Prime-Offset Formulation
This note records the strongest Route 2 formulation that currently looks mathematically promising in this project.
Setup
Fix n >= 1. Let
q = lastIncludedPrime nrbe the next prime afterq, soConsecutivePrimes q r
The existing Route 2 hypothesis in the formalization is the upper-interval statement
∃ p, Prime p ∧ q# + 1 < p ∧ p < q# + r^2.
In Lean this is IntervalPrimeExistsAtPrime q r, and it is equivalent to the
plain offset form
∃ k, 1 < k ∧ k < r^2 ∧ Prime (q# + k).
In Lean this is OffsetPrimeExistsAtPrime q r.
Stronger Formulation
The more interesting research target is the prime-offset strengthening
∃ m, Prime m ∧ q < m ∧ m < r^2 ∧ Prime (q# + m).
In Lean this is PrimeOffsetPrimeExistsAtPrime q r.
The point is that this does not merely ask for some short offset producing a
prime translate of q#; it asks for a prime offset above the whole range of
primes already built into the primorial.
Why This Is Useful
This formulation is stronger than the interval statement already used in the formal reduction chain, so any proof of it would immediately feed into the existing machinery:
PrimeOffsetPrimeExistsAtPrime
-> OffsetPrimeExistsAtPrime
-> IntervalPrimeExistsAtPrime
-> Route2ReductionBound
-> FortuneConjecture.
The corresponding Lean theorems are:
offsetPrimeExistsAtPrime_of_primeOffsetPrimeExistsAtPrimeintervalPrimeExistsAtPrime_of_primeOffsetPrimeExistsAtPrimeroute2_reductionBound_of_primeOffsetPrimeExistenceroute2_fortune_of_primeOffsetPrimeExistence
So this is not a competing reformulation. It is a strictly stronger target that plugs directly into the route we already formalized.
The next refinement of this strategy is the sifted-candidate formulation in
sieve-target.md, which isolates the analytic gap in a
form closer to sieve methods.
Why This Looks More Natural Than Raw Interval Primes
The interval statement
q# + 1 < p < q# + r^2
is mathematically correct but somewhat indirect. It quantifies over an ambient
prime p and only recovers the offset after subtracting q#.
The prime-offset version instead talks about the arithmetic object we actually care about:
- the offset is itself prime
- the offset already lies above
q q# + offsetis also prime
This interacts well with the primorial congruence structure. If m is prime
and m > q, then for every prime ell <= q we have m mod ell != 0, so
q# + m ≡ m (mod ell)
cannot vanish modulo any included prime. In other words, prime offsets above
q automatically evade all divisibility obstructions coming from the primorial
base.
Why Generic Short-Interval Prime Results Are Not Enough
The current interval has length about r^2, which is about q^2. Since
q ≈ log(q#), this means the interval near q# has length on the scale
O((log q#)^2).
That is far shorter than the ranges controlled by standard unconditional prime-in-short-interval theorems. So even though Route 2 is phrased as an interval-existence statement, the live research problem is not a generic prime gap question near a large number. The special structure of primorial translates has to do the real work.
This is one reason the prime-offset formulation seems like the better target: it keeps the problem focused on the structured family
m, q# + m
rather than on arbitrary primes in a short interval.
Heuristic Interpretation
The prime-offset formulation asks for a prime m such that both members of the
pair
m, q# + m
are prime, with q < m < r^2.
For primes ell <= q, the condition ell ∤ m is automatic once m is prime
and larger than q, and then also ell ∤ (q# + m). So the obvious local
obstructions from the included primes are already removed. That does not prove
anything, but it suggests that the right heuristic object is a structured
two-prime pattern rather than a bare interval.
Status
This note is a strategy document, not a proof. The repository currently proves:
- the interval criterion imported from the literature in
Fortune.Literature - the equivalence between interval and plain offset forms in
Fortune.IntervalExistence - the implication from the stronger prime-offset formulation to
FortuneConjecture - the sifted-candidate and full-sieve formulations in
Fortune.Sieve
What remains open is the actual existence statement PrimeOffsetPrimeExistsAtPrime
in any uniform infinite form.
References
- Čejchan, A.; Křížek, M.; Somer, L. “There are no sign changes in formulae
involving prime numbers.” Rocky Mountain J. Math. 48 (2018), 1165-1178.
This is the source of the primorial interval theorem formalized in
Fortune.Literature. See alsoreferences.md. - Křížek, M.; Luca, F.; Somer, L. 17 Lectures on Fermat Numbers: From Number Theory to Geometry. Springer, 2022. Section 2.4 surveys Fortune-related primorial interval results. Open chapter PDF: https://cs.uwaterloo.ca/journals/JIS/VOL25/Krizek/krizek3.pdf
- Guy, R. K. Unsolved Problems in Number Theory. 3rd ed., Springer, 2004.
This remains the standard problem-source reference for Fortune’s conjecture;
see
references.md.