1) How many events are there in the ALCARECO with BH triggers?
In the first 3796 events of the data set, there are 165127 tracks.
When selecting on the trigger bits, only 181 events (4.8%) survives, and 9124
tracks (5.5%).

2) The bit '0' (crossing bunches) should be removed.
By asking:
 - good track: pMin=2.5, nHitMin=8, minHitPerSubDet.inTECplus=1, minHitPerSubDet.inTECminus=1: 27 selected tracks,
 - L1 tech. trg: '36 OR 37 OR 38 OR 39',
I have 27 tracks in the usual 3796 events. Only 22 are left when also bit '0'
is requested in AND.

3) How many tracks are there when no trigger selection is done?
In the usual 3796 events, with the usual track selection:
 - pMin=2.5, nHitMin=8, minHitsPerSubDet.inTECplus=1, minHitsPerSubDet.inTECminus=1,
 68 tracks are found. Therefore the efficiency goes done from 22/181=12%, to
68/3796=1.8%.

4) Use the TrackRefitterBHM
In the usual 3796 events, with the usual track selection:
 - pMin=2.5, nHitMin=8, minHitsPerSubDet.inTECplus=1, minHitsPerSubDet.inTECminus=1,
 83 tracks are found (they were 68 with the standard TrackRefitter)
 - with also trigger selection: '36 OR 37 OR 38 OR 39',
 31 tracks are found (they were 27 with the standard TrackRefitter)
 - with trigger selection, but looser cuts: pMin=1.0, nHitMin=5,
   minHitsPerSubDet.inTEC=1,
 263 tracks are found.