(Feel free to skip)
In May, I released an article titled “Upside Wins Championships” which introduced a draft-strategy describing essentially that: when it comes to winning fantasy championships, UPSIDE IS EVERYTHING. A player’s bull-case projection matters much more than his base-case projection which matters far more than his bear-case projection. Leagues are won and lost, not by balanced teams drafting a number of good-to-great ADP-beaters, but by teams who correctly identified a few key league-winners and rode those players all the way to a Championship title.
In practice, an “Upside is Everything” strategy is basically this: after about Round 4 of your draft (give or take based on personal preference), upside should be the determining factor behind every pick you make.
If not quite so revelatory a concept, at least not as much as I might have hyped it up to be, it’s still one of the most important concepts to grasp. Understand this one point, and you’ll have a massive edge over your opponents.
This article was met with a mostly positive reaction on social media, and even – at least to spite the haters – spurred an influx of “5 players with league-winning upside”-type articles. Still, I wanted to address the most common criticisms, which were:
1) “Yeah, right! Drafting a league-winner is so easy? You wish, buddy!”
2) “I wish you spent more time detailing what league-winning upside looks like, which players would have qualified as league-winners in recent years, and how those players could have been identified ex-ante.”
Okay, fair. Well, you asked for it… and you got it. In tomorrow’s, I’ll respond to the second point directly, with an incredibly in-depth article titled“Anatomy of a League-Winner.”
In today’s, I’ll respond to the first point.
I never said drafting a league-winner was easy, I merely stated that that’s the correct object of the game. That drafting the right league-winning players was crucial to, well, winning your league. A team that correctly identified Christian McCaffrey and Lamar Jackson as league-winners last year and whiffed on most other positions was still significantly better off than a team that drafted a number of players that merely successfully beat their ADP.
In today’s article, we’ll be defining league-winners by their rate of occurrence on ESPN playoff teams and ESPN championship-winning teams. Our data only goes back to 2017 and will be unique to ESPN’s specific settings, but it’s also still the best and most robust sample we’ve got.
And, again, correctly identifying those players ex-ante is not easy, but it’s also not impossible. To prove my point, we’ll be looking at the top-five league-winners over the past three seasons (both on playoff teams and championship-winning teams) and see if we couldn’t have correctly identified those breakout players before the fact. In each instance, we’ll only be using my pre-season (ex-ante) analysis on each player in each section below, and then looking for some overall takeaway that might be applicable to the 2020 season.
How difficult is it to spot a league-winner? Certainly not easy, but also not impossible. That’s clearly evident by my preseason analysis on each player below. Burying the lede: 8 out of 10 ain’t bad at all
Ex Ante Player Analysis – Playoff Rosters
1. Christian McCaffrey , 2019
McCaffrey was a revelation in 2018, breaking the single-season record for receptions, averaging 5.01 yards per carry, and leading the position in snaps (by 76). Remarkably, there’s a good chance he could be even better in 2019. McCaffrey suffered a shoulder injury in Week 12 (which ultimately required surgery) and played through it. Volume should be just as good this year, as head coach Ron Rivera is projecting a similar touch-count for McCaffrey. He’s well worth a top-three pick in 2019 drafts, and with just a little more passing-down volume or efficiency, could post a 1,000/1,000 season.
July 2019, ADP: RB2
Okay, this one was easy. But still, McCaffrey did record a 1,000/1,000 season, becoming only the third player ever to accomplish this feat.
The bigger takeaway was just how ridiculous McCaffrey’s season was. Just by rostering this one player – who, mind you, you probably drafted with a top-three overall pick – you were twice as likely to make the playoffs. Looking ahead to 2020 and beyond, if McCaffrey really is the second-coming of Ladanian Tomlinson, then it might be time to switch over to an auction-style draft. Otherwise, the first-overall pick really might be just too unfair of an advantage.
2. Alvin Kamara , 2017
During the Saints Day 2 press conference, GM Mickey Loomis said “[We're excited] to get a player we coveted… that I would expect to fill the role that Reggie Bush and Darren Sproles had for us in the past.” Loomis didn’t need to say it, it was clear they coveted him when they traded a second-round pick to the 49ers to move up and grab him in the third. By adding a running back on Day 2, it appears an already complicated New Orleans backfield only gets more confusing. However, if we can take Loomis at his word, Kamara’s role might not be.
The Bush/Sproles role Loomis had alluded to was a very attractive one for PPR fantasy leagues. Bush averaged 16.7, 17.4, and 17.2 PPR fantasy points per game during his first three seasons in New Orleans. Darren Sproles averaged 17.0, 16.5, and 11.8 PPR fantasy points per game during his first three seasons in New Orleans. Kamara never saw more than 110 carries or 50 targets during his college career, and likely won't see the same rushing workload Bush saw early in his career, but if he were to match their 7.3 target per game average, he could make his mark as a perennial RB2 in PPR formats.
May 2017, ADP: UDFA
In 2017, Kamara averaged 20.0 fantasy points per game, which was the fifth-most by any rookie running back over the past 35 seasons and the fourth-most by any running back that year. At a price-tag of “free,” when every running back above him was drafted in the first two rounds, it’s not hard to see how he swung leagues.
The analysis was right, we said he’d draw a much more valuable role than fantasy drafters were envisioning, and he did. Darren Sproles and Reggie Bush averaged 101.0 carries and 99.3 targets per season in their first three years with Sean Payton. Off by 19 carries and 0.7 targets, Kamara totaled 120 carries and 100 targets in his rookie season.
The one thing we missed on was, oh boy… Kamara is awesome. In 2017, Kamara averaged 1.59 fantasy points per touch, which is the most by any 100-plus-carry back ever. And it wasn’t a fluke either! In 2018, Kamara averaged 1.29 fantasy points per touch, which is the 13th-most all-time.
3. James Conner , 2018
Le’Veon Bell is currently in the middle of a holdout and has missed 15 games over the past three seasons. Bell is still well worth a top pick in fantasy drafts because few other players can come close to his expected volume. What many are forgetting (at least given Conner’s ADP), is that when Bell misses time, his backup typically commands a near-equal workload. In 2015-2016, in games started and finished, DeAngelo Williams saw a nearly identical workload to Bell and averaged 22.9 fantasy points per game. For perspective, 22.9 fantasy points per game would have ranked third-best at the position last year. Conner is my favorite handcuff but has standalone value as well.
July 2018, ADP: UDFA
I mean, Le’Veon Bell’s decision to hold out for a full season was extremely unpredictable, and especially because it’s still not clear to me that that decision didn’t cost him millions of dollars in the long run. And any partial holdout would have capped James Conner’s upside, as he’d be worthless during your most important weeks (the fantasy playoffs).
But still, the broader point was absolutely correct: Mike Tomlin typically prefers a bell cow approach at the running back position. For running backs, fantasy points are driven far more by volume than efficiency. And therefore it wasn’t hard to imagine “Pittsburgh RB1” being one of the most valuable players in fantasy regardless of who that player ultimately was.
And yes, it still is – we’re probably sleeping on James Conner this year too (ADP: RB22).
4. Dalvin Cook , 2019
Dalvin Cook has just two touches this preseason. On one of those touches, he scored an 85-yard touchdown. If Cook sees similar usage to what he saw last year, he has legitimate top-five upside in Gary Kubiak’s offense. In 12 seasons as an offensive coordinator, Kubiak’s offense has ranked top-five in rushing yards nine times and top-12 12 times. In his last year in the league, Kubiak led the league in outside zone runs and outside zone run percentage. Over the past two seasons, Cook ranks fourth-best in yards per carry when running in outside zone.
August 2019, ADP: RB9
Despite battling injuries last season, Cook averaged 0.27 missed tackles forced per touch, which led the league and ranked fifth-best by any running back this past decade. From Week 11 (his first week off the injury report) until the end of the regular season, he played on 76% of the team’s snaps while averaging 15.8 fantasy points per game…
Cook has significantly less injury risk this year. His injuries last year were related to ACL surgery, but running backs typically make a big improvement the year following ACL surgery.
September 2019, ADP: RB9
Right again. Cook was a massive injury discount – running backs typically suffer in efficiency and with compensatory injuries immediately following their return from an ACL injury, but are essentially back to full health the following season. And on top of that, clearly, drafters were sleeping on Cook’s talent (historic missed tackles forced per touch numbers despite playing hurt) and situation (Gary Kubiak).
5. Lamar Jackson , 2019
Among all Konami Code quarterbacks to ever enter the NFL, Jackson might be “the chosen one.” Throughout his college career, Jackson averaged 108.7 rushing yards per game, or, the most by any Power-5 quarterback since (at least) 2000. In 2018, he totaled 147 rushing attempts, the most by any quarterback in any season in NFL history. And keep in mind he wasn't named the team's starter until Week 11. In games started, he averaged 17.1 rushing attempts per game, which would have amounted to 274.2 rushing attempts over a full season — or, the most ever by 133.
Jackson has been (rightfully) criticized for poor passing efficiency in his rookie year, and also had to contend with poor overall volume. He ranked last among all quarterbacks in dropbacks per start (27.0), while Baltimore ran the ball at a historically high rate. Last season, once Jackson was named the starter, Baltimore passed on only 36% of its plays, while the league-average rate was 58%. Still, I’m optimistic Jackson improves on both counts.
Last season, Jackson effectively played in Joe Flacco’s offense, only earning playing time after Flacco suffered a hip injury. Now, Baltimore will have a full year to build its offense around Jackson. They’ve already done this via the draft, adding speedy playmakers in Marquise Brown, Miles Boykin, and Justice Hill, to better complement Jackson's style of play. In terms of play-calling, Greg Roman has been announced as the team’s new offensive coordinator. Fellow mobile quarterbacks Colin Kaepernick and Tyrod Taylor both had their best fantasy seasons under Roman.
Though Jackson struggled as a passer last season, he was far more efficient in college, finishing his career with a depth-adjusted completion percentage that was 3.2% over his expectation (same as Patrick Mahomes).
Perhaps the best metric to prove my ultimate point — that Jackson is a screaming value at ADP (QB21) with legitimate league-winning potential in 2019 — is this: Jackson averaged 0.83 fantasy points per dropback in 2018. Not only is that the best mark this past decade, but it’s the best by 13%, and 21% better than Mahomes’ number last season. From Week 11 (once Jackson was named the starter) until the end of the regular season, Jackson ranked eighth in fantasy points scored. The way I see it, that might be closer to his absolute floor in 2019.
June 2019, ADP: QB21
Jackson is a screaming value with legitimate league-winning potential? Bingo.
Maybe you’ve noticed a running theme here, but fantasy drafters were underrating Greg Roman’s impact on Lamar Jackson, just like drafters were underrating the impacts of Gary Kubiak (for Cook), Mike Tomlin (for Conner), and Sean Payton (for Kamara).
On top of that, drafters were severely underrating the importance of the Konami Code, Baltimore’s other offseason moves, and the fact that quarterbacks tend to make a significant efficiency jump in their sophomore seasons (e.g. Carson Wentz, Jared Goff, Andrew Luck, etc.)
Ex Ante Player Analysis – Championship Rosters
1. Christian McCaffrey , 2018
Last season McCaffrey totaled 175.1 receiving fantasy points, the 10th-most by any running back this past decade and what would have been the 29th-most among all wide receivers last year. As impressive as that was, he underwhelmed as a runner. His 3.7 yards-per-carry average — while still better than the efficiency of Le’Veon Bell (3.5) and LaDainian Tomlinson (3.6) in their rookie years, and though his yards-per-carry average looked much better relative to blocking advantage and to the efficiency of Jonathan Stewart (3.4) — was at least a disappointment relative to expectations coming out of college.
McCaffrey was one of PFF’s six highest-graded runners in each of his last two seasons in college and totaled 6.1 yards per carry over that stretch. He was a do-it-all bell cow in college, and Carolina coaches appear set on using McCaffrey in a similar capacity this season. Head coach Ron Rivera said in June he sees no reason why McCaffrey can't reach 200 carries in 2018. In July, Rivera said getting McCaffrey the ball 25-30 times per game would “be ideal.” In August, OC Norv Turner agreed that number was realistic. For perspective, McCaffrey totaled just 117 carries and averaged only 12.3 touches per game last season.
As it stands, McCaffrey is priced at just about his floor – RB12, after finishing RB13 in fantasy points per game last year. Given the likelihood of a larger workload and better efficiency, he’s a glaring value with league-winning upside.
August 2018, ADP: RB12
He’s a glaring value with league-winning upside? Nailed it again.
This call was easy. McCaffrey was elite in college, and both of his coaches had said multiple times across multiple months that they were intent on giving him a generationally-good workload. Remember, rushing is nowhere near as important as receiving for fantasy running backs, and also, it’s not uncommon for a rookie running back to struggle with efficiency before breaking out in their sophomore year. Actually, it’s pretty common.
David Montgomery is by no means the next Christian McCaffrey, but I do think he’s being unfairly slept on and for similar reasons. His rookie season was highly underwhelming from an efficiency standpoint, but his volume was good (averaging 18.5 touches per game from Week 8-on). He averaged just 3.67 yards per carry, which was troubling, but also not far off the rookie seasons of Christian McCaffrey (3.72), LaDainian Tomlinson (3.65), Joe Mixon (3.52), and Melvin Gordon (3.48).
2. Patrick Mahomes , 2018
Like Garoppolo, we have an extremely small sample to work with when discussing Mahomes. He was our highest-graded quarterback last year during the preseason, and in his lone start and snaps of the regular season (Week 17), he was our fifth-highest-graded passer. In Week 17, he managed 22 completions on 35 attempts for 284 yards, zero scores, and one interception, as well as seven rushing attempts for 10 yards. That’s pretty good and the rushing prolificity is important, but it’s hard to base much off of such a small sample. Chiefs' general manager Brett Veach said of Mahomes earlier this offseason, “He is one of the best players I've ever seen.” Mahomes was very raw coming out of college but has always had incredible upside. He may be more hype than anything else, but a breakout could be coming, especially after the team upgraded their receiving corps (signing Sammy Watkins) in the offseason.
April 2018, ADP: QB16
Despite writing Mahomes up as one of my top breakout candidates of the season, I still wasn’t nearly as high on Mahomes as many of the other names on the list.
Understandably, the argument in favor of Mahomes is much easier now in hindsight, but we did hit on some of them in the analysis above. He flashed in a small sample the year prior and did have one of the best supporting casts in football (including offensive genius Andy Reid). Chiefs GM Brett Veach said in March of that year, after nearly 15 seasons in the NFL, that Mahomes was “one of the best players” he had ever seen. He backed that up by quickly jettisoning Alex Smith to Washington after a career year where he the league in AY/A and passer rating.
Clearly, the writing was on the wall. It was just a little less legible than for some of the other players on our list. As for the major takeaway here? Sometimes it’s not just “coach-speak” or “GM-speak,” sometimes coaches are telling the truth and pointing the way to a potential league-winner. We saw that with our last name, we’ll see this again with our next name, and we’ll provide a few more examples in tomorrow’s article.
3. Todd Gurley , 2017
In the entire history of the NFL, there are 320 instances of a running back accumulating at least 275 carries in a single season. Among these 320 seasons, Gurley’s 2016 season ranks fifth-worst in yards per carry. So, that’s the bad news. The good news, if we can call it that, is that four of the 12 worst seasons came from a running back coached by Jeff Fisher. This is likely not a coincidence.
I broke down running backs by a number of unique metrics earlier in the offseason. Gurley was not at a serious disadvantage when it came to strength of schedule or stacked boxes. He was, however, hurt by a poor offensive line. Last season, Los Angeles ranked fifth-worst in yards before contact per attempt, suggesting his offensive line didn’t give him very much room to run. I’m optimistic that this improves with the arrival of Andrew Whitworth – who ranked among our six-highest-graded offensive tackles in run blocking in two of the last three seasons.
While I’m excited about Fisher’s exit and the improvement of the offensive line, I’m still concerned by Gurley’s lack of efficiency, tough 2017 schedule, and likely reduction in target-volume. I’m not sure he’ll break off as many long runs as he did in 2015, but I am optimistic for an improvement on his 2016 season. Gurley leads the league in rushing attempts since his first NFL start, and should again see a high rushing workload, but is a much less attractive option in PPR leagues (in part because the Rams brought in pass-catching back Lance Dunbar this offseason). I’m fine taking him in the late second of standard leagues, but will likely wait a full round in PPR leagues.
June 2017, ADP RB13
Okay, we can say this one was an ‘L.’
“I’m fine taking him in the late second- or third-round” wasn’t good enough when Gurley finished the season with 25.6 fantasy points per game, or, the 15th-most by any running back all-time.
Our analysis was at least partly right, but here’s what we got wrong:
First, although we were right that any coach would be a massive upgrade on Jeff Fisher, we didn’t know Sean McVay would also become known as one of the best offensive minds in football.
Second, we were way off on Gurley’s receiving expectations. Though, to be fair, I don’t know how much of that was our fault. McVay called Lance Dunbar his third-down back in March, but then picked up an injury that kept him off the field until October. In August – two months after the above analysis was written – McVay started hyping up Gurley as an every-down bell cow. He told reporters:
"Regardless of whether Lance is available or not, we always knew that Todd would be the lead dog. And it's just kind of us figuring out what is that fine line between him being at his best, while making sure that he gets his touches necessary, and then also being mindful of having somebody that can spell him so that when the fourth quarter rolls around, and to finish the game, he feels good.”
After a quote like that, maybe it wasn’t so surprising Gurley finished the season second in weighted opportunity points per game.
4. Travis Kelce , 2018
After ranking eighth, 12th, and 19th in expected fantasy points per game, Kelce catapulted to the top of our list in 2018. He set new career highs in targets per game (7.8), expected touchdowns (5.8), and deep targets (17, matching his prior career total), which should help explain the jump in expected fantasy points. In terms of actual fantasy points per game, Kelce also led the position and would have ranked 10th among wide receivers.
Unfortunately for Kelce, the team parted ways with Alex Smith, who is the only quarterback this past decade (of 42 qualifying) to target tight ends on at least 25 percent of his throws. Patrick Mahomes, meanwhile, never once targeted the tight end position in college (Texas Tech’s Air Raid offense). The team also made Sammy Watkins the fifth-highest-paid wide receiver in the league this offseason and has promised more receiving work for Kareem Hunt. Andy Reid’s history suggests at least one of Kelce, Watkins, or Tyreek Hill is currently overvalued by ADP.
That’s the bad news. The good news is, Kelce is historically awesome. Among tight ends with at least 30 games over the past four seasons, Kelce ranks behind only Gronkowski in fantasy points per game over expectation (+2.2). He also leads all tight ends in missed tackles forced and yards after contact over the past two seasons. I’m still taking Kelce as my No. 2 tight end in drafts, despite these concerns and the lack of elite volume in prior seasons but there’s a sizeable gap between him and the next player on our list.
August 2018, ADP: TE2
Okay, so this was certainly one of our more lukewarm recommendations. However, even with the benefit of hindsight, all of the analysis above seems accurate. We just missed on one key point, and it was the same as the one from our Mahomes write-up.
Yeah, I definitely did not expect Mahomes to throw for 5,097 yards and 50 touchdowns, negating my concerns related to an increase in target competition.
5. Adam Thielen , 2018
Last season, Thielen ranked ninth in PFF grade and ranked fifth in total receiving yards. He also ranked seventh-best in yards per route run (2.33), trumping Stefon Diggs for the second consecutive year. While this was a surprise to some, it wasn’t to me. In 2016, Thielen was one of only two wide receivers (along with Michael Thomas) to rank top-20 in yards per route run, yards per target, WR rating, and drop rate. This year, he should be spending more time in the slot, which is more good news, considering new quarterback Kirk Cousins has been one of the league’s most efficient passers when targeting slot wide receivers since entering the NFL.
August 2018, ADP: WR12
Though certainly less impressive than calling his breakout season the year before (when he finished eighth in fantasy points with an ADP of ‘free’), I suppose we can count this as another at least lukewarm ‘hit’.
What was most interesting was that Thielen found his way on 33% of all ESPN Championship teams in 2018 (fifth-most), despite not even appearing on our first chart (ESPN Playoff teams). Keep in mind, Thielen’s production totally fell off of a cliff in the fantasy postseason. During the fantasy regular season (Weeks 1-13), Thielen averaged 22.6 fantasy points per game (most among wide receivers). During the fantasy playoffs (Weeks 14-16), Thielen averaged just 10.0 fantasy points per game.
Perhaps this was just an outlier. Perhaps this was merely due to variance. Or, perhaps, at least on an anecdotal level, this supports the notion that one’s primary goal should just be to make the playoffs and that whatever comes after that is in the hands of fate/variance. We’ll discuss this more in-depth in tomorrow’s article.