4 ms·

I think it's saying that peoples' friends are better predictors of illness than just people.

• rdlecler1 10y ago
  But isn't that just a function of the larger distribution that gives you more resolution on the tail? I assume I'm missing something here because that seems obvious.
  • Scarblac 10y ago
    I assume they meant that _on average, per person_, friends became ill before the random people. Because friends have more contacts on average, and contacts give you a higher chance of becoming infected.
  • pizza 10y ago
    No, it samples a different probability distribution