View Single Post
Old 01-07-2013, 07:24 PM   #45
krosero
Legend
 
Join Date: Dec 2006
Posts: 5,357
Default

Ratios for Federer against Nadal at Wimbledon:
2006 3.34
2007 4.15
2008 3.40

Ratios for Nadal against Federer at Wimbledon:
2006 3.12
2007 5.17
2008 5.81

Yesterday we were debating whether to use margins or ratios. We talked a lot about winner/error ratios, but I wondered what would happen if I used the broader category of Aggressive Points rather than just the Winners that are usually reported in the media. Aggressive Points are all the points a player wins either by striking clean winners or forcing his opponent into errors.

Above are the resulting ratios for Nadal and Federer and Wimbledon. For example, in the 2006 final, Federer's Aggressive Points were 3.34 times more numerous than his unforced errors.

But the ratios do not reflect the results accurately. Federer's ratio is just slightly better than Nadal's in 2006, implying a tight victory, when in fact it was a four-setter.

Nadal barely won the 2008 match, 9-7 in the fifth, but his ratio is much larger than Federer's, implying an easy victory.

Finally in 2007, Nadal lost the match but he has a higher ratio than Federer.

When I sorted all my data in Excel according to ratios, some strange things came up.

The best example is Wilander's five-set loss to Sampras at the 1989 USO. Wilander had an Aggressive Margin of 29.8%. That was #74 among all the AM's I had calculated, that is, the 74th highest AM.

But when ratios are used, Wilander's performance jumps to the 5th best performance in all my data. His ratio of Aggressive Points to UE is 12.38. It's not only better than Sampras' performance in the same match (Pete drops down to #110 on the list, with a ratio of just 2.88 ). Wilander's performances also looks better, by this measure, than any performance by Federer, Sampras, Nadal or Djokovic.

The reason is that Wilander made only 8 unforced errors in that match. Nobody was better at keeping his unforced errors down.

And if you have your UE's down in the single digits, you're going to generate extremely high ratios of Aggressive Points to UE's. It's easy to see why: if you make 50 Aggressive Points and 2 UE's, your ratio is 25.0. Drop down just 1 UE, and your ratio jumps to 50.0.

Sampras/Wilander was the most extreme example, but there were others. It's clear that ratios disproportionately favor great defenders, and give inaccurate pictures of results.

The Aggressive Margin, by contrast, has Sampras at 31.5%, Wilander at 29.8%, an accurate representation of Pete's narrow five-set victory.

Anyway, I'll get back to posting more AM's -- I have several more now, including some very high ones for Nadal.
krosero is offline   Reply With Quote