In previous seasons reliable scores projections could be made with less data, but – apparently – not this season. There are key benchmarks that the model needs to hit (straight-up win/loss percentage and margins of projected victory), before I can call the posted projections reliable, but it’s looking like it will need a full half-season’s data before it’s ready.
![[Linked Image from i.imgur.com]](https://i.imgur.com/0pFObFc.png)