Minutes of the 3pi meeting 21.3.94
 
Present: E. Aslanides, R. LeGac, T.Ruf, G.Polivka, A.Ealet, E. Hubert,
         M. Dejardin, M. Fidecaro, O. Wigger, P. Fasnacht, P. Weber,
         P. Bloch, P. Pavlopoulos, R. Rickenbach
 
- Ann presented a comparison of her old analysis with the new official
  cradle. Differences were a stronger cut on the secondary opening angle,
  the 3c fit (new version of Marc) and the cut on the angle between the K0
  momentum and the vertex separation is space (plus some othe cuts that showed
  no influence on the CP parameters). This cuts will have to be studied in
  more detail and possibly implemented into the official cradle.
  Ann also had a closer look on the anti pi0gold fit and found that it
  causes asymmetric effects with resepct to the magnetic field, x and K+/K-.
  Also here a closer investigation is needed.
- Rene reported on a comparison between the data and MC. There is a big
  discrepancy between data and MC in the lifetime and decay radius radius
  distributons at short life times. On the other hand a comparison of data
  and MC looking at primary momenta and primary invariant mass show reasonably
  well agreement at short and at long lifetimes. One possibel conclusion
  is that our background at short lifetimes is golden. Comparison with MC
  shows (from the shape of the distribution and from absolute numbers (even
  if there is some disagreement between Edwige and Rene) that it cannot be
  semileptonics. From the MC acceptance and the branching ratios also
  goldpi+pi- is excluded. This is confirmed by the time dependece of the
  ratio of 1 neutral shower/2 neutral showers which is 2 rather independant
  of the lifetime for our data and is 4 for goldpi+pi- as determined from
  normalisation data. A look at the infuence of Marcs e/pi separation shows
  that it cuts on our data independant of lifetime. If the bachground would
  come from gamma conversions one would expect a stronger effect at short
  lifetimes. This is also confirmed by a comparison of the secondary opening
  angles between data and MC. There is a good agreement at short lifetimes.
  A strange thing is a clear discrepancy at long lifetimes. This however
  goes in the wron direction.
  Rene also presented a small study on the effect of NNKIN on our date.
  The outcom is that NNKIN cuts on our data before the cut on pi0gold in
  the same way as on MC data for good events. The conclusion is that we
  loose about 15% of good events at short lifetimes which has no dramatic
  effect on our errors, but it is a loss without a gain for us as it does not
  change the signal to background ratio. This result is confirmed by an
  independant study done by Patrick.
- A discussion with Eli showed the need to define clearer the work to be done
  by the various people from Marseille. He will inform the people concerned.
 
  I propose Monday April 18th 14:00 h for the next 3pi meeting. Please tell
  me soon whether this is ok.
  Until the next meeting I think Ann should give a detailed study on the
  effects of the new 3c fit on data as well as MC as a function of the
  lifetime. If she has time left the same should be done for the other cuts
  she looked at. If possible we should take a decision on what to implement
  in the official cradle at the next meeting.
  Edwige should have a look at the differences between her an my numbers
  concerning the differences we find in the MC acceptances.
  There is some disageement between the Lifetime distributions presented
  by Andrej at the collaboration meeting and what Ann and I find. Andrej
  should check whether he possibly has mixed up the different triggers.
  Andrej should also report on the status of his work. (If he cant come
  in writing).
  Frederic should give a report on his work concerning the phase space
  acceptance. Looking at the edges of the phase space could give information
  on background by backscattering and gamma conversion. In view of the
  obvious contradiction in the conclusion about our background (I shoud be
  golden but it can't according to MC) he shold try to find deviations
  between MC and data.
  The influence of the trigger on our life time distribution will be checked
  by a student of Marseille with the help of Renaud.
  Attached to the minutes you will find a small study on the errors. I hope
  this will help to study systematic effects caused by our knowledge of the
  background and the acceptance for good events.
 
                                                cheers
 
                                                    Rene