Originally Posted by mightyrick
Sure, I agree with what you're saying completely. You've put a strawman out there for critique, I think it is a valid start for sure.
So, I think the first thing to do would be to look at the results of the model and determine "how close" the results come to what we think is reality. For example, Rod Laver or Ken Rosewall did not come up at all. While we all may disagree on where Laver/Rosewall fall on the Top-10 list... we all can agree that they should at least appear on the list, correct? Even in the Open Era they should appear.
So now we either need to add another variable to the model which would show these results... or modify the existing coefficients (weights) on the current variables to make that happen. The question is... if we add a new variable... what is it? And what coefficient do we apply to that variable... and why?
Laver and old Rosewall (older than Emerson) dominated not only the pro events 1963 to 1967 (they won actually ALL of them) but also the first part of the open era (winning 8 out of the first 10 GS tournaments where they participated). Emerson did nothing then.