A team presents a new construction of starting scales that makes Born spreading publicly available for NLO event generation matched to parton showers, including samples merged with the FxFx scheme. The improvement reduces the fraction of negative-weight events in W+jets production with up to two jets at NLO without affecting the physical predictions, and speeds up event sample generation sixfold compared with the previous implementation.

Negative weights are a major cost in NLO event generation matched to parton showers at the LHC, as they increase the number of events needed for a given statistical precision. Born spreading addresses this by redistributing the Born contribution over the radiation variables. However, the technique introduces no bias only if everything passed to the parton shower together with S events depends on Born-level quantities alone. The authors show that the same condition is needed for the MC counterterms to reproduce the first shower emission, independently of how the radiation phase space is integrated.

In MadGraph5_aMC@NLO, the S-event starting scales did not fully satisfy this condition, which so far prevented the public release of Born spreading. The researchers present a new and consistent construction of the starting scales for ordinary MC@NLO, FxFx merging and MC@NLO-Δ, with which Born spreading is now publicly available, including for merged samples. They also study modified formulations of the MC counterterms.

As a proof of concept, they apply the method to W+jets production with up to two jets at NLO merged with FxFx. Born spreading reduces the fraction of negative weights after merging without affecting the predictions, and the event samples are generated six times faster than with the former implementation. The result covers both standard generation and the merging of jet multiplicities, a setting relevant to the precision simulations feeding LHC analyses.