Reinhold Matay-USA TODAY Sports
Thanks to the nflFastR project and NFL NextGen Stats for the timely sources of data. Commentary will be brief but feel free to let me know in the comments that stats aren’t everything. (click charts for larger view)
For those of you new to this, I will publish key QB stats each week judging how well the Colts passing game performed. I’ll put a glossary at the bottom for those not familiar with the statistical measures.
Wins and losses tend to cloud one’s view of performance, and so many will find it hard to believe that this was actually an improvement over last year’s Colts passing.
Even with the 2 picks, Philip Rivers had good efficiency, recording a 0.21 EPA per dropback and a 53.2% success rate, which were much better then his 2019 numbers (0.09, 49.7%) and far, far better than Jacoby Brissett’s 2019 (0.01. 44.9%).
The following charts are pretty bare with just a single week, but you can see Rivers was above average in the 4 key passing stats most related to wins. QBs that are above average in these 4 stats win about 80% of their games.
Yes, the picks are a killer and the game was primarily short passes, but those short passes were far more productive then the Colts 2019 short passes. In 2019, on passes thrown <=5 yards, Jacoby Brissett had a 74% completion rate, averaging 5.5 yards per attempt. On Sunday, Rivers completed 100% of those passes and averaged 7.6 yards.
Don’t fool yourself. However ugly it was, this was a step up.
You would rather your wide receivers lead the team in receiving yards, but Jonathon Taylor put up a lot of YAC on passes from behind the line of scrimmage.
You can see in the following chart that most of the valuable completions came from shorter passes. (the 2nd chart displays the same numbers but with a different scale).
The average completed pass only traveled 4.2 yards in the air and the average depth of target was a paltry 6.2 yards. Without a high completion rate and/or YAC, that’s not going to generate much value. Rivers will have to stretch the field for any continued success.
With such short throws, Rivers was very quick. His 2.34 release time was second fastest of all Sunday QBs and he took 0 sacks.
Even accounting for the primarily short passing depth, Rivers 78% completion rate was far above expected (11% CPOE). He was perfect from 5 yards or closer, but that is likely unrepeatable, so he’d better learn to get comfortable throwing father downfield.
You can see the middle of the field was problematic and deep shots were almost non-existent.
Dropbacks: Attempted passes, Sacks or QB scrambles with QB spikes removed.
EPA per Dropback (EPA/db): For any given down, distance and field position, there is an average amount of points the team is expected to generate from that spot (EP). After the play the added EP (EPA) is calculated. EPA/db is the average of EPA for all QB dropbacks.
Passing Success Rate (PSR): This is the % of passes that were successful. Success is determined by dropbacks that have an EPA > 0. This is useful, as it evenly weights all dropbacks, so that extreme EPA events can’t bias the measure.
First Downs per Dropack (1st %): The percent of dropbacks that result in a first down.
Net Yards per Attempt (NY/A): Passing yards plus scramble yards minus sack yards divided by dropbacks.
Air Yards (ayd): The distance a pass travels through the air past the line of scrimmage. Passes behind the line of scrimmage have negative air yards. Batted down passes have 0 air yards.
Yards After Catch (YAC): The distance gained by a receiver after catching a pass.
Completion Percent Over Expected (cpoe): The percent of passes completed above expectation based on air yards, field placement and other variables. This is a proprietary stat provided by nflFastR.
Time to Throw (TtT): The number of seconds from snap to release on passing attempts.