Last week, the model crashed and, if it didn’t burn, there was significant smoke. When past-performance numbers suddenly drop or, just as suddenly, explode, an entire season’s data can look like shit – and is the shit the model relies on. That said, against just a handful of recent games’ numbers, the data of the previous 82 games still has weight, and I’m not about to pull the plug. The model’s has had, in past years, its share of poor Playoffs starts and has come around, and I’m counting on that happening again. Hopefully sooner than later. GLTA
