I don't think that any change first makes to the alliance selection process can possibly even out the alliances. The new draft system would work if each team had great scouting. At Chesapeake I saw people pick much weaker robots than we did for their last pick. Our 3rd robot was a really great machine, and it was the last robot chosen. I suppose that Chesapeake is not a good example of the serpentine drafts supposed effectiveness since there were no upsets.
I think that in all years it depends on the quality of your robot and the quality of your scouting. Last year at BAE we were a 7th seed which moved up to sixth seed. We ended up making it to finals, even without the serpentine draft. Why? We had a good robot and we made good picks. But we still lost to the team of 121 and 126 whose final alliance partner was mostly (if not entirely, during actual game play) defensive.
Someone last year made the prediction that the winner of each devision would be the highest seeded good robot with good scouting. This sound obvious but it is true. I think that, in this game especially, the winning combination is of two teams that do all things well, rather than one thing exceptionally.
So, my feeling on the serpentine draft? I think that it is really good for the lower seed alliances, especially if they have great scouting. However, I have seen many 8th seeds pick their first robot and then have no time to confer before making a second pick. I think that we will be able to see it's effectiveness more when we get to nationals. If a system works well for 25 teams and well for 85 teams, then it works. I would rather see interesting matches, rather than looking at the 1 vs. 8 match and thinking that 8 is just going to get owned.
