But don't you see that it doesn't matter what I've owned, or what you have? These are just little stories about personal success or failure. To get an accurate picture of what's going on, you'd have to look at the big picture, at the thousands of PSUs shipped out across the country in the last several years. What are the failure rates of generic PSUs versus that of name-brand units? And do the failed units damage other system components?
You're right about that, actually. Anecdotal evidence is useless. These little tales of horror and explosions are little use to anyone, and to use them alone as evidence would be logically fallacious (hasty generalization). I can, however, provide real evidence, both practical and theoretical.
Theoretical: Cheapo units (I don't mean cheap units, I mean cheap crappy ones) use lower-quality components, meaning that there is a normal tendency for many to operate out of ATX spec (especially wrt AC ripple) even if QC is perfect and there are no duds. If, operating purely on calculations (a pain in the ass, to be sure, but it counts), it is shown that there is a high probability (there's always that tinier-than-tiny chance that each ultra-low-quality crappy part will perform perfectly at spec, variation and all that) of a good chunk of each batch operating out of ATX spec in the real world, then many will make it out onto the market.
More Theoretical: It seems obvious that cheapo PSU makers will skimp on whatever they can. It's clear that they skimp on components (clear if you open 'em up, anyway), and it follows that if they skimp on the most important part of the thing, they'll certainly aim to cut money in QC-- meaning more duds.
Practical: Buyers of cheapo generics experience more issues than buyers of higher-quality PSUs. If you add up all the anecdotal evidence and good/bad reviews out there, you do eventually reach a usable sample size. If you're not comfortable with the amount of effort needed to do that sort of aggregation, ask a medium-size reseller how many returns they have for Brand X and how many they have for Brand Y, where one brand is widely thought to be crappier than the other (say, Powmax and Fortron). As long as you find a human, you will get a nice big sample size.
I did present an argument, and I believe it makes much more sense than yours. That you don't agree doesn't mean it's not "valid."
Also, can you make just one post without tossing around insults?
Your argument was "that's a pretty good argument you've got there, but I still don't believe it." That's not an argument, much less a valid argument. It is an unqualified refusal. (To make things clear: in this context, I mean unqualified as the refusal itself not being qualified or explained, not any kind of personal attack on your credentials.)
What, exactly, do you find offensive? Do you feel that calling your argument invalid qualified as an insult, or do you find an insult in my explication of the possibility that you don't know of a relevant counter-argument? I'd like to know this. (Side note: these really are the only logical reasons for not raising a valid counter-argument. I sincerely apologize if either offends you, but one of the two really must be a fact.)
