Fantasy Football: Don’t Overreact to Week 1 (Unless…)
We made it back to Week 1 of the NFL Season!
If you’re like me, you were enveloped in a feeling of profound serenity as you settled in to watch the early afternoon slate of games. Hopefully your months of analysis and draft prep paid off and you are starting your fantasy season undefeated.
Still, even the best-laid plans go awry. There were tons of surprises this week: Elijah Mitchell taking the reins from the injured Raheem Mostert and a (shockingly) benched Trey Sermon, Jameis Winston tossing five TDs, including a pair to TE Juwan Johnson, Mark Ingram being the workhorse of the Texans backfield…the list goes on! On the flip side, Ezekiel Elliott managers were left tilting after backup Tony Pollard outscored him on Thursday, and Saquon Barkley was given a meager workload as he continues to recover from injury.
I am writing this article to remind you not to overreact (at least, in some cases). As we will see below, many of the ‘week one wonders’ are just flashes in the pan. Remember Sammy Watkins‘ and Malcolm Brown’s week one dominance of yesteryear? How did they fare for the rest of the year? All fantasy data, unless otherwise specified, comes from nflfastR and is in Half PPR format. ADP data comes from the fantasy football calculator; we will be looking since 2015.
Week One Wonders
Just like Christian Kirk with his two scores in the rout of the Titans, there have been plenty of players in the past that impressed in Week 1 despite floundering at the end of fantasy drafts (or being left on the waiver wire). DeSean Jackson, with his 154 yards and two TDs to kick off 2019, is a notable recent example. These are all players that were drafted after the 8th round (97th pick or lower) and scored 12 or more points in Week 1.
The next question is natural: how did players perform from there?
If we take a look at where players finished at their position for the rest of the season – not including their week one performance – and compare it to where they were originally drafted, we can get a better picture. The straight line here is just the unit line: players above that line outperformed (their final positional rank is lower than their drafted positional rank) and players below that line underperformed. I also took out 2018 James Conner on this chart (and a few others), partly because it was such an extenuating circumstance (Le’Veon Bell sat out the entire year) and partly because it would skew the visuals of this chart.
Although there are definitely some players who continued to deliver value – 2018 Austin Ekeler was drafted as the RB60 and finished much higher – there are also many players that fell back to earth. Darius Slayton in 2020 was a great example: after catching two TDs in Week 1, he just found the end zone one more time for the rest of the season. He ended as about the WR60 (from Week 2 on) despite being drafted as the WR40.
In general, we see 58% of Week 1 wonders actually outperform their ADP for the rest of the year (not including Week 1). This number makes a lot of sense: the first week of the season definitely gives us more information – it’s an NFL game, after all – so players that perform well have a higher than 50% chance of actually being better than they were drafted to be and outperforming ADP. However, the probability isn’t that high: a great performance in Week 1 is far from a guarantee that a player will even return on investment relative to their late-round flier status for the rest of the season going forward.
In fact, only two players in this sample beat their ADP by a really significant amount (40 spots in a positional rank): James Conner in the wild 2018 year and DeAngelo Williams’ 2015 comeback year with the Steelers. In this sample, just 13% end up being a ‘first positional player’ (top 12 RB, WR, or TE) for the rest of the season; 39% end up being a ‘first or second’ (top 24 RB, WR or TE) for the rest of the season. As you can see in the chart above, half of the players that end in the top-12 at their position are actually tight ends, which is far, far less valuable than a top 12 RB or WR.
What’s happening here is a very, very significant regression to the mean. After the first week of excellent performance, most of these players fall back to the mediocrity they were drafted at. Just look at this chart of these players’ performance over the first three weeks: there is a very strong downward trend, with many of the lines even going to zero.
The two outliers here – with lines that jump up going into Week 3 – are both Falcons: Tevin Coleman (who had three touchdowns in his Week 3 outing) and Austin Hooper (two touchdowns). Both finished solidly on the season: about 1,000 yards for Coleman and eight TDs, nearly 800 yards and five TDs for Hooper. Still, as noted, they are outliers.
Week One Disappointments
What about the other side of the coin? Are first and second-round players that, to put it simply, stink in Week 1 a cause for worry? Should the Zeke and Saquon – and potentially, Derrick Henry – managers be concerned?
Here’s a chart of some recent notoriously ‘scary’ first weeks from top draft picks (selected in the top-20 in a draft and scored six points or less in week one). Michael Thomas is, of course, a memorable example from last year:
How do these players end up for the rest of the season? If we recreate the same chart as before, the results are a bit disconcerting: it looks like most of the players end to the right of the line (underperformed their ADP for the rest of the year).
Now, this chart shows a bit of a flawed metric: if you are a highly ranked player, it’s hard to beat your ranking by much, and there’s often nowhere to go but down. Still, it’s important to note that after a Week 1 disappointment, only 20% of stud players end up at the ADP they were drafted (or better). Naturally, the Week 1 disappointment hurts – it indicates the player might not be who we thought they were – but injuries add to these ranks as well (Joe Mixon in 2020, Adrian Peterson in 2016, etc.).
Just 33% of the Week 1 disappointments end as a top-12 option at their position; 53% end up as a top-24 option. Overall, this sample size is far smaller (since there are simply far fewer players drafted at the front of a draft than the end!), and there isn’t that much that top players can do to return excessive value (as opposed to players drafted very late), but it’s still interesting to note that the ‘regression’ is not as strong. Consider the same chart through the first three weeks: there is definitely a strong upward slope, but there are a decent amount of players that keep scoring low amounts (or score nothing at all!).
The first lesson here is simple: don’t go crazy chasing the Week 1 breakouts. While Mark Ingram is certainly more valuable than we once thought, the most likely scenario is that he regresses significantly. Try to get exciting players like this on your roster, but don’t go nuts, either with FAAB, waiver priority, or trade pieces. At the end of the season, we will likely look back and laugh at the solid outing of Jalen Reagor in Week 1. Don’t be left holding the bag!
Now, of course, don’t hear what I’m not saying. We certainly learned a lot this week, and you should be targeting the players that broke out. A sneak peek from a future article: although Week 1 waiver wire breakouts do regress, they actually perform better than later week breakouts by about 1 point per game on average (likely because we have the least amount of information coming into the first week of the season). The point is that you shouldn’t overspend and expect greatness from these players. Certainly, try to roster Elijah Mitchell if he is available, but don’t dump the majority of your FAAB or offer a ridiculous trade. For someone like Juwan Johnson – potentially not even the top TE on his team – I might not even spend a waiver priority.
On the flip side, while Ezekiel Elliott will probably return to his high-scoring expectation, there is some reason to be wary. This is especially true for Saquon Barkley, who was affected by a potentially nagging injury. Keep an eye on his play in the coming weeks and don’t be afraid to add depth at the position or even look for a trade (for a high price, of course!).
Did I miss anything? Message me on Twitter.
*Because of some ‘hashing’ issues – I don’t have a good way to merge the fantasy data with ADP data – there might be some confusion between player names (i.e., many players have the first initial “D” and the last name “Johnson”; I just removed these from the data, since it was too difficult to match up). Keep this in mind when you see a data point and think of a specific player, and let me know if you notice anything strange.
**No, you didn’t miss James Robinson: I didn’t look that deep in drafts (he went undrafted in many leagues) because the data isn’t as reliable! However, I’m willing to bet that the trend is the same: undrafted rookie RBs regress heavily to the mean, James Robinson is just an extreme outlier.