Its not my fault that there is just one article available in the
public which talks about numbuer of subscribers of EQ.
And it was meant as an example.

Meldur

I'm also inclined to beleive that the author (who of course is not a
Sony employee, merely a web-columnist, and the figure in question was
only tangentially related to the subject matter so good 'fact checking'
can hardly be assumed) took half a million divided it by 52, got 9615
and then guessed that the servers were supporting up to 10,000 players.

Suggesting that each server only supports 10,000 accounts has some major
inconsistencies. EQ lets each player have 8 characters PER SERVER. Now
very very few people have done THAT. But practically EVERYONE I know has
created a characte or two on at least one other server. (be it a couple
weeks on a Zek, or FV, or a testbuff on test, or a server someone they
met in RL plays on...) sure in many (most) cases those characters are
essentially abandoned, but they are still there, with the accounts paid
in full each month.

10,000 accounts per server with 52 servers leaves almost NO wiggle room
for the number of accounts with characters on multiple servers. I mean,
I have abandoned characters on 5 servers (even post merge). I'd be
counted 5 times under meldurs assumptions, and I'd wager that a very
very large number of accounts would be counted at least twice.

I'd fully expect the number of "sever-accounts" to be at least 50% again  
or even double (or beyond) the total number of subscriptions.

So at its PEAK, with 500,000 subscribers I'd have expected 500,000 -
1,000,000+ server accounts. Which is a *heck* a lot more than 10,000 per
server with 52 servers.

-cheers

If a server can support 10K people, that doesnt' mean it has 10K people
on it.  There were around 50 servers, but it was easy to see that some
were more crowded than others.  10K isn't a number to conjur with, its a
"at most" number.  50 servers with 10K people each, though, would still
be a mere 500K people, yet EQ supposedly peaked out at 2M people.

The merger arguement which suggests that if you take two servers which
support 10K people each and combine them, then current servers support
20K people (I know, its not your arguement, I'm being too lazy to go
back and put this where it belongs) is also incorrect; the servers still
likely support a maximum of 10K people, but were so depopulated that
they seemed empty; combining the populations into half as many servers
mean each has a better population... probably still substantially under
the 10K limit, as after a 9 month absence I returned to a newly (to me)
combined server environment which still seems emptier than the previous
non-combined servers I left.

If the 10K figure is accurate, it must mean active population, not
causual barely ever playing population.  Even so, the server merge
suggests that the former figure of less than 500k active subscribers is
now at best 250K active subscribers... probably substantially fewer.

Of course its also possible that the 10K cap is now substantially
larger, what with the addition of so many new zones, and of
instancing... the relative emptiness I'm feeling could in part be due to
that.  But whether this is the case or not has nothing to do with the
mergers themselves.

Lance

That could well be.  For all we know, the figures on the chart also come
form an SoE insider.  In both cases, we don't actually know.  One thing
we do know is that the chart very closely tracks what was in that article
you pointed to as far as active accounts up to the last available
datapoint in the article, which was over 500k accounts about a year ago.

/quote from another post:

... took half a million divided it by 52, got 9615
and then guessed that the servers were supporting up to 10,000 players.

/end quote

Whether your columnist did that math or SOE did it really doesn't
matter. It was certainly true at some point in time, that just over half
a million players were served by 52 servers, and thus approximately 10k
/ world server.

The EQ worlds of 2005, run on newer hardware, with dynamic load
balancing accross world server clusters, severing an ever expanding game
with more and more realestate for people to spread into, both interms of
zones, and zone-instances. Its likely that world servers can support
much larger player counts than they could "once upon a time".

Especially if non-trivial quantities of players are splitting their time
with other titles now.