The Lonely Runner Conjecture is a problem in mathematics about runners moving at different speeds around a circular track. Imagine several runners on a circular track of length 1, all starting at the same point. The conjecture says that at some moment in time, every runner will be far from the starting point, specifically at least a distance of 1/n away, where n is the total number of runners including a stationary one at the origin. A set of speeds is called "tight" if the best separation you can achieve equals exactly this predicted minimum of 1/n. The conjecture, still unproven in general, has been verified for small numbers of runners, and researchers study special families of tight speed sets to better understand its structure.
The paper focuses on a specific construction: take the simplest possible set of speeds, namely the integers from 1 to n-1, and swap out one of those speeds for a different value. Previous work by Goddyn and Wong showed that if you replace a speed r with an integer multiple of r, you can classify exactly which replacements keep the set tight. They also proved that non-multiple replacements can only be tight in finitely many cases, which hints at why two known exceptional tight sets involving the speed 2 replaced by an odd number seemed to have no cousins. The new paper makes this finiteness precise and computable, by carefully analyzing which portions of the circular track are left uncovered when one speed is removed, and calculating the exact lengths of those gaps.
For the specific case where speed 2 is replaced by an odd number, the authors give a complete answer: the only tight sets of this form correspond to n equal to 5 or 6, matching exactly the two known sporadic examples and confirming there are no others. They extend this approach to handle speed 3 as well. The paper also reports extensive computer searches using exact arithmetic, and points out a surprising counterexample to a natural guess that all speeds in a tight set should be smaller than twice the number of runners. A known example due to Goddyn and Wong, involving 32 runners with the speed 90 in the mix, shows this intuition fails.