Page 1 of 6

Are the rolls stacked in the computers favor, absolutly not.

Posted: Sun Sep 17, 2006 11:55 am
by Jonas
There seems to be an interesting effect in backgammon that folks feel like the computer gets special rolls to its benefit.

I'm here to tell you that not only it's it not the case, but that ethically we wouldn't do that. That just not how we roll. :) (sorry couldn't resist)

Anyhow, it seems that just about any backgammon implementation has suffered though similar accusations, and their authors have come to their creations defense.

So beyond standing on one foot, blindfolded, and doing a jig to help sell the point that we're not' horrible mean people out to get you, I present some interesting post on "Computer dice" in backgammon that you may find revealing:

It's a good read about randomness, and psychology.

have fun

Posted: Mon Oct 16, 2006 10:27 am
by FredrickVegas
It has taken me sometime to find this forum but well worth it. I have been playing backgammon for about 35 years. I rarely meet my match, till I bought hardwood. Your AI does cheat. It rolls exactly what it needs to win. Not every game (which surprises me) but more often than not. I have played more games than I can count in my many years of playing and I am an odds player, meaning even with random dice, certain rolls over time have more odds than others. Your AI knows that and uses it to it's advantage. Also, when playing in XBox live, I have noticed that when a player hits the A button twice in his roll (using 360) in a certain time frame, the dice come out doubles at least 6 out of 10 times. Other players online have tried this and agree. Did some programmer put in a back door for his own game play? I read your other forums on this subject but still believe what I see and experience. Not looking for a conspiracy theory but the facts. You may not even know if the programmer never told...

Posted: Wed May 02, 2007 2:43 pm
by Kurt Lang
I have to agree completely with FredickVegas. Playing on Easy is the only way to play the game without wanting to throw your computer from the nearest cliff. After playing hundreds of games on Normal, I really wish I could get a refund. The rolls the computer gets are not just convenient, they go far beyond mystical. At the same time, the rolls you get are so bad so often, it makes you want to scream.

I can't tell you how many times I've had the computer's stones off the board with only one spot to get back on with, and it does so in no more than two rolls every time. While if it's you trying to get a stone back in play, you'll be lucky to get back on in less than 6 to 8 rolls, even if there's more than one spot open. Also during that time, the computer continues to get an unbelievable number of doubles and rolls that block any opening you have in very short order.

If you have any stone unprotected, the computer will hit you with almost 100% certainty. Especially if the computer is coming off the bar and you have even just one unprotected stone in your home row. While you on the other hand, get numbers that can't touch any unprotected stone the computer has open.

In short, there is a very SEVERE lack of true randomness in the rolls of the dice. So much so, that I really would like a refund. The game is absolutely no fun to play.

Posted: Thu May 03, 2007 12:25 am
by kissmad
I've been wanting a Harwood Backgammon game for ages but due to the "cheating" I will not be purchasing it.

Posted: Thu May 03, 2007 12:14 pm
by Jonas
There is no special rolls for the computer, and we've entirely rewritten the dice rollers, not because it was wrong, but to dispel the rolling issue.

The sad truth is unless we stack the rolls in favor of the player, this perception will continue. We've spent countless hrs looking into the perception, and thats all it is, a perception.

We remember the extraordinary and forget the mundane. The more you look for rolls we think we see, the more we see them. Kinda like when you get a car, and then suddenly you notice the same style car all over the place. Or it always seems like what ever line at the store you pick is the slowest. Same thing, we remember the interesting (or frustrating) stuff and we just don't remember the uneventful stuff.

Posted: Thu May 03, 2007 12:57 pm
by Kurt Lang
Hello Jonas,

I know that's how you feel, and I don't blame you. Any programmer knows of at least one randomizer function. You then expect the computer to give random results. But in this case, after keeping track (sometimes on paper) how often the computer got just the roll it needed to win, and the junk you got that would guarantee you lost, you couldn't help but see the pattern.

The very good news is now that I've downloaded the new version, there is a drastic improvement in play. You can't believe how different it is from previous versions. I no longer feel like playing any game on Normal or above is simply a lesson in frustration.

Years ago, I used to have a card game that acted very believably on one computer, but would cheat like mad on another. It would do ridiculous things in Bridge, like dealing all of one suit to each player! And it would do this every 50 deals or so. I can't imagine the odds of that ever happening even once in real life. Cribbage was no better in that set. The cut cards would almost always double the computer's count, while you would get nothing time after time.

Anyway, I just thought it be fair after writing the above to also give praise where it was deserved. The new version allows me to play this game without feeling like the computer is using loaded dice.

Posted: Thu May 03, 2007 1:01 pm
by Jonas
Glad its feeling better now. Enjoy! ;)

Dice Are Biased - Statistically Proven

Posted: Sun May 20, 2007 10:13 am
by From Ohio
I wish Silver Creek Entertainment would spend the time developing in a Backgammon game that would play WELL, rather than play LUCKY. Regardless of the commentary from Silver Creek, the game makes some poor opening moves and has runs of simply improbably luck.

For example, if the computer opens 6-4, it plays 8-2, 6-2, making a point on the 2 pip, which is too deep to defend. If it opens 5-3, it plays 8-3, 6-3 which is worse than 13-8, 13-10. These are simply examples, there are many moves when the computer opts to build (or land on) an anchor when leaving a blot with the intention of building is better.

Despite these errors, the machine wins 90% of the time. For instance, today I was ahead. I had a computer piece on the bar and both of us had built primes on the 1, 3, 4, 5 and 6 spaces. I had a lousy roll and had to put a blot on the 7 position, which opened me to a 2-5 (or 5-2) shot. I grumbled because I figured the computer would roll a 2-5 - which it did. A 5.5% shot.

Then I couldn't get in on the computer's 2 spot for the next 11 rolls. The chance of me getting in on any roll is 11/36 or 30.5%. The probability of me staying out for 11 rolls is 30.5% "raised to the 11th power", or 1 in 461,326.

If you average 50 rolls each in a game (this is a reasonably long game), you would expect to see this happen about once in 9,200 games (or on average pop up about the 4,600th game). I have only played a couple of hundred games with the computer so this is not very likely.

Get this - the chance of the computer rolling a "2" followed by me not being able to get in for 11 rolls is far worse, 1 in 8,387,745...

I have also documented a NUMBER of times when I can't get in when there is one spot open - for 8 consecutive rolls. By the same logic this only happens 1 in 13,160 times. Again, once every 263 games - but I've seen this happen a dozen times.

The machine cheats.

If Silver Creek would like me to provide a Hypothesis Test asking what are the chances of seeing this behavior and having honest dice, I can provide another post. It takes a little time - but the probabilities are much less than 1 in 9,200 games or 1 in 263 games.


Posted: Mon May 21, 2007 4:34 pm
by Kurt Lang
Sorry to say, but I have to report back that things have not been fixed. The new algorithm changed things slightly. Such that it felt "different". But overall, From Ohio has hit the nail on the head. The game still cheats like a bandit.

As he mentioned, if you get stuck on the bar with one open spot to get on, you'd better do some serious praying to the deity of your choice that you can get back on the board in less than 8 rolls. Where if the computer only has one spot to get back on, it will do so on either the first or second roll, almost without fail. Not only that, it's almost always a doubles that it rolls to do so.

I could go on and on about how predictable the game is. I'm not kidding in the least when I say that I can accurately tell you what rolls I or the computer is going to get in a given situation more than 70% of the time. This simply shouldn't be possible for myself, or anyone else.

It's almost as if you left the old algorithm in and the game uses either the old or new method at the beginning of each game. Some games are not so hard to believe, others bring tears to your eye with how blatantly the rolls are in favor of the computer opponent.

Sorry again, but still thoroughly disgusted with this game.

Posted: Mon May 21, 2007 5:24 pm
by Jonas
sigh, the game does not cheat.

If it did, I sure wouldn't want to make a statement weather it does or doesn't.

So I'd ask that not imply that it "cheats".

We've check and rechecked and the rolls are fine. They we're fine before, the are fine now.

The AI (the part that actually plays the game) and the die rolling are not connected in any way. The AI has no knowledge of what roll will come next. and the Die rolling doesn't have a clue what the AI needs to make it's next play. two two programmers who have never met have these components.

I'm sorry you have to take my word for it, and sadly don't think my word is enough.

Hardwood Backgammon DOES NOT CHEAT. Sheesh.

Sorry to get so grumpy about it, but when you say we are cheating our players, its an attack on our honor.

Is is possible there could be an error, sure. But to say we cheat is going over the line IMO.

From Ohio post 2

Posted: Mon May 21, 2007 8:08 pm
by From Ohio
I'll put some additional statistics together to make my case. One one hand we have mathematics saying one thing, on the other "the Illuminated One" making the same statement repetitively with no proof.

When several hypothesis tests say what happens in Hardwood Euchre is 1 in the billions, the issue will be put to rest.

I have a Masters in CIS, and years of experience programming. This is conjecture, but I'm pretty certain I know how and in which situations the game triggers "luck".

Jonas, can you approve refunds?

Posted: Tue May 22, 2007 7:16 am
by Kurt Lang
Sorry Jonas, but I'm just relaying my experience with this game. I wouldn't be wasting your time or mine if it were simply a matter of being outplayed. But the rolls the computer gets go far, far beyond "fortunate".

I'm sure you and your staff have put a lot of time into programming these games. As such, you have more of an emotional attachment and take it personally when others complain about how the game plays.

While I haven't done much programming in my life, I do understand enough of how it's done. I would imagine the roll of the dice is done in a separate function and the results then passed to the AI (or yourself) to decide what to do with them. Neither the computer or human player should have any idea what random numbers are going to come up.

But that's the problem. They aren't random. Not when I can sit here and call out the rolls of the dice before they are thrown so often. I have seriously thought of setting up a video camera and recording a couple of games along with commentary. You would be astounded.

Re: Are the rolls stacked in the computers favor, absolutly

Posted: Wed Dec 26, 2007 6:47 am
by Wanderer
Jonas wrote:There seems to be an interesting effect in backgammon that folks feel like the computer gets special rolls to its benefit....
I'm here to tell you I don't believe you. I have absolutely ZERO confidence in what you say is true. I've played this site for a long time and the longer I play the less confident I am. It is simply impossible to get the runs of doubles and 2 out of 36 chance rolls that your randomiser throws so regularly. I've taken it on the chin long enough and I now give up in disgust. I hope all who stay enjoy themselves but I'm off. Bye.

Re: Are the rolls stacked in the computers favor, absolutly not.

Posted: Sun Jun 29, 2008 8:31 am
by Death_Row
I use Zone Alarm for my desktop firewall. Whenever I start any of my games... I have to give the OK for any of them to contact the Internet (why is that). If I didn't have ZA I'd not know the games are attempting to reach the Internet. I hope you guys are monitoring the dice rolls with the information being collected from people. I don't think your random number generator is a true RNG. I think if you monitor the states you'll see what people have and are still talking about. I've never seen so many doubles and just right rolls in my life for the bot. If you're monitoring anything monitor your dice rolls (RNG)


Re: Are the rolls stacked in the computers favor, absolutly not.

Posted: Sun Jun 29, 2008 3:12 pm
by Death_Row
The RNG is out of wack. Why don't you guys collect information (Stats) to see if there is a difference between PCs. I just started a game where bot got 4 - 5 doubles back to back. It seems the faster one click when the bot finshes the more doubles are thrown. There may be a correlation. What kind of RNG (algorithm) are you using? There is something wrong!!!!! collect the stats I think you'll agree. I actually stop playing the game because of this very issue. I start back playing and it's the same thing.