2 ms·

I think it's one of those things where a few point shift in the average of the distribution really changes some numbers near the edges of it, though.