Hello vub-recoilers,
we had an AWG reading of BAD 1212 this morning and a couple of the things
that came up are of interest for the Vub analysis as well. So here they
are:
* Marcello said that they observed a shift in the Qtot distribution in
more recent Runs compared to earlier Runs. Cutting on Qtot==0 would
then introduce an inefficiency if the signal is shifted from picking
up background tracks. I think I remember seeing a broadening of the Qtot
distribution (and this actually makes more sense to me, background
tracks could have either charge). But I guess we should have a look
at this, although I am not sure that there is anything we could do about
it as opening up this cut will bring in a lot of bkgd and poorly reco'd
signal and worsen the resolution.
* Why are we using the muMicroTight selector? The NN based muon selector
should give us higher efficiency and (substantially) lower fake rate.
* Are we using a cut on the impact parameter when performing the track
selection? This should reduce bkgds from material interaction in the
beam pipe. Should be answered by a look into the tcl files.
* Partial D* reco veto: I recall that Virginia tried extending this to
neutral slow pions, but that this did not yield any improvements
(in S/B?). I'd just like to cross check that that is correct.
Are there any plots and/or numbers from this study?
* We are not scaling the MC such that the relative statistics for
MC with conditions for the different Runs matches the statistics for
those Runs in data. Has anyone looked into that at some point already?
* What exactly are we doing for PID weighting/killing/tweaking?
Should also be answered by a look into the tcl files.
Also, I think we hadn't decided on when to have the next vu-recoil
meeting...
Cheers,
Kerstin
|