2026 Fantasy Football Lab is Open for Business
The episode covers three core themes: filtering preseason noise, live draft decision-making across multiple leagues, and AFC player what-if analysis using Dennis's forecast/upside MVP framework. John and Dennis walk through real draft picks in real time, demonstrating their shared decision-tree process and metric-driven approach to cost versus upside.
Overreacting to one preseason game is the primary warning — one game should represent no more than ~1% of total data weighting. 1
Fight-or-flight instinct drives poor fantasy decisions; evolutionary bias toward pattern recognition causes managers to overvalue small sample sizes. 23
Filtering framework: Trust analysts who provide factual, actionable breakdowns (e.g., Greg Kellogg) over entertainers chasing clicks or last week's news. 45
Cost is the key variable — being wrong on a player at the right price is far less damaging than being right on a player at the wrong price. 6
Pick 1.5 (JSN vs. Cook): Chose JSN for safety and WR premium value; noted Puka Nacua's combined injury + suspension risk as a reason to avoid. 78 Cook ranked RB3, but WR scarcity at the top justified the WR-first approach. 9
Pick 1.2 (Cook vs. Bowers): Chose Cook, targeting McBride or A-Shane on the turn as the optimal pairing. 1011
Third-round start (A-Shane + Sun God): Sun God fell to 1.09; stacking with Goff as QB2 noted as a best-ball upside play. 1213
Mid-draft (Egbuka, Deebo Samuel): Egbuka taken as WR protection; Deebo selected in round 10 as an early-season contributor with Kittle's uncertain health creating extra value. 141516
Late-round RB strategy: Target Zach Charbonnet as a late-season ramp-up play in Seattle as Derrick Henry wears down; note Woody Marks as an alternative. 1718
WR-heavy early opens a path to 3 strong RBs in rounds 3–5 (Kyron Williams, Scatabo, Swift/Montgomery/Bucky). 1920
Stacking preferred in best ball over redraft; redraft stacks risk a lean lineup if the stack fails. 21
Adapt or Die / Better Than Zero: Take long-shot value late — hitting one of three 30-to-1 RB dart throws (Singleton, Katron Allen, Claiborne) can be season-defining. 222324
Bankroll management parallel: Bet the minimum on long shots across many entries; law of large numbers collects the edge over time. 2526
Decision tree process: Before each pick, assess whether the roster needs safety or upside; text-based collaboration between John and Dennis resolves close calls. 2728
A-Shane (De'Von Achane): Forecast 120 MVP (down from 131 last two seasons) due to Miami's offensive uncertainty; floor still RB10, upside top 3. 293031
James Cook: Forecast 115; upside 135–150 if passing volume increases; 5–10% chance of overall RB1 if touchdowns also rise. 3233
Travion Henderson: Conservative forecast 75 (RB21); upside 120 if second-year leap (~45% avg improvement) materializes or Stevenson misses time. 3334
Travis Kelce: Floor TE8; cost TE11 — value play given safety profile; more under-center usage in 2025 sets up play-action routes he excels at. 353637
Mark Andrews: Upside 90 MVP but health/age risk; pair with a safer TE2 if drafting him. 38
Cam Ward gains credibility from Brian Daboll (OC) — the coach who turned Josh Allen around — now running the offense. 39
Carnell Tate (WR, Ohio State, top-10 pick): Forecast 70 MVP (flex-worthy); upside of top-12 WR if he hits like a rookie Jamar Chase; cost WR30. 4041
Nicholas Singleton: Target if Pollard (historically under 80 MVP) is replaced; new regime pattern strongly favors the incoming RB. 4142
Exposure limit: Don't stack Tate + Robinson + Ward on the same roster; spread Tennessee shares across teams. 43
Dion Burks (Colts): 7th-round pick; waiver-wire watch only — activates if Josh Downs and Alec Pierce both miss time. 444546
Aidan Mendoza (Raiders): Third QB in Superflex; went 10-of-16, 97 yards, 1 TD in preseason Week 1; new regime may start him earlier than Week 6–7 consensus. 474849
Cyrus Allen (Chiefs): Rising from undrafted to ADP ~194 on FFPC after Mahomes praise — approaching the price point where track record matters. 50
John: Publish Game Script Predictive Index (GSPI) season-long test; pre-season week 1 showed 60–70% directional accuracy. 2651
Dennis: Write a decision tree process article tying together Adapt or Die and Better Than Zero frameworks. 5253
Both: Continue background testing of new metrics (Boom Score, CSI Index, MAG, High-Low) across ~50 leagues. 54