1. Aleksandra: Triggering efficiency drops near the feet and the large and small muon sectors are also visible.

2. Efficiencies:
   a) How do we calculate them?
   Tag-and-probe method. You need a clean resonance, signal with low background that gives muons. You define tag as the muon that is medium muon, isolated and that triggered your event. Probe is a muon that gives the invariant mass within the resonance of interest. Efficiency is defined as ratio of number of probes that pass the selection to the total number of probes.

   b) Why do we use different samples for efficiency calculation and momentum correction?
   To avoid correlations and also for efficiency calculation, you need to trigger on low pT events so you use 2muon triggers where the rate is smaller than single muon triggers for efficiency. You can go down to 5 GeV muon efficiency with 2muon triggers.

   Trigger prescaling: You cannot trigger on every event, so what you do is do prescaling where you select 1 in 10 events but then you don't know how to scale it correctly.

3. Figure 8, why is there an efficiency drop in pT of 50-100 GeV?
   Because of the sharp drop in muon pT at around 50 GeV. There might be event migration in neighboring bins.

4. Momentum corrections:
   Formula 5 and formula 6 were written on board.
   You want MC to look like data so you correct MC pT to match data. Corrections are applied to ID and MS separately. The numerator which is scaling shifts the momentum right and left. The denominator (un)smears it depending on the coefficeints.
   s0 is the energy loss term. Does it depend on the pT? No, because this is the correction term not trying to model the calo energy loss

5. Plot that shows relative momentum resolution contribution.
   Cyan color: Energy loss correction goes down as pT increases. Eg: If your correction is 5 GeV then it will effect 10 GeV much more than the 1 TeV muon
   Multiple scattering: You don't model everything that you have inside the detector like the hybrids, silicon,etc. So this correction takes care of that. It corrects for the matter we missed to model in Geant. So this leads to 2% correction for muon pT resolution. The contribution is flat while for the energy loss there was dependency.
   Chamber alignment: Increases for high pT stuff because there is no curvature for high pT muons.
   Spectrometer entrance: Corrections are sensitive to perpendicular. Models where in the MS the muon hit

   You use the J/psi and Z boson to get the resolution to cover the entire pT range. They perfectly match in the overlap region.
   Higgs mass discrepancy working group: Tried to resolve the mass difference in 4l and the gamma gamma channel. Smallest pT muon from H->ZZ is 6 GeV coming from virtual one so you definitely need J/psi. There was discrepancy due to energy loss modelling

6. Table 3 features:
   s1: Double for endcap because of misalignment and the error is large because t is ID and beyond 2.2, there is less events.
   r2: Misalignment corrections more important for high pT. Look at units, it is TeV-1

7. Table 4 features:
   r0=0 assumption. Assume corrections for energy loss from MS is really small.
   s0: Energy loss correction. Modelling of material in barrel is worse than endcap.
   s1: Worse in barrel and endcap transition region.
   r2: Same story as ID.

8. Why do you use different transforms for ID and MS? Like Hough transform and Kalman filters?
   Because there are fewer points in the MS.

9. Why do we take measurements when the toroid is off?
    To get r2 i.e. misalignment and this is dominant for high pT muons. This will be important as we increase energy.