Anyone play there? Anyone have any comments?

I see the forthcoming IIBGF World Cup is to be played on there. I have registered and played a few games to try to get used to it. If you think Fibs dice are bad..................etc etc.

Something pretty strange going on there though.

Anyone had any similar experiences?

The standard of play is much higher on GridGammon than it is on Fibs, with a number of world class stars to be seen in action. If fibsters play there, they can expect to have a rating about 140 points lower than their normal fibs rating. Shouts isn't quite as lively and their tournaments less interesting. They don't currently have a league.
Good server though.

I haven't the faintest idea how a random number generator works, nor do I care much, but I do know that on either server there is no way to forecast what number will come up next with a greater degree of accuracy than plain guess work. There is no pattern. Anybody who tells you different is talking through his hat.
Do you seriously imagine that Stick or Kaz or Mochy or Wells or any of the other mega stars that we see play on GG would play on a server with defective dice? It's laughable.

Gentlemen,

I am no newbie - though often play like one so comments like 'it's not the rolls but how you play them' are wasted on me, and verging on insulting.

I don't know any of the people you have mentioned, dorbel, so couldn't comment on what they would or wouldn't do.

The point I was trying (poorly it would seem) to make, wasn't that the dice were 'defective' or indeed 'unfair' - just, possibly, different. If different random number generators provide equidistribution of rolls over differing timescales, then is it possible that the 'patterns' of, say, doubles, jokers etc etc will be different on different servers. I agree that it would appear random to any human, but, for instance, gridgammon may be prone to throwing up more consecutive doubles than Fibs or vica versa.

I would add that, while I could expect ridicule from the likes of dorbel, I am dissappointed in Tom.

Regards,

Garp

The four gridders I named are four of the best players in the world, the first three of which I thought would be instantly recognisable to anybody who played backgammon. The fourth, David Wells, earns a comfortable living as a travelling money player at $500 a point minimum. They are good judges of fairness.
A random number generator doesn't provide equidistribution of rolls over a timescale. If you think about this, you can see that if its time scale was 36 billion rolls, then as it began to reach the end of its list, it would have to artificially increase the ratio of those numbers that up until then it had underrepresented. It would also mean that once a number had reached 1 billion, it couldn't appear again!
All random generation means is that the chances of any number appearing on any roll are exactly 35 to 1 against. At the end of 36 billion rolls this means that some rolls will have appeared less often than 1 in 36 and some more. It would be astonishing if one occurred exactly one billion times!
Neither GridGammon nor Fibs has any recognisable pattern to the rolls. How can we be sure of this? Both servers have their share of dice conspiracy theorists, often claiming to have observed their particular bug over thousands of recorded games, but when offered the opportunity to bet on their theory at favourable odds, they invariably decline!
Ridiculous theories attract ridicule. That's a fact of life, so don't take it to heart garp. Take the time to learn how to play. It will be far better spent than trying to discern patterns to dice rolls that don't actually exist.

I agree that it would appear random to any human,

That is actually the main flaw with this line of thinking. Humans are not 'built' to appreciate random. A survival technique is to look for non random events, patterns and predict what might happen.

They approach everything they do this way. That is why every single dice generator has been accused of cheating by humans who have played, usually fewer matches.

At a table with real dice, they will fixate over which dice to use, how to throw them, whether someone is finding a way to cheat because their rolls appear flukey...etc.

At a Vegas tournament, one of those big names [I forget which, donz would probably remember better than me] has a quirk that after he loses a game in the match, insists the dice are redistributed [which most players would only do at the beginning of the game].

Humans have a very odd relationship with random, and therefore dice..online or in real life. It is really a very small proportion of them that truly understand random..and that to take 2 dice and roll 44 12 times in a row, exists in the field of possibilities, therefore if you do it enough times, it will happen at some point  

The main way to get some understanding of this basic human trait..which is not a failing btw, but what has ensured the survival of the species.,..is to look at the gambling industry. Those who are frustrated trying to understand it, and those who become addicted and lose everything..to those few who truly understand it and the fact that the only right side to be on is to own the casino...

In terms of establishing if dice are truly random or not - it really isn't possible for one single person playing a few games every day - you need a computer recording all rolls and matches for millions of events.  To spare yourself having to do all that, and to understand that every single dice generator has been drawn into question at some point...read this..

http://www.bkgm.com/rgb/rgb.cgi?menu+computerdice

Since you are one of the good people on Fibs, Diane, I will not reply.

sh**!

But seriously - let me put this point in words of one syllable for dorbel and tom (I tried to get an algorithm to make words less than one syllable but failed) - I had a quest-ion. I think it was val-id, Ab-out diff-er-ent ser-vers hav-ing diff-er-ent dice.

Dorbel made the point that the dice MUST be totally random on Gridgammon, not because they use the Mersenne Twister or any other mathematical theory - no!  The dice on Gridgammon are, officially, random and fair because some guys who are good at the game play there.

That's good enough for me and I will stop searching for God now too, if these guys say it's good.

I will close this 'debate' now - from my side anyway.

Almost all online bg sites use a pseudo-random dice generator. The pattern of spread of rolls differs from the natural type, just check it on google. In a figurative presentation the natural random spread looks like a soft mosaic, compared to the pseudo random that looks like a rough mosaic. As far as i know only the server of "sorrytiger" uses rolls coming from another server that uses a  natural generator based on earth's vibrations or something.

In my opinion there is no significant difference between the two for the purpose of having equal chances Vs your opponent.
Assuming of course the dice generator has not been hacked.There are so many cheaters on online sites, particularly in yahoo, playok even on Microsoft's msn zone, play65 (may it rest in peace ) that I cannot exclude this possibility.

Anyway coming back to the point of discussion, no it is not pattern of spread rolls of the dice generator that gives a regular Fibs player an advantage over someone coming say from gridsgammon and vice versa. In my opinion it is the style of play that dominates a specific site.

FIBS for example has too many bots.They play excellently of course, by always making the best move that will give them maximum chances. Human players at FIBS try their best to play similarly, so i would say there is a specific style of playing at FIBS. Similarly there is a different style of playing at every other site for example at playOK there is what i call the "Romanian" style.At play65 people tended to play a lot of "arabic" style. Perhaps the various styles of play  could be the subject of a newa new discussion.... The point is that a human player need some time to adjust to a site where players play a particular style, and during that time he or she will lose a lot of matches vs players of equal strength.

Furthermore the what i call "fibs style" which is based on playing like the bots -always trying to make the move that will maximize your chances, lacks a very specific element: That of human psychology.
You cannot imagine how much confusion you can cause a human opponent by playing alternative moves, or setting up TRAPS.
The scope is to win isn't it? And you can win Vs a human by judging his way of playing/his estimated response/his psychology.

You seem to revolve all your arguments on the issue of world class or expert players dorbel while at the same time forgetting that these people are actually a minority. Surely there is no distinguishable style of play among those top players regardless of the site they play, but I insist there is a distinguishable style among average players. There is no style at the lower classes either, other than waiting for the doubles to come 

It is an opinion coming from a subjective observation. We are all allowed to express our opinion in here, just like you did , even if that is based on our personal subjective observation aren't we?
Would you be kind enough to tell me which bg software ever tells in its analyses of any "styles" of play?

Of course expression of opinion is allowable, desirable even. I only asked if you had any evidence to support it.

Backgammon analysis programs don't of course include an assessment of style, probably because it doesn't exist. They can only measure what they see. The point that I was trying to make is that having spent more hours than any human should, looking at humans and bots play backgammon, is that I haven't seen any signs of these styles that you say exist. If this is because I am being obtuse, please tell me what I should be looking for! If they are obvious to you, it shouldn't be too hard.

I would advice you to join all available sites have an open mind, don't rely so much on bg software and find out yourself. Try to figure out why for example while the bg software evaluates your level of play by 90% of the times being higher to much higher than your opponent and you still lose. You will be surprised.