I'm doing several studies on random trading in the markets and wish to make sure I'm getting a true proxy for randomness. Currently using perl's built in rand function like this: $number = int(rand(100)); Gives a random number between 1 and 100. Using this function over a large number of simulations, a random trader's average trade converges to the vig/trade or to 0 with no vig - as one would expect, so it seems I'm getting good randomness. However I realize the limitations of the rand function and am wondering if anyone here has a better method. Or is this type of rand function sufficient for an analysis of random trading? Andrew Moe