The XFP Report: Week 4

Hello, and welcome to the Week 4 XFP Report. If you’re unfamiliar with XFP, I’ll get to that in a little bit.

But basically, every week we’re going to be telling you which players are seeing the best volume for fantasy, as measured by Expected Fantasy Points (XFP). We’ll be telling you who the best buy-low and sell-high candidates are, as measured by Points Above Replacement (PAR), or the differential between actual- and expected fantasy points. This is an especially effective approach in DFS, where players are typically priced by production rather than volume, though PAR will regress to the mean. And (at the end of the article) we’re going to be telling you who the best volume-per-dollar DFS plays are.

What is XFP?

Premium subscribers can access XFP (and other advanced stats like air yards, deep targets, end zone targets) here.

Expected fantasy points (XFP) is flat-out the best and most comprehensive way of measuring a player’s volume. It’s telling you – based on a player’s unique usage – how many fantasy points that player should have scored. It’s telling you how many fantasy points a perfectly league-average RB, WR, or TE would have scored with that same exact volume. It looks at every individual carry by down and distance and distance from the end zone and every individual target by depth of target and distance from the end zone, and then cross-references each carry and target to each carry and target with those specific qualifiers over a multi-year sample to tell you what exactly those carries and targets are worth (historically).

Expected touchdowns (XTD), same thing. RBs score from the one-yard line on 54% of their attempts. RBs score from the 17-yard line only 3.6% of the time. So why ever use “red zone carries,” which treats both carries the same, as a fantasy stat? I have no idea.

Why doesn’t everyone point to XFP in their fantasy research? I have no idea. Once you have XFP and XTD you can contrast that with a player’s actual fantasy points or actual touchdown total to tell you how efficient a player has been (PAR). This is especially useful in highlighting regression candidates, buy-low targets, and mispriced players for DFS.

Through 3 weeks of action, here are the top-25 players in expected fantasy points (XFP) per game:

The Top 25

Najee Harris, RB, Pittsburgh Steelers

[FPG: 18.0, XFP: 22.3, Diff: -4.3]

Pittsburgh-era Le’Veon Bell-type usage is what you were promised. And Pittsburgh-era Le’Veon Bell-type usage is what you have received.

Harris saw 19 targets in Week 3, falling one target shy of the NFL record for a RB. Add in 16 carries and 4 opportunities inside the red zone, and this was worth an astounding 37.6 XFP. Or 20% more (+6.3 XFP more) than any other RB in any week this season.

He now ranks 2nd among all RBs in XFP per game (22.4). He leads all RBs in Snap% with 96%, while the next-closest RB ranks a sizeable distance behind (Alvin Kamara, 80%). 14.0 carries per game, 9.3 targets per game, and 99.5% of the backfield XTD.

Yes, the QB looks cooked (3rd-worst in PFF grade). The offense is bad (5th-worst in points per drive). The offensive line is very bad (4th-worst in yards before contact per RB attempt). And I don’t care at all.

This sort of usage yields RB1 production 98 out of 100 times regardless of how talented the RB is. And, in this case, the RB is very good.

Joe Mixon, RB, Cincinnati Bengals

[FPG: 14.6, XFP: 16.9, Diff: -2.3]

Mixon isn’t quite seeing the target-volume we hoped for (27th in target share), but he’s still no doubt a high-end bell cow and an easy RB1 for fantasy.

He ranks 2nd among all RBs in XFP market share (30%) and 3rd in snap share (79%). He ranks 2nd in carries per game (23.3) and 10th in XFP per game (16.9), but just 19th in FPG (14.6).

Look for all of those numbers to pick up this week, favored by 7.5-points against a Jaguars defense that is giving up the 6th-most FPG to enemy RBs (despite only facing Houston, Denver, and Arizona).

7 of Cincinnati's 8 offensive touchdowns have come through the air this year. That’s going to regress to the mean, and, I think, regress in a big way this week.

Mike Williams, WR, Los Angeles Chargers

[FPG: 25.2, XFP: 20.3, Diff: +4.9]

I think Joe Dolan summed it up perfectly in the Week 4 Market Report:

We don’t have a category above “upgrades” on the Market Report, but if it did exist, we’d be firing Williams into that category. He has been an absolute monster. Like, Cloverfield level terror. On 9 targets in a huge win over the Chiefs, Williams posted 7/122/2 receiving. His 33.2 fantasy points were the second-best of his career and easily his best game since 2018… when he put three touchdowns on this very same Chief squad. But it’s the consistency that is now defining Williams. After comments from OC Joe Lombardi this off-season that Lombardi (erstwhile of the Saints) envisions Williams as a Michael Thomas-like player… Williams has dominated like Michael Thomas. The 9 targets he saw were actually his fewest of the season, and through three games, only Cooper Kupp has scored more fantasy points at the WR position. The funny part is Williams isn’t even tops on his own team in targets — his 31 rank behind Keenan Allen’s 33. But Justin Herbert is playing well enough to distribute the ball to two fantasy stars. One thing that also makes Williams (and Allen) great for fantasy is how frequently the Chargers throw the ball in tight — both Allen and Williams have 4 targets inside the opponents’ 10-yard line, with Williams scoring 2 TD to Allen’s 1. But Williams also has the deep-threat ability to make him a truly dynamic playmaker for an aggressive QB. If Williams stays healthy, he’s going to win leagues. Simple as that.

I don’t really have anything to add, but I’d just like to agree that I can envision Justin Herbert supporting two WR1s for fantasy. Although Williams has been more productive than Allen (25.2 FPG to 17.6 FPG), both are seeing high-end WR1-levels of volume. Among all WRs, Allen ranks 2nd (22.4) and Mike Williams ranks 5th in XFP per game (20.3).

Marquise Brown, WR, Baltimore Ravens

[FPG: 17.0, XFP: 14.0, Diff: +3.0]

Here’s what I said last week:

Is it… Is it possible that I was not just dead-wrong on it being #HollywoodSZN last year? Perhaps I was just a little bit off on the timeline, and #HollywoodSZN actually began in Week 12 of last season. Since then (postseason included), Brown averages 7.3 targets and 17.1 FPG (WR9). Over this span he’s either caught a touchdown or eclipsed 80 receiving yards in 10 consecutive games. This year, he ranks 7th in FPG (21.4), on the back of a massive 31% target share.

And, of course, once Hollywood sucks me back in and onto the hype train, he finds a way to let me down. I’m still bullish long-term, but, yeah, Week 3 was absolutely painful to watch.

Alvin Kamara, RB, New Orleans Saints

[FPG: 15.3, XFP: 18.0, Diff: -2.7]

Kamara is inarguably seeing the best usage of his career. And he’s arguably seeing the best usage of any RB in fantasy. Well, maybe only if you squint really hard, because Derrick Henry has 28 more carries and 2(!) more receptions. But…

Kamara is now playing on 80% of the team’s snaps (2nd-most), up from 65% last year. He’s hit 20 carries in two of three games this year, after reaching that mark just once across his prior 60 career games. He ranks behind only Najee Harris in target share (20%), though he’s averaging just 4.7 targets per game (down from 7.1). Among RBs, he ranks just 16th in FPG (15.3) and 8th in XFP per game (18.0), but he leads all players at all positions in XFP market share (32%).

Now, before we get too excited, it's important to remember the impact of pace. Kamara ranks highly by any market share metric, but the Saints are the second slowest-paced team in the league (30.1 seconds per snap) and have run the 2nd-fewest plays per game (56.0). Hopefully that’s just a function of them having played in three blowouts, but if not, that could at least partially offset Kamara’s market share jump. For what it’s worth, New Orleans played at the 4th slowest pace last year (28.4) and Kamara still finished as the overall RB1 (25.2 FPG).

Further, it’s odd to see Kamara falling short of his expectation by 2.7 FPG. Especially considering he’s (no hyperbole) the most efficient fantasy running of all time. That’s a number that’s going to regress closer to his prior career average (+4.4).

D’Andre Swift, RB, Detroit Lions

[FPG: 20.0, XFP: 20.2, Diff: -0.2]

Swift is sort of the reverse of Kamara. He ranks 19th Snap% (63%) but 11th in total snaps (125). He ranks 21st in positional XFP market share (59%), 9th in team XFP market share (23%), and 3rd in XFP per game (20.2).

He's averaging 11.7 carries and 8.0 targets per game, but that's only 2.3 carries and 3.3 targets more than Jamaal Williams. And Swift has only captured 59% of the backfield XTD (near-end zone usage).

Still, he ranks 3rd among all RBs in target share (19%), and that’s huge for a team that’s unlikely to be favored in any of its remaining 14 games. (Remember, outside of the red zone, a target is worth roughly 3.13 times as much as a carry in PPR leagues.) Detroit also ranks top-7 in plays per game and pace of play.

Despite being stuck in a 60/40 committee backfield, I still view Swift as a near-lock to finish the season as a fantasy RB1.

Odell Beckham Jr., WR, Cleveland Browns

[FPG: 13.7, XFP: 18.3, Diff: -4.6]

OBJ is back? I’d say so.

Or, at least definitely from a usage standpoint. Beckham earned 18.3 XFP in Week 3, which led the team and ranked 17th-best among WRs on the week. In total, he earned 10 targets, 1 rushing attempt, and 175 air yards (3rd-most). He might not be fully healthy, coming back from ACL surgery, but I’d expect volume to remain about as good as it was this week for however long Jarvis Landry remains sidelined. Because, keep in mind, this good usage came in a 26-6 beatdown where the Bears totaled just 68 passing yards. Beckham gets a pillow-soft matchup this week, against a Vikings defense that has given up the 4th-most FPG to opposing WRs.

Brandon Aiyuk, WR, San Francisco 49ers

[FPG: 8.1, XFP: 12.4, Diff: -4.3]

Brandon Aiyuk is back? I’d say so.

Or, at least definitely from a usage standpoint. Aiyuk’s route share jumped from 58% in Week 2 to 89% in Week 3. And he led the team in XFP, with 21.6, which also ranked 7th-best among all WRs. He wasn’t very productive (14.5 fantasy points), perhaps still a little hampered by a hamstring injury he suffered in the preseason, but this good usage is the most important takeaway.

Saquon Barkley, RB, New York Giants

[FPG: 11.3, XFP: 15.2, Diff: -3.9]

Saquon Barkley is back? I’d say so.

Or, at least definitely from a usage standpoint. Barkley played on 84% of the team’s snaps (2nd-best), earning 16 of 19 carries and 7 of 7 targets out of the backfield. (For perspective there were only 14 instances of a RB reaching at least 16 carries and 7 targets in a single game last season.) In total, Barkley scored 21.4 fantasy points on a 23.3-point expectation (5th-best).

If you drafted Barkley you have to feel pretty good about this usage, though admittedly his 3.4 YPC average suggests he’s still not quite 100%. He also gets a brutal matchup this week, against a Saints defense that’s ranked top-7 in FPG allowed to opposing RBs for four-straight seasons. But still, if you own him, keep playing the long game. When fully healthy, Barkley is the closest thing in the game to Christian McCaffrey, and now he’s back to seeing McCaffrey-levels of volume.

Emmanuel Sanders, WR, Buffalo Bills

[FPG: 14.1, XFP: 14.0, Diff: +0.1]

Last week we wrote up Sanders as one of our top positive regression candidates, and the regression hit in a big way last week. Through the first two weeks, Sanders averaged just 8.0 FPG on 15.4 XFP per game. In Week 3, he scored 26.4 fantasy points on just an 11.4-point expectation. So, through three weeks, he now ranks 29th in FPG (14.7) and 34th in XFP per game (14.0).

Who is the No. 3 tight end?

Clearly, there’s a big two at the TE position.

Darren Waller is pacing the position in XFP per game. His 22.0 XFP per game would rank 4th among all WRs, and is 45% better (6.8 more XFP per game) than the next-closest TE (Travis Kelce).

Travis Kelce ranks 2nd with 15.2 XFP per game, but is averaging 22.3 FPG, which leads the position and would rank 3rd among all WRs. Over the past two seasons, he’s reached at least 21.5 fantasy points in 76% of his games. The next closest receiver is Davante Adams, at only 53%. Yeah, he’s a freak.

But who is No. 3? Truthfully, I’m not too sure. But here’s who you’re looking at:

George Kittle might be the best tight end in football, but he’s undoubtedly not the best tight end in fantasy football. He ranks just 12th in XFP per game (9.0). Waller leads the position in routes run per game (44.0), followed by T.J. Hockenson (38.0), Travis Kelce (35.3), and Kyle Pitts (34.3). Kittle ranks 12th (27.0)… Kittle leads the position in a more unfortunate stat — pass blocking snaps per game (7.3). None of Waller, Hockenson, Kelce, or Pitts average more than 2.7 pass block snaps per game. That’s a massive disadvantage for fantasy. Unfortunately, unless you’re an absolute weirdo, your league probably doesn’t award fantasy points for pancake blocks. Still, Kittle ranks 3rd among TEs in target share (19%), behind only Waller and Kelce. Mark Andrews ranks 4th (18%) and Hockenson ranks 5th (18%).

Rob Gronkowski ranks 3rd in XFP per game (13.3) and 2nd in FPG (19.5). Among all WRs, 19.5 FPG would rank 11th-best. Like Kittle, Gronkowski is at a disadvantage for fantasy, pass blocking on 6.0 snaps per game (3rd-most). But that liability gets more than offset by the fact that Gronkowski is on one of the most-potent passing attacks in football. And maybe one of the most-potent passing attacks in football. He’s also arguably the greatest tight end and greatest end zone weapon of all-time. He averages 7.0 targets per game, which ranks 4th-most among all TEs, but he also leads all TEs in XFP per target (1.90). That’s because in addition to seeing good target quantity, he’s seeing phenomenal target quality, averaging 1.0 deep targets (3rd), 1.0 10-zone targets (1st), and 1.3 end zone targets per game (1st).

T.J. Hockenson ranks 4th in XFP per game (12.7) and 4th in FPG (15.8). I’m not quite sure what to make of last week’s dud (2 targets, 10 yards), but he’s still probably my bet for TE3 of the 2021 season.

Kyle Pitts ranks 5th in XFP per game (11.2), but just 16th in FPG (8.3). He’s a phenomenal “buy low”, but his ADP was undoubtedly probably too high by multiple rounds this offseason.

Mark Andrews ranks 8th in XFP per game (10.1) and 10th in FPG (10.5). He ranks 3rd among TEs in XFP market share (15.1%), just barely ahead of Gronkowski (14.8%) and Hockenson (14.4%).

Ja’Marr Chase, WR, Cincinnati Bengals

[FPG: 18.9, XFP: 9.6, Diff: +9.3]

Technically, Chase is your top negative regression candidate of the 2021 fantasy season, outscoring his volume-based expectation by a whopping 27.9 fantasy points. Among WRs, Chase ranks just 66th in XFP per game (9.6) but 13th in FPG (18.9).

On one hand, he’s due for a massive efficiency regression. On the other, this is extremely encouraging. Chase, arguably the greatest WR prospect since A.J. Green and Julio Jones, is playing out of his damn mind. And keep in mind, rookie WRs typically see a massive jump in both volume (51%) and production (55%) in the second-half of a season.

So, if you own Chase in season-long leagues, congratulations, that looks like one of the best picks you could have made (especially if you got him after his ADP “dropped” due to a poor preseason performance). That said, I wouldn’t be surprised if a healthy Tee Higgins out-produces him over the next 4-6 weeks. (Higgins ranks 25th in XFP per game with 15.7.)

Top Regression Candidates

- 1) Elijah Moore: Moore has fallen 22.8-points below his volume-based expectation, making him the No. 1 positive regression candidate on the season. Among WRs, he ranks 43rd in XFP per game (12.5) and 103rd in FPG (4.9). Yeah, not at all a great start for one of my most-hyped players all offseason. I’m definitely far more pessimistic now than I was earlier in the offseason, and I don’t love that he’s playing nearly full-time out-wide rather than in the slot (where I think he belongs), but here’s the good news: he’s just 1.4 XFP per game behind Corey Davis for the team-high. And they both have near-ideal matchups this week and next. Zach Wilson has been under pressure on 46.7% of his dropbacks (2nd-most), but his next two matchups (Tennessee and Atlanta) both rank bottom-6 in PFF Pass Rush Grade. The Titans have given up the 3rd-most FPG to opposing WRs, and the Falcons have given up the 12th-most. But here’s more bad news: Moore is currently in the concussion protocol, in danger of missing this week’s soft matchup. If he’s out, pivot to Corey Davis in DFS.

- 2) Jonathan Taylor: Taylor has fallen 22.7-points below his volume-based expectation, making him the No. 2 positive regression candidate on the season. Everything we said about him last week, remains as true as it was when I wrote it. So, I urge you to check that out here.

- 3) Mike Davis: Davis has fallen 21.4-points below his volume-based expectation, making him the No. 3 positive regression candidate on the season. Like with Taylor, everything I said about Davis last week seems just as relevant today. Except now the bear-case argument I laid out seems far more likely than the bull-case narrative. In Week 3, Davis’ usage was only barely better than Cordarelle Patterson (14.1 XFP to 13.1), who out-targeted Davis 7 to 4. Patterson was also, again, far more efficient than Davis. Through three weeks, Patterson has out-scored his expectation by 11.6 points (4th-best), while Davis has fallen short of his expectation by 21.4 points (2nd-worst).

- 4) Dyami Brown: Brown has fallen 18.7-points below his volume-based expectation, making him the No. 4 positive regression candidate on the season. He’s a player to monitor, but not one to get too excited about for season-long leagues or DFS. He’s been hyper-inefficient, but his volume hasn’t been especially great, ranking just 76th among WRs in XFP per game (8.5).

- 5) Darren Waller: Waller has fallen 17.6-points below his volume-based expectation, making him the No. 5 positive regression candidate on the season. Yeah, it’s really scary to think of how good Waller has been, and then realize he should still be about 5.9 FPG better than that.

- 6) Mark Ingram: Ingram has fallen 17.5-points below his volume-based expectation, making him the No. 6 positive regression candidate on the season. But I’m betting Mark Ingram is just your 2020 A.J. Green of the 2021 fantasy season. He may be a top regression candidate on paper, but he’s still very unstartable. He’s 31-years-old and the Texans might not be favored to win another game all season.

- 7) Keenan Allen: Allen has fallen 14.4-points below his volume-based expectation, making him the No. 7 positive regression candidate on the season. Through three weeks, Allen ranks 2nd (22.4) and Mike Williams ranks 5th in XFP per game (20.3). Allen ranks 15th in FPG (17.6) and Williams ranks 2nd (25.2). Not to take anything away from Williams, who really might finish the season as a fantasy WR1 in that Michael Thomas-esque role OC Joe Lombardi warned us about in the offseason. But Allen is no doubt a top regression candidate, and I’d bet he starts to gain some ground on Williams over the next few weeks… And here’s a dope bonus stat: Allen currently leads all players in targets (5) and receiving yards (62) called back due to penalty. If not for those penalties, he’d be leading the league in targets (37) and averaging 106.7 receiving YPG, which would rank behind only Cooper Kupp, Deebo Samuel, and Brandin Cooks.

- 8) Stefon Diggs: Diggs has fallen 14.2-points below his volume-based expectation, making him the No. 8 positive regression candidate on the season. Among WRs, Diggs ranks 3rd in air yards per game (147.0), 5th in targets per game (11.7), and 7th in XFP per game (19.4), but just 27th in FPG (14.7). Yeah, he’s a screaming “buy” in season-long leagues. He’s seen 8 deep targets this year (259 total air yards) but has caught only one, good for a 13% conversion rate. But he caught 49% of his deep targets over the prior two seasons. That’s something that’s going to regress to the mean, and regress to the mean in a very big way. For instance, if he caught his three deep targets last week, he would have added a minimum of 10.9 fantasy points and a maximum of 35.1 fantasy points.

- 9) Michael Pittman: Pittman has fallen 13.9-points below his volume-based expectation, making him the No. 9 positive regression candidate on the season. Unbelievably, he ranks 13th among all WRs in XFP per game (17.8). Yeah, if Carson Wentz was playing like the 2017 version of Carson Wentz, and not like a middling QB with two bum ankles, Pittman might be a league-winner. But alas…

RB Team XFP%

1. Alvin Kamara (33%)

2. Joe Mixon (29%)

3. Derrick Henry (28%)

4. Najee Harris (26%)

5. Christian McCaffrey (25%)

6. David Montgomery (24%)

7. Aaron Jones (23%)

8. Dalvin Cook (23%)

9. D’Andre Swift (23%)

10. Darrell Henderson (23%)

11. Jonathan Taylor (23%)

RB Team Pos XFP% (The Bell Cow Stat)

1. Najee Harris (99%)

2. Darrell henderson (85%)

3. Saquon Barkley (81%)

4. Joe Mixon (81%)

5. Dalvin Cook (80%)

6. Derrick Henry (79%)

7. Alvin Kamara (78%)

8. Aaron Jones (75%)

9. Chris Carson (74%)

10. James Robinson (72%)

11. David Montgomery (72%)

WR / TE Team XFP%

1. Davante Adams (31%)

2. Diontae Johnson (29%)

3. Cooper Kupp (29%)

4. Anthony Miller (29%)

5. Brandin Cooks (26%)

6. DK Metcalf (25%)

7. Odell Beckham Jr. (25%)

8. Tee Higgins (25%)

9. Keenan Allen (24%)

10. Deebo Samuel (23%)

11. Tyler Lockett (23%)

12. D.J. Moore (22%)

DFS Values (DK)

1. Anthony Miller, WR (4.46X)

2. Mike Davis, RB (3.65X)

3. Diontae Johnson, WR (3.33X)

4. Michael Pittman, WR (3.30X)

5. Najee Harris, RB (3.28X)

6. D’Andre Swift, RB (3.25X)

7. Elijah Moore, WR (3.20X)

8. Evan Engram, TE (3.17X)

9. Odell Beckham Jr., WR (3.15X)

10. Collin Johnson, WR (3.12X)

11. Brandin Cooks, WR (3.02X)

12. Cole Beasley, WR (2.98X)