theprof00 said:
trend matters because it will show other things for example here is an imaginary data set
01/09 360 17% PS3 24% 02/09 360 18% PS3 22% 03/09 360 13% PS3 17%
 this kind of data would have different implications.
Of course that data could be written as actual units:
01/09 360 57(85+) PS3 43(85+) 02/09 360 59 PS3 44 03/09 360 60 PS3 44
but then you have to do the percentages in your head, or at least, there is a second step to seeing how those percentages are changing.
It's all about what you are trying to prove. Say,.. ps3 85+ games were increasing in frequency whereas 360's was falling. You can always show this by showing the actual units but why make someone do the percentages in their head? It is stronger to show the result with as little extra steps as possible.
For your technology!
|
Bolded the word trying, because in the end you can't actually prove anything with percentages like this, only spin numbers. I agree with Kowenicki wholeheartedly - just because a game was released a year ago does not make it an unplayable antique that has no business being compared to current releases.