Hi,
I tried to compute the systematics from the mes fit.
First let me remind you how the fit is performed. After all cuts a fit to
the mes distribution for Mx<2.5 is performed with all parameter floating
in order to get the signal parameters. Then those parameters are fixed in
all the mes fit in each Mx bin.
I repeated the fit varing those signal parameters within their error.
This is the change in the fitted BRBR value:
* +1 sigma in mean gauss (cry ball) +0.0 * 10^-3 (+0%)
* -1 sigma in mean gauss (cry ball) +0.01 * 10^-3 (+0.6%)
* +1 sigma in sigma gauss (cry ball) +0.20 * 10^-3 (+1.2%)
* -1 sigma in sigma gauss (cry ball) -0. * 10^-3
* +1 sigma in alpha param (cry. ball) -0.50 * 10^-3 (-3%)
* -1 sigma in alpha param (cry. ball) +0.48 * 10^-3 (+2.8%)
* +1 sigma in n param (cry. ball) -0.30 * 10^-3 (-1.8%)
* -1 sigma in n param (cry. ball) +0.88 * 10^-3 (+5.2%)
--------------------------------------------------------------------
total +1.02 +6%
-0.58 -3.5%
As a crosscheck I ran the fit leaving floating the cb parameters (alpha
and n) in each Mx bin fit.
I got
BRBR = 0.0178748 +- 0.00488316(stat) +- 0.00146808(MC stat)
that is +5.8%, then is within the systematics.
This numbers, together with the crystall ball vs gaussian fit
difference(1.2%), will give the systematics from the mes fit.
+6.1%
-3.7%
We are also forgetting the error from the fit for Nsl. It has a stat
error of 1.6%.
Daniele
