Players Who POP: Regression and Bounce-Back Candidates for Week 9 (Fantasy Football)

Welcome back to Players Who POP for Week 9 of the 2024 season, where we try to identify bounce-back players who can give our fantasy teams a much-needed boost or players who may be due for a bit of a fall from grace from a production standpoint. What a brutal week for injuries. It’s been so tough with so many high-profile players out for the year. Still, we must persevere, and the POP model can help us a ton. Let’s dig in to find the players who are under or over-valued heading into the second half of the fantasy football season.

Methodology

Each week, we use machine learning to predict rest-of-season average fantasy scoring. You can read the complete strategy behind POP, and why it’s different than other expected points metrics here but here’s the tl;dr: The two best-performing metrics for predicting player fantasy points per game were player share of team opportunity value and teammate fantasy points per game. This model understands that team strength – measured in teammate fantasy production – does in fact play a factor in an individual’s fantasy success, albeit at roughly one-third the importance of a player’s opportunity according to variable importance tests. What is opportunity value, you ask? Without getting too far into the weeds (the linked article goes deep if you’re curious), we’re using a machine learning strategy called “stacking” – using several sub-models (like expected receptions, expected rushes, etc.) as predictors in a new model. Each sub-model is also a machine learning model, but they are rolled up (or “stacked”) into one metric in order to improve predictions within this final model.

Note: The original article refers to this roll-up metric as “expected fantasy points,” but that’s a bit of a misnomer. Instead opportunity value is a more appropriate renaming for this volume metric created in the feature engineering phase of the model-building process.  The metric is distinctly different than pre-existing expected fantasy points metrics like the one expertly analyzed in our Expectation Report series.

Running Backs

A Note on Top Running Backs — “Not All Is as It Seems”

This note is here to say that things are a little bit different according to the POP model than in seasons past, as we have significantly more divergence from expectations in the RB1 tier of players compared to the WR1 tier, and it’s usually the opposite. Maybe it’s model “drift” (deviation from former expectations), maybe it’s a change in how players are used, and both ideas are worth exploring in a separate article, but this is what you need to know: there are a ton of running backs when you look at a full-season sample that appear to be drastically different than expectation, which means we may still see some alterations in that area of the rankings before the season’s said and done. The top six to eight backs seem to have distanced themselves, but beyond that, everything’s up for grabs. At the wide receiver position, I’m starting to feel like we’re pretty locked in, and it’s only injuries that will alter the landscape drastically. I think the biggest wide receiver trades are behind us as well.

Regression Candidate: Kyren Williams

Williams just scores and scores and scores. He has only played in four football games where he hasn’t scored a touchdown since Week 1 of 2023. You really can’t fault someone for scoring touchdowns, it’s literally the most important part of fantasy football, but they’re supposed to be somewhat random. They are supposed to distribute to different players in different weeks, especially when there are not one but two premier wide receivers on your team. It’s supposed to be difficult to score, but comically, it seems to be easier and easier. Williams isn’t jumping off the page from an efficiency standpoint this year, ranking outside the top 45 running backs in yards per touch, though he has been elusive of late, evading 21% of would-be tacklers, inside the top 25. But he’s still not generating a ton of yards after that elusiveness, ranking outside the top 40 in yards created per touch. Williams is always going to look a little bit better than expectation because of the touchdowns, but again, I might look like a fool based on just how frequently Williams finds the end zone. With all three of their star players back in tow, the Rams are two-point favorites on the road against Seattle in a wide-open NFC West.

Regression Candidate: D’Andre Swift

Swift had yet another impressive day for the Eagles in their Week 8 loss, but the usage was a bit concerning. We don’t want to take anything away from the fact that he’s now topped 100 total yards in four straight contests, but not getting a single target? That’s concerning to me. His 18 rushes were the second-most he’s seen in a game this season, so I guess you could say that’s a good thing, but target share is what helps us with floor when the efficiency and the touchdowns aren’t there. In the same four-game stretch where he’s topped 100 yards, he’s also scored a touchdown in each game, and (unless you’re Kyren Williams, apparently) most touchdown streaks tend to come to an end around the three or four-game mark. The Eagles get a winnable game against the Cardinals this week, but I still think Swift’s opportunity is a bit thinner than we thought.

Regression Candidate: Tank Bigsby

Bigsby has probably played his way into the starting role at this point, but Travis Etienne is going to play, and when he does, it’s going to inevitably eat into Bigsby’s workload. It was incredibly encouraging to see him get three targets last week, the most he’s seen all year, but he was not efficient, gaining only 86 yards on 21 total opportunities and was held without a score. Plus, the Jaguars are touchdown underdogs on the road against the Eagles, whose defense ranks fifth in yards per play allowed and seventh toughest against running backs in adjusted fantasy points allowed this season.

Unlock Exclusive Tools + Bonus Episode

Join the FootClan

Wide Receivers

Regression Candidate: Justin Jefferson

This is always the worst. The model says a superstar is overperforming, and everything your eyes tell you says that the player is going to keep performing just as well as they have been, if not better. But this is one of those situations where if you are a “trust the numbers” guy, you have to say that Jefferson may be due for a little bit — and I stress a little bit — of regression. He has caught 15 of his last 16 targets and, more importantly, has not seen the same type of consistent target ceiling as we would have expected him to have. He’s had one game with double-digit targets, and as good a player as Jefferson is, an eight to nine target expectation is simply very different from an 11 to 14 target expectation as we have come to know and enjoy for Jefferson. Of course, he can dominate on just a small sample, and with Joe Flacco starting for the Colts this week against the Vikings, this game has all the makings of a sneaky shootout. But the biggest thing for Jefferson is volume right now; let’s hope that the Vikings get him some layup targets in the weeks to come to boost the opportunity floor, and all should be well.

Regression Candidate: CeeDee Lamb

I hemmed and hawed about including Lamb on this list as well because, on one hand, of course, a player who earned 17 targets and scored twice is going to be a regression candidate the next week. But on the other hand, Lamb had been lacking in the huge-target game department, and now suddenly he’s earned 31 looks over the Cowboys’ last two games. This is really about the fact that he’s not going to average over 15 targets per game the rest of the season, and not about the fact that the Cowboys have made a clear shift in the passing game, ensuring their star player gets even more opportunity. Of course, it hasn’t helped the Cowboys on the scoreboard much, as they’ve dropped each of those games against tough opponents, and this week they’re underdogs on the road against the first-place Atlanta Falcons. But Atlanta is just a league-average defense and struggles against wide receivers in particular, so there is a chance that even if he sees fewer looks (very likely), the regression looks more like a slow slide compared to a massive drop-off.

Bounce-Back Candidate George Pickens

Pickens had not one but two touchdown plays overturned in Week 8, one by a penalty and another that was overturned after review. But there’s no doubt the Steelers have a serious outside threat and a quarterback who can get him the ball consistently. Pickens still managed 74 yards on four receptions, but he was oh so close to a monster game on the heels of his top-three fantasy performance in Week 7. Most importantly though, I think we end up seeing Pickens hover around the eight to 10 target mark with Russell Wilson at the helm when the season is said and done. The team has not upped their pass rate significantly since Wilson has been efficient, but the overall volume should still increase. From a fantasy standpoint, that would drastically raise Pickens’ floor; we know the ceiling is sky-high already.

Bounce-Back Candidate Jalen McMillan

The Buccaneers are going to take some time to find their identity in the wake of their massive wide receiver losses, but you can’t help but be encouraged by McMillan’s usage in his first game with a new role. But you’re also allowed to be disappointed in the fact that he’s now earned 15 targets over the last two games, has caught less than half of them, and hasn’t topped 50 total yards receiving on those looks. We know that the catch rate is going to be slightly deflated for a player like McMillan whose average depth of target is nearly 14 yards, further downfield than all but a few wide receivers in the NFL. But it also means the Buccaneers’ burner has a chance to impact games with just a few catches, and a long touchdown score would go a long way toward the Buccaneers having any type of chance at an upset in Arrowhead, as they are 8.5-point underdogs against the Chiefs this Monday night. Kansas City is top five in the NFL in adjusted fantasy points allowed to wide receivers and in yards allowed per play.

Tight Ends

Regression Candidate: Mark Andrews

Okay, Mark, we get it — you are still here. After an abysmally slow start to the 2024 season, where he was held under 52 yards receiving in each of the first four games and had two games in that span without a single catch, Andrews has now scored touchdowns in three straight games. And this was the thesis for Andrews heading into the season: he’s always had a great connection with Lamar Jackson in the red zone, and even with the threat of Derrick Henry taking away touchdowns, we knew the end zone upside for Andrews wasn’t going entirely away. But now the pendulum has swung the other way, and we should expect the scores to be distributed differently going forward, especially with the team trading for stud wide receiver Diontae Johnson this week. I’ll change my tune if we see a quintessential target game for Andrews, but I’m really not holding my breath for that type of usage on a team that also includes Zay Flowers, the aforementioned Johnson, and two of the most potent rushers in NFL history in the backfield.

Unlock Exclusive Tools + Bonus Episode

Join the FootClan

Regression Candidate: Kyle Pitts

It only took four targets for Pitts to rip the Tampa Bay defense to shreds last week. His 91 receiving yards were the most he’s had all year; in fact, it was the most receiving yards he’s had in a game since his rookie season. But he’s still not going to score two touchdowns every week, especially not on four targets. It’s great to see Pitts back doing what we know he’s capable of, but this is pretty cut and dry: he’ll need to get back to a much larger target volume in order to keep up his double-digit scoring, something he’s now accomplished in four straight games. And it’s not like his matchup this week, Dallas, has been very effective defensively this season, as they rank 27th in yards per play allowed and 30th in points per play allowed.

POP Scores (Previous Opportunity-Based Projection) – Week 9