How many games for specific ratings?

I just started Age of empires around 4 month ago and played around 50 single player games and 400 team games ( of which maybe 100 were DC games).
I just got to around 1250-1300 elo in single player.

I was wondering, how many games does a normal player need to get from 0 to… Let’s say
And 1900?

I’m really wondering, because I think the climb from 1250 to 1900 will be maybe some thousand games.

I know the progress of players depend on many things, but:
Sometimes I see players with over 1000 games and ratings below 1200. So I’m wondering if they are just playing extremely casual? Or is it normal to need 1000 games + for 1200 elo?

I’m just wondering how long does a player who watches streams, is talented and wanting to work alot on his mistakes needs for, let’s say 1400 elo. And comparing that to the average player. And that to a very casual player who plays only for fun.

Thanks in advance!

1 Like

I started in January, have a few hundred games, and hover between 1300-1400 Elo. I watch tons of content and study every update, and have pretty much memorized every civ bonus.

It takes time, practice, and most importantly game sense (which is my weakest area, I think, and I’m constantly working to improve it). Every game is different, and every player learns differently. Might sound like a cop-out, but it is true.

Remember that even 1400 is top 5-10% of the competitive 1v1 ladder. There’s obviously a much higher skill ceiling than this, but it’s very good in the big scheme of things.

Keep learning, and watch your recorded games for mistakes. You will improve at a consistent rate if you continue applying yourself to improving.

The number of games isnt related to Elo. Elo is based on your skill level. Some players are talented players and will reach the top in no time. Others just stay around the middle. Therefore some users will stay even below 1000 elo forever, while others will be 1900 within 100 games. There isnt really any rule of thumb.

1 Like

Another aspect is that the game itself is not only 1 year old.
There are players, who have been playing for decades. So their DE-account will show a small number of games, but the real number is in the thousands.

I dont think you can get to 1900 (or even close to it) with only 100 games. You still can climb pretty fast though, as you can see with player like MarineLorD, who is a SC2 pro, started AoE2 after DE came out and now is already quite high on the ladder.

I think 1300-1400 (don’t know the exact number) should be pretty easy to reach, if you have access to good ressources/coaching. After that it get’s more difficult, as it’s not so much about improving very simple mistakes anymore.

I think someone like Capoch, who recently came back could have reach 1900 in 100 games. But he played this game a while back and was still playing similar games at high level. These kind of users are really exceptions. Normally you wont get to 1900 in under 100 games.

People have been looking at this before:

From 500 until around 1700 there seems to be a decent correlation of games played vs elo rating. After that it becomes pretty damn random.

1 Like

I’ve gone from about 800 to 1100 1v1 ELO over the course of 300 games.

But I think it depends a lot on how much you’re trying to improve. I could have been significantly better if I’d bother to rewatch my games, and I would be significantly worse if I hadn’t practiced hotkeys and practiced with the Art of War scenarios and build order guides.


This depends on your wins/losses, lets take a pro start playing with his new smurf account as example. The rate/elo will start after 10 games, if the pro with his smurf account won his first 10 games i think he will start with 1400 elo in the ladder, from now on the more game you win in row the faster you will go up, the pro with his smurf can reach more than 1900 or even 2000 with 100 games less or more a little depend on his wins in row

As far as i know streaks doesnt influence your elo in any way. I havent noticed something like that anywhere.

So i would love to hear your source for this claim. Could you proof this game?

Not sure about AoE2, but many games have mechanics where your MMR and your displayed elo do not match, the MMR beeing “ahead” of the ELOnumber. So, if you win a few games in a row you get to stronger opponents quickly and therefor will gain lots of Elo. This mechanic is used to ensure people get to play players on their level quickly, without having huge fluctuations in the displayed elo.

If that was the case for AoE2, the winstreak would indeed make you gain elo more quickly, although its not actually the streak that causes this.

1 Like

Yeah, i do understand how such system could work, but based on my observations it looks like AoE2 DE doesnt have such system. There is not a hidden elo which is used in match making. It just use the visible elo which is used.

So if someone claim we do have such system, then i want to hear there proof, for example a dev saying something about this subject.

I mean if you win in streak the more points that you will get, but if you just keep win and lose and win and lose it will become harder to go up

I read this post a while ago and found it quite interesting. I recently wondered if the number of games a player had to play to reach a specific ELO has changed since DE came out because I suspect the skill level to have risen quite a bit.

Therefore I wrote a small script to plot some statistics I extracted from data collected via’s API. I think the plots might be interesting to others as well, so here they are:

Number of games by elo:
The number of ranked 1v1 games the average player played at a specific elo is shown by the line marked “50 % quantile”:

The expected number of games (probability = number of games played by player / total number of games played) is way higher than I expected and does not align with my experience, so take that one with a grain of salt.

Elo by number of games:
The elo an average player has after a specific number of games (50% quantile):

Elo Distribution:
The elo distribution given the approximate number of ranked 1v1 games that were played:

The statistics are not filtering out secondary accounts, and the number of played games are only games in ranked 1v1 mode and not considering if someone has played a lot of unranked / ranked team games.


Amazing data, thank you.

What is really surprising to me is how weak the correlation is. There is a learning curve, but its incredibly flat.

Just one thing i dont understand, you have a lot of datapoints at 0 Elo, with very low number of games. How does that even work, considering you start at 1k?

1 Like

I think my histogram plots look a bit misleading: at the X axes I have no tick label for the leftmost column; that column is actually summarizing all players with 700 Elo or less, and I do the same on the rightmost column for 2500 Elo and above. The “0” next to it belongs to the y axis, so the data points actually belong to players of 700 or less Elo.

On the right we only have the top players with more than 1k games and maybe some secondary accounts with a couple hundred games.

Another reason for the corellation between the number of games and the Elo might be that players who are already good at the game tend to stick with the game; maybe a lot of the players who are 2k now might’ve been 2k after 100 games already. I did not analyse how the Elo of particular players changed over time, that would be really interesting as well (taking into account the entire sequence of Elos and number of games per player (i.e., taking into account a player with 1000 games 1000 times, once for each game they played and the Elo they had at that point in time).

1 Like

I have to say: nice stats.