Posted by Jason Lisk on September 2, 2009
For those that haven't been checking the blog on a regular basis through the off-season, this is part of a series on the AFL versus the NFL. For previous posts, we have now created a category and if you click here you can read all the others.
Let's now turn to trends developing in the middle part of the decade.
In the NFL, total offensive yards per game declined from the bubble of the early 1960's. Yards per attempt also decreased from the historic peak following rapid expansion, going from 7.9 in 1962, to 7.5 in 1963, to 7.2 in 1964, with a correction back to 7.5 in 1965 followed by a drop to 6.9 yards per attempt in 1966.
The way in which those yards were being gained was changing. Yards per reception had dropped dramatically in the post-war era, but had then held pretty steady for fifteen years, fluctuating between the low 13's and high 12's. Yards per reception (or if you prefer, yards per completion) hit an all-time low of 12.2 in 1964, and after a slight rebound, dropped to 11.9 in 1966. Of course, by today's standards, that is high, as the trend would continue after the merger. No season since 1970 has seen an overall league yards per reception of over 11.8.
Scoring in the NFL peaked in 1965, and as we will see, scoring in the NFL actually surpassed the AFL for that season for the first time since the rival league began.
Over in the AFL teams continued to pass the ball more than they ran it, with the peak difference occurring in 1964. Points per game in the AFL dropped in 1962, from 24.5 to 23.1. The next big drop in scoring occurred in 1965, when it went from 23.2 to 21.5. While scoring rebounded the next year, the AFL never again averaged over 23.0 points per game over the course of a season.
Some of you may have read my post a few weeks ago on passer personality types. One of the characteristics I talked about was the Gambler/Holder trait. That trait looked at the interplay between sack rate on the one hand, and interception rate and completion percentage on the other. If a quarterback holds on to the ball under pressure, his completion percentage is going to be better (he's not throwing harder to catch balls as often or throwing it away) and he's going to throw fewer interceptions.
Well, if you look at the AFL and NFL during this period, you see that the AFL always had the lower completion percentage and the higher interception rates. We don't have reliable individual sack data, but we do have team sack data for the entire decade. And lo and behold, the NFL had the higher sack rate in every season of the decade.
Here is a chart showing the league-wide completion percentages, followed by a sack-adjusted completion percentage (completions divided by attempts + sacks), and net passing yards per attempt (passing yards - sack yards lost divided by pass attempts + sacks).
Completion Percentage Sack-adjusted comp% Net Yards Per Attempt NFL AFL NFL AFL NFL AFL 1960 50.2 48.5 45.8 45.2 5.9 5.6 1961 52.1 46.8 47.4 43.4 6.1 5.6 1962 53.3 47.6 48.7 44.3 6.5 5.7 1963 51.5 48.7 47.1 44.7 6.1 5.8 1964 51.4 49.0 46.3 45.4 5.6 5.9 1965 51.3 45.3 46.8 42.5 6.1 5.4 1966 51.6 46.3 47.0 43.6 5.6 5.9 1967 51.0 47.6 47.0 44.0 5.7 5.5 1968 51.6 47.5 47.3 43.8 5.8 5.7 1969 52.6 49.8 48.4 45.9 5.8 5.7
As we can see, the NFL still completed a higher percentage of drop backs throughout the decade, though the gap closes some with sacks considered. While AFL quarterbacks were throwing incompletions and interceptions (and occasionally making a big offensive play), their NFL counterparts were taking sacks at a higher rate. When we consider sack yards lost along with positive yards gained in the passing game, we see that the NFL held a big advantage in the early part of the decade in net yards per attempt. But in 1964, the AFL actually led in net passing yards per attempt for the first time. Between 1960-1963, the NFL held a 6.15 to 5.68 advantage in NYPA, almost a half yard per pass. From 1964 to the end of the decade, those numbers are 5.77 for the NFL and 5.68 for the AFL--a difference of less than 0.1 net yards per attempt. By the middle and late parts of the decade, the two leagues were arriving at roughly similar passing game yardage per play production in different ways.
ROOKIE AND SECOND-YEAR STARTING RATE TRENDS
In the early decade trends, I looked at young starter rates, so we'll just continue that here. If you recall, the AFL was an extremely young league in 1960 (younger than the average expansion team). By 1963, the NFL and AFL rookie starting rates were tightening up, 9.4% and 11.9%, respectively.
Here are the rookie and second year player starting rates for 1964-1966:
AFL young starter rates
rookies 2nd year combined ================================================ 1964 0.091 0.142 0.233 1965 0.114 0.114 0.227 1966 0.091 0.182 0.273 ================================================
NFL young starter rates
rookies 2nd year combined ================================================ 1964 0.091 0.127 0.218 1965 0.071 0.114 0.185 1966 0.064 0.121 0.185 ================================================
There are a couple of things here. First, the rookie starting rates in 1964 were equal-exactly equal. However, we saw that the NFL ended up winning the 1964 draft battle pretty handily, so we would expect the NFL rate to actually be higher (if the two leagues were equal) since they had a better class that year.
On the other hand, if the AFL was still in a state of rapid talent improvement through the middle of the decade, we should expect the rookie and second year starting rates to be decreasing over time, regardless of the quality of the draft class, because earlier classes are forced into action sooner than ready, at a higher rate than later ones, to replace the sub-replacement level players already in the league. We see that in 1961 and 1962, but we do not see that here. Remember, the 1963 and 1965 AFL draft classes were much better than the 1964 and 1966 versions. And we see that the 1963 and 1965 draft classes had higher rookie starting rates, and higher second year starting rates (well, you can't see the 1966 class' second year, but I will tell you it was 12.1%)
ALL-PRO AGING TRENDS
I also looked at early aging trends for all-pro selections, and the NFL experienced a predictable age increase in its all-pros during the early part of the decade, as a result of one-third of the incoming talent and depth departing to the new league. The NFL was oldest in 1962 (average all-pro age: 29.1) and I don't think it is a coincidence that year was the peak year in terms of offensive passing yards per attempt. Those numbers rebounded in the middle part of the decade, though the AFL continued to be a much younger league since it had few stars from the drafts prior to 1960.
Here are the average all-pro ages of the two leagues:
NFL AFL 1964 28.2 26.2 1965 28.3 26.4 1966 28.3 26.8
Here is the breakdown by age group:
24 and under 25 to 30 31 and older AFL NFL AFL NFL AFL NFL ========================================================================== 1964 0.274 0.167 0.661 0.619 0.065 0.214 1965 0.226 0.13 0.726 0.638 0.048 0.232 1966 0.176 0.113 0.735 0.625 0.088 0.263 ==========================================================================
The 1964 draft class really re-vitalized the NFL. From 1961-1963, there were only six players age 23 or under who made an all-pro team in the NFL. Nine players age 23 or under were named all-pro in 1964, and that class would continue to produce the highest number of all-pros each year through the rest of the decade.
PLAYER MOVEMENT BETWEEN LEAGUES
Tim Truemper sent me an e-mail about a month ago inquiring about player movement between the leagues. Clearly, a lot of the early AFL players moved from the NFL. 31.2% of the starters in 1960 had some previous experience on an NFL roster.
I put together a list of all players who debuted between 1960 and 1966--basically, from the start of the AFL to just before the merger--and switched from one league to another between 1963 and 1969. I'm not going to post the full, mostly unknown list of 75 here, but if anyone is interested in those names, I can provide upon request. So, my list does not include the early "NFL rejects" who moved to the AFL in 1960-1962 or began their playing careers before 1960.
Most of the guys are fairly non-notable. Generally, we had backups from one league going to play for one of the worst teams (at the time) in the other league, generally also as a backup. Occasionally, there were surprises.
There were 50 guys who debuted in the NFL in 1960 or later and then joined the AFL from 1963-1969. Of those, Oakland signed 14 and Denver signed 13. Al Davis signed a lot of NFL "rejects", and for every Ben Davidson, there were many others you haven't heard of. The Jets were third with 9 players. Kansas City was least likely of the AFL teams to sign an NFL player after 1962 (of course, Len Dawson was a big addition that year), and only signed one after that, Jim Kearney.
Here is a summary of the player movement from 1963 to 1969:
to AFL to NFL ===================== 1963 15 1 1964 7 0 1965 5 0 1966 10 3 1967 5 8 1968 4 10 1969 4 3 ===================== total 50 25 =====================
The trend of NFL veteran movement to the AFL continued but slowed throughout the middle part of the decade, but we see that there was no movement the other way. Stunningly, though, that was not true once the merger was approaching. Check out 1966-1969, the Super Bowl era. 23 AFL signings of NFL players compared to 24 NFL signings of former AFL players. 5 of the 23 transplanted AFL guys and 6 of the 24 transplanted NFL guys stayed on a roster for more than 2 seasons after switching, very similar percentages. Thus, it wasn't as if the NFL guys moving to the AFL in the Super Bowl era were suddenly experiencing career revivals at rates greater than the players moving the other direction.
INCREASE IN PLAYERS FROM HISTORIC BLACK SOUTHERN COLLEGES
I think the evidence is pretty overwhelming that there was a decrease in the overall average quality of play in professional football, including the established NFL, when the AFL entered and the NFL expanded in 1960. Going from 12 to 22 teams in the span of two years will do that. On the other hand, I think that football was ready to expand by 1960, and the conditions were ripe, and the AFL owners took advantage of those conditions when the NFL was slow to do so without the push of competition.
The extent to which professional football recovered after that rapid expansion in 1960-1961 depended on whether the population of talented young males interested in football could meet the demand. To meet the demand, the league needed to find additional talent sources. This could have been met in any number of ways:
--an increase in the overall population pool. Population was increasing over time, but the biggest jump, thanks to the baby boomer post-war generation, was not going to arrive until the end of the decade.
--an increase in the football-playing population. Football was becoming more and more popular, and so it stands to reason that a higher percentage of young males were attracted to football than ever before. However, this would have been a gradual increase over time and would not have completely filled a sudden need for twice the manpower.
--a new and relatively untapped source of football talent.
Which leads us to players from historical black colleges and universities (hereafter, HBCU). I'm limiting my study here to players from HBCU's not because I don't want to look deeper into the issue, but for the sake of the blog post, it would be too time consuming to track down a complete record of every African-American player who entered the AFL or NFL. I can, however, fairly quickly, track down players by school affiliation and can presume that players from the HBCU were African-American. So, keep in mind, these numbers are NOT the overall number of African-Americans to enter the league, only those who attended a HBCU.
Sean Lahman, in his book "The Pro Football Historical Abstract", does break down the number of African-Americans in the NFL, MLB, NBA and NHL from 1946-1959. He lists 121 debuts by African-Americans in the NFL and AAFC from 1946-1959. As we will see, most of those players were not coming from the South. They were players from universities in the Northeast, the Great Lakes region, and the West Coast.
The Southern schools were still segregated, so the top African-American athletes from Virginia to Texas attended HBCU's unless they moved outside of their region. Today, players like Buck Buchanan and Willie Brown would likely be helping LSU win the SEC instead of playing for Grambling, and Bullet Bob Hayes might play for Florida instead of Florida A&M. But not so in this time period. The NFL wasn't completely ignoring the HBCU's, but the numbers were small. They would explode once the AFL joined, and competition created demand for a new source of talent.
Here are the numbers on players from Historical Black Colleges (HBC) to debut in certain leagues and periods:
NFL and AAFC, 1946-1959: 37
NFL, 1960-1966: 63
AFL, 1960-1966: 76
NFL, 1967-1969: 63
AFL, 1967-1969: 64
As we can see, only 37 of the 121 (30.6%) African-American players who debuted before 1960 attended a HBCU. The AFL had a reputation for being the more proactive league in pursuing talent from these schools, and by the raw numbers, we see a slight edge to the AFL. Now, let's convert those numbers to account for the number of teams in each league and number of seasons to get an estimate on debuts per team per year.
NFL and AAFC, 1946-1959: 0.2 per team per year
NFL, 1960-1966: 0.6 per team per year
AFL, 1960-1966: 1.3 per team per year
NFL, 1967-1969: 1.3 per team per year
AFL, 1967-1969: 2.2 per team per year
The AFL had fewer teams, but more players from HBCU's. Of course, you have to consider that they also had a lot more job openings because there were virtually no veterans, particularly early in the 60's, so that's going to skew the numbers up. But it should have been pretty even by 1966. The AFL still was more proactive in the late 60's, even with a common draft by that point.
Now, for my purposes, I'm interested in a couple of questions. First, how much did the introduction of greater numbers of players from HBCU's offset the talent dilution that occurred in the early 60's? Second, how much does the under-drafting of players from HBCU's skew my draft value numbers? It's pretty clear that both leagues were still under-drafting players from HBCU's as the decade progressed. Consider how many African-Americans from the ACC, SEC, Big XII South and Conference USA are drafted in the first 100 picks today. Compare that to the number of players from the HBCU's drafted in the top 100 of either leagues draft as the decade progressed.
1960 2 1961 4 1962 4 1963 5 1964 7 1965 14 1966 16
34 of the 52 drafted in the top 100 were drafted higher by an AFL team. Deacon Jones lasted until the 14th round of the NFL draft in 1961, and was undrafted by the AFL. Willie Brown and Emmitt Thomas were undrafted by both leagues--you don't typically find two hall of fame cornerbacks who go undrafted in the span of three years. Of the ten undrafted successes in the AFL from 1963 to 1966, only one attended a major college power. That was George Sauer, who went to Texas and had a father who was an executive with the New York Jets. Six of the remaining nine attended HBCU's.
I mention this because, to the extent African-American's from the HBCU's were under-drafted relative to their actual talent, my draft value numbers are going to understate the talent the leagues were getting. And since the AFL signed more players from HBCU's and had more free agent stars, it will undervalue the AFL's talent more.
The AFL came along at the right time, because there was a talent pool that was severely under-utilized by the NFL, and without those players, I'm not sure the AFL would have become the serious competitor to the NFL that it was. And the increased competition caused both leagues to give exponentially greater opportunity to young African-American athletes from the South as the decade continued.