Not much hope for run 3 with regards to detecting supersymmetry. The improvements are somehow logarithmic, so that the bounds on supersymmetry that could be extracted from the full run 3 data with respect to those extracted from the full run 2 data are not much different.

Two exceptions:

• Supersymmetry exactly lies in a region where one is still limited by statistics. More collisions will make a difference.
• Supersymmetry is slightly too heavy for the run 2 collision energy. Therefore, run 3 with an slightly increased energy can make a difference.

In my opinion, we should instead investigate deviations from the Standard Model for observables where one is currently statistically limited. This is our best handles on new phenomena.

For instance, asking a lot of missing energy and verifying that the shape of the missing energy spectrum is the Standard Model one. The rate for producing a lot of missing energy on the Standard is small.