Computer Dice

Forum Archive : Computer Dice

Does Motif cheat?

From:   Rick Kiesau
Date:   16 March 2004
Subject:   MOTIF BG question

I have been playing on the Motif backgammon site for about four month
now and after exhausting ALL benefit of the doubt, I have concluded that
it cheats.  I can guess with pretty good accuracy what dice it will roll
(mostly when it's in trouble).  If it needs an 6-1, then double sixes,
then another double to make an amazing comeback, more times than not, it
will do just that.  Has anyone else had this experience, and which other
sites have compareble graphics and features? (except for the cheating
feature)  I appreciate any input


Frank Berger  writes:

Dear Rick,

What kind of motivation should Tom Keith have to let Motif cheat?
Here in r.g.b. every fortnight one claims that x, y, or z cheats,
especially bots. But till now, no one has ever come with an evidence.
Even GNU-BG, where you can look at the source, is sometimes called to
be a cheater.

I have made a lot of bot-to-bot comparisons and bots of equal strength
sometimes have very strange results (e.g. laying in front with 0,4
points per game on 1000 cubeless games. If that happens to one on the
losing side, he/she stops playing BG altogether probably). So BG is
very much influenced by luck.

Further it is related to human perception. Everytime Motif rolls a
joker your impression is reinforced, but if *you* roll a joker this
didn't make the same impression (you will often not regard this as
joker). And when Motif plays stronger than you (and AFAIK Motif plays
pretty well) there are more perceived lucky rolls on Motifs side......

ALternatives for you are
  - Fibs to play against Humans
  - or BGBLitz, GNU-BG if you want to play against bots.


Rick Kiesau  writes:

Thanks for the response, Frank.  I don't think there is any motivation
for Tom's game to cheat, and like I said I gave the game four months to
prove me wrong, and I fully understand that luck is a big component of
the game, as I have been playing for twenty four years.  But when the
game consistently and regularly rolls mind boggling combinations of dice
to come back from huge defecits and win, the law of averages doesn't
permit it.  For example;  at the end of the game, when I am on the bar
or in Motif's backboard waiting for a shot,  I have on many occasions
missed more than fifteen times in a row, often while having two or even
three shots per roll, but when he is in the same position,  I will get
hit almost every time, and after thousands of games of that happening
and me often guessing what his rolls will be, I have given up on the
idea that the game is fair.  When I play with another person, there are
bizarre streaks, but nothing even close to that magnitude.  Who knows,
maybe I am just having the worst streak of luck in history.  Thanks.


Bob Newell  writes:

I can't comment on your direct experience, but I do know that one tends to
forget one's own good luck or the opponent's bad luck, but to remember
one's own bad luck or the opponent's good luck.  Human nature.

Couldn't all of the cheating business be resolved by recording a few
thousand games and doing a statistical distribution on the dice rolls?

Tom Keith  writes:

> Couldn't all of the cheating business be resolved by recording a few
> thousand games and doing a statistical distribution on the dice rolls?

In the case of Motif, I recorded 100,000 games and published statistics
of the most commonly proposed ways the dice might be biased.  See

This reduced the number of complaints I got, but didn't eliminate them.
People would now write to say that it's not the *total* number of doubles
that mattered but the *timing* of the doubles.  (Players presumably get
their doubles when on the bar and can't use them.)

I can do an analysis to test that theory, but then somebody else would
come along with a more sophisticated theory.  It's a never-ending cycle.
There is no way to test every possible way the dice might be biased.

Tom Keith
Author of Motif

Rob Adams  writes:

Hi Rick.  In one post you mentioned not knowing how to statistically test
the dice... that is understandable as unless you've studied statistics
you probably wouldn't.  But many people here do... so if you ask
politely they may help.

I'll start.  To begin with you need to rationally decide exactly how
you think motif cheats.  Does it roll too many doubles?  Does it roll
too many 1,6's to escape a prime?  Does it give you 6,6 from the bar
when it has just a 1pt board too often?  You have some experience with
the program now and have some strong feelings that it cheats, but to
progress you need to decide HOW it cheats.  Come up with a good thesis
and write it down.  An example would be "When I'm on the bar against a
3pt board I dance at least 2/3 of the time"  If this were true then
the dice would be clearly unfair since you should only dance 1/4 of
the time.

Ok, then you need to examine and clarify your statement.  For example,
do you mean only when you have 1 checker on the bar not 2 or more?  Do
you mean only a 3pt board, not a 2 or 1 pt board?  Does any other
aspect of the position matter?  Do you really mean at least 1/2 the
time or maybe 40%?  Clarify exactly what you mean before you proceed.

Ok, then re-write your statement exactly... something like "When I'm
on the bar with one checker against a 3pt board, I dance at least 40%
of the time"  Ok, this is a little less serious as you might expect to
dance 40% of the time occasionally, but still if that's true, the dice
would be improper since you "should" only dance 25% of the time.

Ok, only at this point should you begin to gather evidence.  It is
very important that you first identify the problem, then state a
thesis, and then gather evidence.  So... play some games!  Every time
you are on the bar against a 3pt board, see if you dance or not.  Do
this for say 1000 examples.  If you dance less than 250 times, I think
you should be able to see that your thesis was off.  If you dance more
than 400, then you have some good evidence that maybe you are on to
something and the dice may be somewhat un-random.  If it is somewhere
in the middle, then you probably can't reject the notion that the dice
are random, but neither can you reject your idea that you may dance a
little too much.

Statistics can help to give stronger meaning to exactly the numbers
you find, and I'm sure someone here could help you at that point.  But
it is most important that you follow these first steps clearly...
First state exactly what you think is wrong and then clarify exactly
what you mean, and then gather evidence.  Analyzing that evidence may
be difficult, but it isn't impossible and likely will be self-evident

Rick Kiesau  writes:

Hey Ron,

Thanks for the response.  I don't have enough computer knowledge to
gather data or keep track of the games I play.  While I am playing, I
always wish someone was sitting there who knew something about the game,
so I could show them what is going on, because it is quite amazing on an
on-going and consistent basis, and even if I had kept track of all the
rolls that have come up in the last five months of playing Motif, it
wouldn't illustrate my point because it is the the context of the rolls.
For example; very often I will get almost no doubles during a game, but
when I am trapped behind a prime, waiting to get out, I will more often
than not get two or three doubles in a row (usually threes, fours and
fives) thereby ruining my prime.  So, just by examining stats, you could
say I got some doubles, but they were at the most inopportune time.
This happens more often than not.  I know it's hard to believe until you
see it, but it happens all the time.  Motif will also, get a 63 if it's
the only roll he can use to jump my prime.  That should only happen 1
out of 21 rolls, but if he needs it, he'll get it at least 1 out of 3
times.  I can honestly predict what Motif is going to roll when he is in
danger on a consistent basis.  I sit there after having guessed the roll
wishing someone had seen it, all the time.  I play at a coffee shop in
Hollywood a lot, and these things just don't happen on anything close to
the scale of Motif.  Not even close.  I'm not some BG newbie.  I have
been playing for 24 years, and I know there are mind-blowing streaks in
the game, but if someone has one for five months straight, they're


Rob Adams  writes:

> "I don't have enough computer knowledge to gather data "

Yes, you do.  Or if you don't, let me help you there.  Keep a pencil
and paper with you while you play motif.  When a situation where you
wish to keep data comes up, write it down.  That's all there is to it.

> "This happens more often than not.  I know it's hard to believe until
> you see it, but it happens all the time. "

I'll assume by "more often than not" you don't mean greater than 50%
and "all the time" you don't mean 100% as that would be extremely easy
to disprove... as it is I think it's fairly meaningless.  But that's
ok... it's still a good start.

> "Motif will also, get a 63 if it's the only roll he can use to jump my
> prime."

Excellent!  This is a theory that you can try to either prove or
disprove.  A good start.

> "That should only happen 1 out of 21 rolls"

Well, that's close.  It should happen 2/36 or 1 out of 18, but you are
in the ball park.

> " but if he needs it, he'll get it at least 1 out of 3 times."

Excellent!  Now you have a theory you can try to prove... namely that
when motif has just a 3,6 to hop a prime, it will roll it greater than
1/3 of the time... as it should get a 3,6 only 1/18 of the time, this
would indeed show that the dice were biased.

Ok, so now what?  First make sure the thesis is stated exactly as you
mean.  This really won't come up a whole lot... only when you have a
5-prime and motif has checkers ideally placed to hop with that roll
and that roll only.  For instance if you have the 5,6,bar,8,9pts and
motif has its back checkers on the 24pt I suppose this would apply,
but what if they have their back checkers split on the 24 and 23pts.
Then either a 3,6 or 2,6 would work, so it wouldn't apply.  And do you
mean no matter how many checkers are back or only when it is the last
checker?  And you are including when motif is on the bar and you have
the 4,5,6,bar, and 8pts right?  This part is up to you.  Make sure to
state precisely what you mean.  Then write it down before you start
gathering evidence.

Then, assuming you stick with something like this... gather evidence.
Play motif and each time you have such a 5 prime and motif needs
exactly a 3,6 to hop the prime make a note of it as well as of the
roll motif gets.  After say 1,000 such times, see how many 3,6's motif
rolled.  If the number is greater than 333, then you have some good
evidence that the dice may not be random.  If it is less than say 60
then that would be good evidence that your thesis is wrong and the
dice are quite random.  If it is in between, then you may need some
help with the math... feel free to post your results here and I'll try
to help... but generally if it is very close to 333 then there would
be some evidence for your theory, if closer to 60 or so, then poor
evidence for your theory.  I suppose I could do this for you, but then
how would you ever believe my numbers?  You really need to do this
yourself.  Have fun!
Did you find the information in this article useful?          

Do you have any comments you'd like to add?     


Computer Dice

Dice on backgammon servers  (Hank Youngerman, July 2001) 
Does Agushak Backgammon cheat?  (Mr Nabutovsky, June 2000) 
Does BG by George cheat?  (George Sutty, Nov 1995) 
Does Backgammon NJ cheat?  (Greg+, June 2010) 
Does Cybergammon cheat?  (Goto Informatique, Aug 1996) 
Does David's Backgammon cheat?  (Joseph B. Calderone, June 1998) 
Does GNU Backgammon cheat?  (Robert-Jan Veldhuizen, Nov 2002) 
Does Gammontool cheat?  (Jim Hurley, Sept 1991) 
Does Hyper-Gammon cheat?  (ZZyzx, June 1996) 
Does Jellyfish cheat?  (Fredrik Dahl, June 1997) 
Does MVP Backgammon cheat?  (Mark Betz, Oct 1996) 
Does MonteCarlo cheat?  (Matt Reklaitis, June 1998) 
Does Motif cheat?  (Rick Kiesau+, Mar 2004)  [Long message]
Does Motif cheat?  (Billie Patterson, Feb 2003) 
Does Motif cheat?  (Robert D. Johnson, Oct 1996) 
Does Snowie cheat?  (André Nicoulin, Sept 1998) 
Does TD-Gammon cheat?  (Gerry Tesauro, Feb 1997) 
Error rates with computer dice  (NoChinDeluxe+, Feb 2011) 
FIBS: Analysis of 10 million rolls  (Stephen Turner, Apr 1997)  [Recommended reading]
FIBS: Are the dice biased?  (Kit Woolsey, Oct 1996) 
FIBS: Entering from the bar  (Tom Keith+, Apr 1997) 
GamesGrid: Too many jokers?  (Gregg Cattanach, Sept 2001) 
GridGammon: Are the dice random?  (leobueno+, Sept 2011) 
Jellyfish: How to check the dice  (John Goodwin, May 1998)  [Recommended reading]
Jellyfish: Proof it doesn't cheat  (Gary Wong, July 1998) 
MSN Zone: Security flaw  (happyjuggler0, June 2004) 
Official complaint form  (Gary Wong, June 1998)  [Recommended reading]
Randomness testing  (Brett Meyer+, Dec 2010) 
Safe Harbor Games dice  (Michael Petch+, Aug 2011) 
Synopsis of "cheating" postings  (Ray Karmo, Feb 2002) 
Testing for bias  (Kit Woolsey, Jan 1995) 
The dice sure seem unfair!  (Michael Sullivan, Apr 2004) 
Too many repeated rolls?  (Stephen Turner, Mar 1994) 
Winning and losing streaks  (Daniel Murphy, Mar 1998) 

[GammOnLine forum]  From GammOnLine       [Long message]  Long message       [Recommended reading]  Recommended reading       [Recent addition]  Recent addition

  Book Suggestions
Computer Dice
Cube Handling
Cube Handling in Races
Extreme Gammon
Fun and frustration
GNU Backgammon
Luck versus Skill
Magazines & E-zines
Match Archives
Match Equities
Match Play
Match Play at 2-away/2-away
Opening Rolls
Pip Counting
Play Sites
Probability and Statistics
Source Code
Strategy--Bearing Off
Strategy--Checker play


Return to:  Backgammon Galore : Forum Archive Main Page