RandomSoccerFan
PREMIER
The thing is the rankings never take into account the time when the star striker or CB was sick or had to go to another event. There is variation and most teams have a couple of players that are difference makers. They don't go to a game and an easy win turns into a toss up.
Of course having additional information about conditions of a particular game can only enhance someone's ability to make an accurate prediction for the game. It's just we humans tend to overestimate our ability to do that, and place more weight on the presence or absence of whatever factor that we have personal knowledge of. That additional knowledge can help shape and augment percentages, when starting already with the information about how the team has performed over a recent/substantial amount of games.
I really like the valuation of the schedule. That can account somewhat for the variance of the leagues.
The schedule rating being shown isn't another variable that can be used to improve predictions. It's the other way around; it's just making transparent data that the model already has so we users can make additional inferences from it. All it represents is the average team rating of all of the teams that the selected team has faced over the past year. It doesn't use this information to predict or affect anything - it is the result of all of the existing calculations. But you're right, it can show some pretty compelling stories about teams that tend to choose leagues and tournaments with other teams who are stronger than themselves, weaker, or about the same.