Touchdown Regression: All But Guaranteed

| June 2, 2011

[Editor's note: you can also find this article, along with the rest of Mike's work, in our PFF Fantasy section.]

Albert Einstein once said that compound interest is “the most powerful force in the universe.” Although the most powerful force in football is obviously the CBA, Touchdown Regression can’t be too far behind.

Back in November (eight weeks into the 2010 season), I wrote a piece that investigated 22 players with abnormally high or low TD rates. I checked up on that piece in December and the results were telling: a 95% success rate. In January, with the NFL regular season in the books, I did a final assessment, which showed a 91% success rate on the season. I went on to investigate the players involved in the 2010 test and followed up by making a few predictions for 2011.

Today, I’ll be adding to the project by taking a look at severely high and low touchdown rates from the last three seasons at the wide receiver and tight end positions. I’ve found that many players become extremely over or underrated (in fantasy and in real life) after a season in which their touchdown rate was abnormal. History shows that this is a mistake.

To start the process, we need a sample size. I’ll be using any wide receivers and tight ends with 40+ receptions in a season since 2008.

In 2008, 86 WR/TE caught 40+ balls

In 2009, 88 WR/TE caught 40+ balls

In 2010, 94 WR/TE caught 40+ balls

These samples include playoff games. Since we are focusing on a rate stat, we can use the playoff games. This will allow an even larger sample size, which is what we want.

Going Down

Next, we will pull out the players in those samples who put up a 13.0% or higher touchdown rate. Considering the league average TD rate for wide receivers and tight ends is just below 8%, this would be considered a rather high rate. Over the last three seasons, 40 players fit the bill, but 19 of them were during the 2010 season. This gives us 21 players we can investigate.

Year Player Pos Rec TD% Next Rec Next TD% Change
2009 Visanthe Shiancoe TE 61 19.7% 47 4.3% -15.4%
2008 Kevin Walter WR 60 13.3% 53 3.8% -9.6%
2008 Bernard Berrian WR 50 14.0% 67 6.0% -8.0%
2008 Calvin Johnson WR 78 15.4% 67 7.5% -7.9%
2009 Robert Meachem WR 49 18.4% 47 10.6% -7.7%
2009 Larry Fitzgerald WR 109 13.8% 90 6.7% -7.1%
2009 John Carlson TE 50 14.0% 35 8.6% -5.4%
2008 Terrell Owens WR 69 14.5% 55 9.1% -5.4%
2009 Lee Evans WR 44 15.9% 37 10.8% -5.1%
2009 Vernon Davis TE 78 16.7% 56 12.5% -4.2%
2009 DeSean Jackson WR 65 15.4% 49 12.2% -3.1%
2008 Justin Gage WR 44 13.6% 28 10.7% -2.9%
2009 Miles Austin WR 92 13.0% 69 10.1% -2.9%
2009 Roy E. Williams WR 43 16.3% 37 13.5% -2.8%
2008 Larry Fitzgerald WR 126 15.1% 109 13.8% -1.3%
2008 Randy Moss WR 69 15.9% 88 14.8% -1.2%
2008 Vincent Jackson WR 61 13.1% 75 12.0% -1.1%
2009 Greg Olsen TE 59 13.6% 47 12.8% -0.8%
2009 Fred Davis TE 48 14.6% 21 14.3% -0.3%
2009 Randy Moss WR 88 14.8% 28 17.9% 3.1%
2008 Visanthe Shiancoe TE 43 16.3% 61 19.7% 3.4%

The chart above shows each of our 21 players’ reception total and TD rate during the season noted in column 1. ‘Next Rec’ and ‘Next TD%’ refer to the reception total and TD rate the player put up the next season. Finally, we have the ‘Change’ column, which shows us the difference in the player’s TD rate from Year 1 to Year 2.

The results are very telling. 19 of the 21 players (90%) saw a drop in TD rate in Year 2. The only exceptions were Randy Moss, whose sample size dropped by 60 receptions, and Visanthe Shiancoe, who enjoyed a ridiculous TD rate with Brett Favre under center in 2009. Still, Shiancoe regressed in epic style the following season, dropping by 15.4% to 4.3% in 2010. Notice that only six (29%) of the players managed to put back-to-back 13.0%+ TD rate seasons together, and 11 (50%) saw a drop of at least 3%. It’s worth noting that 16 (76%) of the players did manage an above-average TD rate the following season, but remember that our goal here is to determine how easy/hard it is to maintain the “extremely” high TD rates. Clearly, it’s almost impossible.

Going Up

Next up, we’ll take a look at players who suffered a 3.0% or lower TD rate. As mentioned, the league average rate at these positions is right around 8%, so anything under 3% is considered to be poor/unlucky. 26 players fit the bill since 2008, four of which qualified in 2010. That gives us a sample of 22, but notice that five of the players managed five or fewer receptions in Year 2. That isn’t quite enough to give us a fair assessment of their TD rate, so we will eliminate them and focus on the other 17.

Year Player Pos Rec TD% Next Rec Next TD% Change
2009 Malcom Floyd WR 48 2.1% 37 16.2% 14.1%
2008 Owen Daniels TE 70 2.9% 40 12.5% 9.6%
2008 Jeremy Shockey TE 50 0.0% 55 9.1% 9.1%
2009 Bo Scaife TE 45 2.2% 36 11.1% 8.9%
2009 Jason Witten TE 108 1.9% 94 9.6% 7.7%
2008 Chris Cooley TE 83 1.2% 29 6.9% 5.7%
2009 Greg Camarillo WR 50 0.0% 20 5.0% 5.0%
2008 Steve Smith WR 59 1.7% 107 6.5% 4.8%
2009 Mike Thomas WR 48 2.1% 66 6.1% 4.0%
2009 Davone Bess WR 76 2.6% 80 6.3% 3.6%
2008 Zach J. Miller TE 56 1.8% 66 4.5% 2.8%
2009 Michael Jenkins WR 50 2.0% 47 4.3% 2.3%
2008 Josh Reed WR 56 1.8% 27 3.7% 1.9%
2009 Danny Amendola WR 43 2.3% 85 3.5% 1.2%
2008 Davone Bess WR 56 1.8% 76 2.6% 0.8%
2008 Wes Welker WR 111 2.7% 123 3.3% 0.5%
2009 Antwaan Randle El WR 50 0.0% 24 0.0% 0.0%

Not coincidentally, we see the same trend here. None (0%) of the players saw a drop in TD rate and all but one (95%) saw an increase. Our only exception is Antwaan Randle El, who defied the odds by going two full seasons (74 receptions) without a touchdown. Only two (12%) of the players managed back-to-back seasons under 3.0% and nine (53%) enjoyed and increase of at least 4%. Only five (29%) players in our sample ended up with a TD rate above league average in Year 2, but that is to be expected.

2011 Losers

Year Player Pos Rec TD% Proj TD%
2010 Rob Gronkowski TE 46 21.7% 15.0%
2010 Kenny Britt WR 42 21.4% 10.0%
2010 Dwayne Bowe WR 72 20.8% 14.0%
2010 Antonio Gates TE 50 20.0% 12.0%
2010 Marcedes Lewis TE 58 17.2% 13.0%
2010 Mike A. Williams WR 64 17.2% 14.0%
2010 Mike Sims-Walker WR 43 16.3% UFA
2010 Calvin Johnson WR 77 15.6% 14.0%
2010 Mike Wallace WR 73 15.1% 14.0%
2010 Mario Manningham WR 60 15.0% 12.0%
2010 Greg Jennings WR 97 14.4% 12.0%
2010 Brandon Lloyd WR 77 14.3% 14.0%
2010 Nate Washington WR 42 14.3% 12.0%
2010 Hakeem Nicks WR 79 13.9% 14.0%
2010 Austin Collie WR 58 13.8% 12.0%
2010 Jeremy Maclin WR 73 13.7% 12.0%
2010 Dez Bryant WR 45 13.3% 13.0%
2010 Santonio Holmes WR 61 13.1% 12.0%
2010 Aaron Hernandez TE 46 13.0% 12.0%

A whopping 19 players enjoyed Touchdown rates that were 13.0% or higher in 2010. If history tells us anything, it’s that near 90% of these players will see a drop in 2011. Included in the ‘Proj TD%’ column is the touchdown rate my 2011 projections currently show for each player in the list. Probability tells us to expect plenty of regression here, which is why 18 of the 19 players show a decrease.

2011 Winners

Year Player Pos Rec TD% Proj TD%
2010 Brian Hartline WR 43 2.3% 4.0%
2010 Tony Scheffler TE 45 2.2% 6.0%
2010 Steve Breaston WR 47 2.1% 4.0%
2010 Chansi Stuckey WR 40 0.0% UFA

The NFL is becoming more and more of a passing league each year and this project just adds fuel to the fire. After 23 players managed a 3.0% or lower TD rate during the 2008 and 2009 seasons, only four suffered the fate in 2010. Expect an increase from each of these players in 2011.

Final Thoughts

Studying this data every day, it becomes progressively more apparent that Touchdown Rate is predictable … to a degree. The league will always have possession receivers who consistently put up low TD rates and big-play/redzone receivers who manage high rates. The rest will almost always end up in the 6-10% range. You’ll have outliers, but the stats don’t lie. A statistical test that manages a 90%+ success rate is one that should be strongly considered in future projections.

Extreme touchdown rates are all but impossible to maintain … a sign that regression is, well, maybe just as powerful as the CBA.

Follow Mike on Twitter: @PFF_MikeClay

  • PaulK

    I suspect that you may be wrong about Gronkowski. He’s a pretty unique pass-catcher because of his height and leaping ability, and the wide receivers mostly excel at quickness so they get wide open. Brady has had little reason to target Gronkowski as his first, second or third option on most of the field, because everyone else is wide open and Gronkowski is often still on the line pass-blocking. In the red zone the defense doesn’t have to play the deep ball, but Gronkowski leaping high is always an option. Also, a catch in the endzone can’t be fumbled and it’s hard for a linebacker to get a solid hit two steps after a player has caught the ball in the endzone. Maybe Gronkowski has his weaknesses and Belichick is trying to compensate.

    • http://www.profootballfocus.com Mike Clay

      I’m willing to guarantee he sees a drop in TD rate. You just can’t repeat a 20%+ TD rate. It doesn’t happen.

    • http://www.profootballfocus.com Eric Yeomans

      I’m going to have to go with Mike on this one. Even projecting a 15% TD rate alone seems almost ridiculously high. Projecting anyone for a 20% TD rate would just be completely irresponsible.

  • tom

    As much as Gronkowski is a unique player, I’m sure opposing defensive coordinators have the same kind of analysis as here and would be devising ways to combat him and others with high TD rates this coming year. I doubt we will be seeing Gronk 1-on-1 up the field against smaller DBacks consistantly, plus Brady is always looking for the receiver that should be open first according to the defense played.