After over 500 games I am done with the BGA version

Forum rules
Please DO NOT POST BUGS on this forum. Please report (and vote) bugs on : https://boardgamearena.com/bugs
User avatar
Tharnduil
Posts: 28
Joined: 22 February 2020, 16:43

Re: After over 500 games I am done with the BGA version

Post by Tharnduil »

Romain, I'm really not sure if your reply is even directed to me as again I stated above I am in fact not trying " to convince anyone who know stat with 'one random game"

I was just adding commentary to the on going discussion as the original poster had asked others experience. I do think it is worth noting, but I do not claim any issue. As euklid did an excellent job of giving examples and info on their process.

Now is it BGA's job to test whether or not the PHP-function random_int() works properly or just use one of the most trusted RNG generators? Well that really would come down to if the community as a whole made a statement against the system they used. (again not even something I am suggesting)

Finally, to euklid, thanks for all the info and links they are appreciated. My only other thought would be that there is the possibility for irregularities to exist in it's implementation of the game Catan. Again not something I am saying is true or isn't true. It would be interesting to have someone run the numbers over a 100, 300, or even 1000 game set just to get an idea of it. Is that the Devs job? Well if there isn't enough vocal outcry for it then I guess not. Still to me it would be interesting to have that data available as something to help waylay any more talk when this conversation inevitably comes up again.

And really, in the end, that is the only thing I am pointing out. This conversation comes about too often whether right or wrong. It would be great to have a data set to show for factual data at what the numbers look like.
User avatar
SwHawk
Posts: 133
Joined: 23 August 2015, 16:45

Re: After over 500 games I am done with the BGA version

Post by SwHawk »

Tharnduil wrote: 18 January 2023, 18:06 Now is it BGA's job to test whether or not the PHP-function random_int() works properly or just use one of the most trusted RNG generators? Well that really would come down to if the community as a whole made a statement against the system they used. (again not even something I am suggesting)
Let me quote the PHP Documentation about random_int():
Generates cryptographic random integers that are suitable for use where unbiased results are critical, such as when shuffling a deck of cards for a poker game.
Are Catan's stakes higher or in need of better randomness than a poker game shuffling? I don't think so. By the way, as stated in the FAQ, should this function have been proved to provided biased numbers, the whole world would have heard of it, because genrating cryptographic grade random number is a much larger use-case than random numbers generated for dice throws for Catan on BGA.
Tharnduil wrote: 18 January 2023, 18:06 My only other thought would be that there is the possibility for irregularities to exist in it's implementation of the game Catan.
So after supposing BGA is untrustworthy, you suggest that the game developer is untrustworhty. Nice comment for someone who donated some of their free time to adapt the game on BGA. Yes they may or may not have been paid to port it to the platform, though I'm guessing they were getting paid. But really, don't you think that BGA admins reviewed the code for such an high profile game ? And in the event they didn't, don't you think that such thread would have prompted them to do so ? And even if such review didn't happen, what would be the point. If the dev didn't follow the implementation, and BGA became aware of that, they wouldn't allow them to develop any more games, so what would be the point ? I mean you'd have to be pretty disgruntled to do such a move, and even so, BGA would have to walz in and a fix would be done in mere hours or days at the maximum.
Tharnduil wrote: 18 January 2023, 18:06 Still to me it would be interesting to have that data available as something to help waylay any more talk when this conversation inevitably comes up again.
The data is available, you can watch through those 100, 300 or even 1000 replays, nobody is preventing you from doing so. But even though such an analysis is conducted, you will have players complaining about the RNG. Because as said earlier, the human mind doesn't handle randomness well, that's why I linked about those two biases. Also we tend to remember when the game worked against us, rather than for us... Another cognitive bias.
Tharnduil wrote: 18 January 2023, 18:06 And really, in the end, that is the only thing I am pointing out. This conversation comes about too often whether right or wrong. It would be great to have a data set to show for factual data at what the numbers look like.

And really, in the end, that is the only thing I am pointing out. This conversation comes about too often whether right or wrong. It would be great to have a data set to show for factual data at what the numbers look like.

What astonishes me the most about this coming up, is that these subjects come up pretty much when dice are involved, but don't come so much when there are cards involved... Even though the same level of randomness is involved...

And this comes up because people have a lot of misconceptions about randomness, not because the RNG is or isnt actually broken...
User avatar
Jellby
Posts: 1349
Joined: 31 December 2013, 12:22

Re: After over 500 games I am done with the BGA version

Post by Jellby »

SwHawk wrote: 18 January 2023, 18:34 What astonishes me the most about this coming up, is that these subjects come up pretty much when dice are involved, but don't come so much when there are cards involved... Even though the same level of randomness is involved...
Probably because cards are more "intuitive". You shuffle and deal a deck, and the distribution of cards may be (will be) uneven, but it's guaranteed that each card will appear exactly once. At least up to that point our intuition is correct. With dice you can roll them 100 times and not get a single 2, and we can't accept that.
Stroom
Posts: 404
Joined: 14 July 2016, 19:10

Re: After over 500 games I am done with the BGA version

Post by Stroom »

Yes. You do not understand randomness and claim all kinds of RNG mistakes. Humans are biased and the things you claim to happen are all biases. Can you predict 3-4-5 dice rolls in a row? From any game situation? No? Then there is nothing wrong with the RNG, you are biased. Stop demanding the developers to explain RNG to you if you won't even be able to understand the explanation.

Just... stop blaming RNG for a game that is heavily based on RNG.
User avatar
Tharnduil
Posts: 28
Joined: 22 February 2020, 16:43

Re: After over 500 games I am done with the BGA version

Post by Tharnduil »

SwHawk wrote: 18 January 2023, 18:34
So after supposing BGA is untrustworthy, you suggest that the game developer is untrustworhty. Nice comment for someone who donated some of their free time to adapt the game on BGA. Yes they may or may not have been paid to port it to the platform, though I'm guessing they were getting paid. But really, don't you think that BGA admins reviewed the code for such an high profile game ? And in the event they didn't, don't you think that such thread would have prompted them to do so ? And even if such review didn't happen, what would be the point. If the dev didn't follow the implementation, and BGA became aware of that, they wouldn't allow them to develop any more games, so what would be the point ? I mean you'd have to be pretty disgruntled to do such a move, and even so, BGA would have to walz in and a fix would be done in mere hours or days at the maximum.


I did neither of those things. The only thing I said is that there is a possibility. To suggest that it isn't possible is just as foolish as saying die rolls aren't random. Humans make mistakes, dice are random, gravity exists.

Some things just are and the fact that there could be an issue with the implementation does exist. Again i am not saying it does or that anything is even wrong.
Some of you on here go from logic sound discussions to acting like you are defending an f***ing fox hole (eh hem stroom!)

I get that all those data points are available if someone wants to go through the numbers and set it up. It isn't that important for me to do so or I would have. I have considered a few times processing a set of my games when it feels that a bunch have seemed pretty far fetched but then I am also rational enough to understand that it probably more has to do with the set of games than a likely outcome over a multitude of games.

But if you go on these forums at least once a month this topic will come up. The arguments will be made and the thing seems to repeat over and over. So again for the final time let's state clearly the only thing I intended with my comment:
1. The original publisher asked a question. I gave my feedback that I can also feel that but also recognize it may have more to do with human nature than an issue on this site
2. it would be interesting to see a data set over a large sample size. Does this need to be done? obviously not, but if it was done then I think it would be less likely for this discussion to come up every month
User avatar
Romain672
Posts: 1014
Joined: 05 April 2016, 13:53

Re: After over 500 games I am done with the BGA version

Post by Romain672 »

The problem is that the claim can be very various, here is a quick list from most commun to less commun:
- the opponent/both players rolled too much double (variants: me / too few)
- we rolled too much a certain number (variants: two/three numbers / too few)
- there was the number X rolled Y time in a row (variants: during a streak of Z rolls there was Z-1 W number / the numbers X & A)
- I started first/second/third or forth/not first during X game in a row (variants: during a streak of Y games, I got those too often/too rarely)
- at criticals moments, the random generation does something weird (define weird, define criticals)
- something weird is happening every game (define weird)

Which leave a total of about ~700 differents claims you could do every game, not counting the weird part.
You could limit your search for the most commun, but you couldn't do all of them.
Stroom
Posts: 404
Joined: 14 July 2016, 19:10

Re: After over 500 games I am done with the BGA version

Post by Stroom »

Again. Human bias and unproven claims and asking for SOMEONE ELSE to do the research. Just... stop. People who claim the RNG is broken, come up time after time and they all eventually end up not understanding how statistics and probabilities work so explaining it to them would not matter. It just keeps happening. You either do the research YOURSELF and prove the claims or go and find another game.
User avatar
SwHawk
Posts: 133
Joined: 23 August 2015, 16:45

Re: After over 500 games I am done with the BGA version

Post by SwHawk »

Tharnduil wrote: 18 January 2023, 19:50 I did neither of those things. The only thing I said is that there is a possibility. To suggest that it isn't possible is just as foolish as saying die rolls aren't random. Humans make mistakes, dice are random, gravity exists.
Actually, there isn't any possibility that the implementation is faulty. Why ? Because that function (random_int) is actually used for cryptographic purposes, which as I said earlier, are way more widely used than dice throws. Should the implementation have been proven faulty, there would be a warning on the function documentation, and anyone would have heard of it, as the consequences are quite dire on financial transactions alone. Tests have already been thoroughly conducted, no by BGA directly, but by the PHP devs, and also a whole host of reputable companies, organizations (governmental or not), and scientific laboratories. So saying that there is a possibility actually implies a certain level of untrustworthiness... whether you want to admit it or not.
Tharnduil wrote: 18 January 2023, 19:50 1. The original publisher asked a question. I gave my feedback that I can also feel that but also recognize it may have more to do with human nature than an issue on this site
As stated time and again, those arguments are always debunked, and always stem from misconceptions and cognitive bias...
Tharnduil wrote: 18 January 2023, 19:50 2. it would be interesting to see a data set over a large sample size
No it wouldn't. There is a whole host of studies about vaccines and their efficiency (and I'm not talking only about more recent installments), but it doesn't prevent people from discarding those studies... Cognitive biases have a strong pull on the human mind...
User avatar
Joëlle_
Posts: 34
Joined: 10 March 2023, 08:28

Re: After over 500 games I am done with the BGA version

Post by Joëlle_ »

[/quote]I didn't wanted to do it initially but since you insist.
I took the % of chance to roll exactly those number of time those numbers for each roll possible (per example there is 35,26% chance to roll exactly no 2), then I added all of them, which gave me 1,589.
And I did 3000 simulation of 2x37rolls with libreoffice.

I find that 90 games out of those 3000 (3%) was below that number.

Remember that the thing you checked was done AFTER you see your game, and it's a pretty weird one. You are not looking for number of doubles or one specific number being rolled too much or two few, but you was looking for something weirder.


But I don't like my post, so I did something else.
I search for the third most lowest number between those 11. It give 6,286% which was your number of 9s.
Out of 3000 others random generation: 22 were below that number so would be even weirder. 11 were equal (so their third number which 'was rolled the most weider' was 9 or 5 and a total of 7 times).
That give 1.1%.
But again, I was looking for something you claimed after your game.

If instead I do something fun to only check the number which was first in 'rolled the most weirder'. Here it would be 10 with 2,11%.
With 3000 simulations, I got 696 which was more unlikely that your game, and 89 equal.
That give 26.2%.

So if I'm choosing the thing we are looking for, I find 26,2. It look pretty much a random number between 0 and 100, isn't it?


Please be precise in your claim. List all the unlikely event you think of. Then come back in 100 games, and show us all the game which got one of those events :) And then we could do something with it.
[/quote]


PwaaAHAHHAH!!!! :lol:
WOW funny stranger I love you
Rufasu
Posts: 18
Joined: 04 March 2023, 07:31

Re: After over 500 games I am done with the BGA version

Post by Rufasu »

I think the only way to solve this problem is to make the code public.
Once the code is made public, we can have a constructive discussion.
Post Reply

Return to “Catan”