≡ Menu

The NFL Draft is this week, which means we have something resembling real football to talk about. But how much impact will the players who hear their names called during the 2014 Draft have on the 2014 season? Here’s the short answer: as a group, they will make up about 10% of games played by all players and 8% of all starts.

What do I mean by that? Each year, every team’s players start 352 games, which is the product of 16 (games) and 22 (starters). Players selected during the 2013 Draft started 27 games per team last year, which is in line with the recent average of eight percent. I also looked at the number of games played by all drafted rookies, and divided that by the number of games played by all players on that team. Take a look: the blue line represents games played by drafted rookies and the red line represents games started; both numbers on shown on a percentage basis for the league as a whole.

drafed rookies

So why the decline in the blue line? My guess is that’s due to the draft being limited to just seven rounds now, versus 17 rounds in the early days of the post-merger NFL. But the red line, representing starts, hasn’t changed too much. In general, the ratio has hovered between six and eight percent. Remember, 8% represents 28 total starts by rookies in the 16-game season per team, so some natural variation over the years is to be expected.  Of course, this is just the average trend: What about the very best team in the league?

While every team is optimistic that the key to their next Lombardi Trophy is just one pick away, it’s unusual for a team to really build a Super Bowl title around rookies; the most games started by rookies on any Super Bowl champion is 48 (can you guess who?). So if it’s not their most recent group of draft picks, which draft class is it that — on average, of course — helps Super Bowl champions the most? That is, if you look at any Super Bowl champion, in which year did they draft the players that most contributed to that title? It’s a complicated question that you can attack from many angles. Here’s what I did.

Pro-Football-Reference has assigned an Approximate Value (AV) to each player on each team since 1950. I looked at which draft classes produced the most AV for each Super Bowl team. For purposes of this article, undrafted free agents were excluded from a team’s draft class.

Let’s look at the current Super Bowl champions. The 2013 Seahawks were built mostly on recent drafts, along with one player each from older drafts (Brandon Mebane in 2007, Red Bryant in 2008, and Max Unger in 2009).  In 2010, Earl Thomas, Kam Chancellor, and Golden Tate were added; in 2011, it was Richard Sherman, K.J. Wright, SB MVP Malcolm Smith, and James Carpenter.  Two years ago, of course, it was Russell Wilson, Bobby Wagner, J.R. Sweezy, and Bruce Irvin.

But last year? Well, the Seahawks did deal their first round pick for Percy Harvin, but last year’s rookies made almost no contribution to Seattle’s Lombardi Trophy.  Christine Michael, Jordan Hill, and Chris Harper were the top three picks, while backup tight end Luke Willson was the only rookie to get in to more than ten games. Seattle is a classic example of how Super Bowl Champions in February are almost never built on the work from the most recent April.

I looked at each Super Bowl champion and graded which draft class provided the most AV during that Super Bowl season. Fourth-year players — i.e., the class from three years earlier —  was the one most commonly producing the most value.  That was the case for 12 of the 48 Super Bowl champs, including, curiously enough, the ’74 Steelers.  Pittsburgh’s draft class that year is perhaps the most famous draft class of all-time — Mike Webster, Jack Lambert, John Stallworth and Lynn Swann all would win four Super Bowls with Pittsburgh and wound reach the Hall of Fame. And that quartet accounted for 19% of the Steelers’ team AV in the ’78 and ’79 Super Bowl seasons.

But the 1974 Steelers relied heavily on the class from three years earlier, as is often the case for Super Bowl champs.  Jack Ham, Larry Brown, Mike Wagner, Dwight White, Ernie Holmes, Frank Lewis and Gerry Mullins were all starters on the ’74  team, and that class produced 22% of the team’s AV that year.

We can look at how much, on average, each draft class contributed to the 48 Super Bowl champions. As you can see, about 5% of the total AV on Super Bowl teams comes from rookies, 8.5% from players in the second season of their careers, 8.7% for third-year players, 9.4% for fourth-year players, and so on. Roughly 38% of all AV can be attributed to players acquired through trades, free agency, or undrafted free agents:


[Note: You can see how that compares to all teams, and not just Super Bowl champs, by revisiting this post.]

Don’t get confused by the draft year/Super Bowl year conflation: players in their 4th year provide the most value, but they arrived three draft classes ago (i.e., the ’71 Steelers for the ’74 champs).  As you can see, homegrown players who are in their 2nd-through-5th seasons stand out as the most valuable classes for the average Super Bowl champ, although 38% of the average champion’s AV comes from sources other than the draft. (Related post: I calculated the percentage of AV from homegrown players for each team in the first 47 Super Bowls last year.)

What’s the best single draft class that helped a Super Bowl champion? That question’s easy, at least according to AV. In 1983, Chicago drafted Richard Dent, Jimbo Covert, Dave Duerson, Willie Gault, Mike Richardson, Tom Thayer and Mark Bortz — all started on the famous ’85 team, and collectively they produced 77 points of AV, easily the most from any one class of any Super Bowl champ. Players from that draft provided 27% of the team’s AV that year, also by far the most from any one class for any Super Bowl champion.

The average Super Bowl champion has given just 11.5 starts to rookies (the Seahawks actually handed out 15 starts split between Wilson and 7th round tackle Michael Bowie).  Which team gave the most starters to rookies? The 1981 49ers, of course. While Joe Montana and Dwight Clark — two members of the 49ers’ 1979 draft class — starred in the most memorable moment for the team, it was the young secondary that helped San Francisco rank in the top three in points allowed, yards allowed, passing yards allowed, net yards per pass allowed and turnovers forced. In 1980, San Francisco ranked 2nd to last in passing yards allowed; the 49ers responded by drafting Ronnie Lott, Eric Wright and Carlton Williamson in the first three rounds of the 1981 draft.  Those three players each started 16 games for San Francisco that year.

Whoever wins the Super Bowl in February 2015 will surely be able to point to several key draft picks that helped the franchise reach its ultimate goal. It’s just that they’re more likely to have made those draft picks in April 2011 than this week.

  • Ben

    Very interesting stuff. Would you be able to take a look at how the superbowl teams differ from teams in general? What would that second chart look like if all teams were included?

  • Richie

    I wonder if there might be more to the higher-producing rookies in the 70s and 80s than just more draft rounds. Weren’t roster sizes a few players smaller back then?

    Is it really possible that players drafted in rounds 8 through 17 were really making the roster more often than all the undrafted free agents are making rosters now?

    • Chase Stuart

      This study looks just at drafted rookies, so what UDFAs are doing is not included.

      • Richie

        LOL I guess I should have waited a few more minutes before looking into this.

    • Richie

      I thought I would investigate this a bit. Unfortunately, the Play Index is kind of weak for doing this kind of research (or I don’t know how to take advantage).

      Since I couldn’t find an easy way to find rookie production by draft position, I thought I would just look at career production for late draft picks. I sampled a few eras.

      In 1975, there were 29 guys drafted after the 7th round who went on to start 16 or more games in their careers. (Incidentally, a lot of those guys didn’t even play in the NFL in 1975.) There were an additional 59 players who were undrafted, first started playing in 1975 and went on to start 16+ career games. So that’s kind of 88 players who theoretically could have been rookies in 1975, and went on to start 16+ career games.

      I did the same thing for 1988. That year there were 25 late draft picks, plus 100 undrafted players, for a total of 125 “Rookies”.

      In 1998, the draft was only 7 rounds. So there are 0 “late draft picks” from that year. There were 134 undrafted players, for a total of 134 “Rookies”.

      So it looks like, adjusting for number of franchises, teams in 1988 (4.4 per team) and 1998 (4.4 per team) were finding about 1 additional player who went on to start 16+ games for their career, and were not drafted in the first 7 rounds, compared to 1975 (3.4 per team).

      The chart above shows production for rookie year, and I pulled out career production. But it seems like these should be somewhat related. Also, I obviously used a small sample size. I picked 1975 because it appears to be the high-water mark of rookie production. I picked 1988 because it was about a decade later, but after the 1987 replacement-player season (that might skew the results), and then I picked 1998 because it was 10 years after 1988.

      Not enough information to really draw a conclusion, but my hunch still stands that the extended draft of the 1970s might not be the reason for the additional rookie production at that time.

      • Richie

        And just as an aside, in recent years it looks like the 2010 Packers were the Super Bowl champ with the most AV from undrafted first-year players. They had 13 AV from them.

  • James

    “Whoever wins the Super Bowl in February 2015 will … more likely to have made those draft picks in April 2011 than this week.”

    [Quickly looks up 2011 NFL Draft]

    So it looks like the AFC Championship game will be between the Bengals (AJ Green, Andy Dalton) and the Broncos (Von Miller, Julius Thomas) while the NFC Championship game will be between the Cowboys (Tyron Smith, DeMarco Murray) and the 49ers (Aldon Smith, Colin Kaepernick).

    That actually sounds very reasonable. (Yes, I’m a Cowboys fan).

  • James

    But more practically, it seems like when a team/fan is evaluating their draft needs they should be focusing not on the current team but what holes the team will have NEXT year and for the three years after that.

    It also gives a lot of credence to the cliche ‘can’t evaluate a draft class until after three seasons’.