McCrea's conjecture

What Lean checks

Complete proposed proof of the unique linear-density crossover for every modulus at least three. Positive coefficient asymptotics handle the floor, and the exponent difference has exactly one zero in the stated interval. Independent internal review and exact profile-word diagnostics passed.

For every integer k≥3k\ge3, there is exactly one transition density between 2−k−12^{-k-1} and 1/(4k)1/(4k) in the comparison of the two fixed-perimeter Franklin statistics. For each fixed density below the transition, the multiplicity statistic eventually has more partitions; above it, the divisibility statistic eventually has more. The proof covers the full interval and every allowed modulus.

This proves Conjecture 9.2 of Marcus McCrea, Fixed-Perimeter Franklin Statistics and an Eventual Inequality, arXiv:2609.12005v2, 5 November 2026, Section 9. The full source PDF distinguishes this conjecture from the earlier fixed-number theorem.

Positive generating-function factorizations give exact local coefficient asymptotics when the marked count is ⌊xn⌋\lfloor xn\rfloor. The two strictly concave growth exponents attain the same maximum at different means. Between those means, one decreases and the other increases, so their difference has a unique zero. The sign of that difference gives the eventual strict inequalities.

Run from the repository root:

python3 results/mccrea-single-crossover/verify.py

The standard-library verifier enumerates profile words directly, compares the two statistics with both generating-function forms, and checks the exact saddlepoint means. These are finite consistency checks; the written Fourier and concavity arguments prove the unrestricted result.

Status: complete proposed proof, independently reviewed internally; unrefereed. No prior crossover resolution was found in the documented search. No statement at the transition, outside the specified density interval, or about a uniform finite-perimeter threshold is claimed. No Lean work was performed.