If you’ve found your way here from the Wall Street Journal, welcome! If you don’t know what I’m talking about, go read what Carl Bialik has to say in today’s paper, and in an online follow-up. I’ve written at length about my rankings and prediction system and published full odds for Indian Wells here.
As you may have read in the Wall Street Journal, my ranking system rates Federer number one. The difference between Fed and Nadal is even more striking if we use my hard-court-specific rankings. However, in the hard-court-specific rankings, Djokovic closes the gap quite a bit.
Before you email me to tell me what an idiot I am for publishing something so blatantly wrong, please read my description of what the system does.
The goal of these rankings isn’t to say who is the greatest of all time, or to say that any player here is guaranteed to beat anyone below him. Instead, they are the result of an algorithm that is better than anything else I’ve seen at predicting the outcome of tennis matches.
Here are the current top 100 hard-court players, along with the hard-court rankings of several other players who are in the Indian Wells main draw:
1 Roger Federer 8579 2 Novak Djokovic 6853 3 Andy Murray 5013 4 Rafael Nadal 4892 5 Robin Soderling 4363 6 Juan Martin del Potro 3624 7 Nikolay Davydenko 3118 8 David Ferrer 2913 9 Andy Roddick 2671 10 Tomas Berdych 2284 11 Gael Monfils 2226 12 Stanislas Wawrinka 2094 13 Marcos Baghdatis 2062 14 David Nalbandian 1967 15 Mardy Fish 1961 16 Marin Cilic 1779 17 Fernando Verdasco 1709 18 Jurgen Melzer 1615 19 Ivan Ljubicic 1602 20 Jo-Wilfried Tsonga 1565 21 Michael Llodra 1475 22 Mikhail Youzhny 1317 23 Gilles Simon 1314 24 Florian Mayer 1312 25 Nicolas Almagro 1305 26 Milos Raonic 1231 27 Alexander Dolgopolov 1223 28 Guillermo Garcia-Lopez 1109 29 Juan Monaco 1102 30 Richard Gasquet 1091 31 Radek Stepanek 1044 32 Viktor Troicki 1021 33 John Isner 901 34 Lleyton Hewitt 883 35 Tommy Robredo 867 36 Albert Montanes 841 37 Jeremy Chardy 840 38 Ernests Gulbis 820 39 Philipp Kohlschreiber 796 40 Feliciano Lopez 787 41 Samuel Querrey 773 42 Janko Tipsarevic 734 43 Fernando Gonzalez 711 44 Julien Benneteau 695 45 Kei Nishikori 686 46 Jarkko Nieminen 638 47 Juan Carlos Ferrero 635 48 Dmitry Tursunov 633 49 Xavier Malisse 588 50 Thomaz Bellucci 578 51 Ivo Karlovic 559 52 Andreas Seppi 507 53 Andrei Goloubev 488 54 Benjamin Becker 487 55 Michael Berrer 466 56 Thiemo de Bakker 457 57 Igor Andreev 455 58 Olivier Rochus 449 59 Philipp Petzschner 447 60 Juan Ignacio Chela 434 61 Fabio Fognini 434 62 James Blake 432 63 Pablo Cuevas 426 64 Santiago Giraldo 413 65 Sergey Stakhovsky 402 66 Denis Istomin 400 67 Ivan Dodig 389 68 Arnaud Clement 375 69 Michael Zverev 367 70 Robin Haase 367 71 Leonardo Mayer 352 72 Robby Ginepri 351 73 Marcel Granollers 350 74 Daniel Brands 345 75 Alejandro Falla 341 76 Daniel Gimeno 341 77 Paul-Henri Mathieu 341 78 Mikhail Kukushkin 330 79 Dudi Sela 325 80 Lukasz Kubot 324 81 Teimuraz Gabashvili 303 82 Victor Hanescu 288 83 Grigor Dimitrov 284 84 Lukas Lacko 282 85 Adrian Mannarino 279 86 Kevin Anderson 275 87 Florent Serra 275 88 Simon Greul 274 89 Potito Starace 270 90 Edouard Roger-Vasselin 269 91 Frank Dancevic 269 92 Horacio Zeballos 268 93 Richard Berankis 266 94 Marco Chiudinelli 264 95 Rainer Schuettler 263 96 Ryan Harrison 262 97 Frederico Gil 261 98 Bernard Tomic 260 99 Nicolas Mahut 259 100 Tobias Kamke 259 102 Yen-Hsun Lu 255 104 Bjorn Phau 248 106 Chris Guccione 247 107 Ryan Sweeting 246 112 Ricardo Mello 240 114 Ilia Marchenko 236 116 Matt Ebden 233 120 Alex Bogomolov 228 121 Michael Russell 226 133 Marinko Matosevic 221 141 Dustin Brown 217 144 Donald Young 216 145 Tim Smyczek 215 147 Somdev Devvarman 215 156 Rik de Voest 212 174 Marsel Ilhan 208 196 Flavio Cipolla 202 261 Rohan Bopanna 109 319 Pere Riba 55 354 Ruben Ramirez-Hidalgo 22
How do you, if you do, other than result, account for injury? I am specifically thinking of Nadal’s knees and hamstring injuries as they definitely account for a large part of his losses vs lower ranked opponents.
I don’t account for injury. You’re right that we can point to certain matches and say that an injury was to blame, and if we are correct about that, it would make sense to discount those results. But, without actual medical data on the players, there’s not much we can do at the algorithmic level.
For one thing, I ignore retirements entirely, so you’re only “penalized” for injury if you play out the entire match. And if a player is injured and plays out the entire match at one time, I suspect it is more likely that he would do so again … thus leading to more results like the ones that are tempting to discount or throw out.
A familiar phrase in baseball analysis circles is that “health is a skill” — in general, guys who get injured a lot keep getting injured, and guys who have stayed healthy in the past are more likely to stay healthy in the future. It certainly does make it difficult to predict the level at which Nadal will play in any given week.