
Originally Posted by
SteveinMN
I don't. No one else should, either.
There are some huge statistical issues with CR's reliability ratings: self-selection, poor definitions of terms (what constitutes a "problem?), poor granularity, little consideration of sample sizes, and the lag between data collection and publishing. I'm happy to go into those issues, but I think most people here don't give a rip so I will do so only on request. Then there's the disconnect between the products they recommend versus the ones they declare are poorly-built based on their collected data; I'm not sure how you recommend to such a readership products that you know has reliability issues.