unless i am mistaken, they had 1600 WD drives against 31,000 Seagate drives. The data is skewed.
Well, if this was a completely random representative sample it wouldn't matter a whole lot. 1,600 is quite a large sample size when it comes to doing the actual math.
The thing is, we don't know if it was a representative sample. Maybe they got a bad batch, maybe they mistreated them, there's tons of factors which we don't know about, which bring not only this one into question but also the previous one that showed that seagate was supposedly much higher in terms of failure rate. To be honest, I don't know what to think!