I’ve had a terrible year of gambling.
After struggling with college basketball for so many months, I decided it was time to throw out my old system and try on a new one. Adapt or die, as they say.
So, I spent the Summer trying to learn more about the numbers.
The result of countless hours of research and self-teaching? The Action Network’s own set of college basketball power ratings, and a simple score prediction model we can apply to any game of the 2025-26 season.
I’ll be updating this article daily with our projected scores, arranged from the biggest projected home favorite to the biggest projected home underdog. Feel free to bookmark this page and revisit while handicapping.
A couple of essential things to note:
- This project is far from complete. I’ll be spending countless more hours tweaking the input data and prediction equation.
- Home-court advantage has been adjusted for, and I have filtered that out for games on neutral courts.
- We’re still working on how to incorporate injury adjustments into the daily projections. For now, all of these projections are under the assumption that both teams are at full strength, so please keep injury reports in mind.
Also, while I’ll mostly be just updating the numbers each day, if there are games that intrigue me and show value, I’ll provide some matchup news and notes at the end.
Best of luck this season, everyone!
Updates
Great news: We believe we have fixed the scaling problem, and projections for games with larger spreads should be more accurate moving forward. In fact, I think the entire process will be more precise, especially since we added more input data overnight. There's still more work to be done, however.
More good news: We've added a start time column, opening market spreads, and opening market totals to our daily projections. This should make this daily column far easier to navigate.
Bad news: North Dakota State's lousy 3-point variance continued last night, as the Bison shot 3-for-27 (11%) from deep in a 12-point loss. To twist the knife, UC Davis shot 8-for-16 (50%) from deep.
