Originally Posted by 757690
This exemplifies why using run differential from season to season is very problematic. Teams change their rosters in the off-season, players get older, rookies and young players break through, and some players fall off in terms of production. And that doesn't even take in account injuries.
I really think we will discover that the Pythag will become less and less useful as we move forward with understanding the game.
It all depends on what you use it for. It's only useless if it's used to blindly predict things without taking into account roster changes.