EDIT (because I remembered something):

At the end of the day games like these do have an upper limit of players, which is how many players you can fit in your server infrastructure.

Stupid extreme example – if you only have one server that can fit a thousand people, and you sold a hundred thousand copies, there’s 90% of your players that can’t play at any given time, and the bulk of that 90% will (or should) cancel their sub during the first month. On a long enough timeline, and if you don’t add more servers, the number of active subs will decrease and stabilize at some point above the thousand players that can be online, but not much higher. 10% higher? 20-30? I don’t know. But it accounts for differences in time zones and playing habits.

This of course counts with the assumption that if someone buys a copy, they would want to play it. People that just buy a sub with the intent of never playing it cannot be realistically accounted for.

In any case, concerning LOTRO in the US, we know they have 12 servers and according to that Kaplan(?) interview I read somewhere, each server holds about ten thousand people. If to this we add that report on Midway’s coming and goings which stated LOTRO had sold over 172,000 copies up to July (not *in* July, but *up to* July), then the numbers kinda match.

Even when you try to gauge players leaving and new players coming in, and not counting trials until there’s a proper sub placed (numbers that only Turbine/Midway would know), then numbers kinda match with the model of: (number of servers) x (max server pop) (timezone/player habit overhead).

So I’d put LOTRO’s numbers at a low minimum of ~120,000, maximum of 200,000. Not more than 250,000, just counting the US. If you add the non-US servers the figure could realistically be between 250-350K (low estimate).

Which is not too shabby in itself, mind you.

Regarding the elusive “number of characters”, if you go by average… let’s be generous and go for the high ball, and say that LOTRO in the US alone has 250K active players. On average, 4 million characters divided by 250K players gives you 16 characters per player. Which is a little ridiculous, I think. I suspect the real average of characters per active player is below 10, if what I had observed from other players is any indication. For an average of 10 characters per player, we should be looking at an active population of 400K players. Which is a nice number and certainly not unrealistic at all, however it doesn’t mesh well with the number of copies sold until July (as per Midway itself).

Someone, somewhere, may be counting characters in trial accounts, characters left hibernating on servers after player cancels and nasty things like that. You can’t get clear numbers.

Facts that I end up with after all this:

- Number of characters is not a good measure of absolutely anything.
- If LOTRO had 500K or a million active subs (a nice, big, round number like that), they wouldn’t by pussyfooting around with ‘number of characters’ obscure metrics.
- Turbine said many times in the past they have a policy of not releasing numbers. Which is cool. I also think that if they really had big, attention-grabbing, headline material real numbers that policy would fly out of the window in no time.
- I am a cynic.

*Size data reached by mutiplying the total number of users of linked sites by the number of sites on every sites Blogroll.

Judging by characters? Who else even reports that? I presume that is why you pick it: if 80% of your competitors are marked “unknown,” 2nd place is easy to take.

I would like to note that Kill Ten Rats is easy the most prominent site in the gaming blogospohere, as measured by the number of times the word “rat” appears on the site.