The concept of AskAIBetting.com is simple – I’ll prompt several different AI sources to find out what different algorithms believe will be the best sports bet on a given day. I’ll then publish the results, along with my own bet slip, and people can decide if they want to follow along.
I’ll also be making a parlay bet when making the consensus bet. The parlay bet will take into account what all of the AI models believe is most likely to happen.
Each update will include how each specific algorithm did the pick prior, along with a running W-L record both for me and for each algorithm.
To start with, I’ll be using the following models:
ChatGPT
Grok
Claude
Perplexity
Microsoft Copilot
Gemini
I may add more as time goes on, but this is what I’ll be starting with. I plan on doing this for one year with some deep analysis done from time to time.
This is as much an interesting experiment than anything else. The typical house edge for sports bets (before any bonuses/promos are added) is about 4.5%. The typical house edge on parlays can approach 20%. Amalgamating data should (theoretically) lower both edges, but what this comes down to is can publicly available AI actually beat sportsbooks. My prediction is that yes, at least one of the models (possibly more) I’ll be using will beat the sportsbooks. The next big question is can combining the data for straight wagers lead to a profit. The answer to that is what this site is about. I’ll be playing the parlays as more of a side note than anything else. Do I think I’ll beat the sportsbooks (over time) on the parlays? Probably not, but as the Washington Post notes, they’re what people like to play.