← All problems

Erdős Problem #390 — Factorizations of factorials into large factors

44 years old · traced to

For a positive integer nn, consider all representations

n!=a1a2⋯ak,n! = a_1a_2\cdots a_k,

where kk is a positive integer and a1,…,aka_1,\ldots,a_k are integers satisfying n<a1<⋯<akn<a_1<\cdots<a_k. When such representations exist, let f(n)f(n) be the minimum possible value of the largest factor aka_k. The question concerns sufficiently large positive integers nn (for which existence is guaranteed by the known bounds): does there exist a constant cc such that

f(n)−2n∼cnlog⁡n(n→∞)?f(n)-2n \sim c\frac{n}{\log n}\qquad(n\to\infty)?

Erdős Problem #390 Erdős, Guy, and Selfridge (1982)

Background

The question belongs to a broader study of expressing a factorial as a product of factors under different restrictions on their size and whether repetitions are allowed. It was posed in connection with Erdős and Graham’s 1980 problem collection and was taken up by Erdős, Guy, and Selfridge in their 1982 paper, “Another Property of 239 and Some Related Questions.” Their results highlight the role of distinctness: for n>239n>239, no factorization of n!n! into distinct integers all greater than nn can have every factor at most 2n2n, whereas if repetitions are allowed, such a factorization exists for every n>13n>13. Erdős, Guy, and Selfridge (1982) Zentralblatt review of Erdős, Guy, and Selfridge (1982) Erdős and Graham (1980)

References

Progress summary

Refreshed
Open

No public discussion or published progress appears to have been found for this problem.

No public discussion or published progress was found.

Current status (as of September 2026): The problem appears open, with no recorded activity.

Solutions 0

No solutions have been posted yet.