Speakers
Description
Recent progress in global analyses have led to pressing questions about precision and accuracy. NNLO PDFs must balance between the precision of experimental constraints and robustness (stability) with respect to the choice of experimental sets and methodological assumptions. We critically compare various strategies for achieving this balance in the Hessian and Monte Carlo formalisms adopted by CTEQ-TEA and other PDF-fitting groups. In typical applications, the differences among these strategies can be as important as missing higher-order corrections. We investigate this topic on the example of NNLO predictions for key LHC processes based on PDFs by the various groups. We comment on timely statistical concepts and confront those to physical constraints.
| Submitted on behalf of a Collaboration? | No |
|---|
