Re: Monte Carlo Data

[Tom Banks <tbanks@socrates.berkeley.edu>]

Beloved Brendan,

after extensive discussion here in Berkeley, we believe that we have
developed a viable explanation for the observed nonuniformity in the decay
rate vs. impact parameter ratio plots.  Basically, we think the "hump" in
the distribution of ratio points is an artifact of the consistent
underestimation of the "true" drift distance, which is itself is an
unavoidable consequence of the electron track/mustop impact parameter
calculations (see my April 12 worklog page for my brief description of
this geometric phenomenon, which I think you said you mu-d MC experts were
familiar with:

  http://weak0.physics.berkeley.edu/weakint/research/muons/private/tbanks_dir/TeleConf/2005Apr5/2005Apr5.html

) My questions for you:

  1. Does your current mu-d MC software take any track reconstruction
     effects into account, or is it 'omniscient'?
  2. If the latter, could we ask you to incorporate these track
     reconstruction effects so that they could be turned on and off?
     Basically, you'd need to generate a random decay direction for each
     mu's displacement, calculate the impact parameter between that decay
     track and the original mustop location, and use *that* calculated
     impact parameter (not the 'true' drift distance) when filling the
     lifetime histograms.  I think the code should be relatively
     straightforward to implement.
  3. Did you fix the error bar problems which you mentioned in
     Analysis-Run8 ELOG entry 47?  If you could do that, implement the
     track geometry effect, and then produce ratio plots where the
     geometry effect were turned on and off, we might be very close to
     explaining this mystery.

The mechanism we have developed is pretty complex and subtle, so we'll
wait until next week's teleconference to give a full and detailed
explanation.  But you can trust us.

XXOOXOX,
Tom