Recoil Studies, July-3-06
Results of recoil calculations, taking into account ISR (July-3-06)

Results obtained on the ttopag sample, M=175, fixed HERWIG.
Require only 4 tight jets, Nloosejets=0

NoISR means that the highest ISR cluster energy is < 5 GeV, code
based on HEPG info by Paul. ISR partons are at eta>2.4.

Unclustered energy recalculated by me, using the MET and undoing the
correction that was used in making the NTUPLE. This correction
subtracted from the unclustered energy the energy of towers adjecent 
to the 0.4 cone. I put this back in and then subtracted the whole 
OOC energy, using the jet corrections factors.
The "range" shown is the width of the distribution that includes the
68% of the events.

The plots are here:  Recoil for nloose=0. 

page 1
  left plots: unclustered energy, my calculation and that in the Ntuple
  right plots: recoil, my calculation (top) 
  range=5.9 and Igor's range=7.4
page 2
  scatter plot for Lina's method. Bottom left, lina' recoil as above.
  Top right: pt(ttbar) from HEPG range=6.3 GeV
  bottom right: prior, range=6.4 GeV
page 3
  same as above for Igv's method. Bottom right: prior range=7.1 GeV

So, the difference in range bewteen LIna's calcualtions and Igor's is
7.1-6.4= 0.7 GeV (about 10%). Assume that we go with Igor's
calculation, because of semplicity.

Questions arise: what are the events with large HEPG pt(ttbar), that
show up at recoil about 0? We suspect ISR.

page 4
  No ISR plots. Top right: HEPG pt(ttbar), range =2.7 GeV
                Bottom left: IGV recoil, range= 5.4
                Bottom right: IGV prior, range = 5.0 GeV
page 5
  No ISR. Lina's recoil (top left) range=4.9 GeV
          Lina's prior(top right) range=4.7 GeV
  Bottom left: pt(ttbar)x-vs-max-ISR clus E; right:Igor's
  recoil-vs-max_ISR clus E

Medians are  8-9 GeV for: pt(ttbar), Igor's and Lina's recoil.
For the no_ISR case the medians are:

                pt(ttbar), median= 3.8 GeV
                Lina and IGV median=~7 GeV

 ISR Studies 

PLOTS OF JET PT_VS_ETA FOR EVENTS WITH ISR and NOISR JETS

Extra jets in events are defined as all the jets after the 4 tight
jets cut. Only events with 4 tight jets and no loose jets are
considered. In the plots shown below all extra jets (including the 
ones below 8 GeV) are shown.
Run was done on ttopag file, which is 175 GeV with the HERWIG bug fixed.
Total number of events that pass the cuts is 35703+7270=42973 events.

An event is considered an ISR event if the maximum cluster energy for an 
ISR parton beyond eta=2.4 is > than 5 GeV. This is a parton that is 
due to ISR according to the HEPG information.

Following two plots show the extra jet distributions for an ISR event
 PT extra jets-vs eta, ISR events.
It has 35703events.
  
 PT extra jets-vs eta, no ISR events .
It has 7266 events. There are a few jets above 5 GeV at large eta that 
are not due to ISR. This means that by cutting on jets at large eta 
we loose some events that are not ISR events).