Admit it, Nintendo totally blew it.

[Melissa]

graphical analysis!!

Wii sales projection in the form of a normal curve, from perspective of April 2007: Standard deviation is f'n high. Let's call it 5 million units. Average projection for LTD sales as of Dec. 31 2007, let's call that 17 million, because that's roughly where sales will be at year end, and I'll assume Nintendo went with the average projection.

Now, with normal distribution, about 63% of the total possibilities will be contained within one SD of the mean. That means that if Nintendo had gone with 20 million units (or increase production by 400K/month more than they did), they would only have had about 27-28% confidence that they'd actually sell that many (that's what the Z table says, at least).

When you think about it, though, 400K/month is a steep hike in production. I don't remember where they were at in April, but let's take current numbers and call it about a 22% increase. They'd have to get 20% more production than they already had. Now, keep in mind that a production increase from April wouldn't have taken effect until at least August, so instead of having 8 months, they've only got (let's be liberal) 5. That means they would *really* have to increase production by 600K/month.

How the hell you going to tell me they should have increased production by over half a million *per month* with less than 30% confidence it would work? That's a 33% increase in production, enacted over 3 months! In preparation for a holiday season that lasts a month and a half! As of January, they'd already need to scale back production to more reasonable levels, or be stuck warehousing those extra 600K/month cumulatively. Hell, even an error of 100K/month would start taking its toll by March, particularly for an inventory-low company like Nintendo.

When Hurricane Wilma hit South Florida, she tore off many roofs. In 2006, it was common to drive down the highway and see blue tarps where Spanish Tile should be. It took about a year before houses were more-or-less roofed. Many buildings had roofs with significant weaknesses for the entire duration of the 2006 hurricane season (good thing Florida only got one tropical storm that year). Who in their right mind would let something like this happen? How the hell could it take an entire year to get 100,000 roofs' worth of tile to South Florida?

Simple. nobody wanted to produce that many tiles.

Far more was at stake here than some piddling stock value, but obviously production channels weren't prepared to take the burden. It happens.