Inspired by the idea of wins above replacement in baseball and real plus-minus in basketball, I created a metric that measures the contribution of each player, position group and side of the ball for every snap in an NFL season. I'm calling it win share to reflect the fact that 11 players are working together on the field for each team at any given time, meaning that the "share" part is a really big factor -- and the interconnectedness took a lot of time to model out.
The metric is built around the number of times a player impacts first downs and touchdowns that lead to wins or losses, either by creating such first downs and touchdowns on offense or by preventing them on defense. On-ball impact (like when a wide receiver is targeted directly) and off-ball impact (like when a wide receiver draws coverage away from other players on the field, resulting in a better matchup for a different pass catcher) are both measured. Win share adds up all of the positive on- and off-ball plays and subtracts the negative ones in order to help attribute a value to each player and phase of the game, as represented by the number of wins each player is responsible for on an individual level in a given season. Personnel on the field, game situations and matchups are all factored in, to the degree that it's possible.
Now the important part: Who will lead the NFL in win share in 2021? Below, you'll find the top five projected win share leaders among quarterbacks, non-quarterback offensive players and defensive players, listed along with their projected win shares for the 2021 NFL season. I have added some notes about units, too. Let me know what you think! You can reach me on Twitter @cfrelund.
Mahomes ranked in the top five in passing touchdowns (35) and passer rating (116.1) when he wasn't under pressure in 2020, according to Next Gen Stats (min. 100 pass attempts). A big driver in his league-leading win share number is that he likely won't be under pressure as often this season, given his retooled offensive line. And maybe teams should think twice about blitzing him. Mahomes has been blitzed on just 20.7 percent of dropbacks since 2018, the lowest rate in the NFL, but he averaged 9.1 yards per attempt when facing the blitz over that same time period, which was the highest average in the NFL. Last season, he earned 13 touchdowns with zero interceptions and a 134.2 passer rating against the blitz, leading the league in those categories.
Computer vision shows that the Chiefs' offense spreads the field vertically and horizontally more than any team in first halves of games, which helps explain why Mahomes' targets averaged 3.8 yards of separation last season, the second-most in the league. However, in the red zone, when space was constrained, his targets averaged 2.9 yards of separation, which ranked 15th. With that in mind, look for the Chiefs to try to create space for their weapons even more frequently to establish the game tempo early on.
Another key takeaway here is that Mahomes has his eye on the quick strike and he knows how to put the ball in the end zone in a hurry. He threw 12 TDs on deep passes last season (tied for second-most in the league), six on deep play-action passes (most) and 13 on the run (most). One last thing that's a bit more predictive: The Chiefs used play-action on 18.6 percent of dropbacks in Weeks 1-6 (ranking 26th) last season, compared with 33.3 percent in Weeks 7-17 (second-highest). My models show that the improved O-line and its run-blocking potential creates the opportunity for increased rushing (or short passes that essentially function as a run) productivity, which will make play-action an even stronger tool in Mahomes' tool belt.
The reigning MVP is set to remain at the top of his game in 2021. Rodgers threw 20 touchdown passes off play-action last season, which is seven more than this total from 2016 to 2019 combined. No QB had a higher passer rating on play-action passes last season than his 136.0. Rodgers attempted a deep pass on 14.1 percent of attempts, the second-highest rate in the league, and earned 12 touchdowns and zero interceptions on such passes, the best TD-to-INT ratio in the NFL. He netted a 128.7 passer rating on deep passes, which was the third-best mark in the league.
When under pressure, Rodgers earned five touchdowns and zero interceptions with a 92.6 passer rating (fourth-highest passer rating). When not under pressure, he led the league with 43 touchdowns and a 126.1 passer rating. As for the big disparity between the totals, Rodgers was only under pressure on 16.8 percent of dropbacks, the third-lowest rate in the NFL. This number forecasts to stay very low -- or even improve -- this season with most of the starting O-line remaining intact and the potential for increased play-calling utilizing multiple backs.
The other reason for the optimism, despite the loss of Corey Linsley in free agency and absence of David Bakhtiari (returning from ACL tear)? No QB has helped his O-line more than Rodgers over the past five seasons. I've been refining a new metric about the impact the QB has on his O-line. It's not perfect yet, so I'll wait to give you the magnitude in terms of win share, but he's the best by a good margin when it comes to making the front five look good.
Remember when there was a debate about whether Brady could still get the job done throwing deep? After some early struggles in 2020, he put that one to rest. His 9.3 air yards per attempt paced the league in 2020, as did his 89 deep passes (20 or more air yards). Brady posted his highest number of completions (31), yards (1,099) and touchdowns (9) on deep passes since NGS began tracking them in 2016. He improved throwing deep as the season went on, going 17 of 30 for 594 yards, seven touchdowns and one interception in Weeks 12-17. With a level of continuity on the coaching staff and roster that is rarely seen in the modern NFL, the forecast for Brady remains strong in 2021. Play-action could be an area of increased use and efficiency this season. In 2020, Brady used play-action on 18.4 percent of dropbacks, his lowest use rate in five seasons, but he earned a 127.0 passer rating on such attempts (second-best in the league) and netted 10.4 yards per attempt (most in the NFL).
You know the metric I'm developing that captures a QB's ability to help the offensive line? Like Aaron Rodgers, Wilson ranks highly -- as in, within the top three -- in this area over the past five seasons. Wilson was pressured on just 24.5 percent of dropbacks, his lowest rate in the past five seasons, per Next Gen Stats, but when he was under pressure, he thrived, notching the best under-pressure TD-to-INT ratio (8:0) and second-highest passer rating (94.7) in the NFL. All 13 of his interceptions came when he wasn't under pressure (that's the highest number of non-pressured picks in the league), likely because defenses that did not bring pressure were able to cover Wilson's pass-catchers more tightly. Wilson paced the league in total deep touchdowns (13) and deep passing TDs while on the run (4), but his success in that facet of the game was front-loaded. In Weeks 1-5, Wilson was blazing hot, with eight deep TDs (most in the NFL) against zero picks, but in Weeks 7-17, he cooled off, notching five TDs against four picks (tied for most in that span) on deep throws. He also completed the same number of deep passes in his first five games last year as he did in Weeks 7-17. Looking to the upcoming season, Wilson's deep passing forecasts to look more like it did early in 2020.
Here is my favorite Allen fact: Not only was he the second-best off-platform thrower (that is, when his lower body and hips weren't in an ideal alignment) in the NFL in 2020, but he improved the most in that area from 2019 to '20. NGS shows that Allen evaded pressure on 26.1 percent of pressures last season, the highest rate in the league. While on the run, he completed 53.2 percent of throws, with eight TDs (tied for third-most) against zero picks and eight deep completions (most in the NFL). On throws of 10-plus air yards, Allen notched 20 TDs (tied for fifth-most) and nine interceptions, with a completion percentage of 56.2 -- +8.2 percent over expectation from that range. His improvement in completion percentage over expectation on deep passes was +17.2 percent, the second-largest gain from 2019 to 2020. On quick passes, Allen averaged 6.8 yards per attempt, with a passer rating of 127.0 (seventh-highest), a completion percentage of 84.7 (best in the NFL), 12 TDs and only one pick. No team passed the ball on first down more than the Bills did last season, and Allen's success on quick passes helped create more first downs and touchdowns. As for pressure, Allen logged the best air-yards-per-attempt average (13.2 yards per throw) while under pressure in the league. So far, Allen has had five or more touchdown passes under pressure in each season of his career. Part of the forecasted success of Allen is due to continuity among his teammates and also his offensive coordinator, Brian Daboll.
NON-QB OFFENSIVE PLAYERS
McCaffrey played a whopping 1,004 offensive snaps in 2019, his last healthy season. He was the only RB to surpass 900 snaps, and he played on 93.4 percent of Carolina's offensive snaps. While it's unlikely he'll be asked to shoulder that kind of load in 2021, it is worth noting that McCaffrey's fatigue (erosion in speed and response time from the first quarter to the fourth quarter) was the lowest for any player in the NFL in 2019. After injuries limited him to three games in 2020, it stands to reason that he falls into a higher probability category to return to top form based on a résumé of maintaining exceptional physical conditioning. Keep in mind that McCaffrey earned the most touchdowns on rushes inside the tackles (10) and gained 5.0 yards per rush (second-most, min. 75 carries inside the tackles) on such runs in 2019, per NGS. He also gained 218 rush yards over expected that season, fourth-most in the league. The ability for head coach Matt Rhule and OC Joe Brady to leverage McCaffrey in both the run and pass game drives his exceptional value. Carolina's efforts to fortify the offensive line this offseason, combined with strong value at the wide receiver position, helps forecast ample space for McCaffrey to work with in 2021.
Henry's 412 rushing yards over expected in 2020 were the most in a single season since NGS started tracking the metric in 2018. This season, between the addition of Julio Jones and the wrinkles new OC Todd Downing could add to the scheme, Henry is likely to face fewer stacked boxes, which is a good thing. However, NGS shows that stacked boxes (eight or more defenders) weren't really a problem for Henry. He averaged 5.9 yards per rush (third-most among RBs with at least 60 carries vs. loaded boxes) and ran for 10 touchdowns (tied for league lead) against them in 2020. Henry also forced the most missed tackles last season (75, per PFF) and his 1,490 yards after contact were the most by any player in a season since PFF began tracking the stat in 2006. As you might expect, NGS shows that rushing inside the tackles is a hallmark of Henry's game. He rushed for 10 TDs (most in the NFL) and 6.3 yards per carry (third most) when running inside the tackles last season, which is part of what drives a strong forecast of continued success for the Titans star.
NGS shows Kamara netted a league-best 6.6 yards per rush inside the tackles (min. 25 such rushes) and 12 rushing touchdowns outside the tackles (most in the NFL) last season. He also reached the end zone five times against stacked boxes in that epic Week 16 performance against Minnesota, the most such TDs in a single game in the NGS era (since 2016). As for his receiving prowess, Kamara was targeted on 33.1 percent of routes run (second-highest; min. 65 targets) and averaged 2.3 yards per route (second among RBs) in 2020. Kamara ran 88 screens, which is almost twice as many as the next-closest player (49 for James Conner). PFF also shows he forced 20 missed tackles on receptions (tied for third most). There will be a new QB1 in New Orleans this season, with Jameis Winston expected to succeed the retired Drew Brees, but Kamara's versatility and efficiency make him a foundational piece with even more forecasted value than he provided last year.
Adams led the league with 3.1 receiving yards per route last season (min. 150 routes), per NGS. He aligned in the slot on 29.9 percent of his snaps -- his highest slot rate in the last five seasons -- and it paid off, as he caught 83.6 percent of slot targets (second-highest rate in the NFL). He also earned seven of his 18 touchdowns out of the slot (third-most). On outbreaking routes from the slot, he caught 23 of 24 targets for 160 yards and five touchdowns, which was both the most receptions and touchdowns on such routes. Adams' versatility, as evidenced by his big uptick in slot alignment efficiency this past season, forecasts to help him match or even exceed his 2020 output with the Packers' continuity on offense only helping matters.
Cook has rushed for 19 TDs outside the tackles since 2019, the most in the NFL over that span, and his +252 rush yards over expected in 2020 ranked third among RBs, per NGS. In fact, Cook and Derrick Henry are the only players to rush for 10 or more touchdowns inside and outside the tackles since 2019. He and Henry were also the only backs to gain more than 1,000 rush yards after contact last season, according to PFF. Computer vision shows that Cook's ability to reach his top speed after contact is third-fastest among backs since 2019, which helps solidify his spot on this list.
No player has a greater delta between his win share and the next-closest player at his position than Donald (0.36 wins). NGS shows that his 71 QB pressures in 2020 tied for league lead with T.J. Watt, and my computer vision shows Donald was double- or triple-teamed at a higher rate than Watt. NGS clocks Donald at a 0.86-second pass rush get-off time, the fastest among interior defensive linemen. Donald has finished in the top three in total QB pressures each season since 2016, and computer vision shows he has been double- or triple-teamed at a rate about five percentage points higher than the next-closest player.
In Weeks 11-17 last season, Young netted 26 pressures (tied for third-most). With a full NFL offseason under his belt, his burst speed (measured as the speed he reaches in the first 3 yards traveled from the line of scrimmage) should be even more of a problem for opposing O-linemen. That has to be a scary thought for offensive coordinators, as he clocked the third-fastest burst speed from Weeks 11-17. He caused four turnovers from pressures as a rookie (tied for fourth-most), and we could see his production improve in that area, too, given his increased comfort with NFL speed.
Garrett posted 56 QB pressures last season (tied for fourth-most in the NFL), and he caused turnovers on a league-high seven of those plays, which doubled his career total to 14 since 2018 (most in the NFL). Garrett already has two games with three turnovers caused by pressure in his career, which helps illustrate one of my favorite computer vision metrics for him: rush style win. OK, I need a better name for it, but here's what it represents: Some edge rushers win with a bull rush but aren't as effective with a pass-rush move like a spin. So I measure hips, arms and shoulders to approximate where there is movement. Then I look to see what happens in sequence (what percentage of each type of rush, along with when and what happens during each rush) and measure the O-linemen, too, to add context. Garrett often bull rushes early. In fact, on early downs with at least 5 yards to go, he bull rushed at the second-highest rate at the position last season. Then on third down, and especially third-and-long, he makes it look a lot like a bull rush (same alignment) but instead uses something like a spin or swim move (anything where his body changes direction). This throws O-linemen off balance and leads to QB pressures that are more likely to be disruptive. I asked an NFL head coach about this, and why other people don't do it as often, and he said not everyone has the ability to control their body to that degree.
The Colts allowed 3.1 yards per carry with Buckner on the field versus 4.9 without him on the field in 2020, per NGS. In the passing game, his 10.0 quarterback pressure percentage was seventh-highest among interior defensive linemen. Buckner's value increases even more with another season of work with DC Matt Eberflus. He could easily be a top-three 3-technique.
In 2020, Alexander averaged 7.5 yards of pre-snap cushion per coverage snap (third-most in the NFL); however, he closed in on his targets and only allowed 1.8 yards of separation per target when in off coverage (when there are 5-plus yards of cushion). PFF counts him as allowing only a 67.0 passer rating (fifth-lowest) and 50 percent completion percentage (tied for second-lowest) in coverage.