Posted by Jason Lisk on December 15, 2009
This is the final piece of evidence before we get to the conclusions and overall team power rankings for the decade. Let's get right to it. Here are the point differentials and the win/loss records (from the perspective of the former NFL teams) for all regular season matchups involving an AFL team and a former NFL team from 1970-1974.
These results include matchups involving Pittsburgh, Cleveland and Baltimore against the AFC opponents. The AFL was not good relative to the NFL in 1970, the first season after the leagues merged. The next set of numbers are weighted by team quality. For example, Houston and Cincinnati played more games against former NFL teams than did Kansas City, Oakland, San Diego and Denver after the merger. Baltimore, Cleveland and Pittsburgh played more games against the AFL (by virtue of joining the AFC) than other NFL members. As Kansas City, Oakland and San Diego were three of the four best AFL teams in the late 1960's, and two teams that had been frequent playoff participants in the NFL were now joining the AFC, this might bias the results slightly in the NFL's favor.
Even considering this, though, the NFL dominated in 1970, and in fact, that year shows as more of an outlier than the raw numbers would indicate. This is because Baltimore and Cleveland show up as worse based on regular season SRS in 1970 (though Baltimore won the Super Bowl), and then they bounced back and Pittsburgh improved steadily over the next five years. Here are the schedule adjusted differentials between the teams from the two leagues during the first five years post-merger.
SRS averages by prior league affiliation:
If this were a standardized college entrance exam, and I gave you the above chart in the Science section, and then asked, "what was the most likely "diff" between the two leagues in 1969?" the answer choice that was more extreme than -7.1 would be the correct one. It would be the correct one, though, only if this was the only information given in the problem. It was the answer I believed to be true when I began this process several months ago and first reviewed the post-merger results.
However, this whole series has been about accumulating as much evidence as possible, about questioning conventional wisdom and trying to look at things from a variety of angles. Now I can say that in looking at all the available evidence, I think it very unlikely that the AFL was anywhere as bad or worse (relative to the NFL) in 1968 and 1969 as they were in 1970. 1970 was the outlier year. It just happened to fall at the beginning of sequence. And thus, we deduce that 1969 had to be similar to or worse than 1970. After all, wouldn't a league like the AFL be constantly improving over time?
The answer is "not necessarily." I'm going to discuss my draft class value based model in more detail in the conclusions. For now, I'll simply say that the most likely shape of AFL team improvement around the time of the merger was rapid improvement/gain on the NFL teams from 1966 to 1968, and then stagnation from 1969 to 1971, with slight improvement until the AFL should have finally overtaken the NFL teams by around 1973-1974. When we look at the actual post-merger results, the only year that doesn't fit that pattern is 1970. The former AFL teams did overtake the NFL between 1973 and 1974 and (along with former NFL bottom feeder Pittsburgh) become the dominant conference for the decade. The AFL teams did show slight improvement from 1971 to 1972 to 1973. When we introduce 1970 in the mix, and then adjust for schedule and the fact that the AFC transplants were improved as well, it looks like it was a case of drastic improvement for the former AFL members(from 1970 to 1971). I think 1971 was actually a case of a bounce back year or "market correction", after a bad year for the AFL in 1970.
At this point, I suspect you are skeptical. Maybe you think I am full of theoretical mumbo jumbo. I'll try to lay out the evidence for why the AFL was better in 1968 and 1969. One way to test this theory is to look at exhibition results. We don't have regular season results from 1968 and 1969 between the two leagues, but we do have those exhibition results. And the AFL and NFL teams also played exhibition games (and lots of them) before the 1970 season. If the AFL was actually worse in 1970 than they were in 1968 and 1969, that might show itself in those results.
And it does. In 1970, former AFL and NFL teams also met fifty times in exhibition games. The former NFL teams went 30-19-1 against the former AFL teams, with a +4.6 scoring margin. Just like previous years, a higher percentage of those games were in AFL stadiums, and once we account for team quality (as represented by regular season ranking) and home field discrepancy, the NFL was +5.6 better than the AFL in the preseason of 1970. Let's compare that to the previous three years before the merger. I list the win/loss record from the perspective of the NFL teams in exhibition games between members of the two leagues, and the point differential, once we adjust for team quality and venue.
Why might the AFL have been worse in 1970 than they were in 1968 and 1969? Here are some thoughts:
1) randomness and team variation. While we think that teams should always improve, that's clearly not the case. Maybe a higher percentage of AFL players had an off year or declined. 1970 would have been a decade after the start of the league, so the original AFL class (where the AFL had a huge numbers advantage over the NFL because of the high volume of 22 and 23 year olds who joined the AFL) might have been hitting a wall in large numbers during this season, and the next generation may have not quite been ready to step in. The effect of AFL expansion may have also hit upon the depth at the same time that some of the stars of the early AFL period were declining. In this way, the AFL may have experienced in 1970 what the NFL experienced just a few years earlier--the decline of elite players who entered about a decade earlier, personified by the Packers dynasty ending in 1968 and the inability of Unitas to stay healthy in 1968.
2) some bad luck. For example, the AFC lost 68% of its fumbles in games against the NFC in 1970, so even though the fumbles were even, the fumbles lost were not. This is just one type of example. We know that luck plays a role in the game of football, and maybe the former AFL teams just had a relatively unlucky season in the small number of games against NFL teams.
3) the quarterback situations and key injuries. Of the ten AFL teams in 1970, only four of them had the same primary quarterback as 1969: Kansas City (Dawson), San Diego (Hadl), Oakland (Lamonica) and Miami (Griese). Some of that might be expected from bad teams changing quarterbacks, but that was the largest quarterback turnover in the AFL since 1962. In 1962, though, it was due to the addition of quarterbacks like Len Dawson and John Hadl, and Jack Kemp moving to Buffalo--in other words, change leading to improvement. That was not the case in 1970. Two of the biggest young stars at the position, Greg Cook and Joe Namath, missed all or most of the 1970 season. If Matt Ryan and Tony Romo both missed all of the 2009 season, don't we think that the NFC East and South would be worse overall? The decline of the 1960 class may have been foreseeable, but the loss of two young star quarterbacks would have been a jolt.
4) Cleveland, Baltimore and Pittsburgh. The three former NFL members consisted of one bottom feeder franchise, and two others that had been among the elites of the NFL. In fact, Baltimore and Cleveland had met in the NFL championship game just two years earlier. Despite having two of the powers of the NFL, these teams combined to go 2-7 against their former mates in the NFC in 1970. If these teams played relatively better against the AFL teams and had their worst games against the NFC, then that would make the whole league look worse. We know that Baltimore beat Dallas in the Super Bowl, so they were likely undervalued due to some unlucky results pulling the AFC teams ratings down in the regular season. I doubt that Baltimore was really more than 7 points worse than Dallas in 1970, just based on previous seasons, and considering they won the Super Bowl despite a -3 turnover ratio, by shutting Dallas down offensively.
5) Week 14. Dallas over Houston, 52-10. San Fransisco over Oakland, 38-7. Houston was a bad team in 1970, so that result isn't too shocking, but the other one is. Oakland had clinched the AFC Western Division the previous week, and as playoff matchups were not based on record, they had nothing to play for. San Fransisco, on the other hand, was in a must win situation to make the playoffs. Oakland had more yardage in that game, but had nine turnovers to the 49ers' zero. A result like that one which happened to occur in an interdivision game could have an impact on the AFL's overall rating in 1970.
6) Home Field Advantage. This gives me a chance to reference research I have done on first time visitors to new stadiums. While the two leagues had not met during the regular season prior to 1970, they had played all those exhibition games, and they were mostly played at AFL sites or neutral venues. Between 1967 and 1969, 46 games were played in AFL stadiums compared to only 15 in NFL stadiums. This continued in 1970. Thus, the NFL players and teams were more likely to have visited the AFL stadiums in 1970, compared with the AFL teams. This should have given a net home field advantage effect to the former NFL teams in 1970.
None of these reasons is going to make a significant impact by itself, but a half point here, and a point there, and the AFL suddenly looks worse in 1970. I think the overall results after the merger show that the average AFL team was behind the average NFL team, and it was unlikely that the average AFL team during the previous decade had caught up, even though the AFL champions won Super Bowls III and IV. Conversely, I think that 1970 may lead us to make some assumptions about the size of the difference between the AFL and NFL at the end of the 1960's that are not justified by all the other evidence.
With that, the evidentiary portion of the AFL versus NFL series is closed. Next up, the closing arguments and demonstrative exhibits (in the form of complete team ratings and league adjustments for every team from 1960-1969).