Turks are the worst, because they miss Elite Skirmishers, they lose so easily to archer civs.
1 Like
Just checked, in TGs turks have the worst pick rate, followed by Italians.
In 1v1 Italians have the worst followed by turks.
Koreans are not very far. Again I am speaking of arabia (70% of the games are played there).
Personally they are, but people do care more about Italians and Portuguese. It seems just few people are complaining about turks. Still both of these 2 civs are very bad (imo Italians are the worst between the two) so I would support a buff for all of them.
1 Like
No, the lowest pick rate is for portuguese.
Not every civ in this game has to be strong or decent on arabia, even tho it’s the most played map.
No dude, stop pushing this “italians are the worst arabia civ” just because you want them buffed.
Turks are way worse just because they do not have trash. Portuguese are considered by many the wrost arabia civ atm, and we have also Bulgarians, Goths and Indians that are arguably on the same level or a lower level than Italians in arabia 1v1
At 1650+ elo the pick rate changes, but at high level people usually play with random civ. Their play rate is therefore debatable.
And you’re not understanding.
Balance didn’t matter because you both have room to improve and get better.
See you just admitted it. You’ve admitted that you can in fact improve. Which means you aren’t being held back by one civ or another, that you can in fact learn from your mistakes.
So should we valance around low skill players who make tons of mistakes? No. We shouldn’t.
This means nothing at all. Look at low elo winrates and what do you see? A bunch of cavalry and a move friendly civs at tge very top and archer civs at the bottom. Does that mean archers need buffs?
That just a strategy. You could do the same thing with pretty much any knight civ with a good boom. Look at low skill archer civs. Almost always lower winrate then knight or attack/patrol move friendly units.
If you include all the elos, Portuguese are the worst, italians are the second worst by what? 0.02? Still both of them are extremely bad as pick rate
As I said here
I agree, but still they are not that far by Italians and Portuguese.
Ofc I want them to be buffed, as I want for Turks since I mostly play random and I am very said when one of these two civs arrive… similar for Portuguese.
The difference is very small at lower elo, we may agree that does not change a lot. Both Italians and Portuguese are very bad, and hence rarely played.
This is something I agree just partially. Just few civs should be strong, but, especially because it is the reference map, all the civs should be decent. Ofc we have specialized civs in every map, and some are specialized in arabia.
Personally they are just slightly better than Italians and Turks, maybe just because they fit more my playstyle. But in terms of strength, I agree that they are weak, so they deserve a small buff (as Italians and Turks do) to become decent
1 Like
I understand what you’re saying, and I disagree.
As for you, I’ve made my point 5 different ways and you still don’t get it.
For the sake of other people reading this, let me explain that if we follow the logic of @MatCauthon3 balance never ever matters. For people like T90 and Nilli, they could beat any one of their peers if they played with Viper-esque skill, regardless of civs. Because the Viper could beat T90 and Nilli with any civ matchup (talking about RM here, not DM). For people like the Viper and Hera, I think that with perfect god/AI-like micro, plus their usual good macro, they could win any civ matchup against anyone. There is always room to improve and get better.
This interpretation of ‘balance’ is patent nonsence.
Thanks for taking my quote out of context. /sarcasm
3 Likes
The statistical model I would use is:
- all players have an inherent strength (ELO) which changes slowly,
str(player) - all civs have an ELO-modifier associated with them,
emod(civ) - players pick civs in a pattern which is specific to the player and the map,
PickChance(civ; player, map) - the modified ELO is
mod_str(player, civ) = str(player) + emod(civ). This modified ELO formula can be made more complicated to catch more complicated behaviour. Eg, themod_strcould also depend on how good the civ is on the map, and on the civ matchup. - the chance of winning is
P("player1 wins") = S(mod_str(player1, civ1) - mod_str(player2, civ2))whereSis the usual ELO-difference-to-win-chance function. - the ELO a player ends up at depends on their inherent strength and their pattern of civ picks, which can be approximated by
ELO(player) ~ str(player) + sum_over_civ(PickChance(civ; player)*emod(civ))but the true formula is more complicated.
If you want to interpret the player elo distribution for each civ, you have to have or assume knowledge about the PickChance distributions, and you’ll likely assume that the PickChance distribution is the same across all inherent strength. This is a strong assumption, and is probably not true. By looking at the win-rates of players you can get rid of most of the effect of the PickChance functions, which is why I suggested it and why only players who have actually played with all civs could be included.
Of course a full regression could give even more information, but that would also be harder to run.
2 Likes
Haha I had a similar lengthy discussion with mat a few months ago too about civ winrates matter uniformly for all elo ranges. But Ou explained it much better.
1 Like
Well, it’s not like Persian have been kicked out of the top-tier club a while ago and are now average at best.
An unexperienced fighter would totally miss all his shots and maim themselves with the gun recoil, while they could at least look dangerous with the knife. This comparison works to explain why people keep picking ez to play civs like Franks, Goths or Huns, and why they avoid stuff like Chinese, but not to say that civilization choice matters so much regardless of the ELO.
He is comparing Persians to Portuguese. Average to worst.
Same thing I want to say.
2 Likes
Portuguese are a really bad civ at every skill level tho, especially on land maps
1 Like
I do not see too many skills needed for Portuguese. They can play easily both meta strategies with the gold discount, the problem is that discount is not strong enough…
Portuguese do need a further buff, but similarly to other bottom civs, it should be a new bonus.
Devs have tried several time to buff Vietnamese improving their tech tree and civ bonuses, but the fix for Vietnamese has been reached just when they got a the new bonus.
Imo, Portuguese should follow a similar path, a sort of early game help, instead of a further buff of the gold discount or of feitorias…
i honestly like how lithuanians are on 50% eventho everyone consider them not even being mearly on 45% or something