[danielweber]

> He made the post mentioning Silk Road on Jan 29, 2011.

which was apparently the first mention. So, as far as the agent could tell, this was the first person to mention Silk Road on the open Internet. That's what's reasonable.

(Also, from a pure Bayesian POV, the fact that it nailed DPR on the very first try goes a lot towards demonstrating its relevance. NB: this parenthetical is not a legal argument; otherwise you could justify any search that turns up evidence.)

[ikeboy]

Along these lines, someone once tried to convince me that governments should be allowed to use evidence no matter how they get it. He basically proposed that police could go into someone's house to search without a warrant, but if they didn't find anything then the police could be prosecuted, which makes them only do it if they have a really strong reason to think they'll find something. It was a surprisingly good argument for something that many people would instinctively flinch away from.

Anyway, my answer was that if the police really had such strong suspicion of someone and they were right, they've got to have enough to get a warrant anyway, which is similar to what you're saying.

[tempestn]

Also if they expected to face prosecution were they not to find anything, you might expect them to often 'find' things whether they were there or not...