Below is the (TOF multiplicity)^{1/4} distribution for the mb, the vpd-mon-tac, and the tof-mon triggers.

Offline, events were selected to have the vertex found, at radius < 2 cm, and with |vz|<75 cm.

The mb came from the physics stream, vpd-mon-tac and tof-mon from the monitor stream.

46% of tof-mon triggers satisfied this condition, 30% of vpd-mon-tac, and 3% of mb events.

In order to put the vpd-mon-tac on the same scale as the mb, needed to scale down 1/N_{vpd-mon-tac} dN/d(TOF^{1/4}) by ~0.7.

In order to put the tof-mon on the same scale as the mb, needed to scale down 1/N_{tof-mon} dN/d(TOF^{1/4}) by ~0.35.

67% of mb triggers (after good event selection) satisfied vpd-mon-tac

23% of mb (after good event selection selection) satisfied tof-mon, but these had different deadtimes.

Looks like the vpd-mon-tac is inefficient both in central and peripheral events.  The tof-mon can compensate the central inefficiency, but not the peripheral.