Yes we used the same technique but got a slightly lower score, but I only saw this post after the competition ended where they say that it’s only human players and not an engine that generated winners. That explains why your approach had better score since you didn’t use a high depth, whereas we use a 20 depth lookahead 
Congrats 