Hi all,

 I send you the plots we discussed in the afternoon. Actually I repeated
teh study decreasing a bit the statistics in order to get exactly the same
error we have on data.

  http://www.slac.stanford.edu/~daniele/vub/testsample.ps

 The first 4 pages contain the summary of all the plots. The remaining
ones are a zoom of the single plots.

 in each page you find this matrix of plots



      ch0 BRBR         ch0 pulls BRBR      ch0 err BRBR


      ch1 BRBR         ch1 pulls BRBR      ch1 err BRBR


      BRBR(ch0-ch1)    pullsBRBR(ch0-ch1)  sigma BRBR(ch0-ch1)
                                           sqrt(sigma(ch1)**2-sigma(ch0))


 *  page 1  ALL

 *  page 2  B0

 *  page 3  Bch


   page 4
   ------

 Here you see the difference B0-Bch both for ch0 and ch1.

 The results are pretty encouraging.

 - the value we get for B0, Bch are not crazy

 - the difference ch0-ch1 is ~1.3 sigma effect

 On the other hand the difference in quadrature of the error of ch1-ch0
does not match with the result from the toy. The difference B0-Bch we see
is a two sigmas effect.

 Daniele