Mathematicians have studied knots for centuries, but a new material is showing why some knots are better than others.

One sunny day last summer, Mathias Kolle, a professor at the Massachusetts Institute of Technology, took a couple of eminent colleagues out sailing. They talked about their research. They had some drinks. Then Kolle noticed something was off: A rowboat tied to his boat had come loose and was drifting toward the horizon. As he tacked across the water to retrieve the wayward vessel, he realized his mistake. In securing the rowboat, he must have tied the knot wrong.

“I almost lost a boat because I got one knot wrong,” said Kolle, a mechanical engineer. “That was pretty embarrassing.”

This slip-up aside, Kolle has become quite the knot wonk. In a recent paper in Science, he and his colleagues used a new way of visualizing the forces inside tangled fibers to revisit an ancient question: What makes some knots stronger than others?

Scientists have a long-standing fascination with knots. More than 150 years ago, Lord Kelvin — working with fellow Scottish scholar Peter Guthrie Tait — proposed that the chemical elements could be represented by different knots. The theory didn’t pan out, but the diagrams they drew of different knots, and their attempts to classify them, jump-started the development of modern knot theory.

In the 20th century, researchers built on this legacy by developing mathematical descriptions of knots that distinguish one from another. Often these descriptions employ topological properties: simple, countable characteristics that don’t depend on size or shape, such as how often strings in a knot cross.

The mathematics of theoretical knots tied in theoretical strings inspired biologists to investigate how real DNA and proteins twist and tangle. Scientists have also developed theoretical models for knots at larger scales, like the hitches that bind ropes to poles. Some have put their models to the test, using titanium wire to determine how much force is needed to pull a knot tight, or using fishing line or strands of spaghetti to explore what parts of a knot tend to break.

“It’s a creative art in my mind, being able to develop an experiment that will capture these properties,” said Ken Millett, a knot theory pioneer at the University of California, Santa Barbara.

But these experiments all tend to have the same limitation — one that makes it difficult for researchers to truly understand how everyday knots operate, said Jörn Dunkel, a mathematician at MIT.