Code review simulation
Have your code reviewed the way a senior engineer would review it, which is about judgement, not correctness, and is the thing courses don't teach.
Some ideas only click when you inhabit a situation, not when you summarise a paragraph.
Role-play a stakeholder. Decide with only period information. Map causes instead of chronology. Interrogate who wrote the source and what they left out.
Choose this family when facts are easy to list and hard to explain.
Have your code reviewed the way a senior engineer would review it, which is about judgement, not correctness, and is the thing courses don't teach.
Be examined out loud, with follow-up questions that respond to what you actually said, which is a completely different skill from writing, and is assessed without ever being taught.
Turn a sequence of events into a structure of causes, which is the difference between knowing what happened and being able to explain it.
Question a document about who made it, for whom, and what it isn't saying, which is what turns a source into evidence.
Decide using only what was knowable at the time, including period tools and values, so hindsight stops masquerading as historical understanding.
Handle a demanding client, patient, or user in simulation, practising translation, boundary-setting, and requirement extraction under pressure, not just solving the technical problem in a vacuum.
Explain the same idea to a novice, a peer, and a hostile expert in turn, adjusting grain and warrants without changing the truth-conditions of the claim.
Question a historical or expert persona about their reasoning, and, crucially, tell the difference between what they actually said and what has been plausibly reconstructed for you.
Read one event through several people's eyes at once, and find out that the disagreement between them is the thing you were supposed to be studying.
Take a sequence of events you already know and convert it into an explanation, forcing every "then" to earn a "because," "enabled," or "irrelevant."
Change one variable in a historical or systemic scenario and reason out what follows: training causal discipline instead of storytelling sprawl.
Talk through a procedure and its failure points before doing it, so the first real attempt isn't also your first time noticing the missing step.
Compare how interpretations of one event have changed, so you learn the argument about the past, not only a single school story presented as fact.
Imagine the project already failed and work backwards to why: catching plan holes while you can still change them.
Respond under time pressure to an unfolding situation: training prioritisation and decision hygiene when information is incomplete and the clock is moving.
Work a case where principles genuinely conflict, practising reasoned choice under tradeoffs instead of hunting for a hidden "correct" virtue.
Play the system or person at the moment it fails and narrate the cascade, so you learn failure as a process, not a moral punchline.
Hear (and then reproduce) a mechanism narrated step by step in time, so static diagrams become causal movies you can rerun.
Hear an idea explained as an insider, then as to an outsider, and name what knowledge was assumed in the insider version.
Argue from the position you find least sympathetic , long enough to feel its internal logic, without abandoning evaluation.
Build a world with someone, historical or invented, in which the concepts you're learning are things people live inside rather than statements you memorise.
Explore the sensory, material, and assumed world of a period, so historical (or literary) actors stop floating in blank space and start making sense as people under constraints.
Put a question to several differing expert personas and harvest the disagreement structure, before you reconcile or pick a side.
Move between molecular, human, and planetary (or other nested) framings of one phenomenon, so you stop being fluent at only one zoom level.
Work with the same expert or character across many sessions so prior context deepens, without letting continuity become uncritical trust.