Consumption Is Not Learning: A Process Protocol for Cutdown Week

Richard Feynman said the first principle is that you must not fool yourself — and you are the easiest person to fool. Fantasy football is a machine for that kind of fooling. Rankings, camp notes, and waiver columns feel like work. They are usually recognition, not retrieval, and recognition does not survive Thursday night lock.

scienceoffantasyfootball.com

The claim is simple and testable: Knowing a role is not the same as allocating a roster spot under a constraint.

The speed of improvement is limited by the speed of being scored on a hypothesis, not by the volume of inputs. You do not need a better forecast of every player’s points. You need a process that beats consensus at the moments that matter: the draft board, the Monday claim, the Tuesday FAAB bid, the Friday start/sit. Beat ADP and beat the market’s delayed reaction to role change. Perfect calibration is optional. Relative edge is not.

youtube.com

The lab error

Most managers run an uncontrolled observational study. Independent variable: hours of content.

Dependent variable: “I feel ready.”

Missing: a pre-registered decision, a leading indicator, and a written reason that can be graded after the fact. 

Sunday points are a lagging, high-variance score. Role is the leading indicator. 

The useful sequence is not narrative → ranking → hope. 

It is:

1. Snap share (is the player on the field?)

2. Route participation or touch opportunity (is the usage real?)

3. Target share/opportunity share (does the offense look his way?)

4. Fantasy points (the noisy outcome)

A wide receiver can post 18 points on two touchdowns and be a fade. Another can post 8 points on nine targets and 80% routes and be a hold. If you grade the first number, you are studying weather. If you grade the second, you are studying climate.

fantasyprojectionlab.com

Hypothesis

Managers who generate a role hypothesis before opening expert lists, then record a one-sentence reason and a kill criterion, will make fewer retail waiver bids and fewer identity-locked starts than managers who consume first and decide second. This is not a vibe. It is an experimental design.

• Treatment: struggle → decide → measure role → update the rule.

• Control: scroll → recognize names → copy the same five adds as the rest of the league.

The treatment shortens the feedback loop. The control outsources cognition and then wonders why the same mistakes recur in Week 6.

Method: a pre-registered weekly protocol

SCIOFF already treats pre-mortems as a way to refuse the story your brain wants. Use the same move in-season. 

Assume the decision failed. List why. Then act only if the remaining case still clears the cost.

scienceoffantasyfootball.com

A. Cutdown experiment (today through Monday)

The first 53-man sheet is not a role. Making the club as a gunner is not offensive opportunity. Monday NFL claims can insert a veteran over the backup you drafted as a handcuff. Confirm opportunity. Do not stamp every survivor as a stash.

footballnationusa.com

For each bench dart, write four answers before you read a tracker recap:

1. Did he make the 53?

2. Did the club add a same-position player after him?

3. Is the job offense, special teams, or both?

4. If the starter misses Week 1, is he next in line for touches or targets?

Failure modes to pre-register

• Role never existed; he was a coverage specialist.

• Waiver claim buried him Monday.

• Preseason usage did not survive regular-season personnel.

• You confused name recognition with snap share.

B. Waiver experiment (in-season Tuesdays)

Before any article:

• List three names.

• One sentence each: what role changed, what two-week path exists if the injury holds, what you drop if you are wrong.

• Assign a bid or priority now.

• Then read one source as a second opinion, not as a menu.

• 
  

Leading indicators worth recording: 60%+ snap share after a starter exit; rising route rate; target share that does not require a touchdown to look real. Box score without those is noise.

scoutcast.ai

C. Lineup experiment (lock)

Write the start/sit as a testable sentence.“I am sitting X because route share fell and the implied team pass volume is compressed, not because a recap said fade.”After the game, grade the reason. A correct process with a bad outcome is still data. A lucky touchdown with a dead role is not skill.

Retrieval, not recognition

Recognition is scrolling a board and thinking you know the player. Retrieval is producing the knowledge with the app closed. 

Five-minute drills:

• Week 1 starters and the first two off the bench.

• Handcuff for each running back you rostered.

• One player you would spend 15%+ of FAAB on if the starter’s role dies Sunday.

• Last week’s three process notes, from memory.

• 
  

If it will not come out of your head at 8:14 p.m. Thursday, it was never encoded. Cheat sheets are external memory. They atrophy the muscle you need when news hits inside the lock window.

Connect before you collect

A new idea earns a slot only if it attaches to a decision rule.

Idea Rule Action

Target share > box score

Do not drop volume for a two-catch touchdown

Hold / add the 8-target WR

Snap share is stickier than rushing yards

Bid the backup who played 62% after the exit

Spend FAAB on role, not on 11 carries

Special-teams survivor ≠ offense

No roster spot without routes or touches

Cut the gunner

You beat consensus, not God

Size the bid to multi-week starter-level role

Do not empty FAAB on Week 1 theater

If you cannot draw idea → rule → action, you are collecting cards.

Compounding residue

Learning compounds only if something persists besides the feeling that you “follow the league.

”One note per week, four lines:

1. Hypothesis I acted on

2. Evidence used (snaps, routes, injury, depth-chart move, price vs my forecast)

3. Result vs process (did the role show up?)

4. Rule I keep or kill

That note makes Week 6 cheaper than Week 1. It also surfaces bias: always fading rookies, always overpaying the first 20-point waiver hero, always starting “your guy” after the routes die. A season-end postmortem is the same instrument at larger scale. Run it while the memories are sharp. Next August you are not a new manager wearing last year’s logo.

What this week is for

Tonight after 6:00 p.m. ET: build your own “made the 53 / role still real?” list for your stashes. No recap first.Monday after NFL claims: update the list. Anyone buried by a veteran is no longer a stash. Then struggle-write three Week 1 watch names — not adds yet — who are one injury from volume. You will be worse at this than the articles for two weeks. That is the point of a lab. The article cannot sit in your chair at lock. The protocol can. Data over narrative. Process over inventory. Measure the role. Write the reason. Let Sunday score the experiment, not the story.