Tuesday, September 6, 2011

Statistics...

...it is interesting how they differ. Blogger reports 71.287 page views, FlagCounter 65.209 and Sitemeter 41.335. And when you click on the Sitemeter icon you see a total of 53.694. Oh well :-)

"Dont' believe in any statistics you haven't forged yourself."

Now I know where this is coming from :-)

In any case... thanks, folks, for the interest, the great discussions and the fantastic ideas. JADE is eternally indebted to your creativity!

5 comments:

  1. I've always looked at statistics with great distance(doin phd from topology-yea its branch of math but quite opposite from probability theory).Its like bikini:show much but not the most important ;)

    ReplyDelete
  2. The reason for those statistics differences are very simple: Ad-Blockers. You have a bunch of JS things linked that are meant to track stats and in the past some of them have misbehaved, i.e. made sites load slowly. So they got blocked. The worse they are/were, the more people have them blocked. (Flag Counter for example just went onto my block list for loading a 300kb PNG.)

    That doesn't mean any of those stats are made up (well, aside from sitemeter maybe, that one is just shitty), just that not all of them can collect all the data.

    ReplyDelete
  3. Well, I have most website scripts disabled by default, except I have Blogger unblocked so I can comment (and a lot of other people probably do to, even if they only comment on other blogs). So I show up on Blogger, but not the other two! Otherwise you basically risk getting malicious stuff installed on your computer.

    But hey, no need to thank us for the interest and comments... Thank you for working so hard and blogging so much about an interesting game that's worth commenting on!

    ReplyDelete
  4. If you created a facebook fan page for JADE (and one for ADOM?) you could have very detailed "social" statistics over there, too.

    ReplyDelete
  5. In addition to script blocking, another factor is whether the measurement in question is smart enough to detect bots and discount them. The upshot is: no specific measure should be taken as gospel, what matters is how the measures change over time (and at least on THAT front, they should more or less agree, modulo noise)

    ReplyDelete