Consider a short piece of rope—could you guess which knots are more likely to form if you crumple and shake it? Synthetic chemists have long been working on a molecular version of this problem, and so far, have succeeded at synthesizing a half-dozen knot types using molecular self-assembling techniques. But which other knot types could be realised in the future? This is the challenging question that SISSA scientists, in association with the University of Padua, have tackled using computer simulations in this new work published in Nature Communications.