Methods · Embodied, practical & performance skills · Short guide

Failure-point prediction

Before performing, list the three most likely places you will go wrong and why, building the internal monitor coaching usually supplies from outside.

When to use this: When something feels wrong
Learning impact: Strong change
Learning evidence: Credible with caveats
Where you are with the topic: Checking readiness
How often to use it: Worth repeating often

When to use it

You are about to run a skill you usually mess up in the same stretch of the sequence.
A coach or teacher will not be watching, and you need your own early-warning list.
You want tomorrow’s practice to target slips, not a vague “try harder.”

What to ask AI

Name the skill or attempt, then list your top three likely failure points and why, before you perform.

Ask it to critique vagueness (“I’ll mess up” is not a prediction) and push until each failure point is observable.

After you perform, report what happened so you can compare hits and misses.

How it helps

A coach standing beside you often says “watch your elbow” before you even move.

Failure-point prediction builds that monitor inside you: before the attempt, you name where you usually slip and why.

It is a pre-mortem for the body. Attention goes to the dangerous stretch instead of floating over the whole skill.

When you later do a post-performance forensic, sharp predictions make the comparison useful, you learn which watch-outs were real and which were noise.

How to run it

Before the attempt, write your top three likely failures and one sentence of why for each.
Optional: paste them to AI and ask whether each is specific enough to notice mid-performance.
Perform with those three watch-outs on a mental sticky note, not a full lecture to yourself.
Soon after, tick which predictions hit, which missed, and what you failed to predict.
Update a personal failure catalogue so next practice starts sharper.

Example exchange

You

I’m about to try a left-hand arpeggio on the acoustic at mid-tempo. Before I play, here are my three most likely failure points, tell me if any are too vague to notice while I’m playing.

Send the three.

You

First, I’ll tense up. Second, timing. Third, wrong finger on the B string.

The first two you can’t catch mid-rep. Where does the tension show, and which beat drifts? Keep the finger one, that one’s already specific.

You

Revised: left wrist collapses toward the neck on the third string; I rush the second note of each group; ring finger lands on B instead of pinky.

Those you can feel. Play once, then tell me which fired, and name any failure you didn’t predict.

Copyable prompt

Skill/attempt: [ ].
I will list my top 3 likely failure points and why. Critique vagueness. After I
perform, I'll report what happened for forensic comparison.

The Tell

Here is how you know this method has flipped: you write predictions, then never check them against what actually happened, so the list becomes ritual, not a monitor.

In that moment the list (or the AI’s polish on it) did the thinking, and you only performed without learning.

Tighten the loop: predict, perform, tick hits and misses, then revise the catalogue before the next attempt.

Principle evidence

Strength of the underlying learning idea, not a claim about AI products.

The underlying learning idea is rated moderate. Pre-performance failure lists are a pre-mortem (prospective hindsight) applied to embodied skill: metacognitive monitoring that aims attention where slips usually happen. That focus is the principle; vague worry is not.

AI delivery evidence

Whether an AI tutor delivers this method well is a separate question.

AI can stress-test whether your predictions are specific enough to notice. How well it does that is speculative. The predictions, the attempt, and the hit/miss check stay yours.