pseudonymous bits

Crooked Timber poses an ethical issue: are anonymous comments fair game for consideration in admission if the committee can crack the anonymity?

I don't know, but I suspect that independent of the ethics, most committees would take the comments into account, and that most would not consider them definitive, they might tip the issue for marginal candidates.

But, this raises the more general issue of anonymity and psedonymous web presence - whether on student web sites, blogs or other forums.

Anonymity is common, some of the best known, and highest quality academic blogs are pseudonymous, including many of the Sci^blogs.
A lot of content relies on this anonymity, some content would be breach of confidence if done openly, other would just be imprudent or offensive - unfortuntely this includes a lot of the most interesting and informative content...

The problem is: academia in particular is a small community, just being a PhD (or grad student) or a scientist, or an academic, sharply narrows down who you could be. In a lot of case identifying field, approximate rank and vague geographic location narrows the number of possible people to less than 10,000.

This means 13-14 bits of information are sufficient to uniquely identify the person, and if you provide any finite amount of content, those bits will be there. Unless you deliberately obfuscate, which reduces the value and fidelity of the content...

So, if anyone actually cares to, essentially any academic blogger, or student commentator, can trivially be outed. And that is without actual effort like tracing IP addresses etc.

So... why be anonymous?
Well, most people in most cases couldn't be bothered, so people are actually safe.
And, there is always the tiny residual doubt - you may think you know who the griping student is, but can you be sure?
Not always, not sure enough.

So there is some residual value in anonymity, as long as it is not relied upon for anything serious.
Security through obscurity never works.