tim wrote in post #11016752
...It's most just to see if any one brand is particular bad, or any brand particularly good.
But it wont. The way the poll is structured is such that no information can be gleaned. Your assuming that everyone has only one drive. For instance Look at the Seagate results. 13 have responded that none of their drives have failed and 9 have responded that one or more drives have failed. So, you have a 40% chance of your Seagate drive failing, right, according to the numbers.
However what if the 13 people only have one Seagate drive and it has not failed but the people who have had a drive fail have dozens of Seagate drives that have not failed but have had one or more fail. You just can't tell.
That's the situation I have. I have, on average, 1 drive fail a week, sometimes two. I manage a test lab and we have somewhere around 200~300 drives in servers and test equipment PCs and laptops. They are of all makes and models. I have had a failure for just about every manufacturer out there at one time or another. So I have to answer that all drives fail.
But If I add up all my drives I have about 1750 operational days a week. That is 250 drives spinning for 7 days. One drive failing a week gives me 1 failure every 1750 days or 1 drive failure every 5 years. So when you hear 1 drive failure a week, it sounds bad but when you take into concideration the sample size, it's not bad at all.
That's whats missing. You have no idea what the sample size is. Are 9 failures good or bad, you just don't know.