In comparing, writing, and running these algorithms I find that there can only be very small improvements made using a brute force method. One interesting thing I did notice was that I increased the order of the first algorithm by n because of the way I stored what I was doing. That is I didn't count in base 3 and have intermediate sums, I counted with an integer and converted and, calculated for each number. The other two algorithms didn't have this problem, but they did have a related problem of storing the lowest schedule. I think it will be very interesting to find ways to speed up the solution to this problem in future versions.