So I'm considering re-upping my betlabs subscription that lets you test out theories and what not. I dug up my old college football stuff from previous years and though it would be interesting to discuss some of them on here. Some are mine, others I found in their database, but just looking for some good open discussion on whether they are causal vs. correlated, dumb theories, etc.
A few baselines I use is that the ROI has to be 20% over the long term, b/c the downside of "systems" is that for them to work, you have to play them all. It's only natural that you're going to miss some, so even if you do, the ROI should still be good. Sample size of the entire population has to be at least 30.
I know these aren't for everyone, and I'm not advocating them at all...just wanted to create some discussion b/c it's a slow time right now and I like hearing different perspectives.
Overall, I think most are rooted in contrarian type things (lots of road dogs, conference vs. non-conference, etc). Some have decent writeups, others don't. Anyway, here's the first one for week 1 college FB (I'll post a few more later)
https://ibb.co/8zw6T6V
Week 1 ML Unranked road dogs with spread <6 (Wait til close)
- Games with lower spreads are likely to be closer
- Public money likely on home team and favorites, artificially driving up the line - hence the road play
- 5% drop in ROI when moving to the key #6
- Have to play line right at close to make sure doesn't move ti 6
Personal take - I think the theory makes sense, especially b/c with the close spread that really eliminates teams who scheduled cupcakes for week 1. Downside is that you have to make sure it doesn't close at 6, eliminating the ROI. Not sure what to think about home field advantage in this situation.
Thoughts?