Computer Dice

 Testing for bias

 From: Kit Woolsey Address: kwoolsey@netcom.com Date: 10 January 1995 Subject: Re: TOO MANY DOUBLES :( Forum: rec.games.backgammon Google: kwoolseyD2617p.60o@netcom.com

```My personal experience from playing quite a bit on FIBS tells me that the
dice rolls are as random as one could want.  Let's face it -- doubles
will come up 1/6 of the time, which means that if you play enough (or too
much as is the case with most of us), you are going to see sequences of
consecutive doubles.  What happens is that we tend to remember the
*remarkable* sequences, while forgetting about the normal sequences where
a bunch of doubles don't occur.  In this way, it is easy to convince
oneself that more doubles occur than there should be.  This is called
selective memory, and can distort a person's picture of reality.

I am not an expert on random number generators, but I would guess that if
there were something wrong with the FIBS dice generator the results would
be more screwed up than just producing more doubles than there should
be.  I'm sure there are some random number generator experts out there

However, it is always possible that something is wrong.  In order to
determine this, it would be necessary to set up an unbiased test.  To do
this, one would have to state IN ADVANCE the bias which they think
exists.  For example, one might state:  I think FIBS produces more
doubles than it should.  Then one would determine a match to start
counting, and from then on keep track of EVERY dice roll (no cheating)
for a large number of rolls, the larger the better (decide in advance how
many rolls you will count).  For example, suppose you decided to count
1000 dice rolls.  It is clear that the average number of doubles one
would expect if things were truly random is 166 2/3.  I'm not much of a
statistician so I can't say a whole lot about confidence levels and all
that stuff, but if more than 200 doubles turned up in the trial I would
think there would be cause for concern.  So, somebody who really believes
that there are more doubles than usual should try such an experiment.  By
the way, for 1000 trials I'm betting on under 200.  Anybody want to bet
over 200?

Kit
```

### 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)
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)
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)
Jellyfish: Proof it doesn't cheat  (Gary Wong, July 1998)
MSN Zone: Security flaw  (happyjuggler0, June 2004)
Official complaint form  (Gary Wong, June 1998)
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)