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.

AFL versus NFL: draft methodology

Posted by Jason Lisk on April 30, 2009

So far, the posts have not really gone into too much detail on the actual teams and players in the American Football League and National Football League during the 1960's. I've looked at how much we can really learn from four championship games, and also looked at expansion teams to get a sense for the rate of improvement we might expect from AFL teams over time, if they were getting equivalent talent to the NFL.

This should be the final stage-setting post before we get to actual details about the AFL and NFL teams from this period. At the outset of this project, I said that I didn't know exactly how many posts would come out of this project, or how it would proceed. And that was true. My plan moving forward as of today is to break the actual drafts up into three periods: early (1960-1963), middle (1964-1966), and late (1967-1969). I'll move chronologically forward, but will not go straight through with just draft discussion. After the early drafts, for example, I will discuss the league trends resulting from the start of the AFL, such as aging patterns, starter retention, and rookie starting rates for the two leagues. I think it will make more sense to do this immediately after discussing the specific players and drafts for the same period.

I'll have similar posts after each period. After the late drafts, when the two leagues drafted jointly, I will close with discussing the pre-season results from the same period, do a more in-depth analysis of the actual Super Bowl teams and the games, and look at the post-merger results, before getting to the final analysis.

Now, to the drafts. As I noted in the introduction, the acquisition of new talent is probably the single biggest factor in how the two leagues compared. It is the one area where they consistently went head to head over a long period of time. The problem is trying to measure that.

I could have come up with some really hokie formula to try to measure the new talent each league was getting. Hopefully, I've avoided doing that too much. The approach I've decided to take for each year is to take a bunch of different approaches, recognizing that no single one is sufficient.

Approach #1: Figure out how much draft value each league got as a percentage of good players drafted.

One problem with the drafts is that there were lots of guys who never started who were drafted--heck, some of them I cannot even tell who they played for, if anyone. I do not want to be relying on non-participants to decide how the leagues compared. So what I decided to do is set a minimum standard for a player to qualify, in terms of seasons started. For guys who signed with an AFL team, it was at least 4 seasons as a starter. For guys who signed with an NFL team, it was at least 3 seasons as a starter. I figure I need to make some slight adjustment in recognition that the early AFL draftees had less competition for starting, with no to few pre-existing starters.

I then took all "good players" who were drafted by either league, and took their highest draft position in either draft, and assigned them a point value based on draft position. I'm not sure that the draft was as sophisticated as today and that the dropoff from pick 5 to pick 100 is as sharp, so I just went with a simple formula of last pick is worth 1 pt, first overall pick is worth X points, where X = number of draft picks in the league that had the most picks that year.

Let's walk through a quick example. In 1961, Tommy Mason was selected 1st overall in the NFL Draft by the Vikings and 2nd overall in the AFL Draft by the Patriots. Bob Gaiterswas selected 1st overall by the Broncos in the AFL Draft. Bob Lilly was selected 13th in the NFL Draft and 14th in the AFL Draft. Mason, who played for Minnesota, would count as 280 points (there were 280 players drafted by the NFL and 240 by the AFL, so 280 is the value for the #1 for both leagues). Gaiters, who played for Denver, is a 0, because he did not start enough seasons to qualify. Bob Lilly is worth 268 points as the 13th pick, even though we know he had a better career than Tommy Mason. We'll get to other approaches to sort that out, but in this approach, once a player qualifies with the minimum starting seasons, their draft value is based on their draft position. This approach is seeking to get a general number on the overall perceived value of the draft for each league, once we discount the bad scouting teams and the draft busts.

Hopefully, that isn't too hokie, or at least makes some sense. And even if you think it is a little hokie, you'll find it to be a good tool so long as it is used with the other approaches.

Approach #2: Head to Head Battles and Steals. The reason I used the highest draft position in the previous approach, rather than the average, is because often, I suspect that the draft position in one league or the other was based on perceived signability rather than talent. Some teams probably knew that certain guys were more likely to sign with one league or the other, and I didn't want those decisions affecting the view of a players' talent.

Under this approach, though, we're only going to look at those cases where a player was drafted in roughly the same range by both leagues, and then see how many each league "won". For example, Gale Sayers was drafted early by both the Bears and Chiefs in 1965. He would be a "win" for the NFL. I'm still debating what should count as roughly the same range, but I'm thinking somewhere around within 20 slots in the two drafts.

The other thing I will look at is steals. This would be a case where a player was drafted much higher in one league, but signed with the other. For example, after missing out on Sayers in 1965, the Chiefs used a late round pick on Mike Garrett out of USC, who was the Rams' 2nd round pick (18th overall selection) in 1966. The Chiefs successfully signed him away, and Garrett would count as a steal for the AFL.

Approach #3: Look at who the best incoming players were for each league
Sounds simple, right. Here, we do categorize Bob Lilly as better than Tommy Mason. Of course, I don't know or have personal knowledge of all of the players from this era. I wouldn't have had any idea who Goose Gonsoulin was, for example. I'll use Doug's pre-merger Approximate Approximate Value to get a snapshot of the best players for each incoming year in each league, and then take that preliminary list and narrow it down to the best eight new AFL players each year, and the best fourteen NFL players each year. (Remember, the NFL had more teams, so they needed to get more good players, relatively speaking).

So, I'll list the top players in each league so that you, the reader, can gauge how good each league was doing, if you want to focus on just the very top players each year. Also, it should help show which positions were going to which league, if there were any differences.

Approach #4: Look at cumulative career totals for the draft class in each league, in things like starting seasons, pro bowls and all-pro seasons.

The issue here is that some adjustments need to be made. It is easier to make a pro bowl when there are 4 teams in a division than when there are 7. It's easier to make a pro bowl, at least early on, when there are no established veterans to compete with, in the AFL. But we'll look at ballpark adjustments, and present the raw numbers. This approach should also let us compare draft classes within each league. For example, we should be able to put the 1963 NFL class next to the 1964 class and see if one was better than the other.

And that's it. For each year, I will summarize the results from using each approach, list the best players each year, and give my interpretation of how each league did relying on those results. I'll also try to summarize the teams that were hurt the most and helped the most in each league by either losing good players or signing more good players than they otherwise would have if all teams were drafting together. Then we can see how those franchises did in the years that immediately followed.

None of this is set in stone, so feel free to let me know what you think. Then we will move on to the drafts from 1960-1963.

This entry was posted on Thursday, April 30th, 2009 at 5:24 am and is filed under AFL versus NFL, NFL Draft. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.