception based upon playing "way too
many games" that gnu is not "on the level". And based upon his
critical thinking skills there is no possibility that gnu is on the
level. I am impressed. Professor Monty sure has demonstrated amazing
critical thinking skills. He is a legend in his own mind.
Piranha
Well, as funny as this thread has become, I can´t resist asking a
serious question:
Monty, have you considered the fact that GNUBG is open source?
Anyone who has a little bit of programming knowledge can look up the
source code of the program and verify that there is nothing but a
clean random generator inside GNUBG which doesn´t consider the
position on the board before it rolls the dice for either side, thus
there is not the slightest possibility for the program to determine
which roll might be useful in any given position for either side
Furthermore, if you believe the program you are using is different
from the code shown in the source files, GNUBG offers the sources for
download and you can compile them yourself, so you can be absolutely
sure, byte for byte that you get exactly what the source code shows
But I guess these tiny details don´t matter much compared to the
subjective impression of a self nominated expert, huh?
Simon Woodhead
roflmao :-)
Monty, this is one of your best ever!
Thanks! And just for more laughs, would you like to
share your "evidence" with us ? Or would that be too
much to ask ?
Simon Woodhead
> But I guess these tiny details don´t matter much compared to the
> subjective impression of a self nominated expert, huh?
Of course they don't. Murat claimed to be a programmer but
wasn't capable of even looking at the gnubg source code, let
alone pointing out the places where it's obviously "rigged".
Troll Monty doesn't even have a concept of what software
does, let alone an understanding of how it works. Which
is fine, because you don't need to know anything to
collect and present evidence of rigging or cheating. You
just need to actually go out there and do it. With all
Monty's vast experience and critical thinking ability,
you'd think he'd come up with something a little better
than the tired old selective perception argument.
Show us your evidence, Monty.
"There are some things that are undeniable facts.
One is that many have posted on this NG about their
thoughts on GNUBG's dice software." -Professor Monty.
Here's another undeniable fact. Despite all the evidence
to the contrary - the earth is round. I know it doesn't
look it, but really it is. I know because I've seen the
source code.
Michael Petch
here is my evidence, and a simple challenge for you and a very lucrative
challenge. The only way you have to lose money in this challenge is if you
are unsure about your claim.
So I provide the evidence code snapshot 20070731 of gnubg (Which I have
downloaded) -
http://www.gnubg.org/media/sources/gnubg-source-SNAPSHOT-20070731.tar.gz
With GNUBG and code snapshot above (Standard non graphical build or
graphical build - doesn't matter) with settings of Mersenne twister, ANSI,
Blum, BSD, Isaac, OR random.org for dice settings (No manual Dice usage,
and no dice manipulation allowed - just straight dice from any of those
random number generators), and no modification to the source files, header
files or documentation files for the linked release - the challenge is this:
Find any code that functions under the conditions above - that produces a
roll based on the "current checker positions for rather player, the cube,
and the score" (GAME STATE) and not any of the random sources above. This
satisfies "rigged dice". The only way to rig dice is to attempt to produce a
roll that is not based on the random source and is produced by some "other"
method that specifically relies on the current state of the game (The
current checker positions, cube and score).
And I will put money on this challenge. I will give you 1 week to decide to
accept this challenge and offer up some help for you.
I give you 1 year from the date of acceptance (starting within the next 7
days) to find such code and inform this newsgroup of where this code resides
in the tarball and sources linked to this post (And it must be the file as
it was on August 11th, 2007 00:00 GMT. If you can produce indisputable
file/module/line numbers that satisfies the challenge then I will pay you
$100,000 US. If you fail to provide such evidence (within the 1 year) for
GnuBG rigging dice then YOU pay me $30,000 USD.
Not only that I'm going to give you a legs up. Here are a list of the
current developers on gnubg who and how to contact them (Just create an
account at savannagh to send messages to them). See this link:
http://savannah.gnu.org/project/memberlist.php?group=gnubg . I will allow
you to seek out any professional help, amateur help, software developers,
house wives, or a lightning striking down from any God that can pinpoint the
code that "rigs" in the conditions above. You can even split the proceeds of
winning the challenge with people that help you (or maybe you do it
yourself). Heck you might want to seek out people here on this newsgroup
that could use extra pocket cash.
This challenge costs you nothing up front. And only costs money if you fail
to provide the EVIDENCE. This is the great thing about GnuBG being open
source. And you have a whole week to decide how strong you feel your claim
is.
Queue up Monty to find some reason why the challenge isn't worth his while
and how irrelevant it is, and how somehow it doesn't answer the claim he is
making (which it does). I would prefer he accept the challenge though,
because it'll be an easy 30,000 USD in my pocket. That's how confident I am
of my claim, and MY evidence.
In the end: My proof is the source code - so my challenge put the source
under a microscope for a year.
Mike
On 8/11/07 4:09 PM, in article
1186870183.1...@22g2000hsm.googlegroups.com, "mont...@lycos.com"
Michael Petch
I have it, but its an amount that I can liquidate in a week and utilize and
mutually agreed upon third party (I prefer a well established escrow
company). However I will ask that Monty and people that support him to put
up 30,000 in collateral. Again I don't care if someone else funds Monty.
On 8/11/07 5:22 PM, in article C2E3A0BF.43550%mpe...@capp-sysware.com,
Michael Petch
"current checker positions for EITHER player, the cube,"
On 8/11/07 5:22 PM, in article C2E3A0BF.43550%mpe...@capp-sysware.com,
"Michael Petch" <mpe...@capp-sysware.com> wrote:
Michael Petch
requiring him to do so unless I am asked to put money in escrow.
On 8/11/07 5:56 PM, in article C2E3A8B8.4355D%mpe...@capp-sysware.com,
Stick
Willing to send the same bet my way =), self appointed "King" of the
"rubber room boys"
"Stick"
http://www.bgonline.org/forums/
On Aug 11, 7:22 pm, Michael Petch <mpe...@capp-sysware.com> wrote:
> Monty you never cease to amaze me :). Critical idiocy is more like it. So
> here is my evidence, and a simple challenge for you and a very lucrative
> challenge. The only way you have to lose money in this challenge is if you
> are unsure about your claim.
>
> So I provide the evidence code snapshot 20070731 of gnubg (Which I have
> downloaded) -http://www.gnubg.org/media/sources/gnubg-source-SNAPSHOT-20070731.tar.gz
> 1186870183.130654.261...@22g2000hsm.googlegroups.com, "monty1...@lycos.com"
>
>
>
> <monty1...@lycos.com> wrote:
> > I deal in critical analysis of available evidence, Michael, not
> > "credentials." Those who can't argue a point often resort to claiming
> > that an "expert" has "great credentials," rather than examining the
> > evidence. You have provided yet another example of someone lacking in
> > critical thinking skills.- Hide quoted text -
>
> - Show quoted text -
Stick
had my original 1 pt. board. He fanned, I made no inner board
improvement, he fanned AGAIN. Insanity ...
"Stick"
Simon Woodhead
And I just played that cheating pile of gnubg. In an even race
it rolled 66, 22, 44 in a row. For me. Giving me the game by a
country mile. Would never happen in real life.
mont...@lycos.com
How can you read the original post and not understand what it is about
- how old are you?
This is about the perceptions of beginners, not about statistical
theory or random-dice software.
I am telling people that they should disregard anything that appears
to be "rigged" in GNUBG and use it in particular ways to improve their
games, as I have in previous posts. I merely provided an example of
something that actually happened (several times over the last few
months) which might tempt some people to think it is "rigged."
I will add here for those who are not bouncing off of white, rubber
padding that thinking only about playing the best GNUBG move has
improved my game tremendously (since I started to do match analysis on
every match several months ago). Put aside the "rigging" concerns and
just let GNUBG show you how to play the game (though the cubing is
often wrong, so do some rollouts when things don't look right).
Yesterday, in a freeroll, I had an opponent whose commentary made it
appear that he was bouncing off walls. He talked about the "great
luck" I had, and how great a player he was. He implied that I was not
good. We both rated "expert," though I was a bit higher. And I did
have better luck than him, but not by a huge amount, according to
GNUBG. It is this kind of thinking that I want beginners to realize
is counter-productive. Use GNUBG's match analysis feature and focus
on raising your rating. Once you get to a high level of play, you can
then forget about the "rigging" issue, because unless it is very
obvious, you will win your share of matches against the typical human
players on the internet.
Michael Petch
On 8/11/07 11:09 PM, in article
1186895395.2...@b79g2000hse.googlegroups.com, "mont...@lycos.com"
<mont...@lycos.com> wrote:
> Michael:
>
> How can you read the original post and not understand what it is about
> - how old are you?
>
Of course Monty resorting to a backhanded personal attack (Usually the sign
of someone who has a weaker argument), but we shall move on.
> This is about the perceptions of beginners, not about statistical
> theory or random-dice software.
Are you a beginner / novice or as you have implied "Near World Class"? I ask
not as an attack but as a basis for anyone who READ your post - to
understand two things (the rest is just an example of these two statements
in your mind)
First the title "Why many think GNUBG is "rigged"". You suggest in your post
to me that YOU ARE NOT among the people who say this, but your OWN words in
your first post of this thread state:
> As I've said before, don't take the automatic dice seriously, but
> instead just go back and play the key the positions without the
> jokers
You claim to have said it before (which is actually the truth since you have
said it before) but more importantly you PREFACE the entire sentence with "
don't take the automatic dice seriously"
First question Monty. If the dice in Gnubg are not rigged as you claim in
your last post to me, how is it that dice that are not rigged can not be
taken seriously. But there is no way of misreading this because you go on to
state the solution "go back and play the key the positions without the
jokers". The implication is that you believe that the dice are not serious
because the bot throws many jokers (and in your example suggest it will do
it quite a few times)
So the dice are NOT SERIOUS (by your words), and I believe you are SAYING -
GNUBG's dice are not random and or rigged, and that when it rolls Jokers you
should go back and manually put dice in to get a more REAL LIFE experience
(Equating REAL LIFE dice with fewer jokers for an opponent).
That's the way most sane knowledgeable (even novice BG players) will read
it. So I ask straight out - why can *YOU* NOT take GnuBg's dice seriously?
Please provide evidence, including the line number/module/file where the
"non serious" rolls take place.
Please don't tell me you didn't say it. You have.
I will tell you this (And I believe Gregg tried to hint at this through his
analogy with the wife beating joke). The title of your post plus the line
about not taking automatic dice seriously, and then suggesting to go back
and use manual dice makes me come to a simple conclusion:
I believe that YOU believe the dice in GnuBG are rigged, but you know that
argument doesn't fly in this forum because its not the supported theory here
(And not one you have been able to prove). So rather than say "I believe the
dice are rigged", you use a title with MANY (Which you then try to distance
yourself from as being inclusive of yourself). You want to fuel the believe
that the dice ARE rigged without you actually coming out and saying it. The
reasoning is to make yourself feel better when some beginner will believe
the same nonsense, and then make posts saying what you are unwilling to say
yourself (Which some will define as a newsgroup Troll)
If you were truly being altruistic and trying to help beginners that might
think "Gnubg's appears to be rigged dice" (As you suggest is your
motivation) you would have had a title to the thread something like "What to
think when GnuBG rolls Jokers". Then YOU tell people the dice are NOT
rigged, and that they CAN be taken SERIOUSLY. Then you proceed to explain
how to use Gnubg to take a position and redo it to see what might happen if
the bots rolls had been different. Heck at this point you might just want to
explain to beginners what a "proposition" is.
So you tell me. Does your method or my method appear to be a better way of
posting to this forum without promoting the nonsense that Gnubg has rigged
dice. I still want an answer to the fundamental question "Why can beginners
not take Gnubg automatic dice seriously".
Mike
Michael Petch
On 8/12/07 1:45 AM, in article C2E416AF.4357D%mpe...@capp-sysware.com,
"Michael Petch" <mpe...@capp-sysware.com> wrote:
> So the dice are NOT SERIOUS (by your words), and I believe you are SAYING -
> GNUBG's dice are not random and or rigged,
So the dice are NOT SERIOUS (by your words), and I believe you are SAYING -
GNUBG's dice are not random and/or that they are rigged,
Gregg Cattanach
<mont...@lycos.com> wrote in message
news:1186866646.2...@k79g2000hse.googlegroups.com...
Can I get a large and substantial WHOOSH???
>
--
Posted via a free Usenet account from http://www.teranews.com
Gregg Cattanach
David C. Ullrich
>I deal in critical analysis of available evidence,
omigodiwasn'tgoingtosayanymmorebutthisistoomuch...
This is one of the funniest thing's you've said.
You've been told many times exactly what would
constitute actual evidence that gnubg's dice
were rigged, and you've never given any such
evidence.
(In case you forgot: The evidence we need is
_statistics_ on _all_ the rolls for a long
period. Talking about anecdotes, this game
gnubg hit five times in a row, as you _do_,
is the exact _opposite_ of dealing in critical
analysis of the available evidence.
Actually, "stats on all the rolls for a long
period" might not convince someone who was
convinced that the dice were rigged - it
could happen that they're fair in the long
run but rigged to give good rolls at
crucial times. If you actually want to
settle the question you need to do the
following - the following has been explained
previously, but not so many times, so you
may have forgotten:
For a _long_ period, do this:
_Every_ time you're playing gnubg, _before_
you let it roll the dice, look at the position
and decide whether you feel it's "crucial".
If you decide it is crucial then do this:
(i) Decide how many rolls count as jokers.
(ii) Let gnubg roll, and record what roll
it got, _along with_ your _prior_ estimate
of how many jokers there were.
Note: If you actually want to give a critical
analysis of the evidence then several things
are important:
You need to decide whether a position is "crucial"
_before_ you let gnubg roll! If you don't decide
in advance that a position is crucial and then
it rolls a joker then it's not fair to count that
roll in your records.
You need to decide which rolls count as jokers
_before_ you let gnubg roll. If it rolls a joker
that wasn't on your list then you have to record
that one as a non-joker, or else you're cheating.
You need to record the outcome for _all_ the
rolls that you decided were crucial.
Let us know when you've done that.)
>Michael, not
>"credentials." Those who can't argue a point often resort to claiming
>that an "expert" has "great credentials," rather than examining the
>evidence. You have provided yet another example of someone lacking in
>critical thinking skills.
************************
David C. Ullrich
muratk
> Of course they don't. Murat claimed to be a programmer but
> wasn't capable of even looking at the gnubg source code, let
> alone pointing out the places where it's obviously "rigged".
Nice to be remembered and talked about... :))
Yes, I have admitted before that I was not proficient enough in
"C" language and that neither did I have time to look through
tens of thousands of lines of coding...
However, I can prove that I was a computer programmer starting
in 1980 and until about mid 1990's. But this is besides the issue.
Nobody needs to be a computer programmer to be a good BG player
and question whether bots like jellyshit and gnudung are rigged.
For years, Murat made fun of you, using amusingly offensive language
to filter out the "ass lisking, daddy sucking, gamblers anonymous
material scums"... But apparently not many passed the test... :(((
To refresh the memories of the old-timer scums and for information
of the new-timer scums, here are some outstanding issues that Murat
had brought in front of this discussion group:
1- Jellyshit cheats! No matter how obsolete jellyshit becomes, its
source code will never be made public! Jellyshit has been obsolete
for about 10 years... Has any cocksocking computer programmer other
than the one who wrote it seen the source code yet...???
2- Gndung is a "neural network" that learns to play along the way.
What dice generator did it use to learn? Mercenne Twister? That or
another, it doesn't matter... What matters is that it will play the
way it learned! That's why it's a great back-game player... While
most humans would shit in their pants, it waits "for the scheduled
train" and more than often enough wins... :))
3- Regardless of the outcome, anybody who argues that the "look and
feel" of the matches against gnudung using built-in vs. random dice
is the same, must be an "ass lisking, daddy sucking, gamblers
anonymous
material scum"... :))
4- Time and again I offered to play against the gnudung on-line and in
real-time (such as using an instant messenger) for two purposes:
a) That gnudung can be beaten by people like me, defying the crude
and cold statisctical method...
b) Based on how gnudung moves or even better makes cube decisions,
one can quess the upcoming dice well above 50% (perhaps in the
70-80%)...!
c) Winning %'s on "a" or "b" can be "reverse-proportional"... Yes,
that means you can bait gnudung and win big! :)
5- Besides all the arguments anybody can make to the "ass lisking,
daddy sucking, gamblers anonymous material scum", which will of
Corse be dismissed by divine will, my advice to the newcomers into
this group is to ignore these scums and "THINK FOR YOURSELVES"...
And don't waste any money for taking lessons or subscribing to
worthless internet publishings of these "ass lisking, daddy
sucking,
gamblers anonymous material scums"...
MK
PS: I will write again when I have time but who were those slimes who
pretented to be me for years...? If you are not sure that an article
published under my name is actually from me, please send me an email
to verify... Thanks and fuck you all.... :)))))
Simon Woodhead
> On Aug 11, 5:12 pm, Simon Woodhead <si...@uniq.com.au> wrote:
>
>> Of course they don't. Murat claimed to be a programmer but
>> wasn't capable of even looking at the gnubg source code, let
>> alone pointing out the places where it's obviously "rigged".
>
> Nice to be remembered and talked about... :))
Sorry, everyone. Monty, meet Murat. Murat, this is Professor
Monty. You two should get along just fine.
David C. Ullrich
wrote:
>> As I've said before, don't take the automatic dice seriously
>
>
>Well, as funny as this thread has become, I can´t resist asking a
>serious question:
>
>Monty, have you considered the fact that GNUBG is open source?
>
>Anyone who has a little bit of programming knowledge can look up the
>source code of the program and verify that there is nothing but a
>clean random generator inside GNUBG which doesn´t consider the
>position on the board before it rolls the dice for either side, thus
>there is not the slightest possibility for the program to determine
>which roll might be useful in any given position for either side
>
>Furthermore, if you believe the program you are using is different
>from the code shown in the source files, GNUBG offers the sources for
>download and you can compile them yourself, so you can be absolutely
>sure, byte for byte that you get exactly what the source code shows
Heh. That wouldn't prove anything - whoever wrote the compiler
could be in on the scam! Perfectly clean source code, but if the
compiler notices that it's compiling gnubg it inserts a few
special little tricks.
So we also need to see the source code for the compiler.
Except that's not enough because...
(There's this classic story from the early days of UNIX
that I don't recall exactly... ah, see
http://dnjonline.com/article.aspx?ID=may06_securecode
(scroll down to the bold-face paragraph.))
>But I guess these tiny details don´t matter much compared to the
>subjective impression of a self nominated expert, huh?
************************
David C. Ullrich
ri...@notyahoo.com
The reality is that GNUBG is not rigged. And that can be proven since
it is open source code.
But every once in a while when playing backgammon you will run into a
string of what appear to be unusual rolls. The more you play, the more
likely you will run into these "strange" sequences that sometimes
result in your losing games which appear you should have certainly
won.
Individuals who are extremely narcissistic, believing that they are
much better than they actually are, and expect to win all or nearly
all of their matches will become very upset when they lose due to bad
luck. They then will try to rationalize their bad luck by making false
accusations of dice being rigged.
And when asked for proof that the dice are rigged these individuals
will then completely deny making such a statement and simply say it is
*just* their "impression".
Although I and others have said that Monty is a troll, I actually
don't believe this. I do believe that he is suffering from a mental
illness however and this disorder results in his very bizarre posts.
Cue Monty to accuse me of making a diagnosis without evaluating him in
person. My response is that his posts make unnecessary direct face to
face contact. Some things are axiomatic.
Stick
"That's why it's (GNU) a great back-game player" If you think GNU
plays great backgames, either for or against, we're done here. Thank
you for participating in my daily humor project.
Stick
paulde...@att.net
Interesting. I didn't know gnubg was a bad back-game player.
If you have examples of bad back-game play by gnubg (and have the time
to dig them out), I'm sure it would be interesting to many here.
I'm a bit puzzled here since there seems to be a near-consensus that
gnubg is either the best or second-best player in the world. (It's a
toss-up with Snowie).
Hard to get that good without being a great back-game player.
Paul Epstein
Stick
Not true.. not true at all. Most of the great players fall into
backgames very rarely. It's a last resort, something you're looking
to get out of at the first sign of light in most cases. The only
times backgames become somewhat appealing is when there is no other
option (esp. no other option + DMP or other match score plays where
you losing a gammon matter not)
There's no question that GNU is one of the best players in the world,
all of the rollouts on my site are done with GNU and the close calls
(TCTC) or interesting ones I also do with Snowie. I don't do this
because I know whether or not one bot is better than the other, but to
get a second *world class* opinion. Truth be told I trust Snowie's
money game play ever so slightly more than GNUs but in general I'd say
they're equal for all intensive purposes. GNU would obviously win
match play with checker play according to score. Anyway... back on
topic.
There are many examples of GNUs backgame play where it's coked out. I
will have to dig for them and again, if I find them I'll post them on
my forums, hate posting examples here, but I may come back and give
you a direct link. I believe Snowie's backgame play to be much better
than GNUs (esp if you toss in the cube, it has been shown how GNU will
fubar the cube) and also believe the top players are in accordance
though I certainly can't speak for them all. If I get time tonight I
will try to toss up some examples of what I mean.
Don't forget BGBlitz ... hell of a bot in itself ;)
Stick
http://www.bgonline.org/forums/
Stick
first find the positions then have at least a GNU rollout and
preferably a GNU & Snowie rollout for said position (because why
bother w/a GNU rollout if it fucks up play that much?) and Snowie
rollouts of backgames take forever and a take. While searching though
I came across one of Neil's postings and thought you may trust his
words more than mine.
Neil says (talking about a backgame position where GNU chooses quite
an odd play): Am doing a Snowie rollout and I don't trust GNU to
handle this far outside prime vs likely advanced backgame
If Neil still checks in here he'll let you know he doesn't think much
of GNU playing backgames ((outside primes))
Stick
Stick
think most any expert-whatever player should be able to similarly
exploit Gnubg's weakness with cube decisions in massive backgames and
snake positions.
NeilKaz: The pro level BG student needs both (GNU & Snowie), since
GNU can be really screwed up in developing backgame situations and
these are very important to understand for top WC play.
Michael Depreli (commenting on Neil's statement above): I agree 100%
with your statement however if you take all aspects of BG and
particularly those that occur much more frequently in day to day play
rather than developing back games, I personally feel GNU has the edge
and in $ play I would put up my own money in a long head to head
session
Again, tough to find specific examples, have found a couple, but
statements from world class players are widespread who have seen GNU
in action.
Stick
mont...@lycos.com
beginners that might
think "Gnubg's appears to be rigged dice" (As you suggest is your
motivation) you would have had a title to the thread something like
"What to
No, I think it was just right, because those who think it might be
"rigged" will be more likely to read it. Your title implies that the
"jokers" are not the result of "rigging." However, harping on such
minutiae is not my cup of tea, but rather precision is. I made my
point precisely, while others are trying to "read into" the post and
replies all kinds of things I clearly did not argue.
I will state it again: it is counter-productive to worry about
whether GNUBG is rigged, because the most important thing to do is to
reach a certain level, then you will know exactly how to use the
software, and you can decide for yourself whether a money site is
"rigged" or not. Can Murat play consistently around the "advanced"
rating? If not, then that is what he should focus on, and not notions
about "rigging." Even if you lose a huge percentage of the time
against GNUBG, so long as you improve significantly, you are getting
what is most important from the software. In poker, beginners are
warned not to "steam," and this is something that I've found to be
important in backgammon as well. Moreover, playing a human and
allowing GNUBG to analyze the match does not bring "rigging" into the
picture.
As to GNUBG's problems with certain kinds of positions, I'll say that
these positions do not come up very often, perhaps because my
competition is so weak, but then again, I do play GNUBG a lot as well,
and these positions don't come up in that context either. Instead, it
is certain cube decisions that I rollout most because GNUBG often
evaluates a double as not quite good enough to double.
I will advise those who play in 3 point matches is not to take them
too seriously, because it is very common to be beaten by someone rated
"awful" or "beginner," no matter how you play. However, over the last
couple of weeks, my record in 9 point matches is 15 wins and 2 losses,
one of which was due to incredibly bad luck and in the other my
opponent played almost as well and got better luck. I am now
tightening up my match equity skills in typical middle game positions
- focus on learning specific things rather than results. GNUBG will
tell you if you are wasting your money by playing those who are too
strong. Of course, if you play for small stakes against someone much
stronger, you might be obtaining a "lesson" at a deep discount. This
brings us back to playing against GNUBG, because it doesn't cost
anything to do so, and by using the analysis feature, you can detect
your weaknesses easily, but if you worry about beating GNUBG, you are
really wasting your time. Early on, I realized that it made more
sense to simply go back and give GNUBG a "nothing" roll, and let play
proceed from there. However, as I said in the original post, I am
amused by some of the weird rolls that occur, though when a 11/36
keeps hitting over and over again, it's important not to get
irritated, especially when you've seen this happen more often than it
appears that it should.
Michael Petch
On 8/13/07 1:00 AM, in article
1186988428....@o61g2000hsh.googlegroups.com, "mont...@lycos.com"
<mont...@lycos.com> wrote:
>I will advise those who play in 3 point matches is not to take them
>too seriously, because it is very common to be beaten by someone rated
>"awful" or "beginner," no matter how you play.
[the rest snipped for brevity]
*Some* of the post is an off topic ramble to my questions (And some sem to
be responses to others posts). I will point out that Monty has an issue with
people "harping on minutia" as a response to anyone who engages in Critical
reading, critical thinking, and critical analysis of his posts.
I didn't ask about your luck, or positions. Monty I read your entire post
and the answer to my fundamental question was not actually answered. The
word "SERIOUSLY" appeared once in your post in the context of " I will
advise those who play in 3 point matches is not to take them
too seriously,". Taking a match seriously, is not what you originally talked
about in your first post.
So on behalf of all the beginners I have one question, the same one I asked
a few times (to get the point home that it was my key question) - "Why can
beginners not take Gnubg automatic dice seriously". This question is in
direct relation to this comment in your opening post of this thread "As I've
said before, don't take the automatic dice seriously"
Please note you said AUTOMATIC, and DICE. I want an answer specific to that
question. I don't care about how much luck you have had this week, how much
luck you had this week, I don't care about positions you have posted, I care
about ONE answer.
Fundamentally, beginners want to know what is NOT Serious about Gnubg's
automated DICE. I look forward to the answer. Just answer the one question
in a clear concise manner that beginners will understand. Thanks
Mike
Michael Petch
On 8/13/07 8:50 AM, in article C2E5CBD0.435AD%mpe...@capp-sysware.com,
"Michael Petch" <mpe...@capp-sysware.com> wrote:
>
> Fundamentally, beginners want to know what is NOT Serious about Gnubg's
> automated DICE. I look forward to the answer. Just answer the one question
> in a clear concise manner that beginners will understand. Thanks
Please do me one favor. As to not lose site of the answer, please answer my
one question in a forum post that is ONLY a response to this question. If
you want to answer someone else's question in this thread please reply to it
separate to your response here.
This will make sure that the answer can be clearly seen and we know what you
are responding to. Thanks.
ri...@notyahoo.com
>As I've said before, don't take the automatic dice seriously
Please explain Monty why people should not take the automatic dice on
gnubg seriously.
This is ONE question. Please answer THIS question and no other.
You clearly warn people not to take the automatic dice on gnubg
seriously. Why????
And don't say that the point of your post was not about whether or not
the dice on gnubg are rigged or not. Your suggestion that people
should not take the dice seriously implies that you think that the
dice are not truly random. But I will let you explain in your own
words WHY the automatic dice should not be taken seriously.
It is my opinion that the dice on gnubg should be taken VERY
seriously. And the reason I think that they should be taken seriously
is because they are random as can be PROVEN since the software is open
source code. Individuals need to understand that given enough rolls,
random dice can "appear" to be nonrandom, fixed or rigged due to
selective perception. That is the lesson that needs to be learned by
beginners when they play against gnu.
But enough of what I think. Why do YOU think that the automatic dice
should not be taken seriously as you DIRECTLY stated in your original
post.
Watch how Monty completely dodges this very straightforward question
to his assertion.
Kees van den Doel
<mont...@lycos.com> wrote:
>With your inability to consider exactly what I was saying, and trying
>to accomplish, Gregg, you have played the role of "troll," actually.
Very funny.
Kees (Very funny, Elzo, however, this madhouse, ou misspelt as English,
I intend to repeat myself, by members at WORK and over.)
Walt
>
> Heh. That wouldn't prove anything - whoever wrote the compiler
> could be in on the scam! Perfectly clean source code, but if the
> compiler notices that it's compiling gnubg it inserts a few
> special little tricks.
>
> So we also need to see the source code for the compiler.
> Except that's not enough because...
>
> (There's this classic story from the early days of UNIX
> that I don't recall exactly... ah, see
>
> http://dnjonline.com/article.aspx?ID=may06_securecode
That's a great story. Of course, one doesn't need access to the source
code to demonstrate that the software "cheated" - one could demonstrate
the exploit by logging in with the "universal" password.
Demonstrating that a bot cheats would be a very easy thing to do. Just
have it play itself, with one instance controlling the dice and the
other not controlling the dice. If it cheats with the dice it should
become apparent after a couple hundred games. And if it cheats as much
as monty et. al. seem to think, it would be apparent after ten games.
//Walt
Wolfgang Braun
> David C. Ullrich wrote:
> Demonstrating that a bot cheats would be a very easy thing to do. Just
> have it play itself, with one instance controlling the dice and the
> other not controlling the dice. If it cheats with the dice it should
> become apparent after a couple hundred games. And if it cheats as much
> as monty et. al. seem to think, it would be apparent after ten games.
GNU is of course programmed so it notices this feeble attempt and always
plays straight when playing itself.
> //Walt
Wolfgang
bcc...@gmail.com
and understand that it is simply using a random number generator for
dice rolls.
What I wonder if other people are considering is just how random a
random number generator really is. And to put it another way, "What
does it mean to say that something is 'random'"?
>From my experience with random number generators, they must be
'seeded' well and they can often produce 'patterns'. Computers follow
patterns (ie. instructions) and even the seeding of their random
numbers is often done with the computer's "wall clock" time. It's
just plain difficult to get truly 'random' numbers out of a computer.
I think that this may be why computers using some sort of random
number generation are often perceived to "cheat". They are not always
going to give results that perfectly mimic the human opponent who sits
across from you. This can certainly be perceived by those who have
little understand of random number generation as a program being
'rigged' and can also be perceived by those unfamiliar with the
specifics of random number generation as "truly random numbers".
So...I'm not expert, but does this solve everyone's dilemma? ;-)
Have a look at the following (I'm pretty interested in this
"atmospheric noise" random number generator...but with all of the RF
we put up in the atmosphere, one has to wonder just exactly how random
even that would be...):
http://www.random.org/randomness/
bcc...@gmail.com
In addition to the "atmospheric" random number generator, here is
another interesting random number generator that relies on the
randomness inherent in radioactive decay. I'd like to stay as far
away from the random number generator inputs as possible for this one:
http://www.fourmilab.ch/hotbits/
Cool stuff, though.
Hardy Hübener
> ..... <many things> ....
No, I don't think that solves the dilemma. The dilemma are the _people_
thinking the random number generators are rigged ;-)
Just three comments:
- If a random number generator is seeded, this is actually a "quasi
random number" generator, because the numbers he is producing are
re-producable. This generator does not depend on PC internal clocks or
anything, just on the seed.
-The numbers this algorithm is producing is not random at all, as this
is produced by an algorithm. Starting the random number generator with
the same seed produces the same numbers. But the pattern of numbers he
is producing is behaving like real randomness.
- Even real randomness (live rolling of the dice) can produce something
that looks like a pattern. So why shoud a random number generator not do
the same. You can roll five 66s in a row ... a pattern? If a random
number generator wasn't able to roll five 66s in a row, he wouldn't be
random any more.
My summary: Most people don't understand randomness and are focused on
their selective recognistion.
Hardy ;-)
--
Hardy's Backgammon Pages --> www.hardys-backgammon-pages.com
bcc...@gmail.com
> thinking the random number generators are rigged ;-)
Oh well, one can only hope. :-)
> - If a random number generator is seeded, this is actually a "quasi
> random number" generator, because the numbers he is producing are
> re-producable. This generator does not depend on PC internal clocks or
> anything, just on the seed.
Correct, "quasi" or "pseudo" random. I'm not sure I get where you're
coming from, however, as the 'seed' for the random generator must come
from somewhere. If you use a table of 'seed' values (as I have done
in the past) in a C program, then (as you mention) you'll always start
with known numbers. Some programs use things, such as the PC's
internal clock (in many cases), to 'randomize' (not the best word, but
you probably get the point) the seed passed to the random number
generator.
> -The numbers this algorithm is producing is not random at all, as this
> is produced by an algorithm. Starting the random number generator with
> the same seed produces the same numbers. But the pattern of numbers he
> is producing is behaving like real randomness.
The key in this point is that the 'pattern' of number behave 'like'
randomness. However, this is why they are termed 'quasi' or 'pseudo'
randomness. Many people believe that pseudo-random number generators
are not necessarily good for things such as gambling and encryption.
This is because these mathematical 'fomulae' have 'periods' or a time
after which the numbers will begin to repeat (granted that this
'period' is often very long).
> - Even real randomness (live rolling of the dice) can produce something
> that looks like a pattern. So why shoud a random number generator not do
> the same. You can roll five 66s in a row ... a pattern? If a random
> number generator wasn't able to roll five 66s in a row, he wouldn't be
> random any more.
I don't understand. A random number generator *could* and *should* be
capable of doing the same thing.
I like the idea of the "true" random number generators mentioned by
the websites I posted, but their true randomness is still somewhat
suspect in my mind. As one site mentioned, you can use various things
about a computer that can be used to generate random numbers but fan
noise, for instance, can produce a repeatable pattern (likely around
60Hz or so, and that's not to mention that AC electrical noise is also
cyclical).
> My summary: Most people don't understand randomness and are focused on
> their selective recognistion.
Very, very true.
Michael Petch
On 8/15/07 8:05 AM, in article
1187186730.0...@b79g2000hse.googlegroups.com, "bcc...@gmail.com"
<bcc...@gmail.com> wrote:
>> No, I don't think that solves the dilemma. The dilemma are the _people_
>> thinking the random number generators are rigged ;-)
>
> Oh well, one can only hope. :-)
>
No, I've been down this road before. You may wish to read one of my
diatribes on PRNG's, much of this is discussed or brought up.
http://groups.google.ca/group/rec.games.backgammon/msg/8e1f99c1c5647c71?hl=e
n&
This article was in response to PRNG's a few months back, but discusses
briefly Atmospheric RNG's among others. Its more of a programmers do and
don't on PRNG usage, but I have received many emails from non programmers
who felt that the non technical aspects of the article was of value to them.
There is an addendum to correct an error about bias but I can't seem to
find the follow-up post.
>> - If a random number generator is seeded, this is actually a "quasi
>> random number" generator, because the numbers he is producing are
>> re-producable. This generator does not depend on PC internal clocks or
>> anything, just on the seed.
>
> Correct, "quasi" or "pseudo" random. I'm not sure I get where you're
> coming from, however, as the 'seed' for the random generator must come
> from somewhere. If you use a table of 'seed' values (as I have done
> in the past) in a C program, then (as you mention) you'll always start
> with known numbers. Some programs use things, such as the PC's
> internal clock (in many cases), to 'randomize' (not the best word, but
> you probably get the point) the seed passed to the random number
> generator.
>
See my article above, there is also discussion on this point.
>> -The numbers this algorithm is producing is not random at all, as this
>> is produced by an algorithm. Starting the random number generator with
>> the same seed produces the same numbers. But the pattern of numbers he
>> is producing is behaving like real randomness.
>
> The key in this point is that the 'pattern' of number behave 'like'
> randomness. However, this is why they are termed 'quasi' or 'pseudo'
> randomness. Many people believe that pseudo-random number generators
> are not necessarily good for things such as gambling and encryption.
> This is because these mathematical 'fomulae' have 'periods' or a time
> after which the numbers will begin to repeat (granted that this
> 'period' is often very long).
>
I don't have the reference in front of me, but if we take the widely used
"Mersenne Twister" that used by Gnubg (as the default PRNG) and other sites
use - the period for repetition is on the order 10 to the power of 6000
(10^6000)
Mersenne Twister has been bashed away at by mathemeticians/statisticians for
some time now. It generates pseudo random numbers that can not be determined
to be non-random by most tests (eg: 12 tests of randomness) which include
patterns, runs, sequencing etc.
From a cryptographic standpoint - Mersenne Twister's weakness is not in the
numbers produced but in the predictability of the numbers (Predictability is
based on actually being able to determine the original seed based on a
reasonably large set of produced numbers). This is a non issue with GnuBg
because one can easily code routines into Gnu to tell you what rolls are
going to be generated in the near future. And Predicting rolls is not the
issue at stake here. This is also a moot point on *most* (Not all) servers
where a pseudo-rng is being used since there is an added element of
semi-randomness in the fact that more than one game will be making requests
of the random number generator.
>> - Even real randomness (live rolling of the dice) can produce something
>> that looks like a pattern. So why shoud a random number generator not do
>> the same. You can roll five 66s in a row ... a pattern? If a random
>> number generator wasn't able to roll five 66s in a row, he wouldn't be
>> random any more.
>
> I don't understand. A random number generator *could* and *should* be
> capable of doing the same thing.
>
> I like the idea of the "true" random number generators mentioned by
> the websites I posted, but their true randomness is still somewhat
> suspect in my mind. As one site mentioned, you can use various things
> about a computer that can be used to generate random numbers but fan
> noise, for instance, can produce a repeatable pattern (likely around
> 60Hz or so, and that's not to mention that AC electrical noise is also
> cyclical).
>
>> My summary: Most people don't understand randomness and are focused on
>> their selective recognistion.
>
> Very, very true.
>
See my article, I comment on this as well. I should point out that GnuBg has
many pseudo-RNG's to choose from, but as well has other mechanisms... Which
include random.org as well (Go to options/dice, and you will see the options
available). You can also generate lists of numbers from other online
services and feed them in as a file to Gnu.
Mike
bcc...@gmail.com
Like I said, I am no expert. I am, however, quickly hacking out my
own (quite cheezy) textual backgammon game (don't ask why...I'm a
nerd, I guess).
I'm currently doing (something like) the following each time a 'roll'
is requested:
int ticks, rnum;
double tmp;
ticks = (int)times(&tms_struct); /* Get system time in clock ticks.
*/
srand(ticks); /* Set seed value...currently each roll - probably not
necessary. */
rnum = rand(); /* Get the random number. */
tmp = (double)rnum / (double)RAND_MAX; /* RAND_MAX is 32767, get a
num between 0 and 1. */
rnum = 1 + (int)(tmp * 6.0)); /* Create RN between 1 and 6 - a die
roll... */
As I mentioned, I am doing this each time a roll is requested. Of
course, I know this isn't up to any server standards (and is probably
pretty crappy to top that off), but I'm just curious what you'd say
about it or tidying it up....
Anyway, do any backgammon games you're aware of actually use or have
an option to use the atmospheric or "true" RNGs?
Michael Petch
Howdy,
On 8/15/07 12:29 PM, in article
1187202540.0...@k79g2000hse.googlegroups.com, "bcc...@gmail.com"
<bcc...@gmail.com> wrote:
>
> int ticks, rnum;
> double tmp;
> ticks = (int)times(&tms_struct); /* Get system time in clock ticks.
> */
> srand(ticks); /* Set seed value...currently each roll - probably not
> necessary. */
> rnum = rand(); /* Get the random number. */
> tmp = (double)rnum / (double)RAND_MAX; /* RAND_MAX is 32767, get a
> num between 0 and 1. */
> rnum = 1 + (int)(tmp * 6.0)); /* Create RN between 1 and 6 - a die
> roll... */
>
> As I mentioned, I am doing this each time a roll is requested. Of
> course, I know this isn't up to any server standards (and is probably
> pretty crappy to top that off), but I'm just curious what you'd say
> about it or tidying it up....
>
I am a code contributor to gnubg, but not a primary developer.
The built in srand in most c++/c implementations is not the greatest but
for what you are using it for it will suffice.
I generally use this calculation for random numbers between 1 and x (in this
case I'll use rand()):
rndnum = 1+(int) (((float)x)*rand()/(RAND_MAX+1.0));
I should point out that your mechanism or the one above has bias, but it is
minimized by the fact that the range of values (1-6) is small enough to not
cause great concern when using rand().
The biggest suggestion I have for you is to call srand(ticks) ONCE when your
program starts. That's fundamentally the only big issue with what you are
doing. Trust me, it makes a difference (So modify your code now!). The rand
functions computes each new random number based on part from the previous
random number. This provides a somewhat uniform distribution of random
numbers. By calling srand each time this uniform distribution is lost and
your random number generator will become significantly less random.
I might recommend you replace rand/srand with something like Mersenne
Twister (There are free implementations one can add to your code for
commercial and noncommercial use) if you want to get real serious. One place
to get C code for Mersenne twister is here (If you feel so inclined):
http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/emt.html
> Anyway, do any backgammon games you're aware of actually use or have
> an option to use the atmospheric or "true" RNGs?
>
Originally TMG among some other sites use hardware devices to generate
random numbers. One I have used in projects a few years back was
http://www.protego.se/sg100_en.htm . It actually was one TMG used
previously.
As for software using other sources, yes. Gnubg for example specifically has
an option to use www.random.org random numbers which as you know is based on
atmospheric noise. This setting is under Settings/Options/Dice/Random Num
Generator.
Many types of software allow you to take a file with random numbers in it
and feed them in as a source for dice. These files can be generated by
pretty much any source.
Mike
bcc...@gmail.com
Thanks a lot for the great info!
I'm fairly new to gnubg, so I'll try out the option to use
"random.org".
Bryan
Guerre
Michael, at least your question managed to shut Monty up - I guess
that's easier than answering a simple question :-)
True troll-like behaviour.
muratk
> Michael, at least your question managed to shut Monty up - I guess
> that's easier than answering a simple question :-)
> True troll-like behaviour.
In this group, it has always been that anybody who refused to be an
ass-licking
(a little different than ass-kissing! ;) conformist has been labeled a
troll...
The only measure of randomness is "unpredictability"!
But the same sequence of dice numbers may not be equally
"unpredictable" in all
situations...
Let's say that one starts using the "Twisted Mersenne" to roll some
dice without
any link to any game. Can you predict the first roll? No. Can you
predict the
second roll? No. Can you predict the third roll? No... I mean, you
have no basis
for making any guesses other than "wild guesses".
Now let's use the "Twisted Mersenne" to roll some dice in a backgammon
game with
a "troll" playing against the "gnudung". Can you predict the dice roll
at any
given turn?
If you are just an "ass-licker" with his head buried in sand, you most
likely can
not. But if you are unbiased enough and a good enough player to beat
gnudung, you
may take a guess based on the assumption that gnudung does cheat and
may be able
to predict the upcoming rolls well over 50%...
It is argued that gnudung always makes the same move in a given
position with a
given dice roll or the same cube decision.
It is also argued that at levels lower than its highest ("grand
master") gnudung
makes proportionately lesser moves, injecting what is called a "random
noise".
However, if the "noise" were truly random, than gnudung would not make
the same
moves in given positions with given dice rolls or the same cube
decisions.
In other words, if the "noise" is not random, how can you argue that
gnudung isn't
cheating??
Yet in other words, if the "noise" is indeed random and gndung doesn't
make the
same moves in given positions with given dice rolls or the same cube
decision, how
can you argue that gnudung isn't cheating??
How about some answers from you "non-troll sick-gambler
slimes"...?? :))))
MK
mont...@lycos.com
play against GNUBG in terms of results, it is still irritating when
sequences that appear "non-random" seem to happen way too often, and
so I can imagine how some people might play GNUBG and also play the 3
point match elimination freerolls on XG/TMG and decide that everything
is "rigged" against them.
I almost always outplay my opponents on the XG freerolls, but
sometimes get knocked out in the first round several days in a row.
This is not uncommon - 3 point matches are not much better than a coin
flip. What I've found is that the 9 point match seems to be the least
amount of points required to "smooth out" the luck factor the
overwhelming majority of the time, so that if you rate at least 200
points higher than your opponent, you win 80 to 90% of the matches (in
my experience).
Because I think backgammon is superior to chess and poker, for example
(though for different reasons), I'm concerned that "newbies" not get
discouraged too quickly. I found that rapid improvement occurred when
I used GNUBG's analysis feature, because it allowed me to detect the
major flaws in my game. This is what newbies should focus on, and not
the "rigging" issue.
donzaemon
If it is the gnubg dice generator that is rigged , why do the bots on fibs
and other places end up with such high ratings , higher than humans , when
they use a different generator ?
Is this maybe a world wide conspiracy with all the online servers involved ?
<ri...@NOTyahoo.com> wrote in message
news:ols0c3p9dfdnnb13q...@4ax.com...
muratk
Anybody home??? Nobody can answer a simple question here anymore??
Do I have to go subscribe to Foolsie's web site in order get and
answer...?? :)))
Nothing more amusing that a herd of dumb asses who thinks they know
what they are talking about... :))))
MK
Raccoon
> >On Aug 9, 11:41 pm, Michael Petch <mpe...@capp-sysware.com> wrote:
> >> Its so difficult.... Errrrrrr.... Don't.... wanna say it.... Feed.... Sits
> >> on fingers... The troll.
>
> >> ;-)
>
> >> On 8/9/07 3:41 PM, in article
> >> 1186695672.961514.56...@g12g2000prg.googlegroups.com, "monty1...@lycos.com"
>
> >> <monty1...@lycos.com> wrote:
>
> >> [removes the stupidity]
>
> On Thu, 09 Aug 2007 21:07:02 -0700, Stick <checkmug...@yahoo.com>
> wrote:
>
> >Are you kidding ... has to be the best humor that gets injected in my
> >day sometimes. I've given up on every explaining anything to him
> >again but still look forward to his posts, always bring a smile ;)
>
> >Stick
>
> If you really want a smile then complete ignore Monty's posts. The
> more you ignore him, the more outrageous his posts will become in
> order to get attention. After a while when no one responds to him, he
> will find another group to troll.
>
> Rich
Yeah! Like www.bgonline.org! Make someone "happy"! Don't be
"irritating." "Bring" a smile to someone's "face"!
Raccoon The reason "many" think anything is
that there are "many" to do the thinking.
The Society for the Preservation of Whoatation Marks
muratk
> Yeah! Likewww.bgonline.org!Make someone "happy"! Don't be
> "irritating." "Bring" a smile to someone's "face"!
I visited that site but I can't quite tell what you incestuous
small circle of sick shit lickers are trying to make fun of..?
Do you have better statistics on expected winning percentages
after the first roll, over 4 billion matches..?
After having reduced backgammon to a single dice roll, why do
you sick gambling slimes still play past the first roll..?
Perhaps you can recommend some BG web sites where I can find
an answer to this also without busting my ass laughing too
hard..?? :)))))
MK
donzaemon
least not convinced any cheating is going on with gnubg.
If you still have your doubts there is a bright light of hope in the fact
that gnubg is open source. That means we can see how it's programmed.
In fact , you can even compile gnubg to use different random generators ,
apparently you can also do it based on a file you put together yourself .
So if you really want to tackle this problem try this : generate your own
file of random numbers and compile gnu to use that file but don't cheat by
memorizing or looking at the sequence , next go and play gnu several matches
using that file for the dice and see if it still kicks your ass , then go
back and verify that it followed the number sequence of the file you fed it.
If this doesn't help you figure out whether it's cheating or not I'm not
sure what will.
muratk
> I think most strong players have seen enough sequences that they are at
> least not convinced any cheating is going on with gnubg....
You don't seem to have clue about the issue at hand, do you...?
You are just parroting back what you have been told...!
Do you have a brain...? Can you think...?
I don't know why did I again post a message in a discussion group
dominated by pathetic herd of stupid shit licking sick gamblers and
the ones who pray on them...
Nice.. Keep going and never mind the question I asked.. You don't
seem to have the mental capacity to neither understand nor respond
to it anyway...
Have a nice day... :))))
MK
donzaemon
Have you ever stopped to think that maybe you're just not comprehending the
information in front of you or how it applies ?
If you want to go through the cycle again , restate the question that is
causing you all this emotional trauma in a clear and consice sentence and we
can and we will give you direct and clear replies.
Repeat this as many times as you need until you begin to comprehend.
See, it's not so hard to get along in here .....
"muratk" <mu...@compuplus.net> wrote in message
news:1188728749....@r34g2000hsd.googlegroups.com...
Marco's Tankless
> I have been tempted to post this kind of position for a long time,
> since it comes up so often, but I have refrained from doing so because
> I don't really care - I just use the software to detect weak points,
> and don't care if I "win" against it or not. However, it is
> irritating when I go back to play a game from a certain position and
> keep having GNUBG roll automatically, and it keeps hitting, time after
I don't doubt the number generators are generating some random dice. However, why is the "seed" always reverts to 8? I can change it to something else, but on the next roll I check again and it is back to 8. Also maybe this is just coincidence but it seems like Gnub gets more "lucky rolls" on a seed of 8, whereas I get either more normal rolls or lucky rolls for myself when the seed is not 8. Hmmmm....
Another thing I do is check who got the most doubles at the end of a match. Most of the time Gnub wins that contest, sometimes brutally. I rarely win that contest. So whereas I get 1 or 2 doubles, Gnub gets 7 or 8 of them. Is this really "random?"
