Today I wanted to continue evaluating the successful play stat for NFL offenses that many follow, including us here at Steelers Depot. The goal for today is to look at the running game using the rates many use currently, see how the results stack up to the ultimate goal of winning, and see if tweaking the stat could possibly give us a more predictive and beneficial data point moving forward.
I provided and tracked success rates through the season, so we will start with those measures. Here are successful play rates defined (with spikes, kneel downs, and aborted/penalty plays excluded) from nflfastR:
- First Down Success = A gain of six yards or more, or at least 40% of the needed yardage.
- Second Down Success = A gain of six yards or more, or at least 50% of the needed yardage.
- Late Down Success = A third or fourth down play gaining the needed yardage for a first down or a touchdown.
Here we see the top five ranks all made the postseason, emphasizing a strong correlation of run success for the top teams to winning, and the Buccaneers and Chiefs also achieving this feat in pass success rates. The top ranked team was the Eagles, highlighting their route to the playoffs considering their much lower 24th rank in passing success rate. The Chiefs also had a very high success rate followed the Packers, Bucs, and Patriots after a drop off. The sixth through eighth ranks missed the playoffs: the Ravens, Colts, and Chargers. Three of the remaining four in the top 12 did punch their ticket, resulting in a playoff prediction percentage of 66.7%. By far the lowest result of my studies of the success rates. For context, the total success rates had a 91.7% result, and passing success rates landed at 83.3%.
The following four teams complete the top half of the league in ranks. The Super Bowl Champion Rams (14th) and the 49ers (16th) were the playoff teams, with the Panthers (lowest win total of this group at five) and Seahawks missing out. The remaining four postseason teams were in the bottom half of the league, including the Cowboys (18th), Bengals (23rd), Titans (24th), and finally the Steelers (25th). This was Pittsburgh’s best rank compared to landing 29th in total and passing success. So, only ten of the top 16 in run success rates (62.5%) were playoff teams.
Now to see if tweaks can make the run success rates better. Let’s see what increasing the first down success rate from 40% of the needed yardage to 50%, considering this sets up the offense for more ideal situations on the remaining downs as opposed to the ‘if you gain four yards on every play’ stance. Let’s call this Adjusted First Down (AFD), leaving the other downs unchanged, and see how this looks in comparison:
The top three ranks remain the same, with the Patriots moving up a spot to fourth. The Ravens moved up as well, highlighting there longer gains in the run game comparatively, and as a result fewer playoff teams at the top. Four of the seven remaining teams in the top 12 made the postseason, resulting in the same playoff prediction percentage of 66.7%. The top half of the league’s results were the same as well (62.5%), with the remaining postseason teams moving down to the final two ranks (49ers-15th, Rams-16th). The rank orders for the other playoff teams stayed the same, with the Cowboys moving down to 19th, the Bengals and Titans holding their spot, and the Steelers moving down to 26th when shorter runs were removed from the success rates (that I recall tracking often during the season). Looking at the numbers for more context, Pittsburgh’s run success rate in the initial view was 50.9%, compared to 44.9% with AFD.
Let’s see how adjusting second down as well to at least 60% of the needed yardage, while keeping AFD. I’ll call this Adjusted Early Downs (AED):
The top two teams held their ranks through the different views, with the Ravens continuing their noticeable climb into the top three. The following five teams made the playoffs, meaning seven of the top eight accomplished this goal. While the playoff prediction percentage stayed the same here, the ten postseason teams landed higher in the ranks for the strongest overall results today. Dallas continues to move down to the 21st spot, with the Bengals and Titans ranks still unchanged, and the Steelers moving back up to 25th.
Next, I wanted to see if there were any predictive trends league-wide on a down by down basis, so here are first and second downs alone with AED for additional context:
Of the 13 teams on the top right, nine of them made the postseason (69.2%) with the Ravens, Washington, Jets (biggest outlier with only four wins), and Colts as the exceptions. The reigning champion Rams were above the mean in both data points as they were in total success rates. The Bengals were above the mean on first down, as they’ve been for much of the studies, but well below on second (27th). The Bills were just below the mean on first down and fared well with the third rank on second down. The Cowboys were near the mean on both downs, with their lower mark coming of first down. Pittsburgh was well below the mean on first down (29th) with a low success rate of 33.5%. They were much better on second down (17th) with a 52.7 number, which bested the following four postseason teams: Packers (19th), Cowboys (20th), Titans (25th), and the Bengals.
Now let’s look at late downs (third and fourth downs) alone, which were unadjusted throughout (gaining the needed yardage for a first down or a touchdown):
The Packers, Bucs, Chiefs, and Titans land at the top four ranks with very high run success rates, followed by two more playoff teams in the Bills and Eagles to round out the top six. The following four teams (Broncos, Colts, Panthers, and Browns) missed the postseason, followed by the Cardinals and Washington to complete the top 12. So, only seven teams in the top 12 appeared in the playoffs, the lowest mark today. Also, the result for the top half of the league remained the same with some shuffling of teams, including the Cowboys (13th), Patriots (14th) and the Steelers (15th). Pittsburgh was tied for the third fewest attempts on late downs though, along with their poor results on first down (the highest usage situation), emphasizing their struggles in the run game last season. The playoff teams who were in the bottom half of the league on late downs were the Raiders (20th), Bengals (28th), 49ers (31st), and surprisingly the Rams at last. Considering the Super Bowl participants results and their few attempts in comparison to the league, this gives context to the results and highlights again that run success rates are the least predictive to winning and playoff success.
With the 2021 numbers, AED success rates seem to be the best measure in the run game (but the least predictive), with the AED total success rates being the best playoff prediction measure, and AFD/late down being the most important in the passing game.
After deep diving into different aspects today, I wanted to provide a table that circles back to run success ranks of teams on all downs, to see where each team excelled to wrap up (SPRK= Successful Play Rank initial view, AFDRK = with ADF rank, AEDRK = with AED rank, OVRRK = average of SPRK, AFDRK, and AEDRK with wins breaking any tiebreakers and stars = playoff teams):
The Eagles held the top spot across the board, but just enough wins to make the postseason (nine), compared to the top rank in total and passing success rates having 12 or more. The Chiefs accomplished the top rank in total success rates across the board, along with the top rank in passing success with AED, and second ranks across the board on the run success table above. The Packers had the third OVRRK in run success, matching this result in total success with AED, and their passing success their lowest mark but within the top ten. The Ravens steadily moved up in the adjusted views in the run game with the fourth OVRRK, but had the lowest win total of the top ten along with their 13th ranks across the board in total success and 17th OVRRK in the pass game. The Buccaneers round out the top five in run success, along with their top two ranks throughout the total and passing rates.
The Patriots were fifth in total success and sixth in run success with AFD their best of the views, and their lowest OVRRK of eighth in the pass game. The Bills best rank in the run game and total success was with AED and were just outside the top ten in passing success. The Colts had the eighth OVRRK in run success, with 14th ranks across the board in total success, and outside the top 20 in pass rates. The Cardinals best view in the run game was with AED, and had top five OVRRK’s in total and pass success. The Chargers round out the top ten, claiming the tiebreaker in wins over Washington, also landing in this range in total success and the pass game. The Raiders complete the top 12, matching this result in total success, and their best mark in the pass game with the 10th OVRRK.
This will be interesting to track further, and especially interested in layering other predictive measures since one stat never tells the whole story (but I’m working on it, lol). The best singular stat to predicting Super Bowl winners in my opinion is a personal and Steelers Depot favorite: Adjusted Net Yards Per Pass Attempt that I love to track. I plan to also tackle the TOX stat (turnovers and explosive plays) with a similar goal to see if adjusting the numbers/layering missing elements can provide stronger prediction tools that can be layered together moving forward.
What are your takeaways from the 2021 successful play results? Thanks for reading and let me know your thoughts in the comments!
