machine learning model: "GET PUCKS DEEP"
type: logistic regression + extra random forest ensemble
took down the @FanDuel NHL Nickel after only its second try. 1st and 3rd on a wild 12 game slate.
#FanDuel#DraftKings#DFS cc @adamlevitan
@BenWolby We do the same thing, but on a team level. We build 4 million teams on gameday with 200-300 features to feed to our model trained to predict probas of that team being a “milly winner.” Have you ever tried that ?
What a great lineup @adamlevitan . Currently 1 and 4 against you on the season. Was watching with a friend and I said “I got Adam again this week, he just lost hurt, sucks.” Immediately after I lose Montgomery then Jimmy G 😂 See you next week. Nice lineup my guy.
Just how well did the chalk do on Sunday?
Only ONE time in the last two years has the $20 Milly winner on DraftKings had higher cumulative ownership totals than yesterday.
Yesterday's Milly Winner
- 167% cumulative ownership / Avg of 18.5% per
- Previous cumulative ownership totals from the $20 Milly winners this year:
Week 1: 118%
Week 2: 81%
Week 3: 132%
Week 4: 105%
The only one that was higher in the last two years? Week 7 of '22 where Brandon Jacobs and the Jets D each pushed 45% for ownership and both made the winning lineup.
���Plenty of nuggets (plus winning lineup grades) in today's newsletter, but that was our favorite.
Sign up is FREE. Details in our profile. Sign up today and we will make sure you get today's issue.
GPP Game Scores combines:
* Salary-based value
* Projected ownership
* Ceiling
* More
Week 4 model results and context are live: https://t.co/V0glzovdqZ