Distribution of VPDL(meas) - VPDL(true)

Last Update: 20-Jan-06

Using the simplified definition of the VPDL:

x = M(B) L_xy / Pt(meas)

gives an error (assuming σ(Pt) is negligible)

L_xy

M(B) / Pt(meas)

Then, for a given event, the distribution of x(meas) - x(true) is a gaussian of width σ(x), unaffected by the distribution of Pt(meas) because, for this event, we are interested in the VPDL distribution for the Pt(meas) specific to this event.