Stat of the Game, Packers-Bears: Understanding how bad Justin Fields was through EPA

All EPA stats courtesy of RBSDM.Com.

You probably have some idea how bad Justin Fields was in Sunday night’s game between the Chicago Bears and the Green Bay Packers. His first of 7 completions in the game was a flea flicker to Equanimeous St. Brown that gained 30 yards. His next 6 completions gained a total of 40 yards. He was sacked 3 times for 22 yards. He threw an interception. He threw the ball while 4 yards beyond the line of scrimmage at one point. He fumbled. He was super bad when called on to pass. And yet you still don’t know just how bad he was.

Let’s dispense with the good first. On his first pass attempt of the game (after fumbling on first down and taking a timeout), the Bears ran a nifty flea flicker to spring EQ downfield. Fields delivered, creating 2.5 EPA (Expected Points Added). The Bears would score 3 plays later (all runs) and take a brief lead. Everything was coming up Chicago!

And then it started. The EPA framework, and the EPA statistic more than anything else, helps us to answer an important question about each individual play: namely, was it a good play or a bad play? When Football Outsiders first put together DVOA, they faced the issue of defining a “successful play,” and settled on gaining 40% of the yards needed on first down, 60% of the yards needed on second down, and converting on third or fourth downs. This is necessary when discussing the value of an individual play because context is incredibly important. A 9-yard gain on 3^rd and 15 is far less valuable than a 1-yard gain on 4^th and 1. In football, 9 is not always greater than 1.

FO did a nice job at creating a shorthand for success versus failure back in the day, and they have refined their own definition of success several times over since then, eventually settling on something that looks quite a bit like EPA. At this point, everyone has more or less settled on EPA as the best way to do this, in one form or another.

EPA models leverage the enormous history of football to calculate just how much, in terms of likely points added or subtracted, each play creates or destroys. When a team gains 5 yards on 2nd and 10 from midfield, we now know historically just how likely that team is to put up points (and what type of points), and so can tell generally whether any given play was a success or failure, and how large of one. When the Bears decided to pass, outside of trick plays, they destroyed an almost inconceivable number of points.

On Fields’ first passing attempt after the flea flicker, on a first down at the start of the second quarter, he was sacked for 5 yards by Preston Smith, costing the Bears -1.3 EPA. On the next play, now facing 2^nd and 15, he hit Darnell Mooney for a loss of 4 yards, costing the Bears -0.93 EPA and all but ensuring they would not convert on 3^rd and 19. They absolutely did not convert on 3^rd and 19, as Fields completed a short 7-yard pass to running back David Montgomery. Remember that Fields only had 40 non-flea-flicker passing yards, and 7 of those went to this 3^rd and 19 pass for -0.34 EPA.

His next pass, on first and 10 with 4:30 left in the second quarter, was incomplete to TE Cole Kmet, costing the Bears another -0.41 of EPA. After David Montgomery was stuffed on 2^nd down for a loss of 5 yards, setting up a 3^rd and 15, Fields hit Equanimeous St. Brown for 9 yards, forcing another Chicago punt. This play only cost the Bears -0.03 points of EPA, which is almost break-even. The reason for this is that Chicago was at their own 10-yard line, and when you’re backed up in your own territory on 3^rd and 15, the EPA model doesn’t expect you to score many points, ergo there are not many expected points to give up. Those 9 yards resulted in a punt, but given the circumstances, it could have been worse.

Of course, we now have 16 of Fields 40 non-flea-flicker yards accounted for on 3^rd down passes that didn’t gain a first down. Only 24 to go! Oh, wait, Darnell Mooney lost 4 yards on his one reception. Only 28 to go!

The Packers would take that punt, and the good field position that came with it, and march right down the field, punching it in on a 5-yard pass to Allen Lazard just before half time. With 32 seconds left and trailing 24-7, Chicago would use the final play of the half for a check down to backup RB Kahlil Herbert, who gained 3 yards and cost the Bears -0.24 EPA. Of course, that number is actually quite high, as the Bears were not trying to score with only 32 seconds left in the half, and this was the functional equivalent of taking a knee. I just wanted to point out that one of Fields’ 7 completions was the functional equivalent of taking a knee. Also, we now have just 25 yards to go.

On Fields’ next passing snap on 2^nd and 3, at the start of the 3^rd quarter, he was sacked by Rashan Gary for 7 yards, costing the Bears another -2.0 points of EPA. That sack would prove quite costly as on the ensuing 3^rd down, now 3^rd and 10, Fields would check down to Montgomery for 7 yards, setting up another punt, and costing another -0.43 EPA. (And we’re down to 18 yards to go!)

And now, finally, with 7:58 remaining in the 3^rd quarter, we have what is both Justin Fields’ first actual positive passing play since the flea flicker, and also his last positive passing play of the game. Fields connected with tight end Ryan Griffin over the middle for 18 yards on first and ten, taking advantage of an overmatched Darnell Savage and moving the sticks, helping to set up an eventual field goal. The pass itself was worth +1.23 points of EPA. Hooray Justin!

As for the rest of the game we had an incompletion to Darnell Mooney on 1^st and 10 (-0.57 EPA), a 10-yard sack by Preston Smith (-1.33 EPA), an incompletion to EQ (-0.34 EPA), and of course, the pick by Jaire Alexander (-2.2 EPA). Top it off with the cherry that was his over-the-line throw (-1.3 EPA), and we really have a game for the ages. If you want to put that first fumbled snap on Justin (-0.8 EPA) it gets even worse, but let’s be generous. When it was all said and done, after the flea flicker, Justin Fields dropped back to pass 14 times. Collectively, those 14 passes/sacks lost 10.3 points of EPA, or about -0.73 per play.

That is hilariously bad. It is a nearly impossible feat, especially when throwing only one interception, and if not for some good work on the ground by Montgomery and Herbert (and some corresponding bad play by the Packer defensive line, De’Vondre Campbell, and the safeties), the Packers likely win this game by 30 or 40.

We mentioned Darnell Mooney earlier, as the young receiver has really struggled this season, and did again in this game. Mooney was targeted twice, catching one pass for -4 yards. Mooney’s EPA per play was -0.75.

Basically, after his first passing attempt, Fields was as valuable, on a per play basis, as a wide receiver who caught half of his targets and averaged minus 4 yards per reception.