nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpq6tx08mwy9vkkjen5s8ahy9e3x5z4dmykefvs7u6wex0s02puuskqv750mt (nprofile…50mt) i think this is 3 only in the limit where v/fps -> 0, unless i'm missing something.
the angle step is u(1) = atan(sin(u0) / (2 - cos(u0)))
having done some horrible manipulation to turn this into a (pretty bad) numeric integral while assuming *only* this term (so no gravity, no friction, no acceleration...), plugging in some numbers, i find that for the first frame, at 0.1 rad/frame, 3.025 is closer, and at 1 rad/frame it would be about 5
so i think you just got lucky