Bracket the Error
Test opposite corrections, then split the difference between them
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 4
- Confidence
- 92%
During Hawk's successful 900 attempt, repeated forward falls revealed a directional problem rather than a reason to repeat unchanged. He shifted his weight toward the back foot and then fell backward. Those opposite misses bracketed the workable position, leading him to split the difference before landing the next attempt. The reusable method is a simple calibration loop: observe the direction of an error, alter one variable deliberately, check whether the result crosses to the opposite side, and test a midpoint. It works because controlled contrast supplies more information than identical retries. The method should be used only where attempts are safe enough to repeat and the outcome responds clearly to the chosen variable; Hawk's high-risk stunt is the origin story, not a general safety standard.
Origin
Extracted from The Diary of a CEO
Core principles
- 01A repeated error contains directional information
- 02Changing one variable makes feedback interpretable
- 03Opposite misses can bracket a workable setting
- 04Small controlled corrections beat blind repetition
How to run it
- 1
Define the directional miss
Observe whether the result consistently lands above, below, ahead, behind, or otherwise to one side of the target. Do not change anything yet.
Pro tip Use a visible or measurable outcome rather than a vague sense that the attempt was wrong.
Watch out Do not repeat a dangerous attempt merely to collect more feedback.
- 2
Choose one control
Identify the single variable most plausibly connected to the directional error. Hold the remaining controllable factors as steady as possible.
Watch out Changing several variables at once makes the result hard to interpret.
- 3
Test the opposite correction
Move the selected variable far enough to see whether the miss changes direction. Record the new result rather than judging the attempt only as success or failure.
Pro tip A miss on the opposite side can be useful evidence that the chosen variable matters.
- 4
Split the difference
If the correction brackets the target with opposite misses, choose a midpoint between the two settings. Retest and refine with smaller adjustments if needed.
Watch out A midpoint is not justified when the system's response is erratic or non-directional.
In the wild
At the 1999 X Games, Hawk fell forward on a 900 attempt without being injured. He then shifted his weight backward and fell backward. Recognising that the target lay between those two positions, he split the difference on the next attempt.
→ He landed the 900 after years of intermittent attempts.
A presenter consistently finishes five minutes late, then cuts too much material and finishes five minutes early. They restore roughly half of what was removed and rehearse again while keeping speaking speed stable.
→ The opposite timing misses provide a bounded next setting to test.
Common mistakes
Repeating without changing anything
Identical retries reproduce the error without testing what causes its direction.
Changing every variable
Multiple simultaneous corrections prevent the practitioner from knowing which change produced the new result.
Ignoring unsafe trial costs
The information from another attempt is not worth uncontrolled physical, financial, or operational danger.
Is it for you?
Best for
Repeatable tasks where one adjustable variable strongly influences a visible outcome.
Not ideal for
It is not suitable for unsafe trials, noisy systems with no interpretable feedback, or problems driven by many changing variables.
From the transcript
“Why not shift your weight to your back foot during the spin?”
“I fell backwards, and that was the epiphany.”
“All I have to do is split the difference.”
From the episode
Tony Hawk: The Man With The $1.4 Billion Name! Burnout, Obsession & Regrets