Recently, many people in the forum think that Chinese is an OP civ. Coincidentally, there is a tournament of AOE3 in China is in progress. So, we can gain some data of Chinese objectively in this tournament.
The table below shows the results so far.
It is easy to find that Chinese only have a low winrate.
Now it is time to analyse the data. So, why i said this data is objective?
Firstly, Chinese have a high number of picks. This can effectively reduce the impact of contingency on data.
Secondly, rank of players in this game is very wide, and they are divided into two groups: 1389-1658 and 1664-2200. As shown in the following table. So this tournament can truly reflect the strength of Chinese in 1400-1700 elo.
Thirdly, the rules of this tournament is very close to the ladder, but players know mappool and which one they will use for every game.
The mappool as follow:
For best of 3 of 1389-1658:
Dakota/Pyrenees/ Saguenay/ Scandinavia/ Sahel
For best of 5 of 1389-1658:
Painted Desert/ Hungary Plain/ Florida/Great Rift/ Mongolia/ Saxony/ Yucatan
For best of 3 of 1664-2200:
Dakota/Pampas Sierras/Saguenay/Arabian/Siberia
For best of 5 of 1664-2200:
Gran Chaco/Hungarian Plain/Pomerania/Nile Valley/Vistula Basin/Ireland/Portugal
Fourthly, players from China in 1000-1300elo understand Chinese better than players from other countries in same rank, most of the time. So at least more of them can use Chinese correctly.
As for Russia’s high win rate, that’s because there are two maps in the bo3 mappool of 1389-1658 group that have the inquisitor outlaw. Which improved Russia’s capabilities greatly.
Although my personal dont think Chinese is OP in this patch, but I still have some doubts.
Will it be possible weaker players choose Chinese more? For extreme example 1664 elo VS 2200 elo, no matter what civ 1664 elo chooses he is not able to win.
Another way Sweden, British and Otto are strong in this patch, but their winrate are all lower than average, does this mean they are UP?
For people considering strong civs such as Haud, German, why do their pickrate so low?
I think it is very hard to know how goid a civ is just by watching tournaments, because players donnot #### # civ at random: there is a correlation between how string a civ is and how often ut gets picked. So the main take away is when a civ has both high pick/ban rate and high win rate.
A civ has a high pick/ban rate because the players believe that the civ is one of the best. A civ has a high win rate because it is better than what it played against.
That’s why the hidden cup makes me believe that british are ine of the best, they are always banned. Players are afraid of British more than other civs, so even if it might technically nit be the best, the player concensus is that there are the best. It is very rare after all that players ban useless civs (happens when the player is very confident and rather want to play string civ vs strong civ)
China here having the highest pick rate and a 40% win rate makes me think that players overestimate China, and pick in on maps where they arent the best. Or when the player uses China as a confort pick and get cheesed or surprised by a new strategy from the latest parch changes.
So for me the stats defenitely tell that China is a very strong pick, but not necessarily OP or the best.
The tournament is played in China so Chinese civ is very commonly used. Chinese players play Chinese civ a lot.
Actually, this matter is common in all tournaments. Even in the same group, the gap between the highest rank and the lowest rank is still very wide. In other words, all data from tournament are affected by it, so it’s no need to worry.
For a weaker player, when he has to face a strong opponent, maybe it is better to choose a lamer nation to gamble rather than a Chinese. This may be one of the reasons why the Swedes have such a low winrate up to now.
For Sweden, one of the reasons is because most of Chinese middle-rank players just know they are OP, but never tried it before. And just like what I said, they want to use it to have a gamble with high-rank players, but finally screw it up.
for Sweden, British and Otto, maybe it’s because they were defeated by Russian in map with Inquisitor. After all, the winrate of Russian in more than 80% is the most anti common sense.
This is because the number of Chinese strategies of Haudenosaunee and German is very small. So few middle-rank players will choose them to compete. In addition, the reason why Haudenosaunee are rare is also related to the scarcity of sea maps in the best of 3 match of the two groups.
actually, the civ pick of this tournament is all in random, and there is no way to ban civ.
Chinese only in China can have a high pick rate, which is inconsistent with global data now and few editions ago. It is mainly influenced by the personal hobbies of Chinese players.
The high ban rate of British in hidden cup is mainly affected by mappool.
As you can see, the mappools of Best of 5 are always consist of maps that 2 x no TP, 3 x TP and 2 x sea. The mappools of Best of 7 are always consist of maps that 3 x no TP, 4 x TP and 3 x sea. That means civ with high map adaptability will has great advantages in tournament. And British, just has such characteristics. All in all, this can only show British have a high map adaptability, instead of they are OP.
One of the reasons why Chinese have been picked so many times is also because of the way of packing civs and maps in the tournament. Players don’t need to worry whether they will be targeted, and they can easily pick Chinese on the map with more wood treasures.
lol, I am the moderator of a Chinese forum. Chinese players are really enthusiastic about Chinese civ. However, there are often some novices really furious because their Chinese are defeated by AI, and they will say that Chinese is too weak 
. so now i always advice novices do not play Chinese in their first game
Hi, is possible for me to join the chinese forum, and tournaments in the future? I am not from china
Thanks for the explanations for the hidden cul and the chinese cup. Everything you said makes sense, except the following which I didnt understand:
I get it that there is no ban, but what do yoy mean about “all in random” ? I dont know rhis concept and players are still picking their civ…
For the second point, do you mean that most chinese citizen main the China civ only and do not know how to play most other civs ?
This makes everything even more complicated to figure out whether China is strong or weak. Some may think everyone knows how to play against China and that people still play China in maps it is not that good on, other will think China is so weak that China mains lose against civs that their opponent do nit really know how to play very well.
Anyways, I am always happy when I see the most picked civ of a tournament having less than 50% win rate.
When I started playing aoe3, I think China (along with Germany and Aztecs) was the civ I had the most trouble figuring out what to do. For me this is really not a beginner friendly civ.
These data do not meet any of the statistical assumptions.
It is not normal (the population is only players of Chinese nationality and the number of picks by civilization varies too much, I recommend taking data from more tournaments or if possible taking the data by groups to control the variability), the variance is too high and the data are not independent of each other.
Your statistical analysis is of little value if it does not meet basic statistical assumptions.
Another thing you can do is validate your assumptions with statistical tests, like Levenne and Shapiro-Wilks and with a QQ plox graph.
Based on the data that I have seen this year, without doing the full analysis, I affirm that China has a winrate of more than 55%. Obviously my statement is a fallacy until I do the corresponding tests, but I’m going to be honest with you, I hated statistics at the University…
I’m afraid that players’ understanding of the game will not vary greatly from different regions.
Players from China may not be able to win the championship in the global tournament, but it is still no problem for them to enter the finals. And these players who have entered the finals can pass on their experience to middle rank players. Therefore, the understanding of players from China of this game will not be much different from that of global players. Therefore, it will have no effect on the objectivity of the data for only Chinese players in the statistics.
Chinese data in this year has been nerfed several times, so data that can show the strength of Chinese best should be this edition instead of this year’s. It’s just like you can’t determine whether you can cross the road rely on the traffic flow data five minutes ago. The tournament that can reflect the strength of the Chinese civ in this edition is nothing more than the Hidden Cup and the tournament I sent (we call it Dream Cup). What’s more, I have analyzed the reason why Chinese have such a high winrate of in Hidden Cup in another topic, but you haven’t given a reply.
In fact, being between regions is very important for the statistical study, because your analysis will only be valid in China and cannot be interpolated to other regions of the world! (but I like the variety of elo, some people only analyze the highs elos and that also limits the statistical study!), Believe it or not, there are many styles of play between regions, the fact that there are videos and guides in different languages influences a lot in the paradigm!
More than nerfed, I would say that it was “Balanced”, that is, changes were applied to weaken certain aspects and also strengthen others, I have a post compiling most of the changes, imo China received the same amount of nerfs as buffs, so we can say that its potency is still high.
I apologize, I’ve been quite busy lately and I receive many notifications every day, sometimes I simply forget to reply to messages, much more if these are not direct answers or posts that I follow very closely xD
of course 
The novice group of the next game (elo1350 -) may be held in January or February, and the group with higher scores will be later. We can add a steam friend first, and I will tell you when we start to register
“all in random”I mean player will pick their civ, when they don’t know what civ their opponents will pick… sorry i can’t find a way to express it better, lol.
For the second point, i mean, they know how to play Chinese civ, also know how to deal with Chinese civ–at least they know the law of Chinese strong and weak. In fact, many players in China said they were afraid that their opponents would use Chinese in daily chat. But judging from their actual combat performance, most of these guys are still very experienced in confronting Chinese.
Chinese is truly a civ that is difficult to get into. But players in China can always find a lot of strategies about Chinese civ. This may be why they can understand chinese civ better .
Hey bro, you may not know that, but the player from China, especially PVP players, has paid much attention to the world.
Although PVP players from China are uncommon on discord and forums, they are very concerned about various tournaments. Especially the big tournaments, like the King of the Old World. At this time, there will be a large number of uploads and anchors uploading game records and rebroadcasting tournaments. A lot of commentary will follow. In addition, players participating in the event will also share their experience with players with lower rank. Therefore, the understanding of players from China is not different from that of international players.
Finally, for educational reasons, China did not remove English from compulsory subjects in primary schools until a few years ago. Therefore, the number of Chinese players who can understand English strategies may be more than you think. At least they know how to use translator.
