Read it fully! It's 11.3% without RROD and 23.7% with!

Anyway, that's kind of shifty data. Really, a 12,000 unit sample out of the 100 million people that own current gen consoles?

Kantor, 12,000 is far more than enough for a statistical sample.

I sample 12,000 London citizens. 6,000 of them are Chinese. Thus, half the people in London are Chinese.

It's not likely, but it's very much possible, and you're unlikely to get an accurate sample of a 100+ million userbase with a 12k sample. In my example, London only has 8 million people, while there are more than 12 times that number who own a current gen console.

Gallup polls usually only have 500-1,000 people even in USA and those are still quite accurate. I doubt any gallup poll has ever had even 10,000 people polled. And 99.99% of the studies have much less than 12,000 people. And your example is less possible than winning lottery 10 times in the row.

Claiming it's "very much possible" doesn't mean your point has any credibility, the point of statistical sample sizes is to reduce likelihoods of such events to non factors. In practice, for something as common as console failures, 12,000 is a huge sample size.

Pwnd.

That's dealt with in the report, actually. Under the heading "Usage is a factor, but Wii is still more reliable".

I would assume the Gamecube was more reliable. But few people had one. It only sold like double the Dreamcast. I only knew like 2 people with a Gamecube compared to a countless number PS2 owners. Naturally I would see more PS2 problems which was still rare.

Are you trying to make an argument based upon personal experience? Ironic, after that other poster doubted the statistical power of a 12,000 sample size.