Not sure how to fix it, but the problem is apparent: the model’s margins of victory are not good enough. It can and does predict straight-up winners, but it likes too many losers (dogs), and that doesn’t bring home the bacon. Got a lot of work to do, but that will have to wait for next summer. Hope it does better in the playoffs. The model’s closing line sweep for the past 6 weeks, ferreted out 6 corrections, 2 against the model, 4 for the model, resulting in a 2 win bonanza for the model’s ATS record, bringing the season up to a whopping 49.2% ATS - wow. Also, I may quit posting totals, unless it hits, at least, 50% this last week. Which will be a challenge given that a lot of playoff team starters will be sitting. GLTA
