Based on Phys.Rev.D 87 (2013) 014010 (arXiv:1208.0010) by Yang-Ting Chien, Randall Kelley, Matthew D. Schwartz, and Hua Xing Zhu

We computed the distribution of jet mass in events with a single jet (recoiling against a photon). The kinematics looks something like this

The calculation was perfomed at the NNLL accuracy using methods from Soft-Collinear effective theory.Here is a comparison of our resummed result and the Pythia event generator:

Our calculation did not include non-global logs (NGLs) at fixed order, or resummed. Insetad, we estimated the uncertainty associated with NGLs by varying the form of the contribution that NGLs could have. With a generous variation, we found that it affected our jet mass distribution in the following way

The qualitative conclusion is that the NGLsonly affect the peak region, where the calculation is most difficult and sensitive to non-perturbative effects. Thus, it seems that resumming NGLs may not be critical to compare perturbative calculations to data.

Unfortunately, there is so far only data on jet mass in dijet events which is folded in with detector effects. But since Pythia seems to be in excellent agreement with the folded data, it is likely that our calculation would agree with the unfolded data.