change the dice generator.

[Henk21]

The suspicion is dice manipulation. It's being reported all the time. So something is going. Let's break it down.

So what people are experiencing is dice manipulation where a certain player receives disproportionately advantageous rolls while statistically much less likely.

The counter argument is basically: it's just an anomaly and that happens in statistics. Proof it! But nobody is crazy enough to put in the effort to actually do that.

If 50% of all games are being manipulated you would slowly see some anomalies in the aggregate data. But if manipulation is obvious but rare, you won't really be able to detect it by looking at statistics of multiple games. That's because all those other games have no manipulation and even things out.

What I would like to know is how the dice generator works. Because I suspect the dice generator has a set of dice rolls he chooses from. Is it possible a hacker can read the dice rolls beforehand and knows the distribution? Or can even pick the distribution for the specific game? Swap distributions with other games in the future?

[Jontia]

Why would any developer do that? Rolling d6 is a common function, you just use the function. It's a one liner that there is no need to make any more complicated than it is.

There's nothing to manipulate or pre-view because providing those leavers would be big chunks of unnecessary code that would objectively make the implementation worse.

[Romain672]

If 50% of all games are being manipulated you would slowly see some anomalies in the aggregate data. But if manipulation is obvious but rare, you won't really be able to detect it by looking at statistics of multiple games. That's because all those other games have no manipulation and even things out.

That can be checked fairly easily.


Ok we got a good claim which is the first step.
Here is the next step and I will help you, define something specific which can easily be checked by statistic.
Here is my proposition:

"In specific games, a specific player will have all of their starting number rolled much more than all other numbers during the whole game. (if a number is present twice, that's counted the same as all other numbers for simplicity)"

Tell me if you agree on that part.
Once you agree, let me see if you play enough... That's fine. Show me here one of your future game, and if there is really anomalies, I will be able to tell you how far from what you expect that game was.

Look like if what you just said was true, we should find some really rare games far from what you would expect.


Just note already on that claim, that on every game you already have 3+ players, which give you already 3 groups of numbers possible. So after 3 games played, even finding something weird which should happen ~12% of the time would be extremely expected.

[Jellby]

What I would like to know is how the dice generator works.

I doubt that. Maybe you do, but 99% of those who say similar things (I can also make up statistics) don't give a damn how it works. No amount of explanation or code will convince them.

[gmrsmrt]

...
But if manipulation is obvious but rare, you won't really be able to detect it by looking at statistics of multiple games.
...

That's a pretty empty statement.

What does it mean "obvious"? If it is obvious it can be proven just quote the table, define a GOOD observable and compute the probability for that outcome. If it turns out that you need to play zillion of times for such an occurrence to happen you have a robust argument for claiming
that something suspicious is going on. If it turns out that something similar can occur several time a day on BGA there is nothing obvious at all.

[Henk21]

What does it mean "obvious"? If it is obvious it can be proven

If someone has villages and later cities around the number 11 and that number is rolled 20 times and 7 is rolled just 5 times during the entire game. You can't prove anything with that. But if these situations occur very regularly as reported by many players, you have a suspicion. But no proof. You can only proof it once you can proof they rigged the system. But no one is willing to investigate because "anomalies occur, nothing to see here".

[Romain672]

Just send me the game.

You can 'prove' it, if you find a game where with my calculations (which take most things into account and reduces biais) you find less than a probability of 1/10k for your future games, that will be weird.
If you find multiples of them about the same thing, and in a short duration of time, that could require further investigations (i read one example in another website where a specific group of players, don't remember which group but potentially with a name which contain X or is early in the alphabet which got slight wrong randomness happening to them, but those cases are extremely rare).
If now you find probabilities of 1/1 000 000 000 (again about your own games, not the game of a random other player who never talked in the forum), then something will be wrong for sure.

But the reality is that you either send me games about various differents things, and either I invalidate them, or I need to take into accout the 1000+ differents things which could be weird for you (and fastly, any calculation don't make much sense).

Or like I said multiple time here, you said something which can be checked fastly. I give you an example earlier, but you seem to really want to take into account numbers after the start of the game, which is much more tricky.
I have something easy, it seem you are saying that the number rolled in the second half of the game should be much more different than the numbers rolled in the first half of the game? If so, that's another thing we could check. (again, if you answer that post and tell me it's exactly that, and some week later show me some games where this is happening, it will be far stronger than staying silent, finding a single game where this is happening, and asking calculations, because already there is 3 differents claims I said here, and I didn't mention if it's about all rolls, or the rolls of a specific player).

[Jellby]

If a player could rig the system in such way (i.e., make sure the numbers they want come when they want them), the highest elo in Catan would be much higher than 500.

[AmongMajesticRuin]

Every single person who complained and has been challenged to show evidence did not show any evidence,

show me the code and I'll tell you where the mistakes are.

[AmongMajesticRuin]

best and only way to determine if the dice have issues is to aggregate all rolls from every single Catan match on this site..
There is no doubt in my mind that if someone did that, the breakdown would be almost exactly
2 /12- 2.78%
3 /11 - 5.56%
4/10 - 8.33%
5 /9- 11.11%
6/8 - 13.89%
7 - 16.67%
One game will never give you any insight.

you're funny. absolutely hilarious, actually. Let me propose a hyperbolic situation to illustrate what I see: Suppose we have a variant of Catan that rolls 1 die. Each number has an equal chance of rolling. Each game consists of 50 rolls. Here's the results:

Game 1: 50 6's and no other number rolls.
Game 2: 50 5's and no other number rolls.
Game 3: 50 4's and no other number rolls.
Game 4: 50 3's and no other number rolls.
Game 5: 50 2's and no other number rolls.
Game 6: 50 1's and no other number rolls.

You would claim, looking at the totality of the rolls in all games that the die rolls operated according to statistical probability. The sane people look at these games and while there IS A MINUSCULE CHANCE this could happen, it's more likely I get imbued with cosmic power and ascend to Godhood.

What I see in many, MANY games is that 1 or 2 numbers are RED HOT!! while the others roll way, way below expectations (like my example above). This is because there's an error in the way they're seeding the random function. Of course these 2 "hot spots" change every game and over 100 games the numbers "look right" but inside any individual game it doesn't "feel right". I think this is the core of the complaints.

When you see a game where 3's and 4's out roll 6's and 8's by a margin of 4-to-1 it COULD be bad luck, but when you see it over dozens or hundreds of games you realize it's just bad programming. Kinda like the person in your game group who doesn't ROLL the dice, they just sorta drop them out of their hands. EVERYONE shouts "roll it again and roll them RIGHT"