BishopLamont said:
It is convincing if you actually believe the dates he uses for "school holidays". They seem to change every time.
I don't think it's "convincing" in the slightest. It's good practice to not be persuaded by statistics without context. The relevant context being all other things that we
know affect hardware sales - shortages, new big software, new big software on competing platforms, price drops for a system, increased/better advertising, big new shipments of a big game (like Wii Fit), seasonal variations, a general increase/decrease in hardware sales momentum (increasing or decreasing good WOM).
The Wii was growing throughout a lot of the year, with a Wii Fit/Mario Kart related spike in the middle of it. If a platform is on an upwards trend, of course the average hardware sales before the holiday are going to be lower, and it's also shows that the average hardware sales in the period after the first holiday are higher, so with the displayed logic, Wii got a boost by entering a non-holiday period - of course this isn't true, just as how if Wii was helped in the holiday weeks, it's not showed by the data.
Even if the data was consistent with some minor correlation with holiday/non-holiday periods (which it loosely is, but we already know the reason why, which most, including doicare has agreed on - software), then that wouldn't be proof. Another relevant point to consider is that Wii suffered a shortage in the opening couple of weeks of 2008, which lowered your average.
The data is too sporadic, has too many inconsistences with such a theory to be provable. The first week back from the first holiday has higher sales for Wii than the first week of the first holiday.
The general outline for Wii last year was, increases in momentum, followed by big Wii Fit/Mario Kart related increases, followed by the typical decreases after the software bump, levelling off in a moderately consistent level of sales, higher than that before the bump (Wii Fit and Mario Kart legs / in tandem with Wii's general momentum increases), then the rest of the year.
That's the general. I see no unexplainable increases at the holiday that buck the trend, followed by a decrease at the end of said holiday, that continued the trend.
What I'm explaining is not that the theory relating to holidays is incorrect, but that it's unprovable because it's at best a minor contributor, and with such variations in weekly data, and all of many the factors which affect sales at different periods, it's not a testable theory. I don't think it's possible to prove this theory, especially as its effect is minor at best, which is why I'm not impressed by the low standards that have been given to make the claim that such a theory has been proven.
I'd imagine that if it is indeed a reality that school holidays make a minor contribution to Nintendo platform's hardware sales in UK/Europe, then the same would apply in other regions. Maybe looking at data from other regions would help in testing the theory, but the statistics must come with context, otherwise statistics are incredibly misleading.