From SLACer John’s blog:
I haven’t reviewed many papers, so I am hoping that I have just had bad luck so far, but it seems that every time I agree to review a paper it has serious flaws.
This is not a matter of luck. Papers that a priori seem likely to suck are disproportionately sent to younger reviewers and/or reviewers at less esteemed institutions. In retrospect, I can’t believe that I agreed to review some of the papers I did as an assistant professor (in terms of papers that were both maniacally written and not on a topic I knew anything about), and I certainly wouldn’t do review those papers if asked now.