≡ Menu

Guest Post: An Argument For HOF Expansion

Bryan Frye is back with another fun guest post. Bryan, as you may recall, owns and operates his own great site at http://www.thegridfe.com/, where he focuses on NFL stats and history. You can view all of Bryan’s guest posts at Football Perspective at this link. You can follow him on twitter @LaverneusDingle.


What makes a Hall of Fame player in your opinion? Is it being in some arbitrary percentile grouping at his position? Perhaps it is a combination of stats and memorable moments. How about playoff performance? Maybe you give extra credit for champions. I certainly don’t know, and my personal Hall likely wouldn’t resemble yours. Any of those criteria you prefer, however, calls for an attendant expansion of the Hall of Fame. [1]Thanks to Adam Harstad, who was a great sounding board for my ideas and who probably helped keep this from being twice as long.

Arbitrary Percentile

One criterion people use to determine if a player belongs in the Hall of Fame discussion is his place relative to his contemporaries. If a quarterback or halfback is at or near the top of the league for a good portion of his career, he is almost guaranteed a bust in Canton. [2]The same can’t be said for some positions. I don’t hear many people talking about the legacies of Kevin Williams, Nick Mangold, or Lance Briggs. I’ve heard some analysts argue that the Hall should be reserved for the top 3-5% of players. If the top 3-5% (or any arbitrary percentage you choose) is your cutoff, then it follows that induction class sizes should increase to accommodate the increase in players. The 90th percentile of twelve starting quarterbacks includes one quarterback, whereas the 90th percentile of 32 starting quarterbacks includes three quarterbacks. Since the league has nearly thrice the teams it had fifty years ago, it makes sense to have a concomitant increase in class sizes.

The inaugural HOF class was in 1963 and included players from as late as 1945 (Mel Hein). From 1920 to 1945, there were enough active roster spots for roughly 6900 players (see the table at the end of this post). For comparison, that is fewer than the number of active roster spots available in a five year span in our current era (32 teams with 46 spots for 5 years means 7360 active roster spots). From the beginning of league history to 1963, there were an estimated 15743 roster spots (I estimate because I have yet to find active roster rules for the AAFC or AFL). That is less than a decade’s worth of players in today’s format. [3]Keep in mind that this methodology includes teams like the Tonawanda (NY) Kardex, New York Brickley Giants, and Muncie (IN) Flyers, each of whom played fewer than five games in their existence.

Of course, looking at all active roster players includes players such as long snappers, second string guards, and emergency quarterbacks, none of whom are considered for the Hall of Fame. If we instead look at only the 22 players on the field, we’ll get a truer sense of the discrepancy.

If we use starters instead of active roster spots (22 times the number of teams each year), there were 7128 starting players from 1920 to 1945 [4]It is important to note that a significant portion of players from this time started on both sides of the ball, as the NFL did not adopt the free substitution rule on a permanent basis until 1949. … Continue reading. That only slightly outstrips the 7040 starting spots up for grabs in a ten year span today. Using 1962 as our cutoff, like before, there were 12782 starting spots available before the first Hall of Fame class was enshrined. That is about the same number of starting spots the NFL has had available in the last 19 years. In other words, there have been more players on the field since 1996 than in the 43 seasons leading up to the first Hall of Fame induction ceremony.

Stats and Signature Plays

In the current NFL, players participate in more games than ever before. This gives them more opportunities to add to both their cumulative statistics and their highlight reels. In order to fully understand just how the league’s schedule has given modern players increased opportunities, we should look at the evolution of the NFL’s schedule structure.

For the first sixteen years of the league’s existence, the scheduling format was utter nonsense. In 1925, for instance, the Frankfurt Yellowjackets played twenty games; the Duluth Kelleys played three. It wasn’t until 1936 that every team played the same number of games (12). Because of the significant disparity in team schedules in the early days of the NFL, I have excluded them from the remainder of the discussion.

There is a fairly common belief that teams run more plays per game now than they did before. That is actually not true. Starting in 1932, plays per game increased dramatically until reaching an all-time high in 1949. [5]1949 – 1952 are, in order, the four highest ranking seasons in plays per game. Since about 1955, team plays per game have fluctuated without ever nearing the apex of 1949 or the nadir of 1932. However, the increase in the number of games per season has significantly boosted the volume of plays per season. Since the schedule expanded to 16 games in 1978, the average team has run about 1000 plays per season. This compares to about 625 in the eleven game era, 770 in the twelve game era, and 870 in the fourteen game era. Thus, there are many more opportunities for modern players to rack up stats or make signature plays. With NFL administration considering a move to 18-game seasons, those opportunities could increase significantly in the near future.

Anecdotally, the prevalence of sports media and the ability of social media to spread ideas virally in a matter of minutes have combined to make highlight reel plays more accessible than ever. I don’t know if this helps or hurts players who are looking to add signature plays to their resumes. For instance, aside from Odell Beckham’s famous catch in an otherwise forgettable game, most other highlights from last season were just white noise. Then again, if supporters of, say, Fred Taylor want to bolster their argument in his favor, they need only turn to YouTube to find him taking it to the house in a playoff game versus Miami.

Playoff Performance

The evolution of the playoff format has given players increased opportunities to make a notable postseason impact. From the establishment of the NFL until 1932, the title simply went to the team with the best record. If you played in the first thirteen years of the league, you had no opportunity to add playoff accomplishments to your legacy.

Afterwards, the NFL was divided into East and West divisions, and the championship game featured the two division winners. [6]In 1941, the Bears and Packers tied for the division lead and had an impromptu playoff game to determine who would face the Giants for the championship. In 1943, the Giants and Washington had a … Continue reading In 1950, the league added three former AAFC teams and renamed the divisions American and National. However, even with the extra teams, the division winner only format remained. [7]However, 1950 did give us a glimpse of the future when both divisions required one game playoffs to determine a winner. Even the progressive AFL adopted the division winner model. In 1963, the 7-6-1 Patriots beat the 7-6-1 Bills for the chance to play the Chargers in the title game. The 10-4 Raiders never got the opportunity, despite winning both regular season games against the Bolts. This format remained until the first Super Bowl, in which the NFL champ faced the AFL champ for the “world title.” [8]With one-off playoff games to establish division champs in 1952, ’57, ’58, and ’65.

In 1967, the NFL split into four divisions [9]Capital, Century, Coastal, and Central. and sent each division winner to the postseason to play for the right to face the AFL champions. When the leagues finally merged, the NFL established the American and National Football Conferences and split each into East, Central, and West divisions. Having three divisions per conference necessitated the addition of a wild card team, so 1970 was the first season to feature eight playoff teams.

In 1978, in addition to expanding the regular season to 16 games, the NFL also added another playoff seed to each conference and gave each top seed a first-round bye. That’s ten playoff teams, for those keeping track. With the exception of the uniquely weird 1982 season, the ten-team model remained until 1990, when the NFL adopted its current postseason model. [10]In the nine-game strike season, 16 of the league’s 28 teams went to the playoffs – including two teams with losing records. The only tweak that has been made in the last 25 years is the 2002 divisional restructuring, which did not have an effect on the number of teams that made the playoffs.

That brief and totally non-comprehensive history lesson provides a look at how the mere structure of the postseason itself can impact the legacies of players. Larry Fitzgerald never would have had the chance to put on a legendary performance in the 2008 playoffs if the playoff format didn’t allow his team to even participate in the playoffs. Eli Manning may one day have a bronze bust in Canton based largely on his stellar performance in two Super Bowls. However, the Giants wouldn’t have been in position to compete for those titles without the expanded playoff format.

Championships

While playoff expansion has given players more opportunities than ever to leave their mark on the game, it has not increased the number of titles awarded at the end of the season. Whether the NFL awarded the title to the team with the best regular season record or to the team that survived a twelve-team playoff tournament, championships have by their very nature been immune to inflation. Like the Highlander, there can be only one. [11]Unless you count AAFC and AFL titles, which — despite the general tone of this site — is not something I personally do. However, I am willing to entertain the idea of a shared title for … Continue reading Thus, the number of title teams on which a player appeared is the only criterion that does not necessitate Hall of Fame expansion.

Resolving the Backlog

Regardless what set of criteria you use to determine a player’s Hall of Fame worth, it is hard to get around the fact that both roster and schedule inflation have given more players more opportunities than ever to make their cases. From both a mathematical and a practical standpoint, the idea of restricting induction classes to just five modern era players becomes an increasingly untenable position. The always great Mike Tanier recently wrote a great article maintaining that snubs should be patient and wait their turn. The plea for players and fans to be patient and avoid overreacting to perceived slights makes sense, and at the surface seems like the most reasonable course. However, patience will not account for the backlog of worthy players and coaches. We can keep telling these men to sit in the waiting room, but the room is eventually going to be too small to hold everyone. Many of them may end up stuck out in the cold.

Because the sheer number of players playing now is so great relative to the number of players playing from the league’s inception up until the establishment of the Hall of Fame, there now exists a backlog of deserving candidates. As recently as this year’s induction announcement, the commentariat bemoaned the absence of Marvin Harrison, Orlando Pace, and Kurt Warner from the induction class. Harrison, of course, had to wait so that Tim Brown could be elected. Brown, for his part, had to wait for Cris Carter and Andre Reed to get in. [12]If Harrison gets the nod next year, it may come at the expense of Terrell Owens. If Owens has to wait, he may then get in at the expense of Randy Moss. This backlog will never end. Pace will have to wait another year because the selection committee likely didn’t want two offensive linemen enshrined in the same year. Besides, Will Shields had waited long enough. Warner didn’t make it in his first year of eligibility. In his second year, Brett Favre will be eligible, which means he may have to wait another year still. Unless the Hall of Fame expands its class sizes, the problem is only going to get worse.

The current backlog will naturally hurt some position groups more than others. Quarterbacks will probably be just fine. Linemen, defensive backs, 4-3 outside linebackers, and wide receivers, however, will not. Offensive linemen will suffer for the same reason they have always suffered: people don’t know enough about what they do to make a fair judgment. Outside linebackers in 4-3 schemes compete with 3-4 pass rushers for Pro Bowl and All Pro nods. No matter how well Lance Briggs, Lavonte David, or Thomas Davis play, they will always have trouble garnering more attention than a player in a position designed to accumulate sacks.

Defensive backs and wide receivers will suffer from position inflation. [13]By defensive backs, I mean cornerbacks. Safeties will continue to suffer because they are safeties. In 1963 there were 22 combined AFL and NFL teams. Those teams generally started two wide receivers, even if they weren’t necessarily called wide receivers. Today, there are 32 teams in the league, and at least half of them feature offenses that start three receivers. To combat this, many teams run a nickel as their primary defensive look. The liquid nature of many offenses and defenses makes the definition of a starter pretty ephemeral, so I’ll be conservative and say half the teams in the NFL start three receivers or three cornerbacks. That means that when the Hall of Fame established its rules, a single season of football would see about 44 starting wideouts/corners. Today, that number is at least 80.

The table below shows lots of information on the number of players in the NFL, among other categories, for each year since 1920:

YearNFL TeamsOther TeamsActive Roster Limit NFLActive Roster Limit OtherActive Player Spots NFLActive Player Spots OtherTotal Active Player SpotsStarter Spots NFLStarter Spots OtherTotal Starter SpotsNFL League GamesOther League Games
20153204601472014727040704160
20143204601472014727040704160
20133204601472014727040704160
20123204601472014727040704160
20113204601472014727040704160
20103204601472014727040704160
20093204601472014727040704160
20083204601472014727040704160
20073204601472014727040704160
20063204601472014727040704160
20053204601472014727040704160
20043204601472014727040704160
20033204601472014727040704160
20023204601472014727040704160
20013104601426014266820682160
20003104601426014266820682160
19993104601426014266820682160
19983004601380013806600660160
19973004601380013806600660160
19963004601380013806600660160
19953004601380013806600660160
19942804601288012886160616160
19932804601288012886160616160
19922804601288012886160616160
19912804601288012886160616160
19902804501260012606160616160
19892804501260012606160616160
19882804501260012606160616160
19872804501260012606160616150
19862804501260012606160616160
19852804501260012606160616160
19842804901372013726160616160
19832804501260012606160616160
1982280450126001260616061690
19812804501260012606160616160
19802804501260012606160616160
19792804501260012606160616160
19782804501260012606160616160
19772804301204012046160616140
19762804301204012046160616140
19752604301118011185720572140
19742604701222012225720572140
19732604001040010405720572140
19722604001040010405720572140
19712604001040010405720572140
19702604001040010405720572140
19691610404064040010403522205721414
19681610404064040010403522205721414
1967169404064036010003521985501414
196615940406003609603301985281414
196514840405603208803081764841414
196414840405603208803081764841414
196314837375182968143081764841414
196214836365042887923081764841414
196114836365042887923081764841414
196013838384943047982861764621214
195912036043204322640264120
195812035042004202640264120
195712035042004202640264120
195612033039603962640264120
195512033039603962640264120
195412033039603962640264120
195312033039603962640264120
195212033039603962640264120
195112033039603962640264120
195013032041604162860286120
194910732323202245442201543741212
194810835353502806302201763961214
194710835353502806302201763961214
194610833333302645942201763961114
194510033033003302200220100
194410028028002802200220100
19438028022402241760176100
194210033033003302200220110
194110033033003302200220110
194010033033003302200220110
193910030030003002200220110
193810030030003002200220110
193710025025002502200220110
19369025022502251980198120
19359024021602161980198x0
193411020022002202420242x0
193310020020002002200220x0
19328020016001601760176x0
193110020020002002200220x0
193011020022002202420242x0
192912018021602162640264x0
192810018018001802200220x0
192712018021602162640264x0
192622018039603964840484x0
192520016032003204400440x0
192418016028802883960396x0
192320016032003204400440x0
192218016028802883960396x0
192121016033603364620462x0
192014016022402243080308x0

By increasing the size of induction classes, the Hall of Fame can account for the significant growth the league has experienced since the Hall was established. Moreover, it can account for both positional discrimination and positional inflation by recognizing a greater number of worthy players. [14]I concede that this view is somewhat limited, as some positions are simply more important than others. There’s a reason why 19% of starting QBs in 2009 will be in Canton one day, while no more than … Continue reading

References

References
1 Thanks to Adam Harstad, who was a great sounding board for my ideas and who probably helped keep this from being twice as long.
2 The same can’t be said for some positions. I don’t hear many people talking about the legacies of Kevin Williams, Nick Mangold, or Lance Briggs.
3 Keep in mind that this methodology includes teams like the Tonawanda (NY) Kardex, New York Brickley Giants, and Muncie (IN) Flyers, each of whom played fewer than five games in their existence.
4 It is important to note that a significant portion of players from this time started on both sides of the ball, as the NFL did not adopt the free substitution rule on a permanent basis until 1949. Thus, the true number of unique starters may be close to half of the 7128 from the table.
5 1949 – 1952 are, in order, the four highest ranking seasons in plays per game.
6 In 1941, the Bears and Packers tied for the division lead and had an impromptu playoff game to determine who would face the Giants for the championship. In 1943, the Giants and Washington had a similar playoff for the right to face the Bears. In 1947, it was the Eagles and Steelers battling to face the Cardinals.
7 However, 1950 did give us a glimpse of the future when both divisions required one game playoffs to determine a winner.
8 With one-off playoff games to establish division champs in 1952, ’57, ’58, and ’65.
9 Capital, Century, Coastal, and Central.
10 In the nine-game strike season, 16 of the league’s 28 teams went to the playoffs – including two teams with losing records.
11 Unless you count AAFC and AFL titles, which — despite the general tone of this site — is not something I personally do. However, I am willing to entertain the idea of a shared title for the 1925 Pottsville Maroons. On second thought, I’d just take it from the Cardinals.
12 If Harrison gets the nod next year, it may come at the expense of Terrell Owens. If Owens has to wait, he may then get in at the expense of Randy Moss. This backlog will never end.
13 By defensive backs, I mean cornerbacks. Safeties will continue to suffer because they are safeties.
14 I concede that this view is somewhat limited, as some positions are simply more important than others. There’s a reason why 19% of starting QBs in 2009 will be in Canton one day, while no more than one 2009 kicker will likely make the cut. Pass rushers are probably more important than coverage linebackers, but they play the same position in name only.
{ 37 comments }