Frequency is basically all I look at. Frequency in full cut results and in wider community conversations.
If you pair the limited tournament data with some basic amount of logic and actual game time its usually pretty accurate. The thing I use the tournament results for the most is actually just to identify player preferences that dont make logical sense. If I try to predict the cut based only on what I consider “optimal” lists its almost always wrong since player preferences lean towards lower ship counts, more hp, and easier jousts. The data is really helpful for identifying popular trends.
Edited by Boom Owl