How many interactions does it take to modify a jet?

It is a continued open question how there can be an azimuthal anisotropy of high $p_\perp$ particles quantified by a sizable $v_2$ in p+Pb collisions when, at the same time, the nuclear modification factor $R_\text{AA}$ is consistent with unity. We address this puzzle within the framework of the jet quenching model \textsc{Jewel}. In the absence of reliable medium models for small collision systems we use the number of scatterings per parton times the squared Debye mass to characterise the strength of medium modifications. Working with a simple brick medium model we show that, for small systems and not too strong modifications, $R_\text{AA}$ and $v_2$ approximately scale with this quantity. We find that a comparatively large number of scatterings is needed to generate measurable jet quenching. Our results indicate that the $R_\text{AA}$ corresponding to the observed $v_2$ could fall within the experimental uncertainty. Thus, while there is currently no contradiction with the measurements, our results indicate that $v_2$ and $R_\text{AA}$ go hand-in-hand. We also discuss departures from scaling, in particular, due to sizable inelastic energy loss.