The Fresh Ears Test
Judge revision 56 against revision 1, not against the hours you spent getting there
- Difficulty
- Easy
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 88%
Rubin's antidote to revision blindness has two mechanical parts. The first is a preserved baseline: always refer back to the thing you are revising, so iteration 56 can be laid directly against iteration one and actually compared, rather than judged against the memory of having worked hard. He warns that craft passes can just as easily make something worse as better, which is why the history of records is full of albums whose demos were better. The second part is engineered neutrality: leave the work and go do something genuinely hard in another area, so you exhaust a different part of your brain. Returning after that gives you a better chance of a close-to-neutral view. He is blunt that this does not save time; you run the hundred experiments regardless.
Origin
Extracted from The Tim Ferriss Show
Core principles
- 01Work time is not progress; more hours can make it worse as easily as better
- 02Every iteration is an experiment, and you may have been right at the beginning
- 03You cannot judge a thing you have seen a hundred times without stepping away
- 04There is no shortcut past the experiments even if you got it right first
- 05Neutrality is a state you engineer, not one you summon
How to run it
- 1
Preserve the baseline
Archive iteration one and every meaningful version along the way so a real comparison is possible later.
Pro tip Date and number versions so you can find the actual first draft, not an early-ish one.
Watch out Overwriting in place destroys the only reference point you have.
- 2
Stop equating hours with progress
Treat each pass as an experiment that may have made the work worse. Rubin notes most people equate work time with progress and that this is simply not the case.
- 3
Leave the work
Step away long enough that the familiarity fades. Familiarity, not the work itself, is what has gone blurry.
Watch out A short break is not enough when you have looked at something fifty times.
- 4
Exhaust a different part of your brain
Fill the gap with something demanding in another domain rather than resting, so the faculty you use to judge this work is genuinely reset.
Pro tip Work on something hard, deliberately not the piece in question.
- 5
Run the head-to-head
Return and compare the newest iteration directly against the first, asking only whether it is in fact better.
Pro tip Compare blind where you can, so version numbers do not bias you toward the newest.
Watch out Comparing the newest against the second-newest hides the drift entirely.
- 6
Act on the answer
If an earlier iteration wins, ship it. If nothing has improved for many passes, the experiments are telling you the piece is done.
Watch out Do not treat reverting as wasted work; the experiments were the only way to know.
In the wild
Ferriss describes revising short stories 47 or 56 times and admitting he is not convinced revision 56 is better than revision 49, while everything has become so familiar it is very hard to look at with fresh eyes. Rubin's answer is to refer back to the thing being revised and compare the latest against iteration one, because the craft passes could just as well be making it worse. He adds that there are no shortcuts: you do the hundred experiments regardless, even if you got it right first time, because there is no way to know otherwise.
→ The decision to stop revising becomes evidence-based rather than exhaustion-based.
Rubin points to the long history of records where people say the album did not really work but the demos were great, the songwriting was strong and the execution went wrong. The demos are the preserved iteration one; the released album is iteration fifty. Without keeping and comparing against the original, the team never noticed that iteration after iteration of craft had been degrading the thing.
→ A permanent public record of what happens when nobody compares the latest version to the first.
Common mistakes
Assuming time invested equals improvement
Rubin is explicit that iteration can make work worse as readily as better, and there is no rule that says otherwise. Hours are not evidence.
Resting instead of switching
The reset comes from exhausting a different part of your brain on something hard, not from idling near the work you are trying to see freshly.
Hunting for a shortcut past the experiments
Rubin says there is no way to save the time you are hoping to save. Trying to skip the iterations just leaves you without the evidence to judge.
Is it for you?
Best for
Anyone deep in iteration on a single piece: writers on draft 40, producers on mix 12, designers on round 9.
Not ideal for
Early exploratory work where no version is yet worth preserving or comparing.
From the transcript
“Most people equate work time with progress, and that's not the case. We're scientists experimenting, and it's all an experiment.”
“There's no way to save the time that you're hoping to save. There are no shortcuts.”
“Work on something hard, not that. And then when you come back to read it, you have a better chance of being closer to a…”
From the episode
#649: Rick Rubin, Legendary Producer — Timeless Methods for Unlocking Creativity, Secrets Hidden in Plain Sight, The Future with AI, Helpful Distractions, Working with Strong Personalities, Breaking Out of “The Sameness,” and More