I set $90 worth of lineups (16 $5 ones and one $10 one) and took back $52 for a net loss of $38. My bankroll was $1050.60, now it's $1012.60.
Despite the loss, I feel I'm getting closer. I was wrong to have Corey Kluber is so many leagues and never even considered Chris Heston, but I had Giancarlo Stanton in eight lineups, Carlos Correa in eight and Yan Gomes in four. I also considered (and passed on) a Reds stack which (along with Heston) was the key to a big score. Neither Heston nor the Reds were that farfetched as the former was facing the Mets and the latter matched up at home against Aaron Harang. I was too rigid in sticking to my ace starter paradigm, though I did have Anibal Sanchez in one lineup.
I still think the key to the big score is imagination - can you envision Harang imploding in that spot or a certain pitcher with a good matchup going off? You imagine plausible paths to winning and submit a bunch of lineups reflective of them. There are far too many permutations to cover exhaustively, but this isn't about hedging or locking in some kind of return. It's about seeing an angle and exploiting it.
I also like Paul Sporer's idea of using cold players with good preseason projections whose prices are cheap. For some research justifying this method, click here. Essentially, preseason projections are better predictors of rest-of-season performance than year-to-date performance by a fair margin. And today's DFS slate is the first day of the rest of the season.