I put together some quick-and-dirty civilization statistics for the second round of Thalassocracy. I thought it would be a good idea to have this data both before Armenians and Georgians were able to be played in the tournament, so these are the stats from before those civs worked on the new build correctly. Obviously nothing here is definitive as there are a ton of factors influencing performance, but I thought it was neat so hopefully some of you do, too!
1 Like
Good to see these stats, but I’d also like to see civ picks and bans. That would reveal what players think are good civs for these maps.
With how low the sample sizes are, the data is pretty weak. However, some of the civs were played more than others, and we can see some hints at likely trends from that. Bengalis with 11/14 wins look like a winner in this (probably because hp regen matters more against melee ships than it does against ranged ships that focus fire). Portuguese also seem to be strong, but sample size is on the small side. Dravidians seem to be doing well, but I’d bet the confidence interval includes 50% (I’m not calculating it though). Italians seem to be the most picked civ, but their win rate is a bit below 50% (reduced dock tech prices along with tech discount reduction is probably more impactful than the addition of new naval technologies). Still, it’s close enough that the confidence interval definitely includes 50%. Vikings, Malay, and Persians also got picked a fair bit. Persians do seem to be performing poorly, but again, low sample size.
But take any conclusions with a lot of salt: the fact civs are picked means that it’s unreliable data even before considering the tiny sample size.
I thought magyars with free forging would be better :(, thats what no eco bonus does i suppose.
Im a bit suprised by the rankings of some of the more common water civs not being as good as I’d expect (malay, italians, Wu, Vikings)
Why? And why no one picked Armenians? They are one of the top 5 water civ right now.
There was a bug in the tournament build which caused negative resources to be gathered after researching techs at the mule cart. That was fixed after round of 64, so the civs are able to be played now.
So they changed Dock and University but somehow Mule Cart is affected?
Great effort in gathering this data. The win % should be adjusted to account for seed gap. Like a win by seed 1 vs seed 64 should get very low weightage while a win by seed 32 vs seed 33 should get close to 1 and a win by seed 64 vs seed 1 should get close to 2. This way you’ll get stats which indicate which civs are only winning when used by higher seeded player and which are too OP that even the lower seeded players are winning often with them.
1 Like