Why do I find it so difficult to rate some players as high as everyone else believes

I think the answer has to be because I think about losses as much as wins.

E.g. If we look at majors matches only quarter finals or better we see the following statistics:

Andre Agassi scores 49 wins and 28 losses or 63.6%.
Jimmy Connors clocks in with 54 wins and 33 losses or 62.1%

I know these players had long careers and played on where perhaps others retired but these volumes of losses at the highest level do seem substantial.

E.g.

Bjorn Borg 44 wins and 10 losses or 81.5%
Rod Laver even with his best years lost to the pro era 46 wins and 9 losses or 83.6%.
Pete Sampras 55 wins and 15 losses or 78.6%

Of course still to complete their careers so these data may well decline as yet are

Roger Federer 60 wins and 11 losses or 84.5%
Rafael Nadal 25 wins and 7 losses or 78.1%

Even some players who are held in much lower esteem, quite undeservedly so in my opinion

Ivan Lendl 55 wins and 26 losses or 67.9%
Mats Wilander 32 wins and 13 losses or 71.1%
Jim Courier 49 wins and 22 losses 69.0%.

To me it seems people are completely fixated on achievements and wins without ever considering failures and losses. They are opposite sides of the same coin and one must balance out the other in any form of evaluation.

Normally I look at things in a far more complex statistical way considering quality of opposition and adjustments for spread of results over time etc but I think the above says enough to bring my frustration to the fore.

Tim

Wuornos said:

I think the answer has to be because I think about losses as much as wins.

E.g. If we look at majors matches only quarter finals or better we see the following statistics:

Andre Agassi scores 49 wins and 28 losses or 63.6%.
Jimmy Connors clocks in with 54 wins and 33 losses or 62.1%

I know these players had long careers and played on where perhaps others retired but these volumes of losses at the highest level do seem substantial.

E.g.

Bjorn Borg 44 wins and 10 losses or 81.5%
Rod Laver even with his best years lost to the pro era 46 wins and 9 losses or 83.6%.
Pete Sampras 55 wins and 15 losses or 78.6%

Of course still to complete their careers so these data may well decline as yet are

Roger Federer 60 wins and 11 losses or 84.5%
Rafael Nadal 25 wins and 7 losses or 78.1%

Even some players who are held in much lower esteem, quite undeservedly so in my opinion

Ivan Lendl 55 wins and 26 losses or 67.9%
Mats Wilander 32 wins and 13 losses or 71.1%
Jim Courier 49 wins and 22 losses 69.0%.

To me it seems people are completely fixated on achievements and wins without ever considering failures and losses. They are opposite sides of the same coin and one must balance out the other in any form of evaluation.

Normally I look at things in a far more complex statistical way considering quality of opposition and adjustments for spread of results over time etc but I think the above says enough to bring my frustration to the fore.

Tim

Click to expand...

Tim,

I love your thinking. I remember years ago, writers were pointing out the awesome achievement of Sampras in winning 14 majors. They were fixated on the number 14. I thought to myself, while it was a great achievement, Sampras accomplished it by enter 52 majors. This was a lot more tournaments entered than many greats so the number 14 in itself wasn't as great as it looked when we realize it was done in a lot more tries than others in the past.

Keep up the great work.

Tim, If you bias your statistical approach to produce the results people want they'll start agreeing with you.

I Connors and Agassi would have called it quits at age 30, how would their stats look?

Henry Kaspar said:

I Connors and Agassi would have called it quits at age 30, how would their stats look?

Click to expand...

In most cases the percentages would be higher but the cumulative totals would be lower also. So Connors probably would have had a higher winning percentage, higher percentage of tournaments won etc but he wouldn't have won as many tournaments and majors.

Agassi, I am not sure about since he dedicated himself to tennis and had some great years when he was older.

Wuornos said:

I think the answer has to be because I think about losses as much as wins.

E.g. If we look at majors matches only quarter finals or better we see the following statistics:

Andre Agassi scores 49 wins and 28 losses or 63.6%.
Jimmy Connors clocks in with 54 wins and 33 losses or 62.1%

I know these players had long careers and played on where perhaps others retired but these volumes of losses at the highest level do seem substantial.

E.g.

Bjorn Borg 44 wins and 10 losses or 81.5%
Rod Laver even with his best years lost to the pro era 46 wins and 9 losses or 83.6%.
Pete Sampras 55 wins and 15 losses or 78.6%

Of course still to complete their careers so these data may well decline as yet are

Roger Federer 60 wins and 11 losses or 84.5%
Rafael Nadal 25 wins and 7 losses or 78.1%

Even some players who are held in much lower esteem, quite undeservedly so in my opinion

Ivan Lendl 55 wins and 26 losses or 67.9%
Mats Wilander 32 wins and 13 losses or 71.1%
Jim Courier 49 wins and 22 losses 69.0%.

To me it seems people are completely fixated on achievements and wins without ever considering failures and losses. They are opposite sides of the same coin and one must balance out the other in any form of evaluation.

Normally I look at things in a far more complex statistical way considering quality of opposition and adjustments for spread of results over time etc but I think the above says enough to bring my frustration to the fore.

Tim

Click to expand...

Is there a particular reason you're only looking at quarter finals or better in majors (as opposed to any match they played in a major?

Thanks for the stats.

This may be true about Agassi but not Connors!

If Connors had retired aged 31, then he would have still won same number of GS tournaments and still won 100 in total. He played on for another 9 years after this, doing quite well sometimes but only winning another 9 tournaments!

And when you think 9 years is longer than lots of players professinal career eg.Borg and many others, Connors stats would have looked way way better if he had retired after the US Open win in 83

akv89 said:

Is there a particular reason you're only looking at quarter finals or better in majors (as opposed to any match they played in a major?

Thanks for the stats.

Click to expand...

Because this provides a measure against a reasonable quality of opposition. To include results for all rounds would just cram the %s up into the 80s and 90s for almost all the very top players and hide the difference I was trying to illustrate.

unless any of the players listed lost lots and lots of matches in the first 4 rounds of majors, which they didn't, the inclusion of these matches would serve little purpose other than to somewhat camouflage their performance against others at the top of the game.

Hope this makes sense

Tim

AAAA said:

Tim, If you bias your statistical approach to produce the results people want they'll start agreeing with you.

Click to expand...

lol :lol:

Tim, do you find it easier to rate Connors now as high as everyone else?

My post above yours explains clearly why you should

yes i have to agree connors was an all time great and i would have him just a tiny notch below borg

Tim,
I think you base your opinions on statistics.

Most of us use statistics (selectively) to back up our opinions.

Wuornos said:

I think the answer has to be because I think about losses as much as wins.

E.g. If we look at majors matches only quarter finals or better we see the following statistics:

A
Bjorn Borg 44 wins and 10 losses or 81.5%
Rod Laver even with his best years lost to the pro era 46 wins and 9 losses or 83.6%.
Pete Sampras 55 wins and 15 losses or 78.6%

Of course still to complete their careers so these data may well decline as yet are

Roger Federer 60 wins and 11 losses or 84.5%

Click to expand...

These numbers would seem to suggest a ranking, such as
Roger Federer -- 84.5%
Rod Laver -- 83.6%
Bjorn Borg -- 81.5%
Pete Sampras -- 78.6%
Mats Wilander -- 71.1%
Jim Courier -- 69.0%
Ivan Lendl -- 67.9%


Could you please post numbers for Rosewall and Tilden as well? (I'd be most curious to see those statistics.)

hoodjem said:

Tim,
I think you base your opinions on statistics.

Most of us use statistics (selectively) to back up our opinions.

Click to expand...

The stats Tim has used above are hopeless when comparing players careers though, he needs to go back to using a more complex analysis!

Borg retired as soon as he started losing, whereas Connors carried on, as I explained above!

And Federer hasn't retired yet!

(Laver is the GOAT, and Connors is about 10th on the list)

hoodjem said:

Tim,
I think you base your opinions on statistics.

Most of us use statistics (selectively) to back up our opinions.

Click to expand...

Hoodjem,

Statistics are great if we all can objectively use the information. I've read your opinions and I think you're pretty objective.

I've often looked at old time players with certain preconceived notions and I've found out the player was far better or far worse than I would have expected. After that you have to examine further to see why he or she was greater or less than you would expect.

Rosewall is a good example of someone who was far greater than I thought at first and I thought he was great to begin with.

And there were a number of players who were far inferior to what I would have expected when I got the information for them.

As one as the person writing the opinions is fair with the statistics that is okay with me.

jimbo333 said:

The stats Tim has used above are hopeless when comparing players careers though, he needs to go back to using a more complex analysis!

Borg retired as soon as he started losing, whereas Connors carried on, as I explained above!

And Federer hasn't retired yet!

(Laver is the GOAT, and Connors is about 10th on the list)

Click to expand...

Don't worry I hve not given up on more advanced aproach. It's just difficult to make a point with them on here as they generate a single number and very few people understand the parameters feeding into them.

The point you make of course is valid but where would this leave agassi who was perhaps more succesful later in his career. You have correctly identified that to take an entire career sells some players short. In strike rate it would be detrimental to Connors. In total career achievement it benefits him.

The very point is of course is the truth is in between or to be specific peak playing standard that can be statistically supported. Those are the ratings I try to calculate. Whichever way you jump with a total career rating you benefit one person or penalise another.

Tim

pc1 said:

Hoodjem,

Statistics are great if we all can objectively use the information. I've read your opinions and I think you're pretty objective.

I've often looked at old time players with certain preconceived notions and I've found out the player was far better or far worse than I would have expected. After that you have to examine further to see why he or she was greater or less than you would expect.

Rosewall is a good example of someone who was far greater than I thought at first and I thought he was great to begin with.

And there were a number of players who were far inferior to what I would have expected when I got the information for them.

As one as the person writing the opinions is fair with the statistics that is okay with me.

Click to expand...

Yes you are right. I don't really have an axe to grind about any particular player. I am only really concerned in the rombustness of the statistical systems I develop. At the heart of these systems must lie objectivity.

I would be most surprised if every output I produced agreed with even one posters opinion. After all even if we agree on who the top 20 players are there are still 2,432,902,008,176,640,000 different permutations of arranging those top 20 players. lol:

Tim

hoodjem said:

These numbers would seem to suggest a ranking, such as
Roger Federer -- 84.5%
Rod Laver -- 83.6%
Bjorn Borg -- 81.5%
Pete Sampras -- 78.6%
Mats Wilander -- 71.1%
Jim Courier -- 69.0%
Ivan Lendl -- 67.9%


Could you please post numbers for Rosewall and Tilden as well? (I'd be most curious to see those statistics.)

Click to expand...

Sorry my database only goes back to 1946. So Tilden is out.

Also Rosewall spent his best years in the amateurs so he is disadvantaged to a certain extent. However his figures are: Won 49 Lost 22 or 69.0%. I suspect this figure would have been higher had he been playing the majors at his peak.

This as you have already said though would need to be adjusted based on the quality of opposition he faced and the spread of the results, i.e, were all the best results concentrated within a short period of time.

Rosewall would actually rank 18th on the list of post WWII players if we take these figures in isolation. If we adjust for quality of opposition and spread, he comes up to 14th with a rating of 2742 on my startistical rating system. If the amateur pro divide had not been present and had he had the opportunity of recording a rating at his peak I suspect his overall position would have put him between 6th and 10th post WWII, but there is no way of knowing. Based on the decline in rating of other players and looking at the rating at the end of his career I feel it is unlikely he would have been top 5. But that is just my opinion and not really backed up on any really firm statistical basis.

One day I will get the chance to roll my database back pre WWII and then we will see about Tilden. As I lost my job as a Statistician and data manager two years ago though, and then suffered a nervous breakdown, this is now difficult as I am doing two full time jobs on minimum wage to try to support my family.

Take care

Tim

obviously federer is GOAT