2018 Passing Plus-Minus
by Rivers McCown
We have run passing plus-minus as a seasonal overview every offseason since 2015. In both 2016 and 2017, Drew Brees was the best quarterback in the NFL in passing plus-minus. This year was no exception.
Passing plus-minus is a stat we annually track to help provide context to completion percentage. Given the location of a quarterback's passes, it compares his completion percentage in each area to historical baselines. This stat does not consider passes listed as "Thrown Away," "Tipped at Line," or "Quarterback Hit in Motion" by Sports Info Solutions charting. How often a pass is completed based on the pass distance, the distance required for a first down, and whether the ball was thrown to the left, middle, or right side of the field. This is a counting stat, so more attempts are obviously a great thing for the purposes of what we're talking about here.
For example, the average quarterback would have completed 69.9 percent of the passes that Drew Brees threw in 2018. Brees completed a single-season record 74.4 percent of his passes. That calculus, plus 464 attempts, put Brees at a plus-minus of +39.6. Updating research that FO did in 2015, Brees now has the following seasons in the top 10:
| Best Single-Season Passing +/-, 2006-2018 | ||||
| Rk | Player | Team | Year | +/- |
|---|---|---|---|---|
| 1 | Drew Brees | NO |
2011 | +48.4 |
| 2 | Drew Brees | NO |
2013 | +42.6 |
| 3 | Peyton Manning | IND |
2009 | +42.3 |
| 4 | Drew Brees | NO |
2018 | +39.6 |
| 4 | Drew Brees | NO |
2014 | +39.6 |
| 6 | Drew Brees | NO |
2009 | +39.5 |
| 7 | Philip Rivers | SD |
2013 | +39.4 |
| 8 | Peyton Manning | DEN |
2012 | +39.4 |
| 9 | Drew Brees | NO |
2017 | +37.9 |
| 10 | Drew Brees | NO |
2016 | +36.5 |
Not only is Brees in possession of seven of the top 10 plus-minus seasons, his 10th-place spot is 2.6 plus-minus points ahead of Aaron Rodgers' 2011 season in eleventh. This is a statistic that Brees has dominated, and it hasn't even been close.
2018 Passing Plus-Minus
These results run 35-deep: I picked every player with 200 or more attempts, and then added Lamar Jackson and Nick Foles because they'll be starters next season and I figured people would be curious. I apologize to Brock Osweiler's fans, but I figured that we know plenty about Brock Osweiler at this stage of all of our lives. I offer similar apologies to Jeff Driskel and C.J. Beathard.
| 2018 Passing +/- Leaders | |||||
| Rk | Player | Team | Attempts | C%+ | +/- |
|---|---|---|---|---|---|
| 1 | Drew Brees | NO | 464 | +8.5% | +39.6 |
| 2 | Kirk Cousins | MIN | 555 | +5.4% | +30.0 |
| 3 | Matt Ryan | ATL | 582 | +3.5% | +20.2 |
| 4 | Russell Wilson | SEA | 385 | +5.1% | +19.6 |
| 5 | Deshaun Watson | HOU | 475 | +3.7% | +17.5 |
| 6 | Derek Carr | OAK | 510 | +3.4% | +17.3 |
| 7 | Jared Goff | LAR | 508 | +2.9% | +14.5 |
| 8 | Carson Wentz | PHI | 386 | +2.7% | +10.5 |
| 9 | Patrick Mahomes | KC | 541 | +1.9% | +10.2 |
| 10 | Philip Rivers | LAC | 481 | +2.0% | +9.8 |
| 11 | Cam Newton | CAR | 442 | +1.9% | +8.2 |
| 12 | Ryan Fitzpatrick | TB | 236 | +3.4% | +7.9 |
| 13 | Marcus Mariota | TEN | 312 | +2.5% | +7.7 |
| 14 | Nick Foles | PHI | 189 | +2.6% | +4.9 |
| 15 | Dak Prescott | DAL | 499 | +0.9% | +4.7 |
| 16 | Nick Mullens | SF | 242 | +1.4% | +3.3 |
| 17 | Mitch Trubisky | CHI | 417 | +0.7% | +3.1 |
| 18 | Jameis Winston | TB | 363 | +0.7% | +2.6 |
| 19 | Ben Roethlisberger | PIT | 643 | +0.4% | +2.3 |
| 20 | Baker Mayfield | CLE | 459 | +0.4% | +1.9 |
| 21 | Andrew Luck | IND | 610 | +0.1% | +0.9 |
| 22 | Tom Brady | NE | 532 | +0.0% | +0.0 |
| 23 | Aaron Rodgers | GB | 541 | -0.1% | -0.7 |
| 24 | Ryan Tannehill | MIA | 253 | -0.7% | -1.9 |
| 25 | Alex Smith | WAS | 301 | -1.4% | -4.1 |
| 26 | Eli Manning | NYG | 549 | -1.0% | -5.5 |
| 27 | Joe Flacco | BAL | 348 | -1.8% | -6.3 |
| 28 | Matthew Stafford | DET | 524 | -1.5% | -8.1 |
| 29 | Lamar Jackson | BAL | 158 | -5.5% | -8.7 |
| 30 | Andy Dalton | CIN | 341 | -3.1% | -10.7 |
| 31 | Case Keenum | DEN | 540 | -2.7% | -14.8 |
| 32 | Sam Darnold | NYJ | 381 | -4.7% | -17.9 |
| 33 | Blake Bortles | JAX | 373 | -5.7% | -21.3 |
| 34 | Josh Allen | BUF | 291 | -8.3% | -24.3 |
| 35 | Josh Rosen | ARI | 351 | -7.2% | -25.4 |
| Minimum 200 passes, plus Nick Foles and Lamar Jackson. | |||||
This season was an interesting result for Brees because he played up to the highs that he had in the past, but for once, he didn't combine that historic efficiency with an overwhelming number of attempts. This was the first time in Brees' entire career in New Orleans that he had fewer than 500 pass attempts. I'm sure Brees will take the tradeoff considering the 13-3 record and the whole "probably should have been in the Super Bowl if officiating buts were made of candy and nuts" thing. I think he'll be OK with these results.
Baker Mayfield's plus-minus of +1.9 may not seem all that impressive on first blush, but when you compare it to other rookies, it was pretty stellar. Marcus Mariota and Deshaun Watson earned +3.0 and +2.9, respectively, in their rookie seasons. Nick Mullens was not technically a rookie -- he was a 2017 UDFA -- but he did have pretty surprising results! However, it was a smaller sample size. Still, pretty good, and this stat probably won't quiet down the 49ers fans who believe that he deserves a real chance to start over Jimmy Garoppolo.
Two of the weirder results to come out of the 2017 season's plus-minus were the big years enjoyed by journeymen Case Keenum and Josh McCown. McCown did not qualify for last year's leaderboard, but was a -9.0 in just 98 attempts. Keenum regressed handily in Denver under a different game plan, and without Stefon Diggs and Adam Thielen. Meanwhile, given the same infrastructure Keenum had, Kirk Cousins stepped right in and finished second to Brees in plus-minus. That's a result that feels weird to the eye test, because the Vikings had games where they simply couldn't score: Monday night against Seattle and the game where they spotted Buffalo a 27-point lead to name two. Yet, Cousins completed 70.1 percent of his passes and the passing attack was generally pretty efficient when Cousins wasn't getting pressured (and fumbling afterwards). Good news for Jaguars fans who recruited offensive coordinator John DeFilippo and who will probably see "only a few games where the team can't score" as an improvement.
However, this year's weirdest-feeling plus-minus has to go Derek Carr. Carr's Raiders took 52 sacks -- those don't count in this stat -- and he was coming off a 2017 season where he finished -1.6 in 491 attempts. Even in Carr's breakout 2016 season, he was in negative territory. So what exactly did Jon Gruden do here? It was partly a huge spike in actual completion percentage -- up to 68.9 percent from a prior career high of 63.8 percent -- but almost all of Carr's plus-minus came on throws that went 10 yards or less past the line of scrimmage. And by "almost all," I mean 12.0 of 17.3. Who drove that? Tune in next week when we run receiving plus-minus!
Tom Brady falling to even on plus-minus surprisingly has a lot of precedent. Brady's 2015 season was actually in negative territory. The shape of the offense changed entirely with Brandin Cooks gone, and given Rob Gronkowski's retirement and the lack of much of anything settled in the passing game besides Julian Edelman, Brady might find himself right back in this territory next season. Then again, Brady has never done particularly well by plus-minus anyway. We have data going back to 2006, and Brady barely cracks the top 10 in plus-minus:
| Career Plus-Minus Leaders, 2006-2018 | ||
| Player | Team | +/- |
|---|---|---|
| Drew Brees | SD/NO | +410.5 |
| Philip Rivers | SD/LAC | +245.2 |
| Peyton Manning | IND/DEN | +211.0 |
| Aaron Rodgers | GB | +174.3 |
| Matt Ryan | ATL | +169.1 |
| Ben Roethlisberger | PIT | +144.5 |
| Russell Wilson | SEA | +129.6 |
| Tony Romo | DAL | +112.4 |
| Tom Brady | NE | +108.9 |
| Matt Schaub | HOU/ATL/Other | +105.6 |
That's right, folks! Brady is barely better in this statistic than the immortal Matt Schaub.
The opposite of Derek Carr's year is this year's big outlier: Ryan Fitzpatrick! +7.9 in 236 attempts, but he had +10.0 on his 100 attempts that went 10 or more yards past the line of scrimmage. Remove those, and Fitzpatrick was actually a below-average starter in the eyes of this statistic. That arm was built for chucking and letting Mike Evans go get it. Or, now in 2019 terms ... DeVante Parker? Yeah, well, uh, the Dolphins will probably try to start Josh Rosen ahead of Fitzpatrick if they can anyway.
Speaking of Fitzpatrick, let's finish this with one more table: the lowest of the low. The worst players by plus-minus since 2006:
| Career Plus-Minus Trailers, 2006-2018 | ||
| Player | Team | +/- |
|---|---|---|
| Jamarcus Russell | OAK | -53.9 |
| Drew Stanton | ARI/DET | -54.3 |
| Matthew Stafford | DET | -54.8 |
| Cam Newton | CAR | -56.7 |
| Ryan Fitzpatrick | MANY | -63.0 |
| Eli Manning | NYG | -68.7 |
| Blake Bortles | JAX | -90.6 |
| Mark Sanchez | NYJ/NFCE | -91.5 |
| Derek Anderson | CLE/ARI/CAR/BUF | -98.2 |
| Blaine Gabbert | JAX/SF/ARI/TEN | -102.5 |
There aren't many days where you get to use the sentence "and if Fitzpatrick can turn in one more good year, he can pass Cam Newton," but here we are. 2018 already got him past Eli Manning! It seems unlikely that anyone will catch Blaine Gabbert to be the worst quarterback in modern NFL history, but Blake Bortles definitely has a lot on his side for the chase: he's real bad, he's just off his rookie deal, and he's still tall enough to convince dumb teams that he could be a good quarterback. On the downside, he did sign with the Rams, and they actually have a good offensive philosophy.
Comments
20 comments, Last at 26 Jun 2019, 3:20pm
#2 by Megamanic // Jun 14, 2019 - 3:42am
Great article. Got to the career leaders, saw Rivers and Rothlisberger, thought "Where's Eli" then I scrolled down and you read my mind - below Jamarcus Russell seems a pretty brutal place to be though :)
How dependent on scheme is this? Are we looking at NO running a very +/- friendly offense or is Brees really just that good?
#3 by Joseph // Jun 14, 2019 - 12:05pm
Plus-minus is designed to give you credit for completing passes more often than every one else does. In other words, check-downs and screens don't really give you credit, but not completing them gives a big penalty. Hit that seam pass 20 yards downfield on 3rd & 15, right before the safety tries to kill the WR? Big credit.
Brees is really that good. His stats aren't what they once were (at least in volume)--but he constantly completes passes that probably shouldn't be thrown. With some guys (e.g., Favre, Mahomes), it's b/c their arm can get it there faster than most. With Brees, (Peyton & Brady too), it's b/c they can put the ball exactly where they want it, when they want it there. They are mentally ahead of everyone else on the field.
Obviously, it helps to have a guy like Michael Thomas that has hands made of glue. There's a reason that for a good part of the year, Brees to Thomas was connecting at a 90% clip--which would be an all-time record by a mile (at least for receivers).
#4 by Theo // Jun 14, 2019 - 12:43pm
It is nice to read that Brees is an accurate qb, but your explanation of this stat is pretty bad.
If you read it again, I am sure you would agree.
"Passing plus-minus is a stat we annually track to help provide context to completion percentage. Given the location of a quarterback's passes, it compares his completion percentage in each area to historical baselines. This stat does not consider passes listed as "Thrown Away," "Tipped at Line," or "Quarterback Hit in Motion" by Sports Info Solutions charting. How often a pass is completed based on the pass distance, the distance required for a first down, and whether the ball was thrown to the left, middle, or right side of the field. This is a counting stat, so more attempts are obviously a great thing for the purposes of what we're talking about here"
Last things first, its not a counting stat.
The rest. Im sure there is a system, but from reading this, there is no way of agreeing of falsifying it.
#5 by Joseph // Jun 14, 2019 - 5:37pm
In the full table, take the number of attempts, times the completion % +/-, and you get the passing +/-.
So, I suppose it is not a counting stat like pass attempts, TD's, etc.--but it is counting tiny percentages every time a QB threw the ball.
Maybe the explanation should read that the total is based upon the number of attempts--but DYAR is a "counting stat" in this same way. I am not a statistics major, so I don't know what the technical term would be.
#6 by LyleNM // Jun 14, 2019 - 6:03pm
A) It appears as though there is a typo above, it should presumably read that an average QB would complete 65.9% of the passes that Brees attempted, not 69.9%.
B) The 74.4% that Brees did complete minus the 65.9% the average QB would complete is shown above as C%+ (8.5%). That rate multiplied by the number of attempts gives the 39.6 number which is therefore no longer a rate but a counting stat.
#8 by nat // Jun 16, 2019 - 6:59am
How much of this is just a stadium-type detector?
It’s well known that completion percentages outdoors (especially northern) drop late in the season compared to indoors for most QBs.
In a 12 year schedule cycle, the Saints played 14 outdoor games in the final 4 weeks of the season, mostly in warm locales. A northern team in the AFC East (Patriots, Jets, Bills) played 45 or 46, mostly in colder stadiums.
Or, to put it another way, The Saints played indoors 34 times to avoid winter conditions, while the Patriots did so just 2 times. Even worse, the Browns did so just once.
You can still tell quite a bit with this stat. You might just have to take it with a truckload of road salt.
#10 by sbond101 // Jun 17, 2019 - 12:01pm
add to that there are definitely non-winter conditions that dramatically lower expected completion percentage that are unevenly distributed among outdoor stadiums (rain & high winds) and you end up back at the essential lesson of football statistics - It borders on impossible to produce large-scale comparable in football, but the quest is endlessly entertaining.
#11 by nat // Jun 17, 2019 - 1:16pm
True, that. Although the effects are most pronounced at a few northern stadiums, and mostly later in the season.
Denver's an interesting case. The thin air there helps all kickers, but only seems to help QBs who practice in it. Makes sense, when you think about it. Kicks go farther and straighter, always a good thing. But passes require accurate distance and speed, not just more of both. If you're used to the thin air, both can be easier, as well as directional accuracy. But if you're not used to it, throws can come out of your hand too fast and go too far. Or you might overcompensate, and throw it into the dirt instead.
One thing is certain: QBs throwing indoors have the absolute best conditions. It's no surprise that they feature prominently in these tables.
Brees is great, as his receivers have been. He'd be in near the top of these lists anyway if the Saints played in Buffalo. Just not so often.
#17 by nat // Jun 18, 2019 - 8:52pm
Retractable roofs are either closed to give perfect passing conditions or open in mild conditions. They’re half way between domes and mild climate stadiums for the purposes of this discussion. They certainly give a similar advantage for completion %. Not exactly the same advantage. But it’s got to be close.
#19 by nat // Jun 21, 2019 - 6:34am
While it is technically northern, Seattle is neither cold nor particularly windy.
Still, Wilson was the only one of this season’s top five to not play most of his games in a roofed stadium.
I don’t think we need to come up with a reason, other than he and his receivers played well.
Was his 5.1% better than expected completion % mostly outdoors more impressive than Brees’ 8.5% mostly indoors? Possibly. The dome effect, although quite real, is hard to assign an exact adjustment value for. I’d put Wilson ahead of Ryan, for sure.
#20 by Scholder99 // Jun 26, 2019 - 3:20pm
How come the world of sports nowadays has data that when viewed within the context of what we actually know to be legitimate and see in the actual field....in other words the game played by human beings in a context of each week, each game, each play, each set of unique circumstances with injuries and momentum etc... No one stops to consider the relevance of data when it so clearly is not even slightly in line with what we know is true of the quality of these players? Can anyone offer a slightly reasonable explanation?