These numbers don't look useful to me. When you have Wilander ahead of Borg you know you have a problem.
Yes the figures do yield some surprising results.
I think the issue you raise isn't so much Borg being low as Wilander being high.
The thing about PR is they calculate an estimated peak playing standard where there is sufficient statistical evidence to support the claim. Many others evaluate total career achievement as their bench mark. If we were to adapt PR to measure total career achievement the I am sure Borg would be much higher than Willander, however, I believe peak performance over a statistically significant period is the better evaluation tool. Without using this approach players like Maureen Connoly and Monica Seles would be underrated. I suppose it all boils down to the time old cry of what does the word 'Greatness' mean. Despite Wilamders Peak rating he only topped the PR rankings for a period of 1 major, while Borg topped the rankings for a period of 16 majors despite his lower peak. It all has to do with the comparitive strength of opposition in the late 1980s compared with the late 1970s.
When you provide your definition of greatness before compiling your list what criteria do you adopt? Do you consider the quality of opposition faced? If so how do you calculate this? Do you measure dominance or would you evaluate a player with 15% strike rate in majors as better than a player with a 50% strike rate if his career was four times as long. Do you look at the losses of great players as well as their wins or do you subscribe to the point of view that greatness is greatness irrespective of consistency?
All these issues can take an evaluation of players a skew them in one direction or another.
Your not guilty of this Cyborg but the thing that really gets up my nose are people who evaluate one player using one criteria and another using another year. E.g 'Greatness can be evaluated by versatility on surface' but then still insist that players who only ever played on Grass are greats because its popular to say that rather than thinking for themselves. At least PR has its definition and applies it objectively.
I suspect that even if we had a magic box that could exactly evaluate every player from history it would yield a few surprises that are counter to popular opinion.
Consequently it is therefore fallacious to judge any rating methodology on a simple comparison between two players.
To obtain an accurate evaluation it is necessary to compare multiple player rankings.
What Coefficient of Rank Correlation would be acceptable between your list and the Performance Ratings to make you accept that the PR methodology has some validity?
Be interesting to compare your top 50 with PRs and make the necessary calculations.
If you post your top 50 post WWII I will happily carry out the necessary stats.
Happy to do both male and female as this would provide a larger data sample.
Regards
Tim
