The 50-Year-Old Sunflower Puzzle in Maths Has a Proof — Posted, Awaiting Checking
Some maths problems sound like games. This one is about sunflowers — not the flowers in a garden, but a pattern made of sets. And after about 50 years, three mathematicians say they have proved it.
Matija Bucić, Fankang He and Hyunwoo Lee posted their paper, "Proof of the Erdős-Szemerédi sunflower conjecture", on arXiv on 19 October 2026. It is a preprint: it has been shared publicly so other mathematicians can read and check it, but it has not yet been through peer review. That distinction matters, and this article keeps it in view throughout.
What is a sunflower in maths?
Take a collection of sets — groups of numbers, for example. A "sunflower" is a sub-collection where every pair of sets shares exactly the same overlap, called the core, and outside that core the sets do not overlap each other at all. Picture a flower: the shared core is the centre, and each set contributes its own separate petals.
Paul Erdős and Endre Szemerédi conjectured in the 1970s that if a collection of subsets of the numbers 1 to n contains no sunflower with k petals, then the collection cannot be too large. Precisely, for any n and k with n ≥ k − 1 ≥ 2, such a sunflower-free collection has size at most c raised to the power n, for some number c smaller than 2. Since there are 2 to the power n subsets altogether, that would mean sunflower-free families are exponentially smaller than everything possible.
The case k = 3 was already known to follow from earlier work on the cap-set problem, by Croot, Lev and Pach and by Ellenberg and Gijswijt, through a reduction by Alon, Shpilka and Umans. What the new paper claims is the conjecture for every k ≥ 3 — the full statement, not just the three-petal case.
Why it matters
Sunflowers appear all over combinatorics, the branch of maths that counts and arranges things. When mathematicians want to show that a large structure must contain a tidy pattern, sunflower-type arguments are often the tool they reach for. Proving a sharp bound on how large a sunflower-free family can be gives everyone working in the area a firmer foundation: it tells them how much structure they are guaranteed to find, and how quickly the guarantee grows.
It is also a lesson in how maths is checked. A posted proof is a claim, however exciting. It becomes a theorem in the full sense when other experts have read every step, found no gap, and the work has passed peer review. Until then, the honest wording is "a proof has been posted" — which is exactly what has happened here, on a named date, by named authors, with the argument available for anyone to examine.
For students, the sunflower is a wonderful example of maths hiding in plain sight. You start with sets and overlaps — ideas met in school — and you arrive at a question that took half a century to settle. Spotting the pattern is the first step; proving that it must always appear is the hard part.
If your child enjoys puzzles where a simple pattern leads to a deep question, a first lesson in maths can start with three sets and a shared core — and end with why mathematicians check a proof line by line before they call it settled. Book your first lesson
Sources:
- Bucić, M., He, F. and Lee, H., "Proof of the Erdős-Szemerédi sunflower conjecture", arXiv:2610.26686, submitted 19 October 2026: https://arxiv.org/abs/2610.26686
London Tutors — post your tutoring request and we will match you with a verified tutor.