Wiki
Wiki

Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.

Updated


Á. Elbert, Über eine Vermutung von Erdős betreffs Polynome I and II, Studia Sci. Math. Hungar. 1 (1966), 119--128, and 3 (1968), 299--324, prove the supremum half of the problem. For ff a nonconstant monic real polynomial with all roots in [−1,1][-1,1] and Ef={x∈R:∣f(x)∣<1}E_f=\{x\in\mathbb R:|f(x)|<1\}, the measure of EfE_f is at most 222\sqrt2; Erdős, Herzog and Piranian [EHP58] had proved that bound when every root is ±1\pm1, where (x2−1)m(x^2-1)^m attains it, and conjectured it in general. The two papers are not held in the library, and the statement is recorded here as the later claimants report it: the manuscript of Darvas, Peng and Tao cites the two papers as the proof of the supremum, in a series of technical papers, and records that Erdős asked in 1976 for a more elementary proof, which Terence Tao's note of December 2025 supplies (Tao's page); Budala's Section 1.3 says that the sharp upper value was proved previously by Elbert. The site's commentary lists sup⁡=22\sup=2\sqrt2 among the known values without naming the papers. No independent check of the proofs is recorded.

Covers. The supremum only: sup⁡∣Ef∣=22\sup|E_f|=2\sqrt2 over the class. The claim says nothing about the infimum, which the three pending full claims of the folder claim to determine.

Refereed. Studia Scientiarum Mathematicarum Hungarica, volume 1 (1966) and volume 3 (1968), as the manuscripts of Darvas, Peng and Tao and of Budala cite the two papers. No reviewer is named: the site's label is OPEN for the infimum, and its commentary credits no one with the supremum.

Depends on. Nothing on the wiki.