Over the last few weeks, Daniel Houston, who goes by the handle @CowboysStats on Twitter, has been tearing it up on social media with his look at the 2018 Cowboys through the lens of EPA (Expected Points Added). I started looking at EPA way back in 2010, and my interest in the metric was rekindled by Houston, so today we’re taking an EPA deep dive to understand what’s working for the Cowboys offense and what isn’t.
What is EPA?
EPA is calculated by taking the expected point value (based on the league average results for that specific down, distance, and field position) before a play is run, and then subtracting it from the expected point value of after a play is run.
The EPA for any given play is a value between -7 and +7. A positive value means the play result contributed positively towards a score, a negative result means the play decreased the team’s odds of scoring.
EPA starts with the basic premise that not all yards are created equal. For example, a three-yard gain on 3rd-and-2 is much more valuable than a seven-yard gain on 3rd-and-8. Any measure of success must consider the down and distance situation, otherwise you’re firmly in the realm of fantasy football.
ESPN explained the concept in a little more detail when they started using EPA back in 2012:
To make the concept more tangible, here are some examples:
• From your own 20-yard line, an 8-yard gain on third-and-10 is worth about minus-0.2 EPA because you don’t get a first down; the same 8 yards on third-and-7 is worth 1.4 EPA for converting a long third down and keeping the drive alive. EPA knows that not all yards are created equal.
• A turnover on first-and-10 at midfield that is taken back to your own 20 is worth minus-5.5 EPA; a Hail Mary interception at the end of the half from midfield is not nearly as penalizing. EPA knows that all turnovers aren’t created equal, as well.
• A 60-yard pass play down to the 1-yard line on third-and-10 is worth 5.7 EPA because it puts you right on the doorstep of scoring. The subsequent 1-yard rushing TD on first-and-goal is worth much less, even though that’s the play that actually gets you the six points. Think about which play is more valuable to the offense (not in terms of fantasy football).
“Success rate” (gaining 40% of the necessary yardage on first down, 60% on second down and 100% on third and fourth downs) is a term that is being used with increasing frequency by NFL observers and probably even some teams. That’s laudable for a sport where a Super Bowl-winning coach can say things like “Stats are for losers. Final scores are for winners,” and still be taken seriously in 2018 by beer-bellied sportswriters who still think it’s 1975.
But relying on success rate leads teams to focus on maximizing success rather than maximizing the likelihood of scoring - which EPA focuses on.
For example, on 2nd-and-short, EPA suggests teams should probably be throwing down the field. But in many cases they do not, preferring instead to run the ball and thereby improving their success rate and not necessarily their likelihood of scoring.
Points, Not Yards
Using EPA (courtesy of Pro-Football-Reference.com), I can tell you exactly what each play so far in 2018 was worth.
- The highest value non-TD pass? Dak Beasley’s 16-yard reception on 4th-and-3 in the fourth quarter against the Seahawks. EPA: 4.36
- The most costly turnover? Dak Prescott’s sack and fumble in the fourth quarter against the Panthers. EPV: - 4.09
- The highest value run? Prescott’s conversion on 4th-and-1 in the fourth quarter against the Giants. EPV: 2.45
- Most costly play allowed? Russell Wilson’s 52-yard TD pass on 3rd-and-9 to Tyler Lockett that put the Seahawks up 14-3. EPV: -6.04
- Highest value defensive play? Damien Wilson’s sack of Eli Manning and subsequent fumble recovery by Taco Charlton. EPV: 4.83
Earlier this week, Daniel Houston published the cumulative EPA for the Cowboys offense over the first three weeks:
Here's the updated season-to-date chart, reflecting Cowboys offensive EPA from weeks 1-3.— Cowboys Stats & Graphics (@CowboysStats) September 24, 2018
It would be easy to describe Elliott as a non-factor in the passing game, but that would be too generous. Elliott is leading the team in pass targets, and it's sinking drives. pic.twitter.com/FAZBOhyjWo
Inspired by Houston’s efforts, I decided to take a look at the 2018 EPA for the Cowboys’ offensive personnel, but instead of looking at the cumulative numbers, I chose to look at the EPA per play for the first three games.
Before we look at the specific results of those games, keep in mind that three games is still a small sample to work with, and if a guy was targeted twice in the passing game but did not catch any of the passes, he’s bound to have a negative EPA, while a guy who catches a 64-yard touchdown pass is almost certain to have a great EPA. But even with that caveat, it’s still interesting to look at the EPA results so far, and we’ll start with the ground game.
Dallas Cowboys running game EPA
Including turnovers on running plays, the total EPA for the Cowboys’ ground game adds up to 14.49 on 63 runs (excluding three QB kneels, three plays negated by penalty, and one two-point conversion). That’s an EPA of 0.23 per run, which makes the Cowboys ground game marginally positive. Here’s how that breaks down per player:
For the ground game, the data here suggests that Dak Prescott should run more; per EPA, he’s averaging almost a point per run. That’s sensational. The Cowboys would be well-advised to add more plays designed for a QB run.
What’s true for Prescott is also true for Tavon Austin: the Cowboys will improve their likelihood of scoring if they give him more touches. Dave Halprin brought up both those ideas in an earlier post this week.
But what’s up with Ezekiel Elliott? At first glance it doesn’t look like he (and by extension the entire run game) is contributing much in terms of points added. But after three games, Elliott leads the NFL in rushing, as my good friend ScarletO points out below, so how does that mesh?
Following the Seattle game, Ezekiel Elliott is leading the NFL in rushing. He has the highest yds/carry average (5.7) among players carrying the ball at least 16 times a game. Only Gurley (20.6 carries/game) gets more 1st downs/carry than Elliott (RB's with 16+ carries/game).— Scarlet O (@ScarletO468) September 25, 2018
Part of the reason for Elliott’s low EPA value is that in terms of EPA, the average combined value of runs for most of the field is close to zero. Here’s why: a first-down play needs at least four yards to be break-even in terms of EPV, and in the NFL this year 54% of runs (653/1215) on 1st-and-10 gained less than four yards.
So if you’re in a situation where half your runs have a negative EPA and the other half has a positive EPA, your total running EPA should be at or around the zero mark. So Elliott isn’t doing that badly with a 0.1 overall EPA/play.
But there’s another reason for Elliott’s low EPA, and that has to do with how the Cowboys are using him. Here’s a look at his EPA values by down-and-distance clusters (excluding his fumble against the Seahawks):
|Ezekiel Elliott||Runs||Total EPA||EPA/play|
|1st and 10||28||4.26||0.2|
|2nd/3rd and less than 5||26||4.03||0.2|
|2nd/3rd and long||12||-1.02||-0.1|
With an EPA/run of 0.2, Elliott is delivering above average points on the ground in 1st-and-10 and 2nd/3rd-and-short situations.
- Todd Gurley of the 3-0 Rams for example has an overall EPA/run of 0.1, and EPA/run on 1st-and-10 just above zero.
- Saquon Barkley of the Giants has an overall EPA/run of 0.1, and a -0.1 EPA/run on 1st-and-10.
- Adrian Peterson, now in Washington, has an overall EPA/run of just below zero and an EPA/run on 1st-and-10 of just below zero.
But what’s not working for Elliott are 2nd/3rd and long situations, where he has a -0.1 EPA/run on 12 runs. Of those 12 runs, one was on 2nd-and-6, one was on 2nd-and-9, all others were in 2nd/3rd down situations with 10-or-more yards to go. Why would the Cowboys choose to run and not pass in those situations?
Ultimately, the Cowboys should look for ways to get more out of Ezekiel Elliott on the ground; right now he’s a little above league average, at least according to EPA. San Francisco’s Matt Breida, who co-leads the NFL in rushing yards with Elliott, has an EPA/run of 0.4. That’s where Elliott should be.
Dallas Cowboys passing game EPA
Same exercise for the receivers, though we’ll change up things a little by first only looking at their EPA per reception (table excludes players with less than three targets).
The data here is pretty straightforward. The Cowboys have three tiers of receivers.
Tier 1: Austin, Beasley, and Hurns all add more than one EPA point per reception.
Tier 2: Gallup, Thompson, and Williams average about half the value of the Tier 1 guys per reception.
Tier 3: Swaim and Elliott do not improve the team’s likelihood of scoring when they catch the ball. This should have all sorts of alarm bells going off at The Star in Frisco.
But looking only at receptions results in an inflated EPA. Which is why we need to include the number of targets for each wide receiver, even if an incompletion is not always the fault of the receiver.
In the passing game, Tavon Austin looks great thanks in part to his 64-yard TD reception, but he’s not going to repeat that every game.
What’s more interesting to look at are the high-volume targets like Cole Beasley (above average), Deonte Thompson (average), Ezekiel Elliott and Geoff Swaim (both below average).
- Cole Beasley is one of the few bright spots in the passing game. What’s interesting here is that Beasley is not a good 1st-down target, delivering 0.0 EPA/play on his six 1st-down targets. But on his ten targets on 2nd/3rd/4th down, he’s averaging 1.1 EPA/play.
- Deonte Thompson looks average at first, but his numbers are hurt by a failed 4th-and-10 conversion against the Panthers. Excluding that one play, Thompson’s EPA/play doubles to 0.4. Also, he’s a good alternative to Beasley on 1st downs, averaging 0.5 EPA/play there.
- Geoff Swaim should not be targeted in the passing game. Every pass thrown his way loses the Cowboys 0.4 points. In fact, the Cowboys should avoid throwing to their TEs altogether. Including Blake Jarwin and Rico Gathers, Prescott has targeted his TEs 15 times and averaged -0.7 EPA/play.
- Ezekiel Elliott is not currently a viable option in the passing game. 18 targets with -0.5 EPA/play is a sure way to end up punting on a lot of drives. Elliott is clearly not an efficient part of the passing game yet. Also, it doesn’t help that he’s Checkdown Charlie on 3rd-and-long.
The Cowboys still have a lot to fix on their offense, and making Elliott a more effective part of the offense is one of the key tasks. Perhaps Elliott’s been hampered by his lack of preseason action, and he’ll round into form in the next few games. But if he doesn’t, then the offense is in big trouble.
The data here also suggests the Cowboys need a receiving tight end much more than they need an extra safety, and if I were in the Cowboys’ front office, I’d trade for a tight end immediately. Which means they won’t.
Other To Dos:
- If you need a TE to help you block, use a tackle instead. If you need a TE to run a pass route, use a WR instead (Hint: you have enough of those).
- Make Tavon Austin the star of this offense.
- The Dak-to-Cole connection is fixed. Use it.
- Don’t throw to Zeke.
- Listen to Joey Ickes
2. PA, read option, or RPO on every 1st down.— Joey Ickes (@JoeyIckes) September 24, 2018
You wanna flow a crap ton of guys to the ball to slow Zeke down? That’s fine I’ll hit you backside with QB runs and quick throws coming over the middle.
This QB plays much better when he gets his legs involved. So do it.