Failing 15% of The Time is Optimal For Learning

Learning is fun for me. I went to a Montessori school in preschool, and it shaped my life because I was taught to learn through play. I still approach every new topic as play and don’t understand doing something any other way; and ever since I’ve had an obsession with learning in general.

Why do some people learn faster than others? How do they learn faster? Is it genetics? Is it skills? Can you get better at learning? How can you get better at learning? What is the optimal way to learn something new? These are all questions that have interested me since I was young.

What we know about learning is that to learn new things we must sometimes fail. Incremental improvement drives progress in the short term, which then compounds over the long-term; fail too much and you lose motivation. With that said, what’s the right amount of failure?

The 85% Rule for Optimal Learning

New research led by the University of Arizona Professor, Jonathan Cohen, proposes a mathematical answer to that question: The 85% Rule.

Jonathan Cohen wrote a paper titled The 85% Rule for Optimal Learning. Basically, when you’re trying to learn something new you want to make the difficulty of what you’re learning such that you’re getting things right about 85% of the time; you’re making errors about 15% of the time.

85%, how did they arrive at that number? To get that number, scientists trained machine-learning algorithms to recognize images of tumors at various levels of difficulty. They found that the algorithms learned most efficiently when the failure rate was about 15 percent.

You don’t improve by always being perfect. If that’s the case then you’re not really challenging yourself. So, make things hard, but not so hard that you’re failing half of your attempts.


Bottom line: If you’re always scoring 100%, you’re probably not learning anything new; the ‘sweet spot’ for learning is 85%.