Hard to find the bright spot in the past three weeks of the model’s performance, and I suspect there isn’t one. Even with its continued solid straight-up performance, it’s lost way too many of its dog picks to blowouts, or 3 point favorites to 2 point wins. Today, I’m blaming the injury module, which is triggered on Wednesdays, and unlike what you see below – which is a strict past-performance projection based on who started the previous week – the injury module is supposed adjust for the replacement value of the projected starters. In fact, initial projections are proving to be more accurate – when they shouldn’t be – so, I’ll be working on the injury module today and tomorrow, and the first adjusted projections will be Thursday - hopefully. GLTA
