Looking Back: 2014 Regression Alert Results

Mike Clay checks on the accuracy of his mid-season touchdown regression article.

| 2 years ago
Randall Cobb

Looking Back: 2014 Regression Alert Results


Gates400x300Following Week 8 of the 2014 regular season, I posted my annual Regression Alert piece, which took a look at players with unsustainably low or high touchdown rates.

Today, I’ll be taking a quick look back at the piece in order to see where I went right and wrong.

Going Up

These players had low scoring rates in the first half and I projected improvement in the category in the second half.

Rushing

  Wk 1-8   Wk 9-17    
Player Att TD %   Att TD %   Change
Marshawn Lynch 111 3 2.7% 169 10 5.9% 3.2%
LeSean McCoy 137 1 0.7% 177 4 2.3% 1.5%
Jerick McKinnon 76 0 0.0% 37 0 0.0% 0.0%
Le’Veon Bell 141 1 0.7% 148 7 4.7% 4.0%
Frank Gore 102 1 1.0% 153 3 2.0% 1.0%
Bishop Sankey 75 1 1.3% 77 1 1.3% 0.0%
Andre Ellington 128 2 1.6% 73 1 1.4% -0.2%

Hit: It was a slow start to the season for Lynch, but he ended up leading the position with 17 total touchdowns…Despite losing goal line work to Darren Sproles and especially Chris Polk, McCoy scored four times on the ground after Week 8…Bell was an absolute workhorse down the stretch, but only had seven more carries during Weeks 9-17 than he had Weeks 1-8. He ran for six more touchdowns in the second half, however…Gore scored three quarters of his touchdowns in the second half.

Miss: Sankey’s first and second half splits were almost identical…Ellington missed a chunk of the second half due to injury, but one score on 73 carries is enough of a sample to call this a miss…McKinnon managed only 37 more carries before suffering a season-ending injury.

Receiving

  Wk 1-8   Wk 9-17    
Player Att TD %   Att TD %   Change
Keenan Allen 63 1 1.6% 49 3 6.1% 4.5%
Jared Cook 45 0 0.0% 46 3 6.5% 6.5%
Vincent Jackson 63 2 3.2% 75 0 0.0% -3.2%
Demaryius Thomas 75 6 8.0% 103 5 4.9% -3.1%
Rueben Randle 56 2 3.6% 66 1 1.5% -2.1%
Andre Johnson 69 1 1.4% 72 2 2.8% 1.3%

Hit: Allen missed San Diego’s final two games, but still managed to score three quarters of his touchdowns after Week 8. He’s a major 2015 regression candidate…Lance Kendricks had stolen a bunch of first-half end zone looks from Cook in the first half, but Cook got it turned around with a trio of scores down the stretch…Maybe I should be thanking Case Keenum for this one, but Johnson ended up doubling his first-half touchdown total.

Miss: It’s hard to put Jackson on here because he had what was likely the flukiest touchdown struggles in the league. Expect a big 2015 rebound…Thomas was a risky name to include because his 8.0 percent first-half rate was pretty high. Usage suggested it should’ve been even higher, but Denver moved to a very run-heavy scheme over the final months of the season…Randle’s playing time dropped in the second half, but he still managed 66 targets. Despite a strong finish to the season, Randle counts as disappointing miss.

Going Down

These players had high scoring rates in the first half and I projected a dip in the category in the second half.

Rushing

  Wk 1-8   Wk 9-17    
Player Att TD %   Att TD %   Change
DeMarco Murray 206 7 3.4% 187 6 3.2% -0.2%
Isaiah Crowell 52 4 7.7% 96 4 4.2% -3.5%
Jamaal Charles 77 4 5.2% 128 5 3.9% -1.3%
Lorenzo Taliaferro 51 4 7.8% 17 0 0.0% -7.8%
Chris Ivory 101 5 5.0% 97 1 1.0% -3.9%
Lamar Miller 95 4 4.2% 121 4 3.3% -0.9%

Hit: As good as Murray was this season, his scoring rate was a bit higher than expected based on his workload. This is a tight one, but his rate dipped and I had him down for five scores (he scored six)…Ben Tate was still on the team and Crowell was battling for the team’s No. 2 job when this article was posted. Crowell scored four times down the stretch, but his rate was still nearly cut in half…Charles actually came up short of my six projected second-half scores, but he, indeed, saw his scoring rate dip on more carries, which was right on target with the analysis…Taliaferro only had 17 carries down the stretch, but I’ll count this as a win since the message was that he couldn’t keep scoring at an eight percent clip as a second/third-string back. He was losing reps to Bernard Pierce and didn’t score after Week 8…Ivory’s carry totals were nearly identical in the first and second half, but his touchdowns fell from five to one…Miller saw 26 more carries in the second half, but scored the same number of touchdowns (four).

Miss: None

Receiving

  Wk 1-8   Wk 9-17    
Player Att TD %   Att TD %   Change
Ahmad Bradshaw 36 6 16.7% 10 0 0.0% -16.7%
Randall Cobb 54 9 16.7% 71 3 4.2% -12.4%
Terrance Williams 37 6 16.2% 25 2 8.0% -8.2%
Julius Thomas 40 9 22.5% 20 3 15.0% -7.5%
Antonio Gates 50 9 18.0% 48 3 6.3% -11.8%
Eddie Royal 40 5 12.5% 46 2 4.3% -8.2%
Dwayne Allen 32 6 18.8% 15 2 13.3% -5.4%

Hit: Bradshaw only lasted two more games after Week 8, but he failed to score on 24 touches. This is a borderline wash, but if we’ve learned anything from touchdown regression, we know he was well on his way to coming nowhere close to scoring six times on his next 26 targets (which would’ve matched his first-half total)…Cobb went from nine scores on 54 targets to three scores on 71 targets. Hope you sold high…Williams was an easy one and a dip in targets helped toward cutting his touchdown total from six to two…Thomas and Allen missed a lot of second half action, but both scored a third of the touchdowns on half the targets…Gates had the exact same touchdown split as Thomas, but did it on roughly the same number of targets in each half…Royal regressed hard for the second year in a row.

Miss: None

Conclusion

In the past, we’ve seen roughly 85 to 90 percent success with these regression calls. This year, we were 100 percent accurate (13 of 13) in projecting rate decreases and (7 of 13) in projecting increases. So that’s 77 percent overall, which is solid, but we’ve done better. It’s always a probability that a player or three won’t regress quite enough since it’s a window of only seven or eight games.

Be sure to check in throughout the offseason for more regression and opportunity-adjusted touchdown (OTD) analysis.

Comments are closed.