Running Small Experiments

Building with Adaptable Discipline is an iterative process. You are rarely handed the right design in advance. More often, you detect a pattern, form a hypothesis, make a change, and then observe what the system actually does.

That is not a weakness in the framework. It is part of how the framework works.

Why Experimentation Matters

A practice can fail for several reasons that look similar from the outside. What feels like a motivation problem may be friction. What looks like inconsistency may be a capacity mismatch. What feels like laziness may really be weak purpose, missing tools, or drift gaining leverage through a channel you have not yet identified.

That is why experimentation matters. It helps you move from vague suspicion to better evidence.

A Hypothesis Is A Working Explanation

In this framework, a hypothesis is a working explanation for what is happening in the system.

It might sound like:

• friction hypothesis: "The return keeps failing because the setup cost is too high."
• capacity hypothesis: "The system only works when I have more energy than I usually have."
• purpose hypothesis: "The practice keeps collapsing because the direction is no longer clear enough to justify the return."
• mindset hypothesis: "The move back gets delayed because each miss turns into proof."

A hypothesis does not need to be perfect. It only needs to be clear enough to guide the next useful test.

What An Experiment Looks Like

An experiment is a deliberate change made to see whether the hypothesis is pointing at the real constraint.

That might mean:

• lowering the number of steps before action
• shrinking the return for one week
• externalizing the next step instead of holding it in working memory
• changing a metric
• protecting one boundary to see whether the practice becomes more coherent

The point is not to randomize your life. The point is to make a meaningful change that lets you observe whether the system behaves differently.

You May Drift More Before It Gets Clearer

Sometimes experimentation creates a temporary increase in uncertainty. You may see more drift, not less, while you are learning what the real bottleneck is. That does not mean the process is broken. It often means the system is becoming more visible.

As long as the observation stays clear and the hypothesis stays falsifiable, that temporary messiness can be worth it. The goal is not immediate neatness. The goal is to find the root cause, or at least the change with the most leverage.

What To Watch During The Experiment

An experiment is useful when it gives you better information. While the change is active, pay attention to:

• whether return gets cheaper or clearer
• whether comeback speed changes
• whether the system holds better under harder conditions
• whether a different bottleneck becomes visible
• whether the intervention solved one problem by creating another

This is where observation matters as much as intervention.

How To Read The Result

Not every useful experiment ends with a clean success or failure. Sometimes the result is more diagnostic than that.

• the hypothesis was mostly right: return got cheaper in the place you expected, and the system held better than before
• the problem moved: one bottleneck improved, but a different one became visible
• the hypothesis was partial: the change helped, but only under good conditions, which means another constraint is still active
• the hypothesis was wrong: the intervention changed very little, which suggests you were solving the wrong problem

This matters because experimentation is not only about finding the winning move. It is also about getting less confused about what the system is actually doing.

A Simple Example

Suppose a writing practice keeps collapsing after three good days. You form a friction hypothesis: the cost of re-entry is too high. So you end each session by leaving the next sentence and the next subsection waiting in the document.

If that change makes the fourth day easier, the hypothesis was useful. If the return is still delayed, but now the real problem looks more like shame after the first miss, the experiment still helped. It revealed that the main bottleneck was not only friction. The system became clearer.

The Return On Iteration

When experimentation is done well, the payoff compounds. You stop making the same vague guesses. You get better at seeing the real structure of a failure. You become more likely to find the change with the most leverage instead of the change that feels urgent but sits on the wrong bottleneck.

That is one of the deeper returns of building with Adaptable Discipline. Over time, you are not just improving one practice. You are improving your ability to understand, redesign, and stabilize systems under real conditions.

Try it: Form a hypothesis for a change you're about to make

Pick a change you've been considering for a practice that isn't holding.

1. Name the diagnosis. What kind of failure is this? One sentence, as specific as possible.
2. Write the hypothesis. "If I [specific change], then [specific outcome] should improve, because [the constraint this targets]." Keep it falsifiable — you need to be able to tell whether it worked.
3. Name what you'll watch. What would count as the change working? Cheaper re-entry, faster return after a miss, less friction on the first step, comeback speed improving? Pick one signal.
4. Set a time window. How long will you run this before evaluating? A week is usually enough to see a pattern.

You're done when you have a hypothesis you could prove wrong.

Where this leads: Knowing If It's Working shows how to read the result once the window closes.