Hi everyone, I'm Edward, founder of IABET. I started building it because I was tired of picks services that only show their wins, so I wanted something where every pick and every loss is public.
How it works: we train separate machine learning models per sport (NBA, MLB, NFL and the main soccer leagues), each fed with player and team data, and then run 100,000 Monte Carlo simulations per game to turn the model output into a probability for each prop or market. Every pick is posted before kickoff, never after.
The record is public at iabet.co/tracker. It shows the last 7 days of graded picks, wins and losses included, broken down by plan, confidence band and sport.
Right now our top tier (Serious plan) is above 85% over the last 7 days: 89-10 (89.9%) for Oct 17 to Oct 23. To be fully transparent, all picks combined are lower: 631-453 (58.2%) over the same window, because the free and lower-confidence picks are included there. Variance is real and these numbers will move, which is exactly why the tracker updates every time a game goes final.
There's a free plan on iOS and Android if you want to check it out yourself. Happy to answer any questions about the models or the methodology.
To remove first post, remove entire topic.
Hi everyone, I'm Edward, founder of IABET. I started building it because I was tired of picks services that only show their wins, so I wanted something where every pick and every loss is public.
How it works: we train separate machine learning models per sport (NBA, MLB, NFL and the main soccer leagues), each fed with player and team data, and then run 100,000 Monte Carlo simulations per game to turn the model output into a probability for each prop or market. Every pick is posted before kickoff, never after.
The record is public at iabet.co/tracker. It shows the last 7 days of graded picks, wins and losses included, broken down by plan, confidence band and sport.
Right now our top tier (Serious plan) is above 85% over the last 7 days: 89-10 (89.9%) for Oct 17 to Oct 23. To be fully transparent, all picks combined are lower: 631-453 (58.2%) over the same window, because the free and lower-confidence picks are included there. Variance is real and these numbers will move, which is exactly why the tracker updates every time a game goes final.
There's a free plan on iOS and Android if you want to check it out yourself. Happy to answer any questions about the models or the methodology.