Earlier quoted context omitted.
It's not like he just started a random new torrent tracker... he took over an old domain that was previously in use by people pirating stuff after observing that torrents were still pointing to the tracker and ran a tracker on that domain. That's a pretty direct line to "he knew this would be used for copyright infringement".
What kind of threshold do courts tend to put on that, for the percentage of illegal activity? Because pretty much any service that connects lots of people together is guaranteed to have some .
For one, a judge/jury does not infer things they are "supposed to know", such as whether torrents are mostly used for piracy or not... they only operate based on the evidence presented.
There is a very large burden of proof in criminal cases, requiring that their intent to facilitate a crime be proven "beyond a reasonable doubt".
Trying to say "everyone knows linux ISOs is code for piracy" or claim that "a judge would see right through that" is simply not how things work... decisions cannot be made based on any type of prior knowledge like that.
The entire point of a criminal court case (as a prosecutor) is to convince a judge/jury that the defendant is guilty using evidence and testimony, which means they must prove that there was clear intent to commit/facilitate a crime, i.e. they knew it was illegal and did it anyway.
Simply running a torrent tracker in and of itself doesn't prove any of that.