OK, thanks. Then how is that any different than what MS posted? He clearly stated it wasn't a certainty that a lower rated player couldn't exceed the ranking, only that the probability was lower.The online scouting community has a real bias. It's a personal preference thing. Obviously a model does not. At no point have I ever said models are 100% accurate. Not even close. I usually word in things like "that will change" when referring to things the model will say. Who watches 10000 kids play hockey every year. No one. A model can give someone easy reference to how 10000 players are performing.
The ranking system that has Willander higher, would on aggregate, be more accurate.