no, not the rounding error: keep in mind, i am a statistic noob:
The total amount of plus and minus in the bar chart should cancel eachother out because otherwise all useres wouldnt be 100% which sounds odd.
But as stated in the other reply i am really a noob and just read the wikipedia on precetnage points.
so the bottom line, i think the quantity (not percentage) of all the changes nets out to zero, so for simplicity’s sake let’s pretend there’s no chrome, mac, or other. since there are so many more windows users a 1.7% drop in windows is (for the sake of this example) the same as the 22.4% increase in linux users. when you convert away from percentages, you’ll see a quantity that represents the same number, but since you have to divide by the starting number of users and windows had so many more… am i making sense?
fuck, i’m high. am i even right? it’s been a few decades.
Yeah i think i get it… the total amount of users is 90 Windows and 10 Linux. At the next survey its 89 Windows and 11 Linux. So linux is up by 10% but Windows only down by 0.9% compared to the last value
I think it’s relative change and not percentage points, from quick mental math that seems to add up. Very weird choice, maybe trying to inflate the perceived increase in linux users. So it would be a change of a bit more than +1 percentage point for linux.
those numbers do not add up to 100% again?
change in share, not total share. or are you referring to the .2% rounding error?
no, not the rounding error: keep in mind, i am a statistic noob: The total amount of plus and minus in the bar chart should cancel eachother out because otherwise all useres wouldnt be 100% which sounds odd.
But as stated in the other reply i am really a noob and just read the wikipedia on precetnage points.
so the bottom line, i think the quantity (not percentage) of all the changes nets out to zero, so for simplicity’s sake let’s pretend there’s no chrome, mac, or other. since there are so many more windows users a 1.7% drop in windows is (for the sake of this example) the same as the 22.4% increase in linux users. when you convert away from percentages, you’ll see a quantity that represents the same number, but since you have to divide by the starting number of users and windows had so many more… am i making sense?
fuck, i’m high. am i even right? it’s been a few decades.
Yeah i think i get it… the total amount of users is 90 Windows and 10 Linux. At the next survey its 89 Windows and 11 Linux. So linux is up by 10% but Windows only down by 0.9% compared to the last value
right right
I think it’s relative change and not percentage points, from quick mental math that seems to add up. Very weird choice, maybe trying to inflate the perceived increase in linux users. So it would be a change of a bit more than +1 percentage point for linux.
Aha i did not know about % points. Why on earth is this the same “%” sign… anyway thanks for clarification