As for the opinions of judges, I moved from law to the sciences so I like to think I have a good perspective on both, and I've found that lawyers and scientists frequently get the other subject wrong, at least in the sense that they frequently don't really understand what the other is trying to accomplish. Approaching it as a science would not be a good idea for many areas of legal reasoning. Judges need to make distinctions between things that are frequently on a continuum. Yes, you can theoretically identify someone from an IP address, just like you can identify someone from a mail address or a telephone number. But it's harder with an IP address. The judge therefore has to decide where on the continuum does it get hard enough to identify someone to the extent that it's not really "personally-identifiable information" as contemplated by the contract; he picked IP address on the "not PII" side of the spectrum and frankly I probably would have done the same thing. And the judge didn't poll law professors, but in a sense they did poll judges -- for the exact point of law you point to, the judge cites another court case (1 judge) and an appellate court decision (at least 3 judges) who came to the same conclusion. I mean, scrolling through the slashdot story you posted, a lot of the commenters (and I would suspect a good percentage of them are network engineers, programmers who work with networking, etc.) agree with the judge.