All I know is that it looks like the new civs need slight buffs at best and all the crying about mayans is unjustified at this time.
1 Like
Thanks to the autor of this for his work and dedication.
He has demonstrated how winrates are bad for balance… These stats “”“show”"" that Incas, berbers and teutons are top3, Indians and bulgarians are better than Aztecs, Mayans and Vikings… Cmon all of that is certainly not true.
1 Like
Everybody knows that there are cerrtain civ which differ in different maps and elos. I’m not surprised that civs with a bit later powerspikes have good winrates in the taken dataset.
And since I always had the opinion civs should be balanced among all Elos, these Datasets are very good for balancing. Especially datasets which differ the elos, so the developers have the oportunity to tweak civs like chinese which only shine in high elos.
Personal opinions are fine, but only scientific data can verify or bust them. We all are biased. I also thought burgundians were OP first because of the coustilier. And there are still many peonple picking britons in arabia, besides they are less than average there. But because these Players ALWAYS pick Britons it looks like britons would be “average”. But they aren’t. You can see it in the playrates, the higher the elo, the less often Britons are picked and they are always around 50 %. WIth Mayans, Chinese and Incas it is the other way around. Also of course because it needs more skill to master these civs, whilst britons are very one-trick.
So these dara can be vers useful for balance, if you know how to read it.
2 Likes
burgs are not OP, just broken. broken as in badly designed and can swing games they shouldnt, but will usually be weak. that civ just shouldnt exist. no amount of balancing can save it.
2 Likes
If this data is reliable it suggests that the game is in a highly balanced state as some have speculated. Perhaps some very slight buffs to the handful of civs from 44%-48% could be warranted. Although it is also possible that some civs are simply more skill-heavy than others (Cumans, Saracens, etc.) in the sense that these civs have specialized mechanics that require dedication to fully exploit and without that dedication they can lag behind.
Perhaps the most important conclusions is that Burgundians post-incredibly overpowered Coustillier are a little weak and Sicillians are pretty well-balanced now after nerfs.
1 Like
I would like to see this charts replicated after m8ning more data.
Korean winrate confidence intervale is very wide now. After a couple of weeks, koreans scores wpuld be more accurate
Quick, someone lose some games with Berbers. I don’t want them to be nerfed.
4 Likes
that will be super sad… but the best part is according to this, cumans will get buffed rofl… and all them peeps been crying about feudal 2xTC… gonna be even worse off hehe
2 Likes
I think they are actually that bad because many people try to boom in feudal with them and get punished hard xD
2 Likes
I wpuldnt mind if their feudal ram rush is buffed.
But maybe they needs something regarding steppe husbandry
3 Likes
well if they fixed the darn semi useless SL, then steppe husbandry would be cooler… or if they gave them a reason to actually make CA…
but #makeSLgoodagain is my vote
2 Likes
Well “according to this” Cuman can be above 45% winrate just fine.
Also the person who made the document didn’t precise anything about maps, which is quite important.
It was OP cos of coustilier with the double attack vs archers, and the initial Charge damage on the coustilliers. But now it’s completely fine after the changes.
Yes
or at least make steppe lancer a unit again xD
no one even considers them now
i’d start with some small buff to it like
reload time of standard steppe lancer buffed from 2.3 to 2.1 and for elite buffed from 2.3 to 1.9
this would increase the dps just slightly (around 9% faster atk speed for non elite, and around 20% faster atk speed for elite) and would increase their micro / hit and run potential
More noticeably, I think it would make the high upgrade cost for elite worth it, because you get that attack speed (although a jump from the new 2.1 to the new 1.9 of elite will only be like a 10% boost)
6 Likes
Interesting analysis.
Once he has more data, I would like this Reddit poster to split up his analysis across different maps. If e.g. a civ is good on Arena and as a result gets picked a lot on that map, that will have a positive impact on its winrate, even if it is bad on Arabia.
There are more reasons why one has to be careful in interpreting this data. Since most players have a winrate of around 50% by virtue of how the matchmaking system works, this means that if you always pick the same civilization, you will bring the winrate of your civ closer to 50%, regardless of how strong the civ actually is. In the case of this specific analysis, if a civ was picked particularly often by players of around 1150 Elo, then this would have a negative impact on its winrate, since the games of these players only count for this analysis when they get matched up with opponents with a higher Elo.
All that being said, I think it is interesting to see Cumans perform so poorly. I guess it is due to the fact that by design, they were supposed to have two viable playstyles: either go very aggressive (with rams in Feudal Age) or very passive (with a 2 TC boom in Feudal Age), but only the passive playstyle is actually sometimes viable, so they are screwed when they cannot wall up properly.
2 Likes
agreed with all you said
these stats favour higher skilled players(you can see the elo distribution on the same page), which means much more aggressive games, and i think a lot of lower elo (aka less aggressive ) players have been complaining about the cuman boom, where it is particularly oppressive
its one of those weird cases where if you buff them for higher elo, they will become even more oppressive in lower elos… probably similar applies for saracens (to make them more viable for lower elos, they might become too good for higher elos)
but i also laughed at the high teuton win rate… since so many people here said they are weak because they lack husbandry (probably just repeating something they heard)
i think because he factored all the maps as is, arabia is always played more at higher elos, so the stats are more skewed towards that, but definitely not 100% accurate for any specific map type
Someone posted a vid with Q&A with a dev. They say the want to have every winrate between 45-55% for every elo and for high elo as well, where they mainly focus on Arabia. That is already a different group as used here.
Also match ups are important. If a civ was unlucky and match up a lot against a counter civ, then it will be put his win rate down. You also have a look at those things if you really look at winrates. All in all i think we can conclude we are currently in a pretty good state for balance.
2 Likes
if anything its weak, and frankly i still see people complaining about them.
I don’t think it’s overnerfed, after all they only touched the charging stat, everything else, including its cheap cost and fast creation time is untouched. Problem is, it’s hard to resist the temptation of castle age cavalier and ez paladin upgrade, so it sees less play.
2 Likes
i meant the civ, not the unique unit, is weak.
3 Likes