It is random, but weighted based on the other countries you've gotten. You are more likely to get a country if you've gotten it less in the past. So be happy, you will likely get your favorite countries soon.

Oh, another diplomat! Yay! nice to meet you.

Come on. Austria and Italy are not all that bad.

I know. I've soloed as each. But i just felt like complaining/starting another conspiracy theory.

Hehe. What good is diplomacy without a sinister force at work?

Thats the point of the game. stabbing, lying, tricking, and power gaining are how the game is played. and on the topic, it would be fun to see the mods all play each other in a game where they could all look at each others moves, bringing a fun new dynamic to the game.

I wish the mods would make you leave, or at least make you stop making stupid threads, please keep encouraging them.

if i disturb you with my non offensive comments there is always a mute button. but id save that for profane, obscene posters.

but if im ever ordered to stop i would, but i dont know if they can.

You're just annoying to everyone here, is his point. Quite true, you know.

I try. But really, im just having a few laughs here, its all good fun, not trying to offend anyone. If a majority of the population told me to stop i would or at least reign in a little bit.

And i do try and help the site by donating and such, to make up for some of my more fringe forum posts.

Diplomat, sites like random.org are not truly random, they are only pseudo-random, just like the country selection here (although I've heard that they weight it so that you get a wider variation of countries). The universe is entirely deterministic until you get into quantum mechanics, and all of that may be able to be explained by a complex set of rules that we are yet to find.

You can look in the code to see exactly how it is done, but I suspect the standard php pseudo RNG routine is called. This is fine for our purposes. A weighting function is used to make it more likely that you play a country you haven't recently. This is actually better than random, as it makes it more likely you'll play different countries.


I don't know the exact definition for true randomness, but random.org uses atmospheric noise to see its RNG, so for all but the most intricate experiments, I'm sure it can be considered truly random.

*seed

If anyone cares, this is the closest to true randomness: http://www.fourmilab.ch/hotbits/

As for random.org, I didn't know that, that's pretty cool. I know that all of that can be modeled by a bunch of equations, so that's not truly random either, but it (like using the least significant digits of unix time of a system, which I believe is how most random libraries for programming languages do it) simulates randomness well enough to make any practical application seem completely nondeterministic.

Diplomat, I wouldn't be worried about a few similar draws, something that simulates randomness well will generate long strings of the same result quite often in large samples.

The random selection is also based on what other people have played recently. So, if you join a game with 6 people who have each played one game and all been England, then the odds that you get England are higher than they would otherwise be (assuming you've played a bunch of games) even if you've had 2 England games lately because the others have all been England 100% of the time.

"Diplomat, I wouldn't be worried about a few similar draws, something that simulates randomness well will generate long strings of the same result quite often in large samples."

QFT

I haven't gotten Italy in so long :'(

Yeah, it's called the Gamblers Fallacy when you think, "I haven't gotten England in a while I must be due to come up with England soon." That's not how randomness works.

Except for the fact that the random country assigner is weighted to do exactly that, so never mind. = /

The gambler's fallacy still applies, I think. I went searching through the code for about 5 minutes and failed to find the actual relevant code (probably just because I'm tired), but I think it probably just weights the choice for each country from the users based on the user's percentage of games with that country. Once you've played more than a few games, the percentage changes will stop being as significant, and the selection algorithm will grow closer to true (pseudo)randomness. Thus the countries you've drawn recently have a negligible effect on a country you draw now.

Maybe they should replace the code if it is as bad as ghug suggests. Find some better software or something. From what they say, there's excess money for years to run the site so maybe spend a little.

Spend money on a random number generator? LOL

@ghug

IIRC, the weighting function only looks at your most recent games (the last 5, say).

But it doesn't look at your current games, right?

Which would explain why, if you start a bunch of games at once, it's likely for you to get the same country more than once.

The last 5 recent game algorithm makes sense, because I've noticed my assigned countries are very streaky; when I first started, I had Italy and Turkey a lot, now I'm getting England and France a lot, but I've only had Russia a couple times, so I guess I have some Russian games to look forward to.

I just wish we had the standard diplomacy vote for country... I'll make my own 'random country', thankyouverymuch. ;)

"I just wish we had the standard diplomacy vote for country"

What's this?

On vdip there is an option where you can pick your country. Wish it was here too, useful for tournaments and such.

Technically, there is no random to anything, especially computers. All computer programs work off of pseudo-random methods, usually basing their coding on the computer internal clock.

That's not true. Plenty of things are random.

Also, using the clock isn't that bad for seeding a RNG. Neither is heat of the CPU.

compute milliseconds since the top of the minute and you get a great random seed.

The 'clock' that computers use to seed isn't a 'clock' like you think, but the pulses that go through the CPU (which is extremely close to random).

@Yonni: "I just wish we had the standard diplomacy vote for country"

In standard boardgame dip, you put down (in order) 3 countries you want to play (or is it in order of all 7 countries?). The game master draws names from a hat, and gives you the top country you have selected that isn't already taken. Voila. No randomness, just get the country you want.
Austria and italy are usually at the bottom, so you'll still end up with them, but it's a nicer way to deal with getting the country you want.

'Technically, there is no random to anything, especially computers. All computer programs work off of pseudo-random methods, usually basing their coding on the computer internal clock.'

Yeah, radiation is random, you could easily define a usb device which picks up radioactive decays..

There is also a webservice which offers truely random data (i think collected from a radio antena listening to the cosmic background - thats the black/white fuzzy noise you might find on a (pre-digital)tv when it isn't tuned in correctly.)

However, pseudo-random number generators are close enough for most uses.

I know, fk, but the millisecodns since the top of the minute produces a nice initial seed and if you reseed every loop through the iterator, you can get even more random than just the first seed and the resulting 'random' pattern from that seed.

I wasn't pointing that comment at you, draug... just informing the masses.

ora, that service you speak of:
http://www.random.org/

thanks FK, i've never seen that site, but i did go to the university it is operated from... must have heard about it there...

also, i may be wrong about the source of their randomness... or at last some of the details.

you could also apply the infinite monkey theorem - as a growing number of players play a growing number of games, the probability of SOMEONE to get a long streak of the same country is growing, and of course that single player often calls foul play - I would actually question the randomness more, if streaks never happened.

Yup, regular humans are notoriously bad at accepting the fact that long streaks happen when randomness is in play. At least the experiments where people had to fake a sequence of coin tosses showed most of the people involved never produced a streak longer than 4 (I might be wrong here). The related issue is people making up figures tend to distribute the digits uniformly rather than accordingly to Benford's law (of course there are cases when that's the valid approach, say a distribution of height in population). This fact received some prominence after an article dealing with faking macroeconomic data was published in German Economic Review. European countries that had to show good results at some point in time (for instance before accepting Euro or Greeks in general) had their statistics significantly less obedient to Benford's law that before or after. It's no proof of anything but it might be a good first sign that's something fishy is going on.

i'm prety sure no following benfords law is proof of tampering. It doesn't prove who tampered with the data, but it's a pretty good bet that the data has been altered.

Ora, you are right about the methods. They use background cosmic noise for their randomness...