Just out of curiosity because I had not the slightest idea what "adikus" is and Janet was, let's see emotional on reddit, I googled, found Adikus is an app and installed it on my iPad and played 2 games.

It's usability is really excellent, the board could be a bit nicer but it is a real pleasure to play. Because the AI plays the move pretty quickly it is a bit difficult to judge every move, but I saw enough moves:

as expected, it is like the hundreds or thousand of other Backgammon apps, the AI is abysmal. He makes some points when he can, but has not the slightest idea of strategy, value of points, timing etc.

The dice rolls overall were o.k. I haven't record it and for s short time Adikus was really lucky when I came in with 2 checkers against a 5point board and hit 3 moves later. The sample is to small for a statistical relevance.

Nearly all testimonials were complaining for the dice. Given how bad it plays it sheds a light of the average playing ability.

@Janet: Use one of the 3 good AIs (Extreme Gammon, GnuBG, BGBlitz) and/or read a good book about BG.

Just once I’d like to see someone post something on threads like these that sounds like this:

BG app/website/bot X is so RIGGED. Every game I find myself losing in the race by 30 pips or more, then I roll like 4 big doubles in a row and win. Every time my opponent has my last checker trapped behind a prime and I need a 61 or 62, guess what roll I always get - yep there it is; I get my joker and win. No way that could happen randomly. The other day I had two men on the bar against a five point board, and I rolled the double I needed to escape! No way all this can happen randomly. Why can’t they just make a site/app/bot with honest dice?

Hey, if a site really is rigged, it is just as likely to be rigged in your favor. Yeah, I know, if these folks actually saw something like this happen (and they most likely do), they would just chalk it up to their superior play.

On game site with money play, there is at least a certain amount of motivation to cheat (give the fish some better dice to keep them playing, or whatever thoughts they might have) so I would never say without some investigation that the dice are fair.

When no money is involved, where is the motivation? But sure the alternative to rigged dice is that I have not the slightest idea how to play. Deep scarves in the ego....
What really makes me mad is that so many complaining about Adikus. So many have no idea about the game....

Dear Marc,

First of all, I don't believe that Backgammon Galaxy (BG) uses rigged dice. I can't find any logical reason for you to do so. However, your attitude all these years is not the right one. Just because you have no reason to use rigged dice doesn't mean that all online backgammon applications don't use them. For many of them, rigged dice are the only way they can profit from the existence of their application.

When did you go out and report them? When did you demand that every online backgammon application be required to provide statistics for each game and player? When did you demand the shutdown of all applications that clearly cheat players and drive people away from backgammon? I suppose you did nothing about all of this; instead, you discredit anyone who complains about rigged dice, whether it's about BG or other applications.

There are two categories of people who complain about rigged dice in Backgammon Galaxy. The first category consists of the innocent ones. They started playing online backgammon on cheat applications like Lord of the Board (LOTB), and they have been cheated numerous times. When they started playing BG, they assumed that BG cheats just like everyone else because of a normal streak of unlucky games. Although their conclusion is wrong, they should be treated with sympathy because they are victims of theft by others.

As for my own experience with BG, I can say that I have lost games that made me want to smash the screen out of frustration, but I have also won games that I didn't enjoy because the dice were outrageously in my favor. The key thing about BG is that both things happen. In contrast, in LOTB, the dice were permanently rigged in favor of the opponent (or the bot), except for the first month after my registration and for one week after the only payment I made. That's why I understand those who complain about rigged dice in BG, although I disagree, because I have also been a victim of cheaters.

The second category of people who complain about rigged dice in BG are agents from other applications. LOTB, for example, has its employees create fake accounts to defame BG, in order to cover up their own scams.

I hope you understand how foolish your attitude is. By only accusing the victims and never the culprits, not only do you maintain an unethical stance, but you also harm your own interests. These scammers have no problem using trolls to harm you. Why do you defend them?

What I would like to see is BG and all the applications that provide comprehensive statistics join forces and demand that the others either add statistics or shut down. They should demand that the WBF stop supporting suspicious applications in exchange for sponsorships.

For example, LOTB until recently claimed that its RNG (Random Number Generator) was tested without revealing by whom. Recently, they started claiming to have the approval of the WBF. I contacted the WBF, and they confirmed it. However, I also saw that LOTB is a VIP sponsor of the WBF. How can you tolerate a competitor, suspected of cheating, becoming a sponsor of its own auditor? They have no hesitation in defaming you for their own benefit. Why do you pity them?

There is another issue that you need to address immediately in your application to dispel any suspicion. You provide excellent statistics after each game, covering everything. But why do you assume that every player knows what they mean and where they come from? We are not all grandmasters to understand what equity is, the difference between error and blunder, what Luck (joker) is and how it is calculated, and how someone can have better PR and Luck (joker) but still lose, among many other things. Therefore, write an article explaining all of these.

Especially regarding Luck (joker), you can see it immediately after the game, and then you lose it. If I go back to old games, I won't find it. I can't understand why this happens. Fix it.

You say that BG wants to become for backgammon what chess.com is for chess. I would suggest studying both chess.com and Bridge Base Online (a similar application for bridge) carefully. To reach that level, you need to add three things that both of those applications have.

Firstly, many frequent tournaments. Allow users to create tournaments. You have only one tournament per day. This way, you lose players who are not available at that time (for example, it's the time I leave for work). Also, since it is a knockout tournament, half of the players will play only one game and have to wait 24 hours for the next tournament.

On chess.com and BBO, you can play many tournaments at any time, and regardless of the outcome, you know that you will play for several time. Therefore, you need to add more tournaments and include the option for a double knockout. Only with this change, all players will be guaranteed to play two games, and 75% of the participants will play more.

Secondly, support for Android devices. The play.backgammongalaxy.com is a good application, but it is different from the original BG. They even have different rankings. Merge the two applications, along with the tournament support I mentioned above.

Lastly, the existence of a community. Chess.com and BBO have groups, forums, and many other features that allow users to organize themselves, discuss everything, and have their own matches or matches between different groups.

If you consider the above points, I believe that you will soon achieve your goal. However, the most important thing is to support the players and expose the scammers take advantage of them. Backgammon loses thousands of players every day just for the sake of a few cheaters. You must not allow that.

Good luck because I truly believe that you love backgammon and want it to flourish.

Spyros

P.S.
I know that it has been almost two years since the initial response. Even if you don't want to reply, I would appreciate a sign that you have read what I wrote.

I am 70 years old and autistic with a particular relationship with patterns and have been playing backgammon since I was about 10, so 60 years. In all my experience I have never seen real world dice behave the way they do at Backgammon Galaxy.
Now I'm not saying the dice are fixed, or at least I'm not saying that the people who reply to this forum are aware that they are fixed, but it is very peculiar indeed how the dice seem to have an unusual relationship with probability. Runs of extreme luck are freakishly long, intense and common and it's almost as if the "luck" factor, rather than being part of the analysis, is set before the game starts. It may be that a programmer is having a good laugh about this, who knows?
It occurred to me at one point that this may only happen when the player is on a free account and that it might go away when he starts paying but being on a state pension and things being tight, I haven't made the step to find out - but it would explain why the top players don't see these patterns.
As things stand I will continue to play at Backgammon Galaxy because I enjoy it, even with its laughable probability foibles. But I look at it more as a BG simulator than as a real game experience. Some day when my personal economy improves I may pull the switch and see if probability normalises for those who pay.

Have you ever ROLLED real world dice nearly as many times as you can roll them online?

That's why online dice look weird. It's easier to play more and longer, so the low-probability events show up more often.

He mentioned being 70 and playing from the age of 10. Would you really bet that he played more online than live?


Don't buy this. Every dice sequence is unusual and I probably don't remember the four 66 in a row thrown 5 years ago whether online or in RL.

Prove it. if it's fishy it should be easy to prove. You sound like a deranged lunatic until you have proof.

Besides that I feel sentences like "deranged lunatic " not helpful, it would be nice if you quoted what you refer too. That makes answering so much easier.....

Ha! I take it you haven't met many deranged lunatics then? I thought I was being quite measured. I was just saying how things appear to me, personally. It's like the dice algorithm is trying hard, but not quite making it. I'm not saying it's cheating, just not quite right.

Humans are known to be terrible to judge or produce random. Our brain is trained to see connections, even if there are none.

I can't speak for the Galaxy programmers, but they probably haven't programmed a RNG of their own (although BG needs not a very good RNG because the value of a rolls depends from the context) but use something standard like Mersenne Twister or the like.
Those RNG are battle tested so there is no thing like "just not quite right" Their might be additional an error in the usage (I'm neither a Galaxy programmer nor a Galaxy user but it seems to me that they lack some quality assurance) but I doubt.

With desktop programs there is no incentive to cheat (unless your AI is so abysmal that is the only way to win or to cheat in favour of the user: a happy user will buy more often). On a server: it depends. When there is no financial incentive like Fibs or Dailygammon there is no reason to cheat. When especially when opponents might be house bots and there is something to buy there is at least a motivation, but it is possible to check the dice: will you risk your complete investment by cheating? I don't know.

The only thing you can do to is collect data: the rolls, dancing etc and then check that (or ask s.o. to check)

But if I have to bet, I would bet on fair dice.

How do you even define an 'unusual' dice sequence?

From "Every dice sequence is unusual" it can be deduced. Let's take "5-1 2-1 2-2 2-4 3-3 5-4 1-5 3-4 5-4 4-4 1-4 4-5 2-2 2-4 1-6 3-6 5-3 3-6 2-4 3-6"

The probability of this sequence is about 0,000000000000000000000000005 (in case I mistyped it should be 5e-27 or 2^16/36^20) so it is absolutely ridicolous that this dice sequence can occur, it is soo unusual! Got it?

BTW it was produced with the Mersenne Twister and seed 482347

And it has the same probability as "5-1 5-1 2-2 5-1 2-2 5-1 1-5 5-1 5-1 2-2 1-5 1-5 2-2 5-1 1-5 5-1 5-1 5-1 5-1 5-1"

As a programmer, what would your attitude be to another programmer who refuses point blank to publish the RNG seed in play and copies and pastes "we provide AI for Microsoft so have no need to cheat" responses to such requests?

Well I would understand if a programmer (on a server, on a desktop it's easy) says it is difficult to calculate the dice from the seed (because you may post process the random numbers, have a HW generator, etc. etc. ), but to answer "Trust us we deliver AI for Microsoft" let's say it wouldn't increase my trust and sounds like bragging.

- as if Microsoft hasn't enough people to do that on their own, so they don't deliver AI but at most they deliver one very small part
- trust and competence are not related in any way
- it costs nothing to sketch what they are doing

so if you have contact to a programmer and get that response I would start to get suspicious

You're doing the wrong math, blitz. You have to sum over all equivalent sequences. That you specifically got 2s where you saw 2s is immaterial.

Further you have to calculate the probability of seeing one of that set of sequences, over your entire sample size.

Please could you explain more detailed where I miscalcalculated, you may use an easier sequence like "5-1 2-1 2-2" where I would calculate the probability of seeing this as 2^2/36^3

My point was BTW that each sequence of dice rolls (ignoring doubles for a moment) has the exact same probability. Just because a human sees a pattern doesn't mean anything, therefore IMHO it is irrelevant whether I say "how probable is it to roll this sequence now" or "how probable is it to roll this sequence in 100 rolls". The numbers are different 4 sure but I don't see why this is important

You didn't mis*calculate*. You did the wrong calculation. You have to add up the probability of *each* equivalent sequence occurring in your entire sample size, multiply by the number of times you've seen sequences in that set, then see what you get.

Good chance you end up with a much larger number than you expect.

One of the members of BG was showing me how to access stats so I could determine (over thousands of matches) if my assumption of my opponents' doubles off the bar at astounding rates was true or not. I was gathering the data to do so (and ready to either prove that the dice are corrupt or to acquiesce to them being fair), but then Backgammon Galaxy went to this "new format". Now stats are not available to "free members". Very convenient.

Also, the new format is not very good. I've lost several matches because the match doesn't start immediately. I guess the program has to match players and sometimes that takes so long that I had to go on to other things on my PC and became preoccupied. Then when the matches started, I didn't know it and my time ran out causing me to lose without even playing. The chat thing is absolutely ridiculous. No "type-in" chat available, just pre-set words and silly icons. And yes, the dice on Backgammon Galaxy are still as unbelievably corrupt as ever.

On other sites, my rating has been close to 2000 (currently at around 1800). As I mentioned in previous posts, I've played in live tourneys in Vegas, Chicago, and Denver and won money and a trophy in a state tourney. On Backgammon Galaxy my rating hovers between the mid 1400's and 1600's). Many of the matches lost are more than laughable at the turn of events against bad players. Anyway, I was ready to prove or be disproven, but was pre-empted out of that with the new format. So for those who want to berate me and criticize me by saying these dice are "normal", save your breath. I was ready to be proven wrong with the actual stats (which I submit wouldn't prove me wrong), but now no stats are available to me. I stand by my other posts.