Skip to content

Erdős Problem #390

Editorial depth: Reviewed explanatory article — evidence qualified.
Snapshot classification: open (not automatically the live result).
Subject areas: number theory, factorials.

Read this status correctly

The status above is reproduced from GCL's immutable external-catalogue snapshot. It is not a new claim of current mathematical resolution, a source-semantic review, an independent proof check, or GCL certification. A published proof and a GCL replay are separate events.

The question in plain language

How efficiently can one express \(n!\) as a product of distinct integers all larger than \(n\)? Define \(f(n)\) to be the smallest possible largest factor \(m\) among products

\[ n!=a_1a_2\cdots a_k,\qquad n<a_1<\cdots<a_k=m. \]

Erdős Problem #390 asks whether there is an asymptotic constant \(c\) such that

\[ f(n)-2n\sim c\,\frac{n}{\log n}. \]

This is open. Knowing only that the difference has the same order of magnitude as \(n/\log n\) is not enough to identify a limiting constant.

A small example

For \(n=5\),

\[ 5!=120=10\times12, \]

and both factors are strictly larger than 5. This gives \(f(5)\le12\). In fact no factorization with maximum factor at most 11 works: a factorization with three or more distinct factors larger than 5 would already have product at least \(6\cdot7\cdot8>120\), while none of the possible two-factor choices below 12 multiplies to 120. Hence \(f(5)=12\).

For every \(n\ge2\), the singleton factorization \(n!=n!\) is formally allowed. It gives a very weak but valid existence bound \(f(n)\le n!\).

What is known

Erdős, Guy and Selfridge (1982) established two-sided order estimates,

\[ f(n)-2n\asymp \frac{n}{\log n}. \]

These estimates allow the normalized difference to fluctuate between positive constants; they do not imply a single limit.

One useful exact identity comes from \((n+1)\cdots(2n)=n!\binom{2n}{n}\). Starting with the factors between \(n\) and \(2n\), deleting some factors and adding larger ones leads to a product relation involving the central binomial coefficient. This is a promising bookkeeping method but not a new asymptotic theorem.

GCL's current state

Source audit #1660 records the literal strict factor-order semantics, the small exact factorization and the literature boundary. The cited 1982 proof and any external formalization were not independently replayed.

What remains

Determine whether the ratio \((f(n)-2n)\log n/n\) converges and, if it does, compute its limiting value. Finite factor searches and two-sided order estimates do not establish that convergence.

Sources and claim boundaries

  • Erdős Problems #390: open problem and historical order estimate.
  • Erdős–Guy–Selfridge (1982), as identified in the maintained registry; full original theorem pages remain to be checked.
  • GCL source audit #1660: bounded factorization identity and explicit source caveats.

Catalogue and evidence provenance

  • Pinned catalogue source: teorth/erdosproblems at 536408716724.
  • GCL issue: #1660 — protected issue/assignment identity; its labels and results are live, not reproduced automatically here.
  • GCL triage: HOLD_STATUS_RECONCILIATION.
  • Formal-statement coverage: A provider statement was included in the pinned snapshot. Presence does not mean the statement or proof has been checked.
  • External mathematical status: consult current problem page and its dated bibliography.

Back to catalogue · Compendium introduction