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 » Sports Reference
For more from Chase and Jason, check out their work at Football Perspective and The Big Lead.
ShareThisPart III: Greatest WRs Ever
On Monday, I described a formula I created to rate every WR season in NFL history. On Tuesday, I listed the best seasons ever, the best seasons each year and the best season by any receiver for each franchise. Today, we'll get to the career list.
Compiling career value isn't that difficult. I took 100% of the WR's best season, 95% of his second best season, 90% of his third best, and so on. Let's use Michael Irvin as an example, who had nine seasons in his career where he ranked above the baseline.
year rec recyd rectd ACY tmatt ACY/A NFL Value Wt CarVal 1995 111 1603 10 2358 494 4.77 2.22 1262 100 1262 1991 93 1523 8 2148 500 4.30 2.35 972 95 923 1993 88 1330 7 1910 475 4.02 2.13 900 90 810 1992 78 1396 7 1926 491 3.92 2.25 820 85 697 1994 79 1241 6 1756 448 3.92 2.18 781 80 625 1996 64 962 2 1322 487 3.95 2.21 582 75 437 1997 75 1180 9 1735 553 3.14 2.14 552 70 386 1998 74 1057 1 1447 474 3.05 2.27 372 65 242 1989 26 378 2 548 513 2.85 2.31 104 60 62 5443
This is the same formula we've used here to rate the QBs and for Doug's Approximate Value method. Once you add up the career values for every WR, you get the following list of the top 100 WRs. Perhaps more specifically, these guys have created the most value for their teams in their careers:
rk name val rookyr 1 Jerry Rice 9095 1985 2 Don Hutson 7335 1935 3 Marvin Harrison 6890 1996 4 Terrell Owens 6463 1996 5 Randy Moss 6260 1998 6 Lance Alworth 6045 1962 7 Steve Largent 5755 1976 8 Michael Irvin 5443 1988 9 Raymond Berry 5367 1955 10 Jimmy Smith 5352 1992 11 Tim Brown 5182 1988 12 Paul Warfield 5141 1964 13 Don Maynard 5061 1958 14 Torry Holt 5038 1999 15 James Lofton 4930 1978 16 Bob Hayes 4909 1965 17 Cris Carter 4894 1987 18 Harold Jackson 4785 1968 19 Isaac Bruce 4752 1994 20 Rod Smith 4687 1995 21 Tommy McDonald 4642 1957 22 Charley Taylor 4594 1964 23 Herman Moore 4564 1991 24 Bobby Mitchell 4377 1958 25 Otis Taylor 4292 1965 26 Henry Ellard 4289 1983 27 Cliff Branch 4222 1972 28 Steve Smith 4211 2001 29 Hines Ward 4197 1998 30 Del Shofner 4146 1957 31 Fred Biletnikoff 4072 1965 32 Stanley Morgan 4059 1977 33 Billy Howton 3987 1952 34 John Stallworth 3961 1974 35 Andre Reed 3897 1985 36 Art Monk 3897 1980 37 Chad Johnson 3884 2001 38 Harold Carmichael 3873 1971 39 Art Powell 3777 1959 40 Sonny Randle 3773 1959 41 Mac Speedie 3765 1946 42 Buddy Dial 3758 1959 43 Dante Lavelli 3707 1946 44 Billy Wilson 3638 1951 45 Jimmy Orr 3627 1958 46 Gary Clark 3599 1985 47 Gary Garrison 3560 1966 48 Sterling Sharpe 3513 1988 49 Derrick Mason 3495 1997 50 John Gilliam 3416 1967 51 Elroy Hirsch 3383 1946 52 Eric Moulds 3377 1996 53 Gene A. Washington 3363 1969 54 Muhsin Muhammad 3362 1996 55 Andre Johnson 3340 2003 56 Boyd Dowler 3331 1959 57 Joe Horn 3324 1996 58 Jim Benton 3259 1938 59 Max McGee 3231 1954 60 Roy Jefferson 3206 1965 61 Tony Hill 3158 1977 62 Pete Pihos 3153 1947 63 Wesley Walker 3131 1977 64 Wes Chandler 3129 1978 65 Gary Collins 3098 1962 66 Drew Pearson 3068 1973 67 Roy Green 3059 1979 68 Red Phillips 3046 1958 69 Anquan Boldin 3040 2003 70 Joey Galloway 3026 1995 71 Hugh Taylor 3009 1947 72 Andre Rison 3000 1989 73 Ken Burrough 2983 1970 74 Keenan McCardell 2975 1992 75 Keyshawn Johnson 2975 1996 76 Cris Collinsworth 2971 1981 77 Homer Jones 2965 1964 78 Harlon Hill 2945 1954 79 Reggie Wayne 2940 2001 80 Charlie Joiner 2912 1969 81 Tom Fears 2834 1948 82 Lionel Taylor 2825 1959 83 Carroll Dale 2807 1960 84 Alfred Jenkins 2801 1975 85 Lynn Swann 2792 1974 86 Mike Quick 2790 1982 87 Frank Clarke 2764 1957 88 Nat Moore 2744 1974 89 Laveranues Coles 2732 2000 90 Gail Cogdill 2694 1960 91 Mark Clayton 2662 1983 92 Drew Hill 2661 1979 93 Charley Hennigan 2632 1960 94 John Jefferson 2617 1978 95 Lance Rentzel 2613 1965 96 Paul Flatley 2562 1963 97 Plaxico Burress 2554 2000 98 Mel Gray 2552 1971 99 Johnny Morris 2536 1958 100 Larry Fitzgerald 2526 2004
Note: Only seasons where the player was a WR were counted. So Charley Taylor and Bobby Mitchell only get credit for their seasons as WRs and not when they were RBs; similarly, Lenny Moore doesn’t even make the list. Carroll Dale doesn’t get credit for his seasons as a TE.
Obviously a few players on that list stand out. Irvin and Jimmy Smith are surprise finishers in the top ten. Bob Hayes ranks ahead of Cris Carter, and he just made the HOF (and so far, Carter has not). Using per-attempt numbers obviously helps some players like Irvin, who did not have the same opportunity to rack up big passing numbers. How much was Irvin disadvantaged (by traditional methods) is an interesting question. I took a weighted look at how often the teams the WRs were on passed the ball (relative to league average) during the best seasons of the WRs' careers.
1.29 Charley Hennigan
1.24 Lionel Taylor
1.22 Don Hutson
1.19 Tom Fears
1.18 Larry Fitzgerald
1.17 Mark Clayton
1.15 Art Powell
1.15 Charlie Joiner
1.15 Don Maynard
1.14 Anquan Boldin
1.14 John Jefferson
1.13 Torry Holt
1.12 Wes Chandler
1.10 Gary Garrison
1.10 Sterling Sharpe
1.09 Billy Howton
1.08 Andre Rison
1.08 Elroy Hirsch
1.07 Harold Carmichael
1.07 Jim Benton
1.07 Isaac Bruce
1.06 Bobby Mitchell
1.06 Pete Pihos
1.06 Marvin Harrison
1.06 Charley Taylor
1.06 Del Shofner
1.05 Reggie Wayne
1.05 Jerry Rice
1.05 Mike Quick
1.05 Lance Alworth
1.05 Raymond Berry
1.05 Johnny Morris
1.04 Cris Carter
1.04 Mel Gray
1.04 Chad Johnson
1.04 Fred Biletnikoff
1.04 John Gilliam
1.04 Tommy McDonald
1.04 Drew Pearson
1.03 Tony Hill
1.03 Terrell Owens
1.03 Gene A. Washington
1.03 Roy Jefferson
1.03 Randy Moss
1.02 Steve Largent
1.02 Joe Horn
1.02 Gary Clark
1.01 Keenan McCardell
1.01 Harlon Hill
1.01 Roy Green
1.00 Alfred Jenkins
1.00 Billy Wilson
1.00 Keyshawn Johnson
0.99 Cris Collinsworth
0.99 Rod Smith
0.99 James Lofton
0.99 Art Monk
0.98 Muhsin Muhammad
0.98 Jimmy Smith
0.98 Derrick Mason
0.98 Ken Burrough
0.98 Mac Speedie
0.98 Eric Moulds
0.97 Tim Brown
0.97 Drew Hill
0.97 Andre Johnson
0.97 Herman Moore
0.97 Henry Ellard
0.97 Harold Jackson
0.97 Frank Clarke
0.97 Homer Jones
0.97 Jimmy Orr
0.96 Wesley Walker
0.96 Sonny Randle
0.96 Plaxico Burress
0.96 Andre Reed
0.96 Joey Galloway
0.96 Cliff Branch
0.95 Laveranues Coles
0.95 Lance Rentzel
0.95 Gail Cogdill
0.95 Red Phillips
0.94 John Stallworth
0.94 Michael Irvin
0.94 Bob Hayes
0.92 Lynn Swann
0.92 Stanley Morgan
0.92 Otis Taylor
0.90 Hines Ward
0.89 Steve Smith
0.89 Dante Lavelli
0.88 Hugh Taylor
0.87 Gary Collins
0.86 Nat Moore
0.85 Paul Flatley
0.85 Paul Warfield
0.83 Carroll Dale
0.83 Buddy Dial
0.81 Max McGee
0.80 Boyd Dowler
This list serves as a nice check to remind you about the shape of some player’s careers. Because of the K-Gun offense, I picture Reed playing on aerial offenses for most of his career, but that wasn’t the case. Steve Smith has put up big time numbers despite constantly playing on run oriented teams. Paul Warfield was an elite talent who simply didn’t play on explosive teams. Charlie Joiner and Torry Holt, of course, go the other way.
Joiner’s an interesting case. Everyone brings up Swann as an overrated HOF WR, but Joiner looks just as bad. At the time of his enshrinement, he ranked 6th in career receiving yards, 5th in receptions and 27th in receiving touchdowns. Now he ranks 16th, 27th and 40th in those categories. I suspect the fact that he was the all time leader in receptions when he retired swayed some voters, but Joiner stuck around for 18 seasons. He was a very good WR, but did not put up the typical numbers of a HOF WR. Consider:
- Joiner’s top two seasons ranked as the 3rd and 10th best in the league those years. He’s got one top five, one more top ten and one more top fifteen season in his career. That’s it.
- He had only four 1,000 yard seasons and never scored eight touchdowns in a single year.
- He played on teams that passed much more often than the rest of the league.
That said, number three cuts both ways. He played with Wes Chandler, John Jefferson and of course Kellen Winslow. So while he gets penalized for being on teams that passed frequently, he doesn’t get bonus points for having to compete with some elite talent. On the other hand, he also got to play with Dan Fouts for the majority of his career. With the Bengals in ‘74 and ‘75, at the ages of 27 and 28, playing with Ken Anderson, he got outplayed by teammate Isaac Curtis, too. Maybe Joiner isn’t the 80th best WR of all time, but it seems to me like he’s no more worthy of enshrinement than Swann.
I don't think this list is perfect, and I'm sure I'll have some good modifications for next time around. Advice and critiques are always welcome. Joiner was just one guy that stuck out to me; I'm curious to see who sticks out for you.
This entry was posted on Wednesday, February 11th, 2009 at 7:57 am and is filed under General. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.
Good stuff. I've actually been working on some of my own WR stuff this week (trying to create a HOF Probability stat like at pro-basketball-reference.com, but just for WR's) and while I'm nowhere close to being done, I am getting some initial results that confirm my belief that Charlie Joiner is the worst WR in the HOF. To play that long with so many quality QB's/passing systems, you have to have more good years than what he did. He rode "longevity" to the HOF.
Nice to see Hines Ward crack the top 30.
Jimmy Smith seems way too high. But for the most part it looks like a very good result. I simply can't imagine a list including Jimmy Smith in the top 10 though. He would not even make my top 20. His 5 pro-bowls is good but far behind Tim Brown (9), James Lofton (8), and Chris Carter (8).
Doug, Smith definitely stood out at first. But now I'm thinking he was just underrated. He didn't score many TDs, he played in a small market, he stunk early in his career and he didn't hang around long enough to have great career numbers. But none of those things -- even combined -- prevent him from being a great WR. He was awesome at his peak -- consider:
He's ranked in the top five in receiving yards in six seasons.
Cris Carter - zero top five seasons.
Tim Brown - five.
Holt - four times, consider his offenses passed 13% more than average and Smith's were 2% fewer than average.
Bruce - four times, but consider his offenses passed 7% more than average.
Art Monk - two
Largent - four
Five top five yardage seasons is really, really good. And his 1999 season was fantastic, and he was a big part of why that team was 14-2.
Thinking out loud: I would love some kind of study of whether throwing more (=defenses focusing on the pass) and throwing less (=Defenses might line up single-man) cancels each other out from the recievers point of view.
Can I make a suggestion you don't hear very often around here? I think this list UNDERrates compilers. Between the heavy adjustment in favor of players whose teams don't throw a lot, and the weight on the player's best seasons, guys like Michael Irvin and Bob Hayes are higher than we might expect.
Look at Andre Reed and Art Monk, tied for 35th. They're barely ahead of Chad Johnson. Does anyone -- and I mean anyone, on the planet -- think Herman Moore (23rd) was better than Monk and Reed?
I can't make a statistical argument for Lynn Swann, but there's no way HOFers Charlie Joiner and Tom Fears should be behind Anquan Boldin or Joey Galloway (sorry, Doug). Joiner IS overrated, but he was the Isaac Bruce of his day, and Bruce is 19th. Tom Fears (81st) and Crazylegs Hirsch (51st) were the two best receivers of their era. Both are probably among the 20 best receivers in history. It would be nice to see them in the top 50.
I wouldn't want to see too many changes in your formula, because it's gratifying to see players like Hutson and Warfield near the top of the list where they belong. But you might want to consider tweaking the adjustment for team pass attempts, and maybe ease up on the season weighting. I think 100/96/92/etc. still gives the effect we're looking for, but might make the list a bit more intuitive.
Brad, your criticisms (while I agree with all of them) show the problem.
Monk and Reed are too low... but so are Fears and Hisrch. Problem is, Fears and Hirsch didn't stick around very long; so if we give more weight to compilers, that would drop Fears and Hirsch.
I'm not sure Irvin and Hayes are improperly rated, fwiw. Irvin was a big part of a very successful team. It's not like he was piling up meaningless numbers.
Re: Brad's ideas:
Chase, I'll say that your bell curve needs to be STEEPER. For example, Rice put up good numbers in his SEVENTEENTH!! yr. So his 10th best year is given 55% credit and is considered a BELOW average year (for him, at least). Without being able to see career numbers from everybody, I would dare say that an average year from Jerry Rice would be better than most guy's peak years. (I am a Saints' fan--I despise SF--but Rice was the GOAT) Thus, for guys who had shorter careers, even their BAD-last-gasp-hanging-on seasons get 50% credit. Jerry's last-gasp-hanging-on season/s get like 10% credit.
So, my curve would be based on the number of years he played--because the longer he played, the more likely that his peak was greater and longer. So, a guy like Rice gets this type of curve: 100/98/96/94/92/90/88/86/84/80/75/70/65 etc. A 10 year player gets 100/98/96/94/90/85/80 etc.
The second greatest season for Rice, Hutson, Largent, Harrison, Owens, et al. is probably twice as good as other players' second best season. This is why they have several seasons in the top 50 of all time. In other words, Rice's 5th best season--good enough for the top 50--gets 80% credit. My guess is that there are receivers in the all-time top 100 who don't have 5 seasons in the top 1000--and their 5th best season, which is probably average--gets as much percentile credit as one of the top 50. I know making a different bell curve for players based on their overall time in the league may make calculating their career value more different. But if a player since the 1970 merger didn't stay in the league for 10 years, he probably wasn't one of the best ever at his position.
[I will readily admit that there are some exceptions like Terrell Davis, Kurt Warner, and Gale Sayers who have put up INCREDIBLE peaks--however, these guys are the exceptions--not the rule.]
Aargh--"I know making a different bell curve for players based on their overall time in the league may make calculating their career value more different." SHOULD READ "more DIFFICULT."
Chase,
First of all, thanks for your patience with all of our suggestions. You said up front that this wasn't perfect, but I think the reason it's generating so many suggestions is that it's an interesting project.
I would guess that one of the biggest factors making parts of the list counterintuitive is the adjustment for team passes. The 1950 Rams and '07 Pats threw a lot because they were good at it, because they had great receivers. It would probably be helpful to soften that part of the formula.
I still like the idea of going from 95% value for the 2nd-best season to 96%, and continuing to drop by intervals of 4% rather than 5%. If compilers like Reed and Joiner are too low, that's a simple way to help them without changing anything too radically.
I'd also look into a more extreme era adjustment. Four of your top five and 12 of the top 20 were active in 1999. No one who was active any time in the late 40s or early 50s made the top 40.
I think if you were to tie together all the "seems too high" and "seems too low" cases, you would find that most of the players who seem too high are recent players, and most of those who seem too low have been retired for a long time.
To me, this looks like the best rating system you have come up with for extablishjing all-time rankings... with the exception of Sterling Sharpe and Cris Carter being too low on the list. Good stuff, but Where is O.J. McDuffie? LOL
This all reminds me again how everything went crazy in 1995. Just to recap:
--Jerry Rice set the receiving yardage record (1,848)
--Second in receiving yards was Isaac Bruce, whose 1,781 yards is second all-time.
--Herman Moore set the receptions record with 123, bettering Cris Carter's 1994 total of 122--which Carter matched in 1995 (which Jerry Rice also matched--Bruce had 119).
--The Detroit Lions had two WRs (Moore and Brett Perriman) catch 100+ passes.
--Michael Irvin had 11 100+ yard games (plus addition games of 94, 90, and 82)
--Two WRs (Cris Carter and Carl Pickens) had 17 TD receptions.
The WR numbers went Tecmo Super Bowl level for one season.
And just another fun passing stat thing from 95: the league's four leaders in TD passes were all in the NFC Central:
Favre: 38
Moon: 33
Mitchell: 32
Kramer: 29
Favre and Moon make sense, but when Scott Mitchell and Erik Kramer combine for 61 TDs, you might have a terrible defensive division. Yet the Tampa Bay Buccaneers threw a total of five touchdown passes.
1995 in football is like 1962 in basketball (Wilt averaged 50-25 while averaging more minutes per game than there are minutes per game, Oscar averaged the triple double with a ho-hum 30-11-12, neither was MVP)--if you try to compare it with other seasons, your brain starts to come undone.
Thank you pro-football-reference.com and basketball-reference.com for destroying hours of my life.
This is great work -- but it’s important to remember the numbers rank “greatest WR seasons” (as in the post texts), not “greatest WRs” (as in the post titles). There’s some correlation obviously, but they are not the same thing. The situations the players were in have a *big* effect on WR numbers. Take Don Maynard and Jerry Rice. . .
Maynard retired with more yards than any receiver in history until then, and still today has the highest Yds/C of all HoF receivers. So he caught at lot of long passes whatever league he was playing in.
You’d think that age 26, at his physical top, playing against inferior early AFL opposition, he’d a been a killer – 2,000 yards! And he needed that much to have a good season then, since he gets only 56% credit for playing in the AFL that year.
Yet he had only 629 yards, because he had an AFL-quality QB, Al “30 Picks” Dorrow, throwing to him, on a team that drew its plays on the infield dirt and didn’t fix its film projector when it broke (literally). And because his numbers are discounted by the quality of the league but not the quality of his QB, he gets credit for only 352 yards for the season.
Maynard’s great streak of >20yd/catch seasons came in his 30s, when he was past his prime, but after Ewbank and Namath arrived. So that “352 yard” season certainly was not a measure of the quality *of him*, but of his situation.
Rice was surely a HoF-great receiver. But he also had the great good fortune to play on a 49ers team that was the premier franchise in the league, making the playoffs 12 of his 14 years there, with the team specializing in a high-percentage passing offense, with two consecutive HoF QBs, Montana and Young, running a passing game designed to throw to, well, Rice.
Rice was almost drafted by the Jets, who after much internal argument passed on him for Al Toon. The Jets then ran through a solid decade of losing with starting QBs like O’Brien, Nagle, Reich, O’Donnell, Lucas and Testeverde, while Rice was with the 49ers.
Imagine that the Jets had drafted Rice, as they almost did. Would he today be #1 all time on the list here with a career value a huge 1600 more than #2, Don Hutson? I very much doubt it. I think he’d be remembered as, with Don Maynard, one of the Jets’ two best receivers all time.
This is no slight on Rice. It’s good to be very very good, and it’s good to be very fortunate. Rice was both very very good and also very fortunate – the unbeatable combination! And he certainly compiled the best statistical seasons of any WR ever, which counts for a whole lot. Nobody else did that!
But before concluding that he was the best WR ever, we’d have to find ways to adjust not just for quality of league (AFL) but even more importantly for quality of QB and quality of team. With 22 starters and more than 30 players making a real impact in every game, football is the “most team” of all sports, every individual’s stats are seriously influenced by the whole team, the passing game players’ stats most of all.
Okay, a possible explanation on why the pass-happy modifier doesn't work in all cases:
The Warner era Rams were certainly pass-happy, that was partly because they had one of the best receiving RB's in history. Faulk averaged 83.7 catches/year over those 3 years. Jimmy Smith's best 3-year stretch had the leading RB with something like 27 catches/year. Holt and Bruce certainly had a bigger passing pie to carve up, but there were more people at the table too.
.
And a possible solution:
One could figure the percentage of completions going to non-receivers, then discount some amount, say half of that, from the raw pass attempts for the team. For teams that rely mainly on their receivers, the receiver values wouldn't change significantly, but for teams who utilize their RB's and TE's as frequent and serious targets, it might make for a better number than simple raw pass attempts. The Chargers threw nearly half their passes to Tomlinson and Gates in 2007, and while their receivers may not have been spectacular, it wasn't their fault they were on a team with those two non-receiver-receiving-monsters. If you threw out 25% of their pass attempts, the Chargers would more correctly look like a team that targets their WR's infrequently. It'd complexify (ha!) things but it feels more "right" to me at least.
The season weighting has always bugged me. I understand why it's there, to discount long, mildly productive careers so they don't drown out reward relatively short, brilliant ones. I think what we're really looking for is for a few stellar seasons, with some smaller bonus for lots of additional productive seasons too, right?
.
I don't have the numbers but for the sake of argument but lets say a "good" season is worth on the order of 300 and a great season is around 750, and 1300 is football-god-walking-the-earth territory. Currently a godlike season is worth 4.3 good seasons or 1.7 great seasons. A great season is worth 2.5 good seasons. To me, there's not enough SPACE there. If you put up among the best numbers in history, even for only a season, it's more significant than 4.3 decent seasons when we're talking about the greatest, no? It certainly is more valuable in terms of rarity.
.
Simple solution - do away with the whole discounting seasons entirely and raise each year's values to some arbitrary power, which is really no more arbitrary than discounting seasons in the first place, yes? Say we raise them all to the 1.5 power. And lets divide em all by 100 just so the numbers aren't obnoxiously large. Rather than 300/750/1300, you've got 52/205/469 -- A godlike season is now worth 2.3 great seasons and 9 good seasons, and a great season is worth 4 good seasons. That feels more reasonable to me -- Everybody gets full credit for every season they played, but a few great seasons puts you out of reach of the compilers, eh? Just a thought.
Luv the list and spot checked some of the raw data. However, I saw Lance Rentzel way down on the list for his 1965 season. But he had no catches. Could we mean his 1969 season? I wonder the same for Bob Hayes--do we mean his 1966 season of 13 TD's? Just a crazy Cowboy fan spot checking the list.
OOPS forget my post I see now that the year listed is for the rookie year. I'm a bonehead. Sorry!
Just wondering why Mark Duper isn't on this list? He clearly put up bigger numbers than many of the guys on here.
I must say, I'm a little biased. My dad Charlie Hennigan had more >200 yard games than most hall of famers. Almost 1/3 of his games were >100+ yards. How often a quarterback passed the ball is certainly relevant, but how accurate the quarterback was is probably more relevant. This is hard to measure. Joe Montana and Steve Young were quite instrumental in making Jerry Rice the best receiver in history. Many receivers are judged more on the basis of career stats which may reflect as much how long they were able to remain healthy as how good they really were as wide receivers.
This list is hilarious.
It has Jerry Rice as the #1 receiver of all time by a significant margin, and there are legitimate, irrefuted arguments that he is underrated by the metrics set out.
And the numbers don't even show the stuff he did without the ball. He's the #1 decoy of all time, and he, despite being the #1 receiver who could get away with not doing so, would throw his body in the way of pursuing defenders to block downfield. If I had directed his HoF induction video, there would have been a montage of "TDs other guys scored because Jerry Rice threw a block".