Another Look at Most Dominant Player

[GameSetAndMath]

Yes, but the 2014 USO was after the streak ended. The streak was from Roland Garros 2010 to 2014.

Oops.

Ya, you are right. Rafa's streak length is 15, Roger's 19 and Novak's 25 using the definitions given in OP.

 

[El Dude]

Oops.

Ya, you are right. Rafa's streak length is 15, Roger's 19 and Novak's 25 using the definitions given in OP.

It is definitely an interesting metric, but like any other has limitations. The main takeaway for me is it illustrates just how incredibly consistent Novak was/is.

 

[GameSetAndMath]

Just giving a bump to this thread as there is a lot of talk about who is best at their peak in the "are we ready to crown Novak as the GOAT thread".

For the uninitiated (aka, if you are new to this thread), please read the blog referenced in the OP of this thread before participating.

The data in the blog of the OP is quite old. Now that we have seven more years of data, we need to recalculate the number of streaks, streak length of each streak and streak strength of each streak for each player. I request our in-house statistician ElDude to get to work on this and report the results. But, let us stick to the definitions given by the author of the blog and not tweak them. I have no vested interest here and I have no idea as to who will come out first when we redo the calculations.

Also, this would only settle the question of who is the most dominant player and not settle the question of who is the GOAT. The interesting point about MDP is that it is little bit more amenable to scientific treatment and a consensus on it is not as hard to achieve as with GOAT (as it is mythological in part at least).

 

[El Dude]

Just giving a bump to this thread as there is a lot of talk about who is best at their peak in the "are we ready to crown Novak as the GOAT thread".

For the uninitiated (aka, if you are new to this thread), please read the blog referenced in the OP of this thread before participating.

The data in the blog of the OP is quite old. Now that we have seven more years of data, we need to recalculate the number of streaks, streak length of each streak and streak strength of each streak for each player. I request our in-house statistician ElDude to get to work on this and report the results. But, let us stick to the definitions given by the author of the blog and not tweak them. I have no vested interest here and I have no idea as to who will come out first when we redo the calculations.

Also, this would only settle the question of who is the most dominant player and not settle the question of who is the GOAT. The interesting point about MDP is that it is little bit more amenable to scientific treatment and a consensus on it is not as hard to achieve as with GOAT (as it is mythological in part at least).

You know I can't resist. Damn you.

OK, here's the first chart - the one that is accumulative, and thus benefits players with longevity. It has, for instance, Roger slightly ahead of Novak at the AO, which is clearly absurd.

Note that the colors are rounded up or down to the nearest 1000, but I think basically even out.

I'll try to get the other ones up in a bit.

 

[El Dude]

Next one.

Grand Slam Points per Slam
1. Borg 1145
2. Laver 1057
3. Djokovic 995
4. Nadal 959
5. Federer 879
6. Rosewall 786
etc...

I like the articles description of Borg as "Sandy Koufax in Wimbledon whites." Clearly this metric goes the other direction, biasing those who didn't play many Open Era Slams. Laver barely played eight Slams after his great 1969 season, despite being probably the best player on tour for another two years, top ten through 1975, and playing a few titles through 1979.

We also haven't seen Rafa and Novak go through a real decline phase yet, so their averages will likely go down - unless they retire soon. But they'll almost certainly remain more dominant than Roger by this metric.

 

[El Dude]

As far as the streaks is concerned, the big difference is that Novak jumps to the top with his streak from the 2010 USO through the 2016 USO - 25 Slams (!). So we have:

I also used the color scheme that Wikipedia and Ultimate Tennis uses, rather than the article. My apologies for not polishing these up - I tried to get them up as quickly as possible.

As you can see, Roger still has the highest average points during his streak.

 

[GameSetAndMath]

Thanks dude. Your third post is the one of most interest to me as it is the one that decides MDP.

Does djokovic has one long streak of 25 or did you combine two separate streaks of him?

 

[El Dude]

Thanks dude. Your third post is the one of most interest to me as it is the one that decides MDP.

Does djokovic has one long streak of 25 or did you combine two separate streaks of him?

One long streak, from 2010 USO through 2016 USO. He never went more than one Slam in a row without reaching a final, with gaps at the 2011 RG, 2012 Wim, 2013 RG, 2014 AO, 2014 USO, and 2016 Wim.

Right now he's on a 12-Slam streak, which would be tied with Connors for 5th best. If he reaches the F of USO and/or AO, he'll have two streaks longer than Rafa's best (13 Slams).

 

[GameSetAndMath]

One long streak, from 2010 USO through 2016 USO. He never went more than one Slam in a row without reaching a final, with gaps at the 2011 RG, 2012 Wim, 2013 RG, 2014 AO, 2014 USO, and 2016 Wim.

Right now he's on a 12-Slam streak, which would be tied with Connors for 5th best. If he reaches the F of USO and/or AO, he'll have two streaks longer than Rafa's best (13 Slams).

OK. So, the strength of Roger's streak is better than strength of Novak's streak. But, the length of Novak's streak is longer than length of Roger's streak. The longer the streak is, the more difficult it will be to keep it stronger and so I would cut Novak some slack about that.

But, what worries me even more is the current streak (which is likely to continue - as our model allows for minor hickups along the way). The second streak may be even stronger than his first streak and Roger's streak (what is the current average in his second streak, btw).

Also, if Novak is the only player having two streaks of length 10 or more, that is again another argument in his favor (assuming Roger and Ralph have only one streak length 10 or more).

 

[El Dude]

OK. So, the strength of Roger's streak is better than strength of Novak's streak. But, the length of Novak's streak is longer than length of Roger's streak. The longer the streak is, the more difficult it will be to keep it stronger and so I would cut Novak some slack about that.

But, what worries me even more is the current streak (which is likely to continue - as our model allows for minor hickups along the way). The second streak may be even stronger than his first streak and Roger's streak (what is the current average in his second streak, btw).

Also, if Novak is the only player having two streaks of length 10 or more, that is again another argument in his favor (assuming Roger and Ralph have only one streak length 10 or more).

Yeah, although I have a hard time imagining that Novak will play at this level long enough for another 20+ Slam streak. That would mean it would have to continue through 2023. I just don't see it, maybe even especially if he wins the USO. At that point, while I think he'll remain the top player until someone emerges, I think he'll drop a notch and be more beatable. I mean, he'll have done something that no one has done in 52 years and pretty much be the inarguable GOAT. Hard to continue at the same pace and level of drive after that.

Roger and Rafa will also play a factor in when he loses that drive. Roger is almost certainly done winning Slams, unless he can find one more burst like he did in 2017 - but it doesn't seem likely. I'm guessing that he retires after 2022, with that year being more of a victory tour. Rafa, I think, will retire maybe a year later, after 2023. I also wouldn't be surprised if Rafa wins RG next year and does a Sampras, especially if it is clear that he can't beat Novak elsewhere. Once he is done, Novak won't be able to sustain it. IMO, of course.

My guess for the final Slam tally is:
Novak 23-24
Rafa 21
Roger 20

 

[GameSetAndMath]

I am not asking for slam tally prediction.

I am asking as to what is the current streak strength of the current streak of Novak (even though the streak is not yet over).

 

[GameSetAndMath]

I am not asking for slam tally prediction.

I am asking as to what is the current streak strength of the current streak of Novak (even though the streak is not yet over).

I just calculated it myself.

Currently, Novak is on a streak of 12 and counting.

His streak strength of the current (unfinished) streak is 1523. The strength of this streak is stronger than his previous streak.

So, this streak strength is also weaker than that of Fed. But, as this is an unfinished streak, there is a potential for him to increase its length and strength.

 

[GameSetAndMath]

As far as the streaks is concerned, the big difference is that Novak jumps to the top with his streak from the 2010 USO through the 2016 USO - 25 Slams (!). So we have:

View attachment 5119
I also used the color scheme that Wikipedia and Ultimate Tennis uses, rather than the article. My apologies for not polishing these up - I tried to get them up as quickly as possible.

As you can see, Roger still has the highest average points during his streak.

Can you double check your work. I think Ralph's streak is of length 15 and not 13 as you have posted above.

 

[GameSetAndMath]

Federer's second longest streak is 6 and that was before his longest streak.

Ralph's second longest streak is also 6, but he had two such streaks.

Amazingly, Novak is the only one with two streaks of length 10 or more. Hats off to him.

This means, so to speak Novak has two peak periods, one before Bollywood and one after Bollywood.

 

[El Dude]

Federer's second longest streak is 6 and that was before his longest streak.

Ralph's second longest streak is also 6, but he had two such streaks.

Amazingly, Novak is the only one with two streaks of length 10 or more. Hats off to him.

This means, so to speak Novak has two peak periods, one before Bollywood and one after Bollywood.

Bollywood? LOL.

Anyhow, you're right - his streak is 15 (2010 RG to 2014 RG) although it is tricky because he missed two Slams (2012 USO and 2013 AO). The article was written at the beginning of 2014, so didn't include the last two Slams of his final. I'll update the chart a bit later.