The Apprenticeship Playbook
Learn a context-dependent skill when no one hands you the apprenticeship.
Generic advice teaches the average case; context-dependent skills only pay out in the specific case, so the 10,000 tips never fire when the moment arrives. This playbook manufactures the two things an apprenticeship gives you (proximity to a practitioner making real decisions, and a tight feedback loop) when you have no mentor, employer, or permission. Six plays run over about 12 weeks: cap inputs at 3 practitioners, reconstruct 1 real deal a week while logging your call before the outcome, then promote only the rules that survive into a checklist that is yours.
- Get proximity to one practitioner making real decisions, then build a tight feedback loop on those decisions. That is the whole mechanism.
- Learn from real deals with money or reputation on the line, not case studies, because only those carry the constraints generic advice strips out.
- Timestamp your call before the outcome is known. A rule you invent after seeing the result teaches nothing.
- Context-dependent skill is built by volume of corrected reps, not insight density: a mediocre loop every week beats a brilliant one every quarter.
- Name the one skill, in a testable sentenceWrite the skill as a sentence with a pass/fail outcome, not a topic. Weak: "get better at treasury management." Strong: "decide, within one trading day, whether a given mNAV premium is a buy, using a rule I can defend out loud." Budget 30 minutes.Pro tipIf you cannot write the pass/fail version, you do not yet know what you are trying to learn.WarningDo not skip this. Every hour after it is wasted if the target is still a topic.
- Find 3 practitioners, not 30 pieces of contentPick 3 people who currently do the real version of the skill and leave a public trail: trades they post, deals they narrate, code they ship, decisions they explain after the fact. Currently-doing beats famous. 3 is enough to triangulate and few enough to actually track.Pro tipA mid-tier operator posting live reasoning is worth more than a legend who only sells the polished story.WarningMore than 5 and you are back to consuming content.
- Reconstruct one real deal end to endTake one decision each of them actually made and rebuild it from the outside: what they knew at the time, the options, what they chose, what happened. Write down the decision you would have made before you know their outcome, then compare. Do 1 per week; after 6 weeks you have 6 reconstructed deals and a written record of where your instinct and theirs diverge.Pro tipThis is the apprenticeship substitute you can run alone: standing over their shoulder on a live deal, retroactively.
- Build the office-hours loopConvert the solo version into a real feedback loop, in order of preference: (1) a standing 30-minute weekly call with one practitioner, one live decision per session; (2) a peer running the same reps, two people each bringing one live decision a week; (3) a public post-mortem of your reconstructed deals with your pre-outcome call attached. The non-negotiable is the cadence and the real decision.Pro tipMost people say yes to "can I show you one real decision and get 10 minutes of reaction?" far more than to "will you mentor me?"WarningA brilliant mentor once a quarter loses to a mediocre one every week.
- Keep a decision journal, call timestamped before the outcomeFor every real decision you observe or make, log four lines before you know how it turned out: what I would do; why (the rule I am applying); what would prove me wrong; confidence (0 to 100). When the outcome lands, log what actually happened and which rules survived. Aim for 3 to 5 logged decisions a week.Pro tipThe timestamp is the point. A rule invented after seeing the outcome teaches nothing.
- Promote surviving rules into your own checklistEvery 2 to 3 weeks, read your journal and pull out the rules that keep being right. Those, and only those, go into your personal checklist, built from decisions you watched or made, in your specific context, with the failure cases attached.Pro tipYou start by rejecting other people's checklists as too generic; you end by writing one that is not.
An operator wants to "get better at treasury management." Applying Play 1, they rewrite it as a pass/fail sentence: "decide, within one trading day, whether a given mNAV premium is a buy, using a rule I can defend out loud."
Instead of asking a busy practitioner "will you mentor me?", the operator asks "can I show you one real decision I am wrestling with and get 10 minutes of reaction?" and brings a single live call, already reasoned out.
Built by Stratapedia on Naval Ravikant's idea that specific knowledge, the kind that cannot be trained or outsourced, is taught through apprenticeships, not schools (How to Get Rich). Naval supplies the diagnosis (why generic advice fails for context-dependent skills) and the one-word prescription (apprenticeship). The six plays are the Stratapedia operational build: how to run the apprenticeship loop when no one has handed you one.