Where bad takes come to die and good ones get receipts.
Every sell list this month moves the incumbent receiver because his team spent a pick on a pass catcher. I ran 583 receiver seasons against the next one, holding target share fixed. The rookie really does take the work, six and a half targets, and that interval excludes zero. The scoring never follows. Sixty percent of the gap the sell lists are pricing was already on the board the season before the pick was spent.
Read the receipts →Every sell-your-RB column this month explains decline the same way, fast-twitch erodes, the cuts get slower, move him at 27. I decomposed 1,826 seasons into the two things that actually make a fantasy total, how many chances a player got and what he did with each one. When a player falls off, 68 percent of the fall is chances he no longer gets. The legs are the smaller half of the story, and the archetype labels everyone argues about could not separate anybody.
Read the receipts →Every August you get handed a list of backups who signed somewhere new and are finally going to get their shot. I ran 2,725 seasons against the next one, holding the size of the player's role fixed. The guys who moved came back worse than the guys who stayed, in every single band. Bit parts who moved reached 10 points a game 2.7 percent of the time. Bit parts who stayed put hit 4.2.
Read the receipts →We tested 13 seasons of contracts against production looking for the contract-year effort bump. It is not there, at any position, under any way of cutting it. What is there is stranger and more useful, and it has an expiration date printed on it.
Read the receipts →Every injury column this month promises you a returning player carries a months-long efficiency dip. I measured 944 return-from-injury runs against matched healthy peers. Perfect knowledge of that dip is worth 0.115 points a game. What is worth something: a knee absence makes another knee absence next season 4.6 times likelier, and a foot absence produced zero repeats in 42 tries.
Read the receipts →Updated · see the follow-up →The July breakout-TE wave sells you draft capital, first-rounders hit and day-three guys do not. I ran ten years of tight end seasons. Round 3 posts the best next-year scoring of any round, and once you know a tight end's target share, a round 1 or 2 pedigree is worth 0.79 points a game.
Read the receipts →Every 2026 dynasty startup piece says pay up for QB age, the rookie-window scarcity math justifies it. I built a study designed to bust that claim. At the same pass-attempt volume, quarterbacks 25 and under beat their older same-volume peers by 0.7 to 1.4 points a game, in every band with enough sample to trust. The premium isn't overpriced. It's underpriced.
Read the receipts →Updated · see the follow-up →The dynasty market treats a new franchise QB as a free lift for his incumbent WR corps, and a downgrade as a death sentence for theirs. I found every clean offseason QB change since 2018 and graded the incumbent WR/TE against my own engine's projection. The upgrade side did not beat the downgrade side. If anything, it lost.
Read the receipts →Twitter wants you to sell James Cook before workhorse-year regression hits, and my own engine's spike detector agrees, docking his 2026 rate 3.2 points a game. Problem is his climb from 5.9 to 16.8 ppg is four straight up years, not one lucky one, and when I replayed the same regression trigger on 154 historical cases, it only earns its keep on real spikes.
Read the receipts →Coordinator usage cards are circulating again, lead-back rates and target spreads pitched as coach DNA you can draft on. I built concentration fingerprints for all 320 team-seasons from 2015-2024, followed 63 coordinators across team changes, and then tried to use coach history to project 2,505 player usage lines. The DNA mostly is not there. Who concentrates the touches belongs to the roster, not the playbook, with one exception worth knowing about.
Read the receipts →The sell-your-elite-WR-at-27 crowd says trade value peaks before production does, so get out now. I banked 434 WR Y0-to-Y+1 pairs at equal target share and found the gap between age 24-26 and age 27-29 receivers is inside the noise floor. Hold Jefferson. Hold Lamb. Hold Brown.
Read the receipts →Dynasty discourse prices WRs who command disproportionate air yards as forward-production value. Two years of data from 139 WR pairs says there is no edge -- at equal target share, the proportional-air-yards WR produced more next season than the deep-ball specialist. Gate 1 fails. Don't pay the premium.
Read the receipts →Dynasty consensus prices pass-catching RBs at a premium because route running ages well. At equal total touch volume, the data produces a +0.48 ppg advantage -- statistically noise. The premium appears only for bellcow receiving backs (McCaffrey territory), and turns negative for the committee-PPR backs most managers are actually evaluating.
Read the receipts →Dynasty consensus says avoid drop-prone receivers, but low catch rate at WR2 production predicted Y+1 breakout not bust in the data; Tee Higgins, Nico Collins, and DJ Moore lived this, and the engine had no term for it.
Read the receipts →Dynasty managers are paying up for TEs who own their team's target tree -- Brock Bowers is the only show in town for the Raiders, Trey McBride ran 694 routes and led every TE in targets last year. We tested whether that team-rank status predicts Y+1 production beyond raw target share. It does not. The aggregate looks real. Inside every comparable band it reverses.
Read the receipts →Elite college dominator plus a young breakout age plus a Day-3 draft slot is supposed to be the Puka Nacua tell. We checked every WR who fit the profile since 2017. The hit rate is identical to a random late-round flier, and the profile doesn't even flag Nacua.
Read the receipts →Dynasty managers are debating Puka Nacua versus Jaxon Smith-Njigba as the WR1 on the basis of efficiency. Both are right to chase the route-runner premium. But the premium only shows up when the volume backs it. At elite target shares, the high-catch-rate WR produces 2.2 more points a week the next year. Below 18 percent targets, the gap closes to nothing.
Read the receipts →Updated · see the follow-up →My own error report flagged the engine for over-rating speed-receiver types, so I tried to prove fast WRs crater. Ten years of pairs say the opposite: hold target share fixed and the 108-plus speed-score guys edge their slower peers by nine-tenths of a point a week the next year. The over-rating is real. It lives in the combine darlings who never earn the targets, not the ones who do.
Read the receipts →The group chat says draft the bell-cow because volume is king. Our own three-year receipts say the rented-volume tax is real in exactly one RB tier, the muddy middle, and reverses at the top. So we tested whether to fade workhorses. The data said no.
Read the receipts →My own error report flagged the engine for over-rating receivers age 23 and under, so I went hunting for the bust. Ten years of pairs say the opposite: hold target share fixed and the young receivers who earn a real role actually edge their older peers by four-tenths of a point a week. The over-rating is real. It just lives in the kids who never see the targets, not the ones who do.
Read the receipts →My own error report flagged the engine for over-rating pass-catching tight ends, and the takesmiths' favorite tell, touchdown regression, looked like the smoking gun. Ten years of pairs say the TD spike does regress hard off its own line. But hold target share fixed and those tight ends land right back with their peers, 0.37 points a week apart. Real regression, no consequence. The over-rating is real, it's just hiding somewhere else.
Read the receipts →Every player in your dynasty league has two numbers. The price the room will trade him at, and the value the math says he'll produce. I built one tool to measure each, and the spread between them is the whole product. VibeRank is live, public, and free.
Read the receipts →My own error report flagged the engine for over-rating big, athletic running backs, so I went looking for the age cliff that would explain it. Ten years of data say it isn't there. Post-peak, the freaks decline 0.16 points a game faster than the field, which is to say, not at all. The over-rating is real. The aging story isn't.
Read the receipts →Updated · see the follow-up →My engine's own outcome audit flagged pocket passers as its single biggest mistake, 57 ranks of under-rating. So I checked the tape. At equal production, rushing QBs repeat as top-12 fantasy QBs 53 percent of the time. Pocket passers, 34. The engine's rushing tilt was right. The flag was measuring something else.
Read the receipts →The smart-money reflex is that running back efficiency is luck and only volume sticks. Band 457 back-seasons by workload and the top-quartile yards-per-carry guys beat the field by 2.2 a game, and by 4 at a true workhorse load, then hold it the next year. The engine already bets on this. The "efficiency is noise" crowd is the one getting fooled.
Read the receipts →The regression crowd is right that a hot passing-TD rate comes down, about 2 points a game off the spike. They're wrong about what to do with it. At the same passing volume those QBs still beat the field by 1.9 a game the next year. The rate regresses. The quarterback doesn't.
Read the receipts →Updated · see the follow-up →Every air-yards column this month wants you to buy the deep threat. I ran ten years of it. At the same target share, the downfield guys score half a point a game LESS the next year, not more. The only place aDOT helps is the low-volume dart-throwers, where it's touchdown variance wearing a lab coat.
Read the receipts →Every dynasty column tells you tight ends fall off a cliff after 32. I paired ten years of TE seasons and controlled for role. At the same target share the over-32 guys lost 0.75 ppg, not a cliff, and in the aggregate they actually scored more. The cliff is survivorship plus job loss wearing an age costume.
Read the receipts →Every sleeper column this month tells you receiving work hands a back a safe PPR floor. On average it doesn't, those backs are a touch worse. Control for total workload and the three-down bell-cow beats the pure grinder by 3 points a game in the 250-touch tier. The floor talk is backwards. It's a high-volume edge.
Read the receipts →Every May the same article tells you year 3 is when the TE light comes on. Ten years of TE pairs say the within-band lift is under a point a game. The engine quietly priced it in months ago.
Read the receipts →Your favorite QB scribe will tell you 100-rush-attempt QBs hit top-12 PPG 91 percent of the time. They're off by 14 points. The real rate is 77. Still huge. Still bet-the-ranch. Just not what the takesmiths are quoting.
Read the receipts →Every August someone publishes a sophomore-WR breakout tier. We tested all 955 rookie-to-vet pairs against it. The "jump" is under a point a game. The engine's been pricing it correctly for months. You can stop forwarding the article.
Read the receipts →An open notebook for an engine that doesn't read FantasyPros. Here's what goes in, what doesn't, and why your group chat is wrong more often than it thinks.
Read the receipts →If you've read one post-draft dynasty piece this month, you've read the phrase "vacated backfield" forty times. Annoying. Also correct. Day-2 rookies inheriting an empty depth chart print +3.5 ppg.
Read the receipts →The dynasty crowd has been screaming "sell before 27" for years. Ten years of cohort pairs say they've been right the whole time. Your 300-touch alpha hits 27 and forfeits 3 ppg to his younger self.
Read the receipts →Updated · see the follow-up →The consensus this draft season: NFL draft capital is the signal for rookie receivers. Round 1 WRs do score more in year two, +2.25 a game. Then you control for how many targets they saw as rookies, and the entire edge evaporates. It was opportunity wearing a pedigree costume.
Read the receipts →