(PS: I think I have a quote fetish...)
patrisaurus wrote:
I really think you are making a mistake going for any sort of average count over a 24 hour consecutive period.
Actually, I thought it was a great way to extend an invitation that
included other timezones. What would happen if we ended up with a massive influx of European players who then became the majority? I'm sure you'd prefer US players to not be left out.
patrisaurus wrote:
it is a meaningless figure in terms of player enjoyment.
If you were one of those playing in the quieter hours with very few people, I'm sure you'd appreciate any incentives offered by the game's administrators to encourage more players during
your playtime.
patrisaurus wrote:
...and then adjust downward...
Adjusting downward when targets are not met is not fun - kinda kills the momentum and morale that we're trying to build. I'd much rather help everyone reach an achievable goal rather than act like some teasing bastard forever keeping a tasty apple out of reach.
patrisaurus wrote:
As a player, I don't give a damn if you find more people interested in playing in the middle of the night while I'm asleep and at work.
That's a rather selfish view, but I can understand how a large peak can be exciting. I have been online when we had 70+ and it's a great whirlwind of fun.
patrisaurus wrote:
If instead I knew you were encouraging lots of people to logon during peak hours means I can come home from work and find a bunch of people already logged in and making things happen, and that's awesome.
The average calculation we're using would benefit from more players being online at ANY time. There is just as much an onus on the players as the administrators when it comes to finding people who might enjoy playing.
patrisaurus wrote:
tl;dr = From a marketing standpoint I'd be more excited to be able to announce there were 40 people playing during peak hours than that there was an average of 13 over 24 hours.
If we get more players, I'm sure I'd probably brag about the higher peaks too - marketing is like that.
grep wrote:
if anything, I'd love for the count command to return a moving 24 hour average. That would be fun.
There's no irregular patterns between days so a rolling window would return a similar average anyway. That is, dropping the 12am-1am sample period on Monday morning only to pick up a similar value for 12am-1am on Tuesday is pointless. A friday PM + saturday AM window
might have a slightly higher average due to people not having to work on Saturday morning and thus playing longer; it's marginal though. Bonanzas have a much more measurable impact in comparison.