The 2026 Science of Fantasy Football Draft Lab preseason guide

The 2026 Science of Fantasy Football Draft Lab preseason guide: mindset, risk, positions, COM rankings, and the articles that apply.

Home: https://www.scienceoffantasyfootball.com/

Lab intro: Introduction to the Science of Fantasy Football Laboratory

This is a working document for draft week. It is not a projection sheet. It maps the Science of Fantasy Football preseason work: how not to fool yourself, where players fail, which positions are scarce, and how the lab’s COM rankings (consensus/composite of metrics) should sit next to ADP instead of replacing it.

1. The first rule: do not draft a story

Richard Feynman’s line is the lab’s draft motto: you must not fool yourself, and you are the easiest person to fool.

Article: Richard Feynman gave science its sharpest self-warning.

Before any reach or fade, write three sentences:

What measurable evidence makes this player better or worse than ADP?
What is the strongest counter-argument, and why am I dismissing it?
If this pick fails, what is the most likely reason?

That last question is the pre-mortem. Do it before the pick, not after Week 6.

Pre-mortems: 2026 Pre-Mortem Tables UpdatedThe tables are living documents for the top 24 at QB, RB, WR, and TE. They assume the pick already failed and list why: target-share / TD regression, the Curse of 350 for workhorse RBs, committees, schedule, usage contingencies. Re-run them after camp news and again around Week 4. Also read: The Winner’s Curse in Snake Redraft Drafts.

A reach versus ADP is an implicit overbid. Mid-rounds (roughly ADP 50–120 in a 12-team league) are peak curse territory because role ambiguity is highest and estimate dispersion is widest. Championship contributors rarely come from that dead zone as often as the board pretends.

Patience article: Delayed decision-making in fantasy football.

Waiting is +EV when the information is real (inactives, camp usage, depth-chart clarity). Waiting is −EV when it is just fear of being wrong.AI vs process: Professor vs AI: Future is here?

Generic AI is fast. Consistency metrics plus value versus ADP is still how you beat a room that all asked the same model the same question.

2. Risk first: AMG and the players who age out of their ADP

AMG (Age and Missed Games) is the lab’s decline-risk index. It mixes age, position risk, career volume, and missed games.

Validated grouping (next-season FPG, pooled): Low AMG ~16.5, Mid ~12.3, High ~10.1. The gap is largest at RB and WR. Young, low-miss, low-volume players cluster Low. Old workhorses and high-miss veterans cluster High.

Use AMG as a tax, not a veto. A High-AMG RB can still be drafted — just do not pay the same price as a Low-AMG peer with the same projection.

Historical RB context: 5 years of FFPC ADP data on the top 24 RBs. Top-24 ADP RBs have averaged roughly 15–16.5 PPR FPG over five seasons. The group mean is pulled down by injuries and committees. The elite still print 18–22. The pack does not.

Consistency career arcs (Expo posters):

3. Running backs: volume inheritance and handcuffs

Article: The Science of Handcuffs: A Data-Driven Guide to the 2026 RB Landscape

The finding that matters: true volume inheritance dominates handcuff value. Bell-cows outproduce committee leaders by a meaningful FPG gap. Own-handcuff replaces your lost starter. Other-team handcuff is a lottery ticket on someone else’s injury.

Elite clear handcuffs (Index 70+): Blake Corum (LAR) leads. Then Kyle Monangai (CHI), Tyrone Tracy (NYG), MarShawn Lloyd (GB).Strong committee/hybrid (50–70): Tank Bigsby, TreVeyon Henderson, Chris Rodriguez, Ray Davis, Isiah Pacheco.

In 12-team PPR, the best risk/reward band is usually rounds 13–15. Paying rounds 6–12 for Corum / Monangai / Henderson is justified only after you already have early-round RB production. Pre-mortem RB themes to keep on the board: Curse of 350, committees, replacement risk.

4. Wide receivers: the 2026 WR1 trinity is not one player three times

Article: 2026 WR1 Trinity — Ja’Marr Chase, Puka Nacua, and Jaxon Smith-Njigba.

2025 baselines were historic. The 2026 question is path, not fame.

Chase: volume floor and script upside in a pass-first offense. OL and Burrow risk are the pre-mortem. Safest volume bet if the price is at ADP.
Nacua: efficiency + scheme + OL + supporting cast. Highest ceiling in many models.
JSN: target monopoly and continuity; slightly more script/pace risk. Best value if he slips.

Pre-mortem WR themes: target-share / TD regression and playoff SOS.

5. Tight ends: premium scoring changes the roster math

In TE-premium / FFPC-style (1 TE + 2 flex, deep benches), simulations point to 3–4 TEs on a 20-spot roster. Do not force a TE in Round 1 unless an elite option is actually falling.

The top tier (McBride, Bowers, then Loveland / Warren / Kraft) is a real gap versus replacement. Kraft’s ACL recovery is the mid-round value story if Week 1 availability holds. In 1QB standard PPR, wait more often. In TEP, the TE slot plus flex is a weekly scoring weapon, not a streaming afterthought.

6. COM rankings section

What COM is in this lab: COM here means the Counts Over Median

Dennis Michelsen and the lab use it instead of a single-year points leaderboard. It is not “who scored the most last year.”

It is a stacked rating of:

MVP Index — weekly outcomes versus a historical positional median: helped you win, kept you in the game, or lost you the week.
Why the MVP Index Works

Folder: DMIC’s MVP Index background

BEST Index — best-ball tilt toward spike weeks rather than weekly floor.
Weekly Value / current-season grade — form, weighted more as the season grows.
Back-class / multi-year MVP — whether last year was a new level or a spike on an old baseline.

In-season example of that composite in action:

Consensus Weekly Value Metric After Week Five: How the same brain ranks leaders midseason:

Leaders of the Pack After Week Seven

At the Half in Fantasy Football

Draft-strategy companion (when to take QB/TE vs stacking RB/WR):

Better Than Zero Strategy for 2024,

Part One

Psychology piece explicitly groups

Consistency Rankings with VBD, Better Than Zero, opportunity indices, and pre-mortems:

Psychology of Fantasy Football Drafting

Homepage catch-all: My Key Draft Points in one place

Index discussion — how to use the board, not worship it

Index is a process ranking. ADP is a market ranking. The edge is the gap.

If COM (MVP floor + spike profile + AMG + role) likes a player above ADP, that is a target — but only after the pre-mortem still looks livable.
If ADP loves a player and COM does not (High AMG, thin spike history, committee, regression flags), that is a fade or a “only at a discount” name.
Index will look “too low” on last year’s narrative stars and “too high” on boring high-floor players. That is the point. Winning weeks are median-plus spikes, not brand.
Do not collapse Index into one number and then ignore format.
BEST-weighted index is for best ball. MVP-weighted COM is for weekly lineups. TEP COM should promote McBride / Bowers / the next tier earlier than standard PPR.
COM without ADP is a preference list. COM versus ADP is a draft plan.
Update COM when role changes. Do not update COM because a player “looked good in preseason Week 3.” Most starters sat. See GSPI notes below.

COM working board (template — drop your personal ranks in the blank column)

Use this table in your notes.

Fill in My COM yourself; the lab articles above are the ingredients, not a stolen expert sheet.

Pos

Market cluster heading into 2026

COM questions to score

My COM rank

ADP / notes

Chase / Nacua / JSN trinity

Volume path vs scheme/ceiling vs target monopoly; OL and QB risk

Trinity article

Bijan / Gibbs / workhorse tier

AMG + Curse of 350 + committee + handcuff quality

AMG + Handcuffs

McBride / Bowers, then Loveland–Warren–Kraft

TEP gap vs replacement; Kraft health

TE Premium + Optimum TE

Dual-threat early, then wait in 1QB

Opportunity cost of the reach; consistency of top-3 vs the pack

Better Than Zero + Pre-mortem QB

If you already have a finished COM list, paste it under this heading and keep the article links above it so the ranks stay attached to the method.

7. Last-look preseason: GSPI, not box scores

Starters were mostly held out. Use the script and who actually played, not preseason fantasy points.

GSPI is a pass/rush multiplier for game script. Week 3 was a rest week before the 90-to-53 cut. Chase backup RBs on positive-script teams and bubble WRs/TEs in games where starters sat. Do not promote a locked starter because he did not play, and do not bury a locked starter because his backup looked fine.

Podcast for the latest lab talk:

SCIOFF Podcast 290

8. Draft-week checklist

Write the Feynman note on every pick two rounds early.
Open the pre-mortem tables for that position.
Check AMG tier. Tax High-AMG names.
Compare COM (MVP / BEST / back-class / role) to ADP. Draft the gap, not the brand.
If you took an early workhorse, queue the matching handcuff in the 13–15 band unless the elite cuff is clearly worth earlier capital.
In TEP, get one of the top-tier TEs if the board offers a clean price; roster 3–4 TEs in deep formats.
Ignore most Week 3 preseason box scores. Use GSPI only for backups and bubble players.
Revisit COM after the first real injury or role change — not after one camp tweet.

That is the lab in one paste: probability over narrative, COM over vibes, ADP as the market you are trying to beat