[dalmo3]

> 1 divided by the number of experiments running on the lab

That would only make sense if you knew there was a leak but didn't know which virus was leaked.

Since we already have the knowledge of what virus (supposedly) leaked, the odds are (1/number of labs experimenting with that virus).

[marcosdumay]

Yes, but isn't that lab the place to go for researching coronavirus? Dividing the odds through all experiments may underestimate it, but it's not very far.