Skip to content

Erdős Problem #218

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

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

List the primes as \(p_1<p_2<\cdots\) and write their consecutive gaps as \(d_n=p_{n+1}-p_n\). Erdős conjectured that each set of indices with

\[ d_{n+1}\ge d_n\qquad\text{and}\qquad d_{n+1}\le d_n \]

has natural density \(1/2\). He also conjectured there are infinitely many equal adjacent gaps, meaning \(d_{n+1}=d_n\). An equality here is equivalent to three consecutive primes forming an arithmetic progression.

A small example

The primes \(3,5,7\) are consecutive, and their gaps are \(2,2\). They witness one occurrence of equal adjacent gaps, but say nothing about infinitely many such occurrences. The primes \(5,7,11\) have successive gaps \(2,4\), so the first gap is no larger than the next.

A useful distinction concerns the density statements: if the equality indices had positive natural density, the two sets defined by the weak inequalities would overlap with positive density and could not both have density \(1/2\) while their union is all positive indices. The conjectured densities therefore imply zero density for equality indices, but zero density is compatible with infinitely many equal gaps.

What is known

The claims remain OPEN unconditionally. Banks (2023) offered a heuristic based on a quantitative version of the prime-tuples conjecture. For fixed \(c\ge0\) and \(\varepsilon>0\) the predicted count for \(p_n\le x\) with \(d_{n+1}\ge c d_n\) is

\[ \frac{\pi(x)}{c+1} +O\left(\frac{x}{(\log x)^{3/2+\varepsilon}}\right). \]

Taking \(c=1\) suggests the half-density assertion for the nondecreasing comparison. This is conditional heuristic evidence, not a theorem resolving prime-gap comparison.

Erdős also conjectured arbitrarily long strings of equal consecutive gaps, which is equivalent to arbitrarily long arithmetic progressions of consecutive primes; see #141. Green–Tao's theorem on arbitrarily long prime progressions does not establish consecutiveness.

GCL's current state

Pinned and current Erdős #218 classify the questions OPEN. Original issue #1487. The modern formal provider contains a statement file, but its translation of density and infinitude has not been independently proved equivalent to these natural-language claims.

The original issue contains an authenticated RESULT/1 source report. This is unadjudicated source evidence only, not independent mathematical certification or editorial admission.

What remains

A bounded task generates consecutive prime gaps through a specified cutoff, computes indices satisfying each comparison, explicitly identifies ties, and checks that the counts satisfy \(A+B=N+T\), where \(T\) counts ties. Acceptance: the \(3,5,7\) tie is found, every comparison uses adjacent prime gaps, and empirical proportions are not called asymptotic densities.

Sources and claim boundaries

No prime-gap density theorem, proof of infinitely many ties or editorial admission is claimed.

Catalogue and evidence provenance

  • Pinned catalogue source: teorth/erdosproblems at 536408716724.
  • GCL issue: #1487 — protected issue/assignment identity; its labels and results are live, not reproduced automatically here.
  • GCL triage: PROTECTED_READY_BACKLOG.
  • 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