In my first recap, I detailed my performance in the $100 bankroll challenge over the first month. It was a mundane month, as evidenced by the +1.1% ROI. For those that missed it, the graphic below shows the summary information through Feb. 26.
I made the decision early in the process to abandon cash games in favor of playing 100% GPPs. I simply don’t think the swings in cash are worth it to a limited bankroll. Your mileage may vary, but when push comes to shove, I want to put my money in the spots with the best chance of a positive return. For me, that equals GPPs.
The results since Feb. 26 have been much better. You will find the results over the past eight days in table below. In those eight days, I am up $11.60, which is also the ROI (+11.6%), as I have wagered exactly $100 in that time. On Saturday night, I played both NBA and MMA contests. On Sunday, I played both NBA slates. (Side note: while it didn’t go well for me personally, we rolled out fighter projections, ownership projections and my personal cheat sheet for UFC 235. While it didn’t work out for me, six of the seven fighters I recommended won their fights. Anyway…)
Honestly, outside of binking one of the tournaments, I couldn’t be much happier with the results. Six of the eight days in the NBA were positive, including the last five. I believe I’m starting to become more comfortable with this challenge. I’m spending more time evaluating the lineups I select. This led me to quite a few top 1% finishes — 14, to be exact.
I have completely modified my process in order to focus more on lineup selection. Instead of playing my top lineups based on projected score, I now rank each lineup by two additional factors. The first factor is my lineup’s likelihood of reaching its ceiling. Each player has a certain percentage chance of reaching a target based on his price. I use those rates to find out which of my lineups has the best chance of a high score. Remember, the goal is to win the tournament. There’s a difference between a lineup with the highest points projection and the highest chance of winning a tournament. Just as an example, on my first crunch this morning, my highest projected lineup on DraftKings ranks 12th in my ceiling ranking. My fifth-highest projected lineup is even more extreme, coming in with the 57th-highest ceiling.
The second factor I’m looking at uses a similar process, but different information. By using our ownership projections, I calculate how unique each lineup is. You can read more about this process in Chris Spags’ Keys to NBA DFS article. We talk about differentiation a lot on the Strategy Show, but this step really helps to understand what we mean. It doesn’t take a lot of changes to get to a unique lineup.
After ranking each lineup based on these criteria, I now have a much different ordered list compared to just sorting by total fantasy points. It’s certainly a bit more work, but so far, it looks as though I’m entering a better package of lineups on a night to night basis. It never occurred to me that I could use this challenge as a sandbox to the rest of my play, but that is exactly what it has become. Keep an eye out next Thursday for the next installment and follow me on Twitter @JoshEngleman for more updates.