At first glance the number of recent earthquakes at Ridgecrest don't appear to be affected by stress index (SI).
But a closer look shows changes for both the activity (A) and the peak magnitude (Mpk) at about the time the stress index was greatest on day of year 212 but they could just be another coincidence.
The changes in behavior is made clearer by the expected values shown in dark blue for the given interval indicated along with their upper and lower bounds. Projections from the baseline values are shown in light blue. For some reason the activity was greater than expected while the peak magnitudes were less than expected.
If one fits a plane through the activity and peak magnitudes when plotted normal to the doy,SI plane one see a slight dependence on SI in both cases. The coefficients were computed using all the available data as standard values with a least squares fit. Note that the doy and SI terms pass through the standard zeroes so their effect is to rotate data in their plane with the vertical and the apparent correlation with it may be due to an imbalance in the data.
An explanation of why we didn't see another large earthquake about the time when the SI was greatest may be that all the available pent up stress had been released.
