I looked at the data more closely. It seems (no proof) that if the number of 3's is a power of 3 where the exponent is a multiple of 3 but not a multiple of 9, the highest power of 3 divisor will be 1.
blackjack
flooble's webmaster puzzle