[ statistics ] [ items ] [ achievements ] [ changes ] [ about ]

So what's all this?

This is my solution to the lack of general server specific statistics. There are several sites that gather data from the armory but most seem to be only providing information about player gear or pve progress. On the other hand, I wanted to know what was the distribution of professions among active players and more stuff like that. That's how I ended up making my own armory crawler to get that data.

What's the "X days" in the stats?

That's a way of limiting who gets counted. If the player hasn't logged within X days that he/she isn't considered active and doesn't get counted. The player activity is base on the "Last Modified" date in armory that seems to get updated almost every time a player logs out normally.

How are the guild averages for achievements and items calculated?

Only the top 25 level 80 characters are counted for those guilds that have at least 25 lvl 80 characters that have been active within the last two weeks.

What's in guild changes?

Again, only level 80 characters. The guild specific changes are as sum of changes within the last seven days.

How accurate are the statistics?

The information for all players is update almost every night when armory is up and responding without too much delay. However, do to limitation in the armory, only players that belong to some known guild (or have been in such during an update) will get updated. The character also need to be over level 60. My wild guess would be that the local database currently has all guilds worth knowing and that should be at least 90% of the population of the server.

Could you add statistics for server X too?

No. The armory seems to have several ways of automatically detecting and banning crawlers by based on ip addresses. Therefore my crawler does things slowly with a single execution thread and takes pauses if the response time from the armory slows too much. Currently updating the information for all guilds and players takes between 10 to 16 hours so there isn't much additional capacity for getting the information of other servers every day.

Technical specs

The crawler is written in Python and the data gets stored in a MySQL database. These pages are generated from the database with PHP that's then cached as plain HTML.

How to contact the author

Vergo @ IRCNet