SITE NEWS: We are moving all of our site and company news into a single blog for Sports-Reference.com. We'll tag all PFR content, so you can quickly and easily find the content you want.
Also, our existing PFR blog rss feed will be redirected to the new site's feed.
Pro-Football-Reference.com ยป Sports Reference
For more from Chase and Jason, check out their work at Football Perspective and The Big Lead.
ShareThisMost Dominant RB seasons
Yesterday, I explained the methodology employed to rank every RB in every season in NFL history. The table below lists the 50 most dominant RB seasons of all time. Why does Simpson's '75 season rank as the greatest ever? Let's run through the formula.
He played in 14 of a possible 14 games that season, and rushed for 1817 yards with 7 fumbles and 1 fumble recovery. That gives him 1667 adjusted rushing yards. He had 23 combined TDs from scrimmage. He recorded 28 receptions for 426 receiving yards, which translates to 468 adjusted catch yards. So how do we compute those values? There are too many things to show in the table, so (not listed) you need to know that Simpson had 119.1 adjusted rushing yards per game while the average starting NFL RB not named O.J. averaged just 53.7 adjusted rushing yards per game. That's a difference of 65.4 ARY per game, over 14 games. That means the Juice added 916 adjusted rushing yards on the season more than the average back; since he did not play in a 16 game season, we must pro-rate his score. As usual, I averaged the number of games on the NFL schedule that season and 16; here, that's 15. So we multiply 916 by 15/14 and get 981.
For TDs, the average RB had 0.68 TD per game in '75. That means O.J.'s 1.64 TDs per game translates to 0.96 more touchdowns per game, or 13.5 TDs on the season. Multiplied by 15/14 and you get his TD value of 14.4. Finally, he averaged 33.4 ACY/G while the average RB was at 28.8 ACY/G; do the math and you get a catch value of 70. By adding 981, 70 and 20*14.4 (since each TD is worth about 20 yards), you get 1339 adjusted yards over average, the greatest RB season in NFL history.
g/nfl ARY TTD ACY RVAL TDVAL CVAL VAL
O.J. Simpson 1975 BUF 14/14 1667 23 468 981 14.4 70 1339
Marshall Faulk 2000 STL 14/16 1409 26 952 452 17.1 475 1269
Jim Brown 1963 CLE 14/14 1688 15 304 1088 6.7 0 1223
Priest Holmes 2002 KAN 14/16 1590 24 777 603 14.4 327 1218
LaDainian Tomlinson 2006 SDG 16/16 1790 31 592 647 21.0 146 1213
Jim Brown 1958 CLE 12/12 1427 18 162 1000 9.8 0 1196
O.J. Simpson 1973 BUF 14/14 1828 12 79 1094 4.6 0 1186
Marshall Faulk 2001 STL 14/16 1357 21 890 437 13.8 472 1185
Walter Payton 1977 CHI 14/14 1702 16 310 992 9.6 0 1184
Terrell Davis 1998 DEN 16/16 1983 23 255 881 13.9 0 1159
Earl Campbell 1980 HOU 15/16 1884 13 64 1063 4.5 0 1154
Marshall Faulk 1999 STL 16/16 1331 12 1179 319 2.9 733 1109
Jim Brown 1965 CLE 14/14 1394 21 379 841 11.9 0 1078
Barry Sanders 1997 DET 16/16 2003 14 355 952 5.0 0 1052
LaDainian Tomlinson 2003 SDG 16/16 1645 17 875 501 6.9 413 1052
Eric Dickerson 1984 RAM 16/16 1855 14 171 907 4.2 0 990
Shaun Alexander 2005 SEA 16/16 1755 28 101 613 17.8 0 970
Barry Sanders 1994 DET 16/16 1883 8 349 963 0.0 0 963
Emmitt Smith 1995 DAL 16/16 1598 25 468 642 15.7 0 956
Steven Jackson 2006 STL 16/16 1478 16 941 325 5.5 506 941
Tiki Barber 2005 NYG 16/16 1860 11 611 722 0.3 205 932
Priest Holmes 2003 KAN 16/16 1395 27 801 243 17.2 337 924
Jim Brown 1959 CLE 12/12 1304 14 226 796 6.1 0 919
Emmitt Smith 1992 DAL 16/16 1638 19 424 708 10.3 0 914
Jim Taylor 1962 GNB 14/14 1374 19 139 711 9.8 0 907
Marcus Allen 1985 RAI 16/16 1734 14 656 738 3.2 92 893
Brian Westbrook 2007 PHI 15/16 1283 12 906 268 3.6 519 859
Emmitt Smith 1994 DAL 15/16 1459 22 416 573 13.9 0 852
Barry Sanders 1991 DET 15/16 1448 17 369 660 8.6 0 831
Jamal Lewis 2003 BAL 16/16 1891 14 244 755 3.8 0 830
Earl Campbell 1979 HOU 16/16 1547 19 118 665 8.3 0 830
Terrell Davis 1997 DEN 15/16 1700 15 350 695 6.5 0 826
Edgerrin James 2000 IND 16/16 1584 18 689 520 7.7 146 820
Thurman Thomas 1991 BUF 15/16 1282 12 724 487 3.4 264 819
Ahman Green 2003 GNB 16/16 1758 20 442 617 10.0 0 817
James Wilder 1984 TAM 16/16 1394 13 813 429 3.1 325 816
Larry Johnson 2006 KAN 16/16 1764 19 472 620 8.6 21 814
Marshall Faulk 1998 IND 16/16 1294 10 1037 168 0.5 631 809
Jim Taylor 1961 GNB 14/14 1282 16 213 626 8.3 0 793
Jim Brown 1961 CLE 14/14 1283 10 528 627 1.4 127 783
Eric Dickerson 1988 IND 16/16 1559 15 431 672 5.1 0 775
Edgerrin James 1999 IND 16/16 1403 17 679 394 8.0 216 771
Jamal Anderson 1998 ATL 16/16 1746 16 360 635 6.7 0 769
Larry Johnson 2005 KAN 16/16 1700 21 393 557 10.6 0 768
Chuck Foreman 1975 MIN 14/14 820 22 801 37 13.3 440 743
LaDainian Tomlinson 2007 SDG 16/16 1474 18 565 409 9.3 145 740
Ricky Williams 2002 MIA 16/16 1703 17 434 603 6.0 0 722
Gerald Riggs 1985 ATL 16/16 1719 10 317 722 0.0 0 722
Emmitt Smith 1993 DAL 14/16 1461 10 500 635 3.4 20 722
Jim Brown 1964 CLE 14/14 1346 9 394 717 0.0 0 717
Jim Brown leads all backs with six top 50 seasons. Marshall Faulk ('98-'01) and Emmitt Smith ('92-'95) had top fifty performances in four straight seasons. Tomlinson and Sanders each have three seasons that made the cut. Simpson has two of the top ten seasons of all time, and Priest Holmes, Earl Campbell, Edge, Dickerson, Jim Taylor, Larry Johnson and Terrell Davis all have a pair of top 50 seasons. Dickerson has two more seasons in the 51-70 range.
Here's a look at the best RB season for each of the current 32 franchises:
year team G/NFL ary ttd acy RSHV TDV CATV VAL O.J. Simpson 1975 buf 14/14 1667 23 468 981 14.4 70 1339 Marshall Faulk 2000 ram 14/16 1409 26 952 452 17.1 475 1269 Jim Brown 1963 cle 14/14 1688 15 304 1088 6.7 0 1223 Priest Holmes 2002 kan 14/16 1590 24 777 603 14.4 327 1218 LaDainian Tomlinson 2006 sdg 16/16 1790 31 592 647 21.0 146 1213 Walter Payton 1977 chi 14/14 1702 16 310 992 9.6 0 1184 Terrell Davis 1998 den 16/16 1983 23 255 881 13.9 0 1159 Earl Campbell 1980 oti 15/16 1884 13 64 1063 4.5 0 1154 Barry Sanders 1997 det 16/16 2003 14 355 952 5.0 0 1052 Shaun Alexander 2005 sea 16/16 1755 28 101 613 17.8 0 970 Emmitt Smith 1995 dal 16/16 1598 25 468 642 15.7 0 956 Tiki Barber 2005 nyg 16/16 1860 11 611 722 0.3 205 932 Jim Taylor 1962 gnb 14/14 1374 19 139 711 9.8 0 907 Marcus Allen 1985 rai 16/16 1734 14 656 738 3.2 92 893 Brian Westbrook 2007 phi 15/16 1283 12 906 268 3.6 519 859 Jamal Lewis 2003 rav 16/16 1891 14 244 755 3.8 0 830 Edgerrin James 2000 clt 16/16 1584 18 689 520 7.7 146 820 James Wilder 1984 tam 16/16 1394 13 813 429 3.1 325 816 Jamal Anderson 1998 atl 16/16 1746 16 360 635 6.7 0 769 Chuck Foreman 1975 min 14/14 820 22 801 37 13.3 440 743 Ricky Williams 2002 mia 16/16 1703 17 434 603 6.0 0 722 Jim Nance 1966 nwe 14/14 1283 11 115 657 2.0 0 698 DeAngelo Williams 2008 car 16/16 1518 20 154 469 9.1 0 651 Stephen Davis 1999 was 14/16 1355 17 146 451 9.0 0 631 Barry Foster 1992 pit 16/16 1515 11 398 581 2.0 0 620 Roger Craig 1988 sfo 16/16 1352 10 648 458 0.0 139 596 Curtis Martin 2004 nyj 16/16 1697 14 307 528 2.4 0 577 Deuce McAllister 2003 nor 16/16 1566 8 620 419 0.0 150 569 Fred Taylor 2000 jax 13/16 1349 14 294 447 5.2 0 551 Ottis Anderson 1979 crd 16/16 1380 10 370 491 0.0 0 491 Rudi Johnson 2005 cin 16/16 1483 12 125 333 1.3 0 359 Domanick Williams 2004 htx 15/16 1088 14 690 - 39 3.0 277 299
Tomorrow, I'm going to post the all time career list. On Thursday, I'm going to bring post-season numbers into the discussion and look at the most dominant playoff performances in NFL history. Friday brings new lists -- a career ranking with post-season numbers included and the top single season stars including the playoffs.
Before we move on, I'd like to address two RBs who won't make much noise over the course of this five-part series. That's why I'd like to focus on them for a minute now.
- Marion Motley: As told by the great Sean Lahman in The Pro Football Historical Abstract, Motley's NFL numbers simply don't tell the story. There are two good reasons for that. First,
Motley spent nearly five years after college serving in the U.S. Navy, costing him most of his prime football years. [Chase note: Although it was here that he met Paul Brown.] The second problem is that when he did turn pro, he started his career in the AAFC, a league that didn't have much competitive balance. Motley was an unstoppable avalanche, completely overwhelming opposing defenses. He averaged 6.2 yards per carry and helped the Cleveland Browns compile a 47-4-3 record and win all four AAFC Championships. [Chase note: In 1948, he led the Browns to a perfect 15-0 record and rushed 14 times for 133 yards and 3 scores in the championship game.] Motley led the NFL in rushing yards in 1950, his (and the Browns') first year in the league. He was already thirty by this time, and injuries were beginning to take their toll.
Motley rushed for just five touchdowns in his NFL career. But in a memorable game against the Steelers in 1950, he rushed 11 times for 188 yards and one score; that's a remarkable 17.1 yards per carry average. He also caught a 33 yard TD pass that game.
One more anecdote from Lahman's book, this time told by Paul Zimmerman ("Dr. Z"): If there is a better football player who snapped on a helmet, I would like to know his name. [Jim] Brown was the best pure runner I've ever seen, but Motley was the greatest all-around player, the complete player." In case you didn't know, Motley was also a devastating linebacker for Cleveland. Here's another great Motley story:
He began playing primarily at fullback when the two-platoon system was generally adopted in 1948, but was still used at linebacker at crucial times. In the Browns' first game in the National Football League, the Philadelphia Eagles had a first and goal at Cleveland's 6-yard line and Motley was put in at middle linebacker. Needing a touchdown, the Eagles ran the ball four times. Motley made the tackle each time. The four plays gained a total of three yards and Cleveland took over on downs.
But most remember him for bringing power football to Cleveland, later sustained by Jim Brown and Leroy Kelly.
- Gale Sayers: Sayers ranks #33 all time on tomorrow's list, higher than on almost any other objective list of career totals. But I suspect he's quite a bit better than the 33rd best RB of all time. From 1965 to 1969, he averaged an incredible 5.1 yards per carry. He also was taking punt and kick return duty, which likely cut down somewhat on his number of carries. The biggest reason Sayers ranks low on the career list and why his best season is just the 64th best of all time is the low number of carries. Barry Sanders and O.J. Simpson both averaged over six yards per carry one season ... and also had enough carries to hit 2,000 yards. Even in Sayers' best year, 1966, he only ranked 4th in carries. (His '66 ranks ahead of his '65 because of his 9 fumbles in '65). The obvious question is, 'Why?' Why did RBs like Bill Brown, Jim Nance and Dick Bass get more carries for their teams than Gale Sayers did for the Bears? Why did teammates Ronnie Bull, Jon Arnett and Ralph Kurek get 207 carries during Sayers' best season, when he averaged over two yards per carry more than them? No one ever called George Halas an idiot, so the two reasons were probably: 1) he didn't want to overuse his special talent, and 2) it was uncommon in that era to have a workhorse back that looked like Sayers. And really, both of those points are true.
We'll never know if Sayers could have handled another 50-75 carries a year and kept up his production, but I suspect he could have, and would have, and would then rank in my all time top ten. On the other hand, consider that three guys who averaged 3.3 YPC got 200+ carries in '66, Sayers saw 200+ carries and averaged over five yards per carry, and the Bears had a losing record. One would think that if the Bears weren't winning many games, they would have given Sayers a bunch more carries. And while maybe Bull, Arnett and Kurek weren't very good, maybe they carried the ball in short yardage situations and Sayers carried the ball in advantageous situations. If that's the case, then you really can't compare Sayers' yards per carry average to the YPC of the do-it-all RBs who are all time greats. But that's just speculation. I will note that it's odd that Chicago had Sayers and Dick Butkus on the same team yet never had much success during the late '60s.
That said, whenever he touched the ball, he sure looked like an all time great. And no one has ever been as good as Sayers was on a cold day in December, 1965. On opening day of that season, the 49ers throttled the Bears at old Kezar Stadium, and cruised to a 52-10 lead by the 4th quarter. Just three months later, the rookie exacted revenge: he scored six TDs, rushed 9 times for 113 yards (and 4 scores), caught 2 passes for 91 yards (including an 80 yard TD) and returned 5 punts for 124 yards (with an 85 yard score).
This entry was posted on Tuesday, April 7th, 2009 at 7:03 am and is filed under History. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.
Great stuff as always! I'm just curious about one thing, and I'm sure you've talked about this before, I just don't remember. When you are correcting for a 14 game season, why do you multiply by (15/14) instead of (16/14)?
Hey Andrew,
The basic reason is that it's easier to be dominant for 9 games, or 12 games or 14 games than it is for 16 games. A simple pro-rating is probably unfair to the modern guys, because of regression to the mean. OTOH, obviously *some* adjustment must be made. I don't think 15/14 is perfect, but it's simple and didn't take a lot of thought.
Thanks for the explanation. Do you guys use the same formula for the other rankings (such as the all-time best QB) too? I've never paid too much attention to the pro-rating, but now that they are talking about going to an 18 game season I am more curious.
Andrew, there's no official PFR stance. What I've done for the Best QB Ever and the Best WR Ever was to do the same thing. So a 9 game season is pro-rated to a 12.5 game season, a 12 game season to a 14 game season and a 14 game season to a 16 game season. Before you think this isn't enough, though, remember this isn't necessarily a cumulative rating. Consider a short example.
Say RB A averaged 100 yards/game in 1965, and the average RB was at 60 yards/game. He'd be +40 for 14 games, which is +560, and pro-rated (15*560/14) +600. So his 1400 yards becomes +600. What would +600 be in a 16 game schedule? 97.5 yards/game. Do you see why?
That basically says he would have averaged 80 yards per game over the next two games. Certainly averaging 100 yards/game over 16 games is more impressive than averaging 100 yards/game over 14 games, so there should be some bonus. If you were at 100/g for 16 games, you'd be +640. So instead of pro-rating the +560 to +640, I make it only +600. I'm not arguing that it's perfect, but it feels right to me. And it's very simple.
I liked the methodology for determining the rankings... but I'm not sure how much I liked how the top 50 RB seasons ever turned out. Jim Brown and Jim Taylor are the only players prior to 1972 to make the list, making up 8 of the top 50 seasons. When you did your WR list, 22 of the top 50 seasons were prior to 1972. Also 12 of the top 50 seasons are from the past 6 years, which also feels a little weird to me.
The first initial thought I had to why this might be is in the past, teams were more run oriented, so it was easier for a few individual receivers to stand out and be dominant compared to the others of their era. Today teams are more pass oriented, so it's easier for a runner to stand out. Also with the past few years, having so many teams using 2-3 backs make feature backs stand out more.
My other thought was maybe the catch value was too high. 22 of the 50 RB's came from 1998-2008, while the other 28 came from 1958-1997. 15 of the 22 from 1998-2008 benefited from the catch value, while only 6 of the 28 from 1958-1997 benefited from the catch value, 5 of those 6 being in the bottom 25.
Good point, Nathan.
As Doug talked about here -- http://www.pro-football-reference.com/blog/?p=1886 -- RBs weren't workhorses before the James Wilder era. So to some extent, changes in philosophy affect this list. A guy who takes all his team's carries is probably more valuable and more dominant than one who splits carries. And before the '80s, almost everyone split carries. On the career list, though, older guys are (slightly) better represented.
For a season in the 60s the Bears had Sayers, Butkus AND Ditka and still couldn't go too far. Not Coincidentally, that was their best season with Gayers, going 9-6.
Just a quick addendum to my previous post. The only other time I can think of that a team had arguably the best RB, LB and TE in the game at the same time, the Chargers did pretty good in the regular season.
I have little doubt that Brown in '58 and '63, plus Simpson in '73 and '75 were the four best RB seasons in history. I think your system overrates players who missed games (like Faulk in 2000-01, and Holmes in '02) and underrates guys in 12- and 14-game seasons. Jim Brown wasn't going to regress to the mean.
In 1958 he tied the single-season record for TDs and had almost twice as many rush yards as the next-best RB. How is that not top-5? I think it's the best season in history.
Your top 50 includes 16 seasons from the 2000s and NONE from the 30s or 40s and most of the 50s. I think this is obviously wrong.
I haven't really studied pre-Modern Era seasons, but where did Beattie Feathers' 1934 season and Steve Van Buren in 1945 fit in?
Brad O.,
Van Buren's best year was '47, which ranked 60th. His '45 ranks '69th, and is not War-adjusted. Like Jim Benton, that's always a concern (http://www.pro-football-reference.com/blog/?p=1362) with the War years. Feathers' '34 ranks 90th.
Let's use a straight pro-rating; none of this adjusted pro-rating business. How does that change the list?
Van Buren's '45 moves up to 41; his '47 is now 43rd. Feathers' '34 moves up only to '80 because he missed two games that season.
As for Brown, his '58 moves up to #2, right behind Simpson. The top 5 seasons are Simpson '75, Brown '58, Brown '63, Faulk '00 and Simpson '73. Payton's '77 moves up to #6.
Why does my top fifty include only seasons from the '50s and forward? Because of the evolution of the game. In 1948, Steve Van Buren led the NFL in carries with 201; his tema rushed 528 times! Fifty years later, Jamal Anderson led the league with 410 carries and Atlanta rushed just 516 times! Is it really so bad that Anderson's '98 ranks ahead of Van Buren's 98? Van Buren may have been the better RB, but that's not enough. As good as he was, he was limited by the philosophy of the day.