I am not sure how much longer I am going to be able to afford dumping money into this NBA model… Good thing my football model is working much better than this one.
Today was insane, I had 3 lineups in the 190 range and barely cashed, my cash lineup hit 170 and didnt cash. And the dollar tournament I maxed wasnt even the one with the $1,500 top prize instead it was the one with only $500. So obviously you do not max enter this tourney as even if you get a top 10 lineup youre only making $10… From now on, I am reserving my entries the night before/when I write this article.
Since it has been a few days and there really wasn’t anything interesting on this night besides another loss I am going to keep this write up simple.
The main thing I took away from this was that even if all of your lineups produce fire next to their name, if you have one person under perform, you will not win. I did want to bring up, I did a new way of optimizing today! For those of you following along I was unable to figure out the genetic optimization problem, I feel as if this is going to take awhile!
Instead here is how I am approaching optimization. I am generating only 5 lineups at a time containing stacks I want in 100% of these lineups. This helps tremendously rather than batch optimizing. I am still working on the optimization problem, however this is going to the back burner as I am going to have to re-scrape and organize upcoming slates since my data provider is getting too expensive while I rack up DFS losses.
Order of priorities now are:
- Get a winning contest strategy on Draftkings for NBA. Play 50/50’s and H2H’s more! But this takes more money and I seem to not have a good grasp on NBA yet. Today I am going to be playing $20 cash $20 GPP.
- Rescrape daily source data from RotoGuru and BasketBallReference.
- Optimize using Genetics.
Thanks for reading! Feel free to hop over to the NFL blog to track progress there as well!
DFS Machine Learning 