hsrob said:
That's what i was going to say, the first part not the second part. |
Yeah, I totally screwed that one up...
Also: uber, I DID know... I was f-ing with you (maybe I should have just said "stick to comedy" instead of *bumps knux*... I expected you to be sharp enough to pick up on it.)
And you are right... taking the linear model at face value is absurd, since you wind up with negatives, rather than the tails of a bell curve extending infinitely in both directions... But it's still a better indicator of trends for analysis then averaging the sales over time (which is what you were offering). My KZ example isn't a strawman, it's more of a micro view of the flaw in your "PS3 has been out for two years and has sold 20m, thus PS3 sells 10m a year" theory. We both know that the high sales in the first year skew the average, much like KZ's high sals in its first week or two. If you just said "KZ sold 1.5m in 9 weeks, so KZ sells about 166,000 a week," you'd be WAY off from a predictive analysis standpoint. However, as flawed as the linear method is, it's going to give you a MUCH more accurate read on where the sales are going.
Other than that, how the hell are you?
Believing in the PLAYSTATION®3......IS.......S_A_C_R_I_L_E_G_E
