← All problems

Multiplicative Merino–Welsh conjecture

For every finite loopless and coloopless matroid MM, with Tutte polynomial \TutteM(x,y)\Tutte_M(x,y), one has \TutteM(2,0) \TutteM(0,2)≥\TutteM(1,1)2\Tutte_M(2,0)\,\Tutte_M(0,2)\geq \Tutte_M(1,1)^2.

References

Source: arXiv

Progress summary

Refreshed
Open · solution reported

A preprint reports that the conjecture is false, while newer work narrows the smallest known counterexamples and proposes a sharp replacement inequality, without independent verification.

The conjecture asserts a product inequality for loopless, coloopless matroids. Beke, Csáji, Csikvári, and Pituk reported its failure in 2023; subsequent work studies the smallest counterexamples and the best universal replacement.

Known results

  • Series–parallel graphs: the inequality holds (earlier work).
  • Lattice path matroids: it holds, with a stronger bound except for direct sums of trivial snakes (2016).
  • Split matroids: it holds, with equality cases characterized (2022).
  • Beke, Csáji, Csikvári, and Pituk improved the universal parameter bound to 2.92432.9243 (2023); later work reported 2.352.35.

October 2026 minimum-size and threshold developments

Chen, Jin, and Ma report a sharp 6262-element minimum within specified parallel-replacement families and structural properties of minimum counterexamples. A separate paper reports the claimed sharp universal threshold c∗=x0c_* = x_0, where x0x_0 is the largest real root of x3−9(x−1)x^3-9(x-1), but both the threshold proof and the minimum-size claim remain unverified.

Current status (as of October 2026): The original conjecture is reported false by explicit counterexamples; the claimed sharp threshold and the 6262-element minimum are unverified, and a complete classification remains open.

Sources

Solutions 0

No solutions have been posted yet.