Bad comparison.
You are comparing averages of maximum values with different time windows. This means that with larger time windows, the average will be higher, not because of a higher player count at the time, but because of an artefact of your sampling.
If you want to make a useful comparison of player numbers, make sure the time windows are equal.
Black Pedro has a nice explanation here:
Since people keep trotting this out, let me try to debunk this mistake once and for all.
The reason there is an apparent decrease in average numbers is completely and artefact of the sampling time of the data. Since the data being returned is the maximum PCU, the shorter the time range, the more data points are from lower activity times of day. This is obvious from the two extremes:
[Screenshot 2020-12-28 at 20.30.43]
In the weekly, you can clearly see the daily low PCU times of day, and eve…
And a bit more here:
I know. This is simple math. I am not sure why you aren’t getting it.
Well, if you didn’t get my explanation before, let me try one last time.
This is a made-up dataset using a sine function:
[image]
The data here is the complete data set and as expected the average is around 0 as you would expect (0.028 to be precise). You can take the average of this it is does reflect the underlying process as you have sampled it fairly.
If however you compress your sampling and start combining your dat…