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.

Ranking the WRs: Methodology Discussion

Posted by Chase Stuart on February 9, 2009

Ranking every WR in NFL history is a daunting task -- rule changes have made comparing across eras very difficult and we have only three main stats to guide us. While I don't think it's perfect, I've come up formula that incorporates the important features of being a valuable WR in any era.

You know I'm worried about selling you guys on a system when I publish the results first. Here are the seven wide receivers that have posted the best statistics in NFL history (according to my formula):

1. Jerry Rice
2. Don Hutson
3. Marvin Harrison
4. Terrell Owens
5. Randy Moss
6. Lance Alworth
7. Steve Largent

I think it's generally agreed that those seven guys are among the very best WRs ever. No one should be upset seeing Rice and Hutson at the top, and Harrison/Owens/Moss in some order deserve the spot right behind them (and Moss should pass Owens before his career ends). Alworth (and Maynard) dominated the '60s and Largent (and Lofton) dominated the '80s; no one really dominated the '70s in the same way, although Harold Jackson and Cliff Branch were very, very good.

Since those seven guys come out on top, maybe my system isn't so kooky. I'll let you guys decide. Let me get a couple of things out of the way first:

1) Post-season stats are not included, yet. This gets complicated enough as it is -- I thought the first time around it was best to just stick to the regular season numbers and add playoff numbers into the system along with other modifications the next time around.

2) Rushing data, passing data and fumble data were also excluded for the same reason. And while important, each receiver's blocking ability was ignored because of a lack of any objective measure.

3) Yes, a player's stats do not perfectly tell the story of how good they were. The quality of the QB, OL and the system a team runs all heavily impact a WR's numbers. So does playing in Chicago compared to playing in Miami. These are all important factors but I chose to leave them out of the system and let each reader subjectively tweak a player upward or downward based on their own thoughts.

Now that the disclaimer is out of the way, the first step is to combine the three WR stats -- receptions, receiving yards and receiving touchdowns -- into one stat: Adjusted Catch Yards. We know that a passing TD is worth about 20 yards and a rushing TD is worth about 20 yards, so it's not surprising that I'm stipulating a receiving TD is worth 20 yards. Doug has argued that at least in modern times, a reception is worth about six yards because of the correlation between receptions and receiving first downs, an important stat that unfortunately was not tracked for most of football history. I don't want to give too much credit to simple possession receivers, though, so I'm going to go with five yards per reception. Making 80-1000 equivalent to 65-1075 seems like we're giving enough of a reception bonus to me. We could quibble about making a reception worth anywhere from three to seven yards, and that might be one of the tweaks for version two of this series.

So we're off to a good start. We know how many ACY each WR has for each season of his career. But there's something else we need to consider: the difference between passing (or catching) and rushing:

Allow me to make some gross generalizations. When you’re a good passing team (measured by ANY/A), you probably won’t need to pass that often. You’ll score points and you’ll be winning, so you’ll stop throwing. If you’re a bad passing team, you’ll throw some INTs, have a bunch of three and outs, and are likely to have to pass a bunch in the fourth quarter. That’s exactly why raw passing yards is a bad way to measure passing production — bad passing teams are likely to have a bunch of passing yards late in the game and good passing teams aren’t going to accumulate many passing yards at the ends of games. That means passing yards doesn’t accurately measure passing ability, since it penalizes good passing teams and helps bad passing teams.

The opposite happens with rushing. If you’re a great rushing team, you’ll keep rushing. If you’re a bad rushing team, you’ll stop rushing.

It is obvious that we have to use per attempt statistics with QBs, although we certainly will give extra credit to QBs who are good and throw a high number of passes. Using just adjusted net yards -- passing yards + 20*PTD - 45*INT - SackYardsLost -- would be an inappropriate way to rank QBs. That's because 3,000 adjusted net yards on 400 attempts is a lot better than 3,000 adjusted net yards on 500 attempts.

The same theory applies to receivers, too. Having a 1200 yard season is more impressive when your team throws 400 passes than when it throws 550 passes. It's also a lot more valuable. Quantifying that difference is a complicated question, but to cherry pick a 2008 example, Derrick Mason had 1537 ACY this year while Lance Moore had 1523 ACY. Moore's team threw 636 passes while Mason's team threw 433 passes. Mason's season looks much more "impressive" to me.

So what do we do? I totaled the number of adjusted catch yards by all wide receivers (and only wide receivers) in the NFL each season. I also found the number of NFL (and AFL) passing attempts in that season. Then I divided the number of WR ACY by the number of team attempts; let's call this ACY/A. For the 2008 season, NFL WRs had 109,351 ACY and NFL teams passed 16,521 times, for an average of 6.62 ACY/A. Remember, that's the sum of all WRs; on an average passing team with an average passing distribution to RBs, WRs and TEs, all WRs will total about 6.62 ACY/A. No individual WR will match that number, of course, but the group of WRs will.

When we did the Greatest QB Ever series, it was easy to use the league average as a baseline -- QBs got credit for being above the baseline and received no credit for their work below the baseline. This system worked well to reward the elite and not merely the compilers. There isn't a natural baseline to use here, though -- we can't use 6.62 since no one will reach that. While arbitrary, I decided to simply divide the league ACY/A ratio by 3, and declare that the baseline for every season.

So now WRs will get credit for their ACY over one-third of the league average for all WRs per team, per pass attempt. To use some real numbers, if WR A's team passes 500 times, he'll need (in 2008) to have 1105 adjusted catch yards to be over the baseline -- maybe 60-705-5. On a team like the Saints with 636 pass attempts, a WR would need 1406 ACY (e.g., 70-896-8) to be above the baseline.

Let's compare Andre Johnson and Roddy White. Johnson had 115-1575-8 (2310 ACY) and White had 88-1382-7 (1962). Clearly, White's raw numbers are not as impressive as Johnson's. On the other hand, the Texans threw 554 passes and the Falcons threw 434 passes. So how do we calculate a score for White and Johnson? Johnson averaged 4.17 ACY/A while White averaged 4.52 ACY/A, a nice edge to White. But our baseline is 2.21 for the 2008 season, which means Johnson averaged 1.96 ACY/A over the baseline and White averaged 2.31 ACY/A over the baseline. While Johnson was less impressive on a per attempt basis, he was still terrific and his team threw more passes which means he was bringing more above baseline value. Therefore, we multiply their value over the baseline by their number of team attempts. That means Johnson has 1088 in Value and White has 1004 in Value. So Johnson ranks ahead of White, but it's closer than their raw totals indicate. I think that's pretty good.

That's pretty much how the system works, at least for modern players. There's one other important calculation -- an adjustment for games played. Steve Smith had 1931 ACY last year in only 14 games played. How do we handle that? The quick and simple fix is to divide the Panthers' number of pass attempts by 1.14 (16/14); since Carolina threw 414 passes, we'd say that Carolina threw 362.25 passes in games that Smith played. Obviously for Smith we can find out exactly which games he missed and how many passes the Panthers threw in the 14 games he played (352, to be exact). But for older players we'll have to approximate, so the formula of (WR games played / Team Games) * Team Pass Attempts will have to be used.

For Smith, he had 1931 ACY and his team threw (we project) 362.25 passes. That's an incredible 5.33 ACY/A ratio, and it's over three yards above replacement. Therefore, his Value for the season would be 1132, best in the league. Here's the list of the top 30 WRs in 2008:

                     tm    g    rec    yd    td   ACY   att   ACY/A  Value
Steve Smith         car   14    78   1421    6   1931   414   5.33   1132
Andre Johnson       htx   16   115   1575    8   2310   554   4.17   1088
Roddy White         atl   16    88   1382    7   1962   434   4.52   1004
Calvin Johnson      det   16    78   1331   12   1961   509   3.85    838
Larry Fitzgerald    crd   16    96   1431   12   2151   630   3.41    761
Greg Jennings       gnb   16    80   1292    9   1872   541   3.46    678
Anquan Boldin       crd   12    89   1038   11   1703   630   3.60    661
Brandon Marshall    den   15   104   1265    6   1905   620   3.28    623
Wes Welker          nwe   16   111   1165    3   1780   534   3.33    602
Derrick Mason       rav   16    80   1037    5   1537   433   3.55    582
Antonio Bryant      tam   16    83   1248    7   1803   562   3.21    563
Vincent Jackson     sdg   16    59   1098    7   1533   478   3.21    478
Hines Ward          pit   16    81   1043    7   1588   505   3.14    474
Muhsin Muhammad     car   16    65    923    5   1348   414   3.26    435
Santana Moss        was   16    79   1044    6   1559   510   3.06    434
Dwayne Bowe         kan   16    86   1022    7   1592   541   2.94    398
Randy Moss          nwe   16    69   1008   11   1573   534   2.95    395
Terrell Owens       dal   16    69   1052   10   1597   547   2.92    390
Reggie Wayne        clt   16    82   1145    6   1675   585   2.86    384
T.J. Houshmandzadeh cin   15    92    904    4   1444   513   3.00    383
Bernard Berrian     min   16    48    964    7   1344   452   2.97    347
Lee Evans           buf   16    63   1017    3   1392   478   2.91    337
Donald Driver       gnb   16    74   1012    5   1482   541   2.74    288
Eddie Royal         den   15    91    980    5   1535   620   2.64    253
Matt Jones          jax   12    65    761    2   1126   537   2.80    237
Justin Gage         oti   12    34    651    6    941   453   2.77    191
Greg Camarillo      mia   11    55    613    2    928   491   2.75    183
Laveranues Coles    nyj   16    70    850    7   1340   529   2.53    173
Santonio Holmes     pit   15    55    821    5   1196   505   2.53    151
Isaac Bruce         sfo   16    60    833    7   1273   509   2.50    150
Jerricho Cotchery   nyj   16    71    858    5   1313   529   2.48    146
Kevin Walter        htx   16    60    899    8   1359   554   2.45    137
Braylon Edwards     cle   16    55    877    3   1212   488   2.48    135
Michael Jenkins     atl   16    50    777    3   1087   434   2.50    129
Lance Moore         nor   16    79    928   10   1523   636   2.39    120

How does that list look to you guys? Obviously the quality of the QB is significant when looking at a wide receiver's numbers, and that's ignored here. So it might be best to think of this as a ranking of the WRs' stats and not the WRs themselves. Even better, we could call this a statistical look at the WRs whose statistics produced the most value for their team. Tomorrow, I'll explain how to rank every WR in NFL history, and provide a list of the greatest WR seasons of all time. Wednesday I'll look at the top 100 WRs of all time, which star WRs have played in the most and least pass happy offenses, and other general thoughts.

This entry was posted on Monday, February 9th, 2009 at 7:14 am and is filed under History, Statgeekery. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.