Random Civs aren't so random. A bug?

[Artifex]

Anyone notice when you pick random Civs..90% of the time the Incas, Aztecs, Amercians, Iroquois, and Mayans are in your "Random" game.

Why is this? A bug?

I play Emperor, 8 civs.

Is there a way to truly make it random?

edit--ARRGH I thought I was in Conquests forum...I hate it when I post to the wrong forum accidentally...sorry.

Will Post in Conquests where I meant to.

[teaster]

I usually play Huge maps with 16 civs, though this does increase the odds I have had all 5 of them in my last 3 random games. I thought this was just a strange coincidence but perhaps not.

[pvzh]

Make sure you play with cultural link OFF. If it is ON you will be placed with the same cultural group predominantly.

All civ you listed are in american culture group.

[geniemalin]

With cultural links 'on' I am the Portuguese.

Is there a European civ anywhere around? Not likely.

My nighbours are the Maya and the Mongols. The Incas and Iroquois are the two powerhouses, and Aztecs and America are also languishing away somewhere to the south of the continent.

I'm definitely beginning to agree there's something fishy going on

[ZargonX]

Artifex: I have the exact same problem. Is something screwed up somewhere that is permanently loading the American cultural group :???:

[vmxa1]

Hum, I am seeing a lot of Inca and Mayan's as well. Do some maps with debug to see what it generated. You can see all the civs at the start.

[SlightlyMadman]

This is really bugging me (pun intended). Almost every single game I've played with conquests, I've had the Mayans on my continent, as well as a couple other American civs (the rest tend to switch around a bit, but always Maya). Only in one out of the 10 or so games I've played was I an American civ (Iroqouis).

It's almost as if the game always thinks the human civ is in the American culture group.

It's possible that this only started with one of the beta patches, as I didn't notice it before then (but it still could have been happening).

It would be nice if you could exclude a culture group from the game. I guess I could always mod them out. I'm just sick of always fighting the same civs in the ancient age.

[UnOrthOdOx]

um...

I haven't seen this yet. However, when I went to start my most recent game, I did notice that the game preloaded some of the above mentioned civs in the opponent boxes, and I had to manually check the 'random all' box before starting.

I think one of the conquests , maybe all of them, messes with your settings and loads them as opponents if you aren't watching. Are you sure all the boxes said random before you started?

[geniemalin]

In my case anyway, yes, I specifically made all 11 civs random.

[DrSpike]

Hmm, some are probably just playing in the same culture group and thus seeing the same civs. For those who aren't it seems strange..........perhaps test it, and turn the cultural link off to see if there is a bug. It doesn't need to be modded as it's an option. I haven't seen the problem, but which patch are people playing? I haven't got around to 1.15 yet.

[Catt]

I actually did a test more than a week or so ago and just forgot to post the results (as the thread slipped off the front page). Version 1.15b, standard map, all random, culturally linked starts off. I launched sixty+ "quick starts" on those settings (just over 500 randomly chosen civs) and immediately checked out the space race on F10 to see who appeared in the game, and noted how many times (1) the Mayans appeared in the game, and (2) how many times the American culture grouping civs came up.

It was pretty clear that the appearance of both the Mayans and the American culture civs was right in line with expectations -- so no unusually frequent appearences.

Catt

[vmxa1]

Great then it is a done deal. Thanks Catt.

[Artifex]

Yes the bug was fixed with 1.15. Big difference. I was talking asbout 1.13. It's definently not a problem now. Wow this post is almost a month old.

[Catt]

Someone at CFC pointed out that the bug, if it existed, seemed to appear when "culturally linked starts" was *on.* I had actually started my tests a week or so back with default rules (including culturally linked starts) *on* before I decided to turn the culturally linked starts off on the theory that the setting could skew a limited number of results -- if the human civ happened to be American and happened to appear on many pangaea or continents maps, I felt it would produce more American results than "normal" and vice versus (i.e. less than "normal"). I had already recorded about 20 starts with linking *on* before I started again. I just ran an additional 55 starts, bringing my total to 75 starts which means 600 random civs, and American civs appeared 135 times -- 22.5% of the time. With 5 American civs of 31 total civs, we should expect approximately 16.1% of civs to be American cultural civs. Although 600 trials is not necessarily a determinative statistical sample, the deviation from expected results certainly seems significant -- based on those results, I'd cast my lot with the view that there is something slightly out of whack with "random" civ choices when culturally linked starts are *on,* even in v1.15b.

Catt

[DrSpike]

Careful though.........is your civ also random? I'm not sure how the cultural link works; if the first civ had a slightly higher than population probability of being American that could snowball with cultural linking on.

Can you be more specific on what you did?