2 ms·
> ... statistically they might be faster than other orders sent original around the same time how about if you take a block of interval (10-150 ms), and mess u
by signa11 10y ago
> ... statistically they might be faster than other orders sent original around the same time
how about if you take a block of interval (10-150 ms), and mess up, err, randomize all orders within that interval ? there goes the statistical advantage. moreover, if the difference in arrival of order between the fastest and slowest upto say 1(or 2)-sigma is what is taken (as the interval block) then it should level the playing field somewhat perhaps ?