AoE3 Population Statistics
Here you are able to see graphical representations of various datasets that we at ESOC have available, usually with a history dating several years back. On certain graphs you can see a number in the bottom-left corner, this number can be edited to produce a more-averaged graph view than the default. You can also zoom into a shorter time period by dragging your mouse over a portion of the graph while holding the left mouse button (double click to return to full view). Data sources for this page will typically be updated once a day.
Steam population data is gathered from http://steamcharts.com/app/105450. The numbers here indicate players who play the game on Steam, which is not a subset nor superset of ESO player population. These players have simply launched the game through Steam and may be doing anything ranging from playing on ESO, over LAN, singleplayer, campaigns, messing with the scenario editor, and so on. As evidenced by further graphs shown on this page, apparently less than half of Steam players spend their time on ESO.
The dataset used for these graphs comes combined from multiple sources to deliver what is likely the most complete ESO population dataset anyone has. It includes data from http://www.agecommunity.com/_server_status_/pop2.txt (2006-11 to 2015-11), plus a no-longer available cached version obtained from XaKOps (2015-11 to 2016-03), plus data mined from http://www.agecommunity.com/_server_status_ (2016-04 to now, mined through two separate servers as backup in case of outage). There are only two somewhat significant gaps in the data, one (from 2010-07-09 to 2010-08-25) was present in the original pop2.txt source, other of about a month around late 2016-03 and early 2016-04; that was when ESO upgrade happened, which broke the original pop2.txt and triggered the need to implement a new crawler.
In the beginning, we see a swift downfall of The Warchiefs (with the release of The Asian Dynasties), followed by a slow decline of all population over several years. The base game without expansions (vanilla) interestingly remains the most populated one above TAD all the way until late 2014, when TAD finally surpasses vanilla population for the first time. From then on, the gap grows only wider as vanilla falls into obsolency and TAD stabilizes.
On the second graph we can observe daily population peaks, again averaged over each month from the same dataset. In a way, daily peaks can offer a slightly more accurate representation of population changes over time, as they are not as much skewed by ESO server downtime/crashes.
The same dataset, this time aggregated into yearly averages. From this point of view, we can more clearly see how TAD population stabilizes after 2014 and leaves nilla in the dust.
Next, we see daily population peaks averaged over whole years and displayed as a pretty multi-bar chart.
Of interest may be that this bar chart argues an overall population increase from the year 2015 to 2016, contrary to the pure average graph. This discrepancy results from an era of high ESO instability around the times of last known ESO maintainer Ryz0n getting laid off and a problematic server upgrade happening in the observed period. Server downtime drove down the pure average population, while daily peaks retained high numbers.
ESO account data gathered periodically from http://www.agecommunity.com/_server_status_.
This chart offers a peek into fairly interesting numbers of daily average unique users logging into ESO, as well as the sum of newly created ESO accounts, all aggregated by month. It comes as no surprise that the number of new accounts rises sharply during Steam sales.
Data from ESOC Elo Ladder, which are in turn mined from https://www.agecommunity.com/query/games.aspx?g=age3y&max=25&md=supremacy (only data from The Asian Dynasties ESO are included). Patch was often not recognized properly until EP 4.1, which means data from before November 2017 are highly skewed towards low numbers on custom patches. Games of any length are included and games played on XPMOD or other variations of ESOC Patch are not differentiated from "ESOC Patch".
Around April 2018 we can see a sudden sharp rise of ESOC Patch usage, caused by an inability to log into ESO through the official patch.
On the second graph, we see a comparison between combined ESOC patches against everything else.
Distribution of maximum achieved player ranks on ESO, as recorded by ESOC Elo Ladder (its recording started in 2017-06).
Distribution of 1v1 matches versus team matches on ESO, as recorded by ESOC Elo Ladder (its recording started in 2017-06). Only rated matches lasting longer than 2 minutes are counted.
ESOC Patch Launcher usage data. This graph allows us to see the number of players utilizing our launcher for any purpose, including singleplayer games or playing on the official RE patch.