The BBC trigger is frequently trigger a bunch crossing early, leading to a peak in the eTOF "digi time to trigger" plot.   For a run last night (20168013) compare the EPD trigger:

To the BBC trigger:

This effect persists at a smaller level even into the HLT-GOOD trigger:

3.4% of the events in the HLT GOOD sample come from only the BBC trigger.   This second peak might account for about 1/2 of those extra events.  

I speculated earlier that because the TAC window in BBC is set from 100 - 2850 that this could be the same feature as we saw when we moved the EPD / VPD tac windows down to 100 and cut into the TAC pedestal values.   However the secondary peak in the BBC does not come from any particular range of BBC TAC values.  Requiring the earliest TAC > 300 doesn't affect the secondary peak.

Also, in the case of reducing the EPD TAC window to 100 there was a strong effect in the etof (Run 20162028) for the EPD trigger:

But the effect was entirely cleaned up by the HLT good trigger: