The kids didn't get the regression memo.
What this means for your roster
Hold Luther Burden at our WR33 and Tetairoa McMillan at our WR14, and do not sell either one into a slow September. Buy Xavier Worthy at WR40 and Keon Coleman at WR43, where our board is already above the market and our own projection is the number that runs low. Hold Malik Nabers at WR20. Buy Matthew Golden at our WR47 price and stash Elic Ayomanor at WR90. And do not carry this over to your young running back or tight end, because the edge stops cold at those positions.
What the consensus says
Week 3 is when a young receiver's quiet start turns into a sell column. DraftKings Network put Luther Burden on its wide receiver sell list because "most of the valuation stems from very bullish projections based on a pretty limited sample." Fantasy Life had already filed Tetairoa McMillan under a sophomore slump, arguing his rookie target share never turned into weekly points. The other side shows up too. FantasyPros says hold Burden, and RotoStreetJournal calls selling a struggling young receiver "the classic dynasty mistake of double-punishing players." But the hold side argues from film and patience. Nobody puts a number on the thing both sides are fighting about: is a young receiver's projection a ceiling you should sell into, or a floor?
The claim, in plain English
The sell case says a young receiver is priced off a projection that trusts one small sample too much. So test the projection. Take every receiver who was under 23 in the last season his projection was built from, and ask how often he beat that projection the following season, compared with every other receiver on the same boards. The comparison is the whole game. Every projection misses, so the only question worth asking is whether young receivers miss in one direction more than everybody else does. The script is here.
How I beat on it
I used our own backtest boards, 2018 through 2024. Each one was built only from seasons up to its year and graded against the season after, in half-PPR with a tight-end premium. That's 824 receiver seasons with a projection on file, 145 of them from receivers under 23, covering 113 different players. The framework lives here.
And this is the hard version of the test, because our projection already does what the sell side is asking for. When a receiver's first season beats the average, it trusts that season about a quarter of the way and fills in the rest from the average fantasy-relevant receiver. Then it trims young players for age. If any projection should be too bullish on a thin sample, it isn't this one.
What the data actually said
First swing: receivers under 23 beat our projection for the next season 68 percent of the time. Every other receiver beat it 54 percent of the time. On margin, the young group out-scored its projection by 1.66 points a game and everyone else by 0.28. That gap is 1.4 points a game, every week, and it runs the opposite way from what the sell columns promise.
Aggregates lie, so I tried the obvious escape hatches. The first is opportunity. Maybe young receivers just got more targets. Split by last season's target share and the gap shows up in every band:
| Target share last season | Receivers under 23 | Everyone else | How much more the young ones beat their projection |
|---|---|---|---|
| Under 10% | 36 | 160 | 1.0 pts/g |
| 10 to 18% | 57 | 238 | 1.9 pts/g |
| 18 to 25% | 41 | 172 | 1.1 pts/g |
| 25% and up | 11 | 107 | 2.0 pts/g |
The second is survivorship. A young receiver who flops gets benched and falls below the eight-game line, so maybe only the good ones got graded. Drop the line to one game and the gap is still 1.3. Young receivers actually clear the eight-game line more often than everyone else, 80 percent against about 70.
The third is the label. In May I told you the "sophomore breakout" tier is mostly noise. The label still is noise. Receivers 23 and older coming off their first season beat their projection by just 0.2 points a game more than the rest of the position, which is nothing. Receivers under 23 coming off one season beat it by 1.1 more, and the ones under 23 with two seasons already on record beat it by 1.9 more. It was never the class year. It's the birthday.
The names make it real. Before the 2025 season our backtest number for Jaxon Smith-Njigba was 9.0 points a game, and he scored 17.8. Puka Nacua in 2024: projected 6.5, scored 15.2. Ja'Marr Chase in 2022: 10.6, then 16.9. All three are still top six on our board today, at WR6, WR3 and WR1, and all three are holds. It misses the other way too. Treylon Burks was projected at 7.6 for 2023 and scored 3.5; he's our WR233 now and not a roster spot. But about two young receivers in three break the kids' way.
Running backs and tight ends under 23 are a different story. They beat their projections by no more than the rest of their positions do (0.2 and 0.1 points a game more, both noise). This is a receiver thing.
What the engine already figured out
Not enough. This is a miss in our own next-season number, and the receipts say where it lives. About three quarters of the gap sits in the base number, the step that trusts one good season only a quarter of the way. The rest is the age trim. That regression isn't wrong in general, since 24-year-old receivers coming off their first season get the same treatment and finish right where the rest of the position does. It's wrong for players who are still growing. For a receiver under 23, the first season works as a floor, not as a noisy draw from the average receiver. The same pattern shows up in reverse at the other end: receivers 31 and older fall 1.1 points a game short of their projection relative to the rest.
I found this by fixing how we grade ourselves. Our accuracy scan had been grading a multi-year value number against a single season, and the age discount buried it. It doesn't move our ranks yet. A fix has to clear the same before-and-after test every change to our board clears, and that test is filed next.
What to do about it
Don't sell a young receiver because a projection says the sample is thin. That includes ours. The sell columns are selling the exact regression that runs low on these players, and the market price they're handing you back was built by that same regression.
The 2026 names
Our superflex, half-PPR, TE-premium dynasty board, receivers who were under 23 last season.
| Receiver | Our board | Call |
|---|---|---|
| Tetairoa McMillan | WR14 | Hold. The slump case is a projection argument, and projections run low on him. |
| Malik Nabers | WR20 | Hold. We sit below the market on him mostly because of how many games we expect after last year's injury, which is a separate question. Per game he's squarely in the group that beats its number. |
| Luther Burden III | WR33 | Hold. The sell case is "bullish projections on a limited sample." Ours is the cautious one, and it runs low on receivers his age. |
| Xavier Worthy | WR40 | Buy. Two seasons on record and still in the group; that subgroup beat its projection by the most. |
| Keon Coleman | WR43 | Buy. Our board is already well above the market on him, and this says our number is the low one. |
| Matthew Golden | WR47 | Buy at our price, not a premium. |
| Elic Ayomanor | WR90 | Buy as a bench stash. |
One honest caveat: over three seasons instead of one, graded against the number our dynasty board actually ranks on, receivers under 23 still beat it by more than everyone else (0.7 points a game), but that margin sits inside the noise, so the next-season edge is proven and the long-run edge is only likely.