It is expected that the JP2 triggerred events should be unbiased, or much less biased for individual tower E/p calibraiton.  However we did see a slope band at the position of BHT1 sample which is the most visible "trigger bias" sample. 

1) check the trigger counting 

if ( (htTrigId==430203 || htTrigId==430223 || htTrigId==430243) &&
event->triggerIdCollection().nominal().isTrigger(htTrigId) )
BHT1 = true;
if( (otherTrigId==430403 || otherTrigId==430413) &&
event->triggerIdCollection().nominal().isTrigger(otherTrigId))
JP2 = true;

The results BHT1 and JP2 and their overlap, at event level.
!BHT1 && !JP2: 70.69%
BHT1 && !JP2: 17.72%
!BHT1 && JP2: 10.70%
BHT1 && JP2: 0.88%
2) tof matching,  with tof matching and cut, the slope band should not be pile up of background. 

3) Effect from the slope band for JP2 sample. 
The slope can be exclucded by removing the "matched tracks", the tracks matched to the tower above the HT threshold. 

gains comparison,