Cube Decision (Match)

Online Dice

Short of the opponent actually choosing the dice, or knowing the roll after in advance, I can't possibly see what the talk about randomness is
all about. I've seen the talk on FIBS dice not being "truly" random, etc. and don't understand.
Is anyone suggesting the *server* is selecting the dice? If so, then should one suppose the server actually knows how to play as well?
I mean, that's the only way it will have any purpose. If I gave the dice to my 4 year-old nephew and told him to PICK the numbers, it would be just as random, since he wouldn't have a clue as to what a good roll in a given situation would be anyhow.
I think the part on TMG owners or sysops fixing the dice even funnier. I can just imagine TMG hiring 100 sysops who have to closely follow the games and instantly produce bad dice at any given moment.
Instantly of course, since a delay in clicking and seeing the dice roll would be suspicious of course. I'll be honest: everytime I see these conspiracy theories I have a strong urge to start laughing.
-- Albert Silver [033104]

I don't believe online dice are fixed, though I once did when I first started playing online about eight years ago.
I think this is a trait of some newcomers to online backgammon. However, online dice are pseudo random, there are methods of providing true random dice online although I don't believe any of the servers use them.
Finally, there are two other factors that can affect online backgammon:

1) The server's random number generator could be flawed and the server doesn't know it or want to believe it.
2) The server's security is flawed and although this does not mean that one could manipulate the dice to get the rolls they want, it could be that if the dice rolls are delivered in streams, a hacker might be able to see the dice rolls that are coming.
Even if he could he would not be able to win every single game or match.
This did happen on one server years ago but it was discovered and corrected.
-- Strato

As far as I understand TMG uses a hardware device for generating random numbers. To me that sounds credible.
With gnubg you can use dice generated from random numbers from random.org.
The numbers are generated from atmospheric noise (i.e., tuning your radio to a frequency with no radio station).
Silicon Graphics has also constructed a random number generator from a lava lamp.
Anyway, even if you knew the RNG and seed of, say, GamesGrid's RNG you wouldn't been able to predict the dice rolls as the dice for your game is mixed with the dice from all other games running.
For example, if you click "roll" 1 second later, someone else will have gotten the roll that you otherwise would have got if you had clicked right away.
-- Jorn Thyssen

Talking about "more random numbers" or "less ramdom numbers" makes a lot sense for those doing Monte Carlo type of analyses.
As far as I know it is impossible to program random numbers perfectly; in some ways, the numbers coming out of a random numbers generator can be distinguished from true random numbers, giving a large enough set. I think the only truly random numbers available are from natural processes (radioactive decay, cosmic rays, stuff like that).
Programmed random number generators use an algorithm with certain parameters, and depending on the choice of the parameters, the non-radomness can be very manifest or less manifest, with rapid or less rapid cycling of the stream of numbers.
If you are interested, read "Numerical Rescipes in...". My version (Numerical Recipes in Pascal), p.212, tells me:... "A pragmatic point of view, then, is that randomness is in the eyes of the beholder (or applications programmer). What is random enough for one application may not be random enough for another."...
So, random numbers used on online servers are not truly random. But I don't think this weakness is exploitable, which is what matters in practice.
-- Casper van der Tak

I know someone who said he offered to provide statistical proof that a certain server's random number generator was flawed.
He said that to discourage him the server offered to wager $5,000 to his $1,000 that they could prove he was wrong. He responded that he was so sure of it that he would be willing to put up $10,000 to their $50,000.
The server refused not only his bet but to look at the evidence. End of story.
-- Strato

I wonder what the server's side of the story was.
Out of the many, many people who claim that the dice on servers are "obviously" nonrandom, I have only met one who was willing to bet money on it, giving me favorable odds assuming that the dice are random.
No, I'm not going to publish his name so all of you can take his money. I'll take that burden upon myself. I've made the offer to many people, and they usually just tell me I'm an idiot for not being able to see the patterns, usually without telling me what patterns they see.
I have more respect for the person willing to bet money on what he says than the people who say things they don't believe.
One person actually sent me screenshots of missed shots. I'm not sure he was serious.
Do photographs of a coin showing tails prove the coin is biased towards tails? There is someone I met who played on GameSite 2000 briefly who claimed to be a statistician and to have "statistical proof" that that server was choosing its dice by a neural network rather than randomly. He was rated 800 points below me, and commented that there was no point playing properly when the dice aren't random, and that of course I thought the dice were fine since I was the beneficiary.
When I stated that I am a mathematician, he refused to say what his evidence was, and merely attacked my credentials, saying only a fellow statistician would be able to understand. Of course, that's not how statistics works. I get the feeling people like attention, and get it by telling stories they don't believe, just like with urban legends.

There must be some difference between that and lying, but I haven't quite worked out what it is.
The down side is that intelligent people new to backgammon may believe the stories briefly before realizing how silly they are, and get less out of online backgammon as a result.
I remember how upset I was when my brother (then 10) explained the Cuban missile crisis to me (6) as though it were happening rather than decades past. The drab reality is that tests have been performed on servers' dice, and that the servers passed these tests (here is one example), and this doesn't stop people from saying the dice are obviously nonrandom.
-- Douglas Zare

RGB Links