There’s a lie embedded in every practice routine, and it took me far too long to identify it.
The lie is this: if you repeat something enough times, it will become consistent.
It sounds self-evidently true. Practice makes perfect. Repetition builds muscle memory. Do it a thousand times and you’ll be able to do it in your sleep.
Except I had techniques that I’d repeated thousands of times and they were at about eighty-five to ninety percent consistency. Not terrible. Not great. Just permanently stuck in a zone where they worked most of the time but not reliably enough for performance.
And no amount of additional repetition moved the needle. Week after week of grinding the same technique, and the success rate hovered in the same range. Eighty-five percent. Ninety percent. Back to eighty-seven. Up to ninety-one. Down to eighty-eight.
The oscillation without progress was maddening. Until I learned about the 90% consistency rule and understood why repetition alone was never going to get me where I needed to go.
The Rule
“Art of Practice” stated it plainly: “In order to make a move or skill consistent, it’s necessary to understand that there’s a ninety percent rule to learning anything. The first ninety percent of learning it is going to be relatively quick compared to how much time it takes to learn it to a hundred percent.”
The first ninety percent comes fast. You pick up a new technique, struggle with it initially, improve rapidly through the early learning phase, and within a relatively short period you’re at roughly ninety percent proficiency. The technique works most of the time. The learning curve is steep and satisfying.
Then the last ten percent arrives. And everything changes.
“That very end, that last ten percent of a skill, can take often twice or as much as three times more time and energy than it took reaching the first ninety percent of learning it.”
Three times. The final ten percent of proficiency can require three times the investment of the first ninety percent. If reaching ninety percent took you three weeks, reaching one hundred percent might take nine additional weeks. For ten percentage points.
Why the Last Ten Percent Is So Stubborn
The reason is adaptation — or rather, the absence of it.
When you first learn a technique, everything is new. The movements are unfamiliar, the coordination is demanding, the cognitive load is high. Your brain perceives this as a significant challenge — a stress that triggers the adaptive response. It builds new neural pathways, strengthens existing ones, and you improve rapidly.
As you approach ninety percent proficiency, the technique is no longer stressful. Your brain has adapted to this level of difficulty. The movements are familiar, the coordination is manageable, the cognitive load has decreased. The challenge that once triggered rapid adaptation now triggers… nothing. The technique is inside your comfort zone.
The rubber band analogy explains this perfectly: when you first attempt a new technique, the difficulty is like a stretched rubber band pulling you toward progress. As you get closer to mastery, the rubber band relaxes. By the time you’re at ninety percent, there’s almost no tension left. No tension means no pull. No pull means no progress.
Repeating a ninety-percent technique doesn’t restore the tension. It just travels the same relaxed rubber band over and over. The brain sees no reason to adapt further, because the current level isn’t stressful enough to trigger the adaptive mechanism.
The Counterintuitive Solution
The solution isn’t to practice the technique harder or longer. It’s to leave it at ninety percent and move to something harder.
No matter how many people believe that the way to make a skill consistent is by practicing it over and over, consistency can only occur by going beyond the difficulty level of it.
Going beyond. Not drilling deeper into the same level, but climbing to a higher one.
The logic, once you understand the adaptation mechanism, is elegant. A harder technique requires all of the skills used in the easier technique, plus additional skills. By practicing the harder technique, you’re engaging the easier skills at maximum capacity as a subset of the harder challenge. This re-stretches the rubber band for the easier technique, because the context has changed — it’s now being performed under the stress of a more demanding overall task.
The somersault analogy from the book made this concrete: “If someone is able to do a somersault, and after learning it, he actually starts to train a double somersault, would you believe that he will lose the ability to do a single one? Of course not.”
Of course not. The double somersault contains the single somersault. Training the double automatically maintains and strengthens the single. And the additional difficulty creates the adaptation pressure that the single somersault alone no longer provides.
The Weightlifting Parallel
This is where the weight-room analogy becomes impossible to ignore.
If your current strength level lets you lift thirty kilograms, and you want to get stronger, your strategy cannot be to lift thirty kilograms over and over. That maintains your current strength. It doesn’t build new strength.
To build new strength, you need to lift more than thirty. Thirty-five. Forty. Enough to stress the muscles beyond their current capacity, triggering the adaptive response that makes them stronger.
If your current skill level is equal to lifting thirty kilograms, and your strategy is to lift thirty kilograms over and over again, you will hit a limit. You won’t see anywhere close to the improvement you’d get if you took forty kilograms and tried to adapt yourself to that weight.
The parallel to card technique is exact. If your current sleight is at ninety percent, repeating it is like lifting the same weight. You need to attempt a harder sleight — one that’s ten to fifteen percent beyond your current level — to create the stress that drives improvement.
And here’s the kicker: when you go back to the original sleight after working on the harder one, the ninety percent has often become ninety-five or ninety-eight. Without directly practicing it. The harder work lifted the floor underneath it.
What This Meant for My Practice
When I first applied the 90% rule, I identified three techniques that had been stuck at approximately ninety percent for weeks. My instinct had been to keep grinding them. The rule said: leave them. Move on.
I selected three harder techniques — each one roughly ten to fifteen percent more difficult than its corresponding stuck technique. Same general skill category, but more demanding in speed, complexity, or precision.
For two weeks, I ignored the stuck techniques entirely. All my deep end practice time went to the harder techniques. I expected the stuck techniques to deteriorate. I expected to come back to them and find they’d slipped to eighty or seventy-five percent.
When I tested them after two weeks, two of the three had improved. One went from ninety percent to ninety-four percent. Another from eighty-eight to ninety-two. The third was unchanged at about ninety percent — not worse, just the same.
No direct practice. Measurable improvement. The 90% rule worked exactly as described.
Why We Resist This
The resistance to leaving a skill at ninety percent is deep and powerful.
It feels irresponsible. Like leaving a job half-finished. Like building a house and stopping just before the roof is complete. Every instinct says: finish what you started before moving on.
This instinct serves us well in many contexts. In consulting, you don’t deliver ninety percent of a strategy document. You deliver one hundred percent. In relationships, you don’t commit ninety percent. You commit fully.
But practice isn’t a deliverable or a relationship. Practice is an adaptive process, and adaptive processes have different rules. In an adaptive process, pursuing perfection at the current level can actually prevent reaching the next level. The grinding becomes a trap, consuming resources that should be invested in the harder challenges that drive real growth.
Leaving a technique at ninety percent isn’t abandoning it. It’s investing in its future improvement by creating the adaptation pressure that direct practice can no longer provide.
The Ninety Percent Audit
I now do regular audits of my practice material. Any technique that’s at ninety percent or above gets flagged for advancement. The question isn’t “Is this technique good enough?” The question is “Is practicing this technique still driving improvement, or has it crossed the adaptation threshold?”
If it’s crossed the threshold — if repeating it no longer creates the stress needed for growth — it moves to the maintenance category. I still practice it, but only in the low-energy tail of my sessions. The high-energy core goes to the next level up.
This audit prevents the routine prison from forming. It ensures that my practice is always weighted toward material that drives adaptation, not material that reinforces the status quo.
The 90% rule isn’t about settling for ninety percent. It’s about understanding that the path to one hundred percent often runs through the next level, not through the current one. You don’t grind your way to perfection. You climb your way there, and the lower rungs get stronger as you ascend.