No Sign-Up Required!
Play Now!
Random number generator glitch
Your random number generator does not appear to be completely random
I took a random sample of 15 games of a&a I've played in over the last month. in 12 of them I have been Japan. Given H0: p = .2 (5 teams implies probability of 1/5 or .2) and Ha: p>.2 I performed a one proportion z-test. The p-value of this test is 3.143259x10^-9. Given a p-value of 3.14x10^-9 there is strong statistical evidence to suggest that the random number used by battleforces to determine which country one plays in axis and allies is biased.
I do not know if this is code that battleforces is responsible for or even can be remedied, but if it can it would be appreciated :D. (I'M REALLY SICK OF BEING JAPAN!!!)
In the last patch they gave you the ability to create A&A games and you can choose which country you would like, or you can choose if you would like to be any ally or any axis team. You should try that.
The random number generator appears to work well in the battles (my statistical analysis of battles has things coming out very well). Now, keep in mind, you're one of several thousand people registered for this site, so that skews things slightly. That said, my Monte Carlo agrees with your results. I get the likelihood of getting a specific country 12 times out of 15 games to be 3.17x10^-7 %.
Are you sure you're not playing games where countries were assigned in at least some of those? I've comleted 23 A&A games since I joined the site. In 2 of those we each picked our country. Of the other 12, I got Germany 7 times, Japan 5 times, USA 4 times, UK 2 times, and Russia 3 times. So it looks like I'm fairly heavy on the Axis side, but not unreasonably so. Is 15 A&A games all you've played in the last month? Could you check to see your country selection for the full complement of A&A games?
One other possibility: There may be some bias in the algorithm based on who starts the game. I don't expect this to be the case, but it's possible.
7+5+4+2+3=12? It seems your statistical analysis is not to be trusted...or you had a brush brief with dyslexia.
I only took a sample....
edit: oh were you talking to brown? I was wondering where those numbers came from...
second edit: yes it would make sense for him to mean 21 as he said he played 23 and 2 were picked the other "12" were so on and so forth...all though I feel statistics is more than just counting :P.
Ok, my comments don't appear to be showing up so I hope they eventually will....I didn't save that count and I don't want to do it again.
Getting a full list would be impractical (i've played near a hundred games of A&A, it's pretty much all I play...) but I got roughly the last 40 games i've played. Every time I've used the default axis and allies where everything is supposed to be random. Japan showed up over 20 times while the greatest occuring country was 7 times USA...
edit: crap.... I stopped on the preview page....that's why...the first 9 were Japan and the next few were a litte more expectable, but then another series of japan, i think japan was like 27, germany only once, russia three times, britain five tiems and usa 7 times...
second edit :(: i also mentioned that i stopped because I noticed team A was axis for the first 40 or so then it switched to allies so I presumed there was an update at that point...and i got tired of counting :P
Hehe...guess what guys. Three new games. One of them I got japan. One of them I PICKED germany and one of them I PICKED allies....go figure :P
Anyways...even though this seems too good to be coincidence I think all that can be said has been said unless someone magically discovers the problem, or it truly is coincidence...