August really only means one thing ...
It's officially prediction season!
Well, actually, I am kind of a stickler for calling the output of my models projections rather than predictions. The distinction, to me, is that I've tried to eliminate as many biases as possible from my mathematical framework, whereas calling something a prediction makes it sound like speculative judgments were involved on some level. The names and numbers you see below are the players who rank first in each category based on their projected performance. There are also some extra informational nuggets on interesting results regarding players who, though they don't necessarily rank first, figure to produce an output that is interesting for fantasy or team (win total) purposes.
As we get closer to the regular season, I'll update my models, and these numbers might change. But let me know if you agree or disagree with today's projections -- and let me know if you're predicting this, or if you've created a model to project your own stats leaders! You can hit me up on Twitter: @cfrelund.
Passing yards: 4,790
The COVID-19 pandemic eliminated in-person offseason activities and continues to present unprecedented challenges as we move toward kickoff. Consequently, we're facing a whole bunch of new factors on the projections front. One of the changes I programmed into my models for this season: to value personnel continuity (prioritized by position and snap count) even more than in prior years. We always see chemistry and playbook learning curves -- my logic is that these will be even more impactful without a normal offseason and preseason (note: the curves do vary by position). The Chiefs have the most returning high-value players by snap count, most notably at the highest-value position of quarterback and among the QB's top passing-touchdown earners. My model projects Mahomes to earn more than 4,500 passing yards in a whopping 67 percent of simulations (think of anything over 60 percent as high). Mahomes ends up atop the league in 18.5 percent of models.
Bonus intel: Rounding out the top five, Tampa Bay's Tom Brady ends up on top in 16.5 percent of simulations (with more than 4,200 passing yards in 63 percent of simulations), and Dallas' Dak Prescott takes the cake in 16 percent. New Orleans' Drew Brees is fourth at 15.5 percent, while Atlanta's Matt Ryan ranks fifth at 12 percent.
Passing touchdowns: 37
Mahomes earns more than 35 passing touchdowns in 57.5 percent of 16-game simulations. Forecasting the 2018 NFL MVP and reigning Super Bowl MVP to lead the league in passing touchdowns is admittedly quite unsurprising, so let's unload the touchdown algorithm here for three other QBs: Seattle's Russell Wilson throws for more than 30 touchdowns in 55.5 percent of simulations, Buffalo's Josh Allen tops 22 in 57 percent of simulations and Tennessee's Ryan Tannehill matches or exceed his 22-touchdown total from last season in 54.9 percent of simulations. One player whose projection surprised me? Bengals No. 1 overall pick Joe Burrow. His median projection is to throw 21 touchdowns. (I use all median projections to order players in fantasy rankings, but there is always a ceiling and a floor.) Last season, Giants rookie Daniel Jones threw 24 touchdown passes. In fact, 14 quarterbacks threw 24-plus touchdown passes. When I look to see how many times in the model Burrow reaches or exceed 24 touchdowns, it happens in 53.6 percent of the simulations. Interesting.
Rushing yards: 1,380
Last season, Henry's 1,540 rushing yards completely unlocked the Titans' success on offense. Expect to see more of the same in 2020, with Arthur Smith's offensive architecture maximizing the strengths of his weapons. Henry earns more than 1,300 yards in a ridiculous 66 percent of simulations and more than 1,500 in 51.6 percent. Cowboys RB Ezekiel Elliott checks in with the second-most rushing yards (and the second-best chance to top the league) with a median projection of 1,290. Don't overlook Seattle's Chris Carson when you're thinking about potential top rushers, as he earned the second-most rushing first downs last season (75) and projects for 1,120 rushing yards in my model. He earns more than 1,100 rushing yards in 55.8 percent of simulations.
My model rarely projects over 130 catches. However, Thomas just set the single-season record in 2019 with 149 grabs. And given New Orleans' offseason improvement to the overall pass-catching situation (namely, the addition of Emmanuel Sanders) and the fact that future Hall of Famer Drew Brees remains at quarterback, Thomas enters rarefied projections air. Thomas earns more 1,500 yards in 51.8 percent of simulations and catches more than 126 passes in 53.8. Again, not very surprising, so I found three other things that stand out from the reception results in my model ...
Fun notes: First, when it comes to Tampa Bay's dynamic duo, Chris Godwin's 91 projected catches are more than Mike Evans' 82. Christian McCaffrey's 116 receptions ranked second in the NFL last season; this preseason, the Panthers running back ranks No. 6 in my model with only 92 projected receptions (sarcasm intended). Green Bay's Davante Adams ranks No. 2 with a forecast of 109 grabs, while Julio Jones earns at least the 99 he had last season in 58.3 percent of results.
Receiving yards: 1,510
The return of the deep ball in Atlanta?! Last season, Matt Ryan only attempted a deep pass (20-plus air yards, per Next Gen Stats) on 8.9 percent of his throws -- the quarterback's lowest rate since 2016. And in the 2019 campaign, despite the fact that Atlanta's intended starting O-line played fewer than 10 percent of snaps together, Jones still managed to earn the second-most receiving first downs in the league (77). Given better health on the offensive line, the offseason addition of tight end Hayden Hurst and the intense, high-flying competition we all expect from the NFC South, Jones slightly edges out Michael Thomas in projected receiving yards. The Falcons star earns more than 1,500 receiving yards in 55.1 percent of simulations.
One other interesting note: Bears receiver Allen Robinson earns at least 1,100 yards in 54.9 percent of simulations.
Total touches: 391
Next Gen Stats show that McCaffrey played 93.4 percent of the Panthers' offensive snaps in 2019, which was the highest among running backs. (McCaffrey was the only RB to log more than 1,000 snaps.) This offseason, Carolina brought in new offensive personnel (most notably, veteran QB Teddy Bridgewater and deep-threat WR Robby Anderson), a new head coach (Matt Rhule) and a new offensive coordinator (Joe Brady). Taking all of that into account -- and factoring in Brady's past play-calling tendencies -- my models don't just project McCaffrey to carry an immense load once again, but he could become even more efficient in 2020.
Fantasy advice: When I dig into my model's touch notes, two backs stand out as players you shouldn't overlook in fantasy drafts: Kenyan Drake and Le'Veon Bell. By now, fantasy fiends have seen plenty of analysis on the top dogs at the position (CMC, Zeke, Henry, Saquon Barkley, Alvin Kamara and Dalvin Cook), but Drake's projected uptick in opportunity is dramatic and makes him my RB8 in PPR leagues. The Cardinals back logs double-digit touchdowns in 55.4 percent of simulations. Meanwhile, Bell is my RB9 in large part due to a lot of questions about the Jets' receiving corps -- and the revamped O-line could help him, too.
Total touchdowns: 13
Kamara narrowly edges out Ezekiel Elliott, 13.1 TDs to 13.0. Kamara earns 12 or more touchdowns in 59.4 percent of simulations, while Elliott reaches or exceeds that in 58.8 percent. Aaron Jones, who tied with Christian McCaffrey for the league lead with 19 total touchdowns in 2019, ranks fifth in my projections with 12.8 TDs, but he has a lot of upside. Jones earns 12 or more touchdowns in 56.6 percent of simulations. Lastly, the Chargers' lead back, Austin Ekeler, earns 10-plus touchdowns in 58.2 percent of simulations.
My model forecasts a lot of passing plays in the NFC South, which helps drive the opportunity for Jordan, who racked up the third-highest sack total in the league in 2019 (15.5). Jordan edges out Danielle Hunter, the next-highest projection, with Joey Bosa, Chandler Jones and Nick Bosa rounding out the top five. (Aaron Donald and Myles Garrett both checked in just outside of the top five.)
The best situations in which to earn interceptions are when defenses face passes in unfavorable down-and-distance plays for the offense. The Ravens' addition of Calais Campbell to the defensive front helps drive more consistent pressure, which is also correlated with greater chances of earning an interception. The other factor working in Peters' favor is his aggressive playing style. And it sure doesn't hurt that he's lining up on the opposite side of the field as Marlon Humphrey, who, on an almost-weekly basis, ranks among my model's top three corners. With the CB duo of Peters and Humphrey, Baltimore boasts the highest pick projection as a total defense, as well.
BONUS ROOKIE LEADERS
Clyde Edwards-Helaire, RB, Kansas City Chiefs: Damien Williams opting out of the season has driven a HUGE surge in fantasy value for CEH. Looking at ADP (average draft position), Edwards-Helaire is currently trading at around RB5 in PPR leagues. While I am fully on board with a ton of value coming from the Chiefs' first-round pick, let's temper expectations a little. My model ranks him as RB10. Not top five, but still a high ranking, especially for a rookie. One way to potentially earn a fantasy edge here: looking at a few other rookies who have strong TD projections and snagging them in the later rounds of your draft. My model likes Colts WR Michael Pittman Jr. to earn five or more touchdowns (occurs in 53.9 percent of simulations). Raiders WR Henry Ruggs III hits paydirt five-plus times in 57.1 percent of simulations.
Chase Young, DE, Washington Football Team: One bright spot for the Washington Football Club: a power-loaded defensive front, with the No. 2 overall pick joining the fold. Young, whom my model projected as the most valuable player in Year 1 prior to the 2020 NFL Draft, earns six or more sacks in 56.9 percent of simulations.