Originally Posted by mattennis
The ATP ranking ( prior to 1990 ) was computed as an average: "total points divided by nš of (computable) tournaments played".
Not all the tournaments played (even some with strong-fields) were computable to the rankings.
This was the system and all players knew it.
Under this system Connors was nš1 in 1978 (and in 1977 ahead of Vilas) because his average "points per (computable) tournament played" was a bit higher than Borg's.
Since 1990, the "average computation" was changed to "the total sum of the best 14 results" (taking into account BONUS points too, that were a good fraction of the total points of any player).
They tweaked things (changing the points awarded, varying some other things) several times during the 90s also.
From 2000 (or so) M-1000 (and GS) started to be mandatory and 18 + WTF tournaments were computable ( instead of 14 ).
Well, there were other changes too that I have not mentioned because I am already tired.
I know in the late 80s you could earn points according to the ranking of a defeated opponent. For example, at the '88 USO Wilander got more for defeating the #1 ranked player (Lendl) than if he had defeated someone else. I don't recall the details but I'm certain wins were weighted this way for a while.
Do you know when that began (and ended)? I doubt it was done back in '78 because Borg had a decisive edge over Connors in terms of defeated opponents (a 21-2 record vs the rest of the Top Ten, compared to 13-4 for Connors).
I think if wins were weighted the way they were in the late 80s Borg could well have been ranked over Connors.