NBA Playoffs -- Sat 4/18 ML Value Plays
Gents,
Just joined SBR to share some of the analysis we've been doing at Team Rankings and debate the pros & cons of math-based systems vs. subjective plays. All our predictions are driven by 100% objective stat-driven algorithms we've developed over the past decade, although recent injuries and things like accounting for player matchups are known weaknesses of the models.
We just ran our money line analysis on Saturday's games, computing expected win odds for each team, translating those odds into implied "fair value" money lines, and comparing to current lines offered. Here are the results for Saturday:
- Detroit +850 (at Cle). The long shot and an expected loser, but our models think the ML on Det should be in the +370 to +550 range, meaning at least a couple hundred points of value.
- San Antonio -175 (vs Dal). Not a very strong call, but appears to be about 20 points of value in that line.
- Chicago looks efficient (maybe about 5 points of value) at +370, although the Garnett situation may provide further upside for the Bulls.
- Houston +160 / Portland -170 looks dead on efficient.
Interested to hear about money line systems folks are using. Our power ratings have done a great job at assessing win odds and we are in the process of developing some ML systems off that success.
|
Linear Mode
