hellfire24 :
^80% of newegg reviews are useless.
I think of it like this:
Any company's / product's reviews are the null hypothesis, and you have to come up with evidence against the null hypothesis.
Assume a store sells WD 500gb blues, and there are 2 reviews, one says "Had it for 2 weeks and it blew up, but I returned it within the 2 weeks, so OK", and the other says "Had it 2 weeks, so far, so good". The null hypothesis is that the failure rate is 0.5. It probably isn't actually 0.5, but that's what the company's website says, so it's 0.5 as a null hypothesis. If you buy one and it does/doesn't blow up in 2 weeks, it isn't evidence against the null hypothesis. If you buy 100 of them and only 2 or 3 blow up, then that is evidence against the null hypothesis.
If there are 20 reviews and none of them say that it failed, then that doesn't mean that the failure rate is zero, but it does indicate that it may be a reliable product.
If there are 20 reviews of a similar, competitive product and people say "don't be fooled that these things are so reliable, because I had to RMA the one I bought", then one may think that the null hypothesis is "reliability A > reliability B". But if they are both reliable, then you would need a ton of data to show that "A > B".
And of course your results may vary.