That.

And also, random generation is what was used before, adn caused the complains in the first place.

I'm sure Ghostmaker will be along soon enough with stats :P

Its actually kind of funny. People didn't like the randomness before. Chances are, anyone who experienced drastic non-equality before are precisely the same people who will experience non-equality as a result of country balancing.

In other words, if you notice you seemed to get Austria and England ALOT in the past, you will now notice you don't seem to get them often at all, making up for your past inequality.

So if you got France a lot in the past, sorry, but now you're going to get Austria alot :P

What they said.

Soon this teething issue will go away.

omg

Here are your country chances kiv:
EFIGART
0.051 0.107 0.124 0.428 0.108 0.063 0.118
So you have 10% chance of being France next time, which will be down from 20% last time. You're most likely to be Germany next

(By the way it is odd that we used to use a simple shuffle to randomize countries, and we moved to a new system because of complaints with a simple shuffle, and now people are recommending a simple shuffle to fix the new system -_- )

Does the new version fix the bug that assigned the same country to people twice in a row?

Not a bug Acosmist. Its randomness.

The players who played Turkey, Russia, and France in one of my games got assigned precisely those same countries in their next game. Clearly something went wrong.

Luck went wrong...
randomness can cause it some times you know...

Really? Couldn't be a programming bug? A quirk?

:/ Occam is crying

I would ask kestas to explain it, but I think he did it so many times before that it shouldn't be necessary.

Note that this wasn't the next game for all of them - but in their next game together, they all had the same country as in their only other game together.

There's the long shot and there's the fact that no program is perfect, but aight. This won't get through to anyone anyway.

The new system has a feature-not-a-bug which means that, if someone hasn't played X very much (relatively speaking) in the past, then they're fairly likely to get X a few times in a row until the numbers balance out. Perhaps that is what is happening in Acosmist's example.

(I wrote my last message before seeing Acosmist's last message, so my message doesn't apply.)

Two of the players have played only two games on here, both as the same country. The "feature" wouldn't do that, because there is no history to weigh. The third player has only played three or four total games, twice as the same country because of this incident.

Pretty sure something got stuck in the analytical engine on this one.

Does this feature start off a player perfectly balanced, or does it seed a history that might skew initial assignments?

It could be a bug or it could be randomness. I don't think based on the example Acosmist provided that we could say that occam's razor favours the bug as an explanation. The simplest explanation would be that this was due to chance.

New players start off perfectly balanced. They should only repeat countries due to pure chance, and chance would be weighted slightly against this.

If this is the only example of such an alleged bug, then I'd say that it probably is chance. But maybe there is a bug; has anybody else seen this? Can we take statistics to see if it happens more often than it should by chance?

>chance would be weighted slightly against this

Sorry, that's not very clear. I mean that the new system will weight the odds so that this is a bit less likely to happen than it would be under the old system (which was pure chance, independently each time).

Improbable events happen all the time (especially those deemed improbable after the event).

If you tell me which players were involved I can tell you their country chances. The game requires that the country you get be random, so there is always going to be the chance of you getting a certain country twice in a row, it's just weighted