Practice becomes improvement when experience is reflected on, abstracted, tested, and used to shape the next attempt.

Summary

Kolbs is the loop that converts experience into insight, insight into a better experiment, and the next experiment into cleaner feedback.

In this wiki, Kolbs sits under Self-Regulation and links heavily with Marginal Gains. Marginal Gains chooses the improvement direction; Kolbs runs the improvement cycle.

Core Cycle

  1. Experience: do the thing and create real data.
  2. Reflection: inspect what happened without turning it into rumination.
  3. Abstraction: form a clear principle, diagnosis, or hypothesis.
  4. Experimentation: design the next attempt based on the hypothesis.

What Good Kolbs Produces

  • A clearer diagnosis.
  • A smaller next experiment.
  • Better awareness of mistakes and conditions.
  • A more accurate model of cause and effect.
  • A practical next action.

Root-Cause Prompt

When a Kolb feels vague, use one Red Team residue question:

Why did this happen, and what deeper condition made that possible?

Repeat only until the answer becomes actionable. The point is to move past the first explanation.

What Bad Kolbs Looks Like

FailureSignalRepair
RuminationThe reflection circles around feelings without changing the next attempt.Force a concrete hypothesis and experiment.
Too broadThe cycle tries to explain everything.Narrow the focus to one mistake, variable, or bottleneck.
No experienceReflection happens before real data exists.Run a small attempt first.
No abstractionThe learner writes observations but no principle.Ask what pattern or rule the experience suggests.
No experimentThe reflection ends with insight but no next action.Define the next attempt before closing the loop.

Relationship To Marginal Gains

Marginal Gains asks, “What is the next valuable 1% direction?”

Kolbs asks, “What did this attempt teach me, and what should I try next?”

Together:

  • Marginal Gains prevents random improvement.
  • Kolbs prevents passive reflection.
  • Marginal Gains prioritizes the target.
  • Kolbs improves the method.

What Kolbs Can Reflect On

Kolbs works best after a real attempt produced data. The attempt does not have to be dramatic. It only needs enough friction, emotion, result, or surprise to teach something.

Good Kolbs candidates:

  • a session that started slowly;
  • a session that went better than expected;
  • a technique that looked correct but did not work;
  • a moment where a shortcut appeared;
  • a repeated failure mode;
  • a meaningful conversation;
  • a workout that revealed a bottleneck;
  • a language session where attention rose or collapsed;
  • an agent workflow that produced a useful or bad result.

Weak Kolbs candidates:

  • abstract planning with no attempt;
  • venting without a next experiment;
  • reviewing too many variables at once;
  • reflecting on something already obvious;
  • forcing a long reflection when a quick note would do.

Priority 0 Reflection Examples

Use Kolbs to improve the way each Priority 0 area is practiced.

Priority 0 AreaExperience To Reflect OnUseful Kolbs QuestionPossible Next Experiment
Agentic EngineeringAn agent produced a hacky fix.What instruction, context, or verification gap allowed that?Give the agent a narrower ownership scope or require a specific verification step.
Agentic EngineeringA coding session moved quickly but felt messy.Which part was true speed, and which part was debt?Add one checkpoint before accepting the patch.
Agentic EngineeringA prompt worked well once.What made it work: role, context, examples, constraints, or output format?Save the reusable part as a prompt pattern.
Learning SystemsA BHS session became note copying.Where did Aim, Shoot, or Skin stop forcing useful thinking?Add one question that forces importance or relationship judgments.
Learning SystemsSIR felt familiar but weak.Was I reconstructing, or recognizing?Add interleaving or closed-book explanation before checking.
Learning SystemsInterleaving exposed a weak problem frame.What variable or relationship did I fail to account for?Reconstruct the problem from one changed angle before executing again.
Learning SystemsA technique felt right but results did not improve.What thinking was the technique supposed to trigger, and did it happen?Change the constraint, not the whole technique.
VietnameseImmersion became passive.What caused attention to drop: content, difficulty, energy, environment, or tool friction?Shorten the clip, change content, or add one noticing target.
VietnameseA video felt more comprehensible than usual.What made comprehension easier?Repeat the same content type or setup.
VietnameseGrammar study became abstract.Did the primer help me notice real input, or did it become rule memorization?Return to input and look for one pattern in context.
中文Chinese reactivation created guilt or distraction.Is this maintenance contact, or am I trying to reopen the full track?Limit Chinese to one phrase, sentence, or clip.
FitnessA workout was skipped.Was the blocker energy, setup friction, unclear plan, soreness, or avoidance?Prepare the first exercise or reduce the entry requirement.
FitnessA lift felt worse than expected.Was the problem strength, sleep, warm-up, technique, or programming?Change one variable next session.
FitnessTraining felt good and repeatable.What condition made initiation easier?Preserve that condition for the next session.
RelationshipsA message was delayed.What made the contact feel effortful?Send a smaller message earlier.
RelationshipsA conversation repaired something.What made the repair possible?Use the same opening move sooner next time.
RelationshipsA relationship felt neglected.Was the issue lack of care, lack of scheduling, or avoidance?Put one next touchpoint on the calendar.

When Kolbs Takes Too Long

Kolbs relies on self-awareness and metacognition at every step. When those are still developing, the cycle can feel slow, vague, or hard to focus. This is a self-limiting property: limited introspection means limited ability to identify what is worth reflecting on, which makes the cycle feel endless.

Common causes and their fixes:

ExperienceLikely causeFix
Reflecting feels tedious or confusingLow metacognitive awareness — unclear what counts as a meaningful event to reflect onTreat metacognition as a high-priority skill to develop separately. Awareness deepens with repeated sessions over 4–12 weeks.
Sessions take too long with no clear focusLimited ability to distinguish signal from noise in the experienceCap sessions at 30 minutes. Because the steps are sequential, limitations early in the cycle affect everything downstream. A capped session done consistently is more valuable than an exhaustive one done rarely.

The cap is practical, not a sign of failure. Speed and focus improve naturally as metacognition develops with repetition.

Full Kolbs vs Quick Note

Use a full Kolbs when the experience has emotional weight, repeated friction, or a meaningful performance lesson.

Use a quick note when the lesson is obvious and the next experiment is already clear.

SituationReflection Size
Repeated shortcutFull Kolbs
Strong emotional reactionFull Kolbs
Technique failureFull Kolbs
Relationship repair or conflictFull Kolbs
Minor logistical issueQuick note
Simple reminderQuick note
One-off low-stakes mistakeQuick note
Obvious next actionQuick note

How It Should Feel

Kolbs should feel like turning a messy experience into a better next experiment. The reflection should move from what happened, to why it happened, to what you will try differently.

Good signs:

  • the experience becomes more specific as you write;
  • emotions and triggers become usable information;
  • you identify a pattern, not just an event;
  • and the next experiment is smaller and clearer than the original problem.

Warning sign: Kolbs has become journaling-only when it records the experience but does not change the next attempt.

Open Questions

  • What situations deserve a full Kolbs cycle versus a quick note?
  • Should Kolbs cycles be stored in daily notes, outputs, or a dedicated practice log?