[Rivet] electroweak results used to tune/validate theory

[Emily Nurse]

Dear CDF and D0 electroweak convenors,

You may already know about this, but incase you don't .... Rivet is a  
toolkit for the validation and tuning of MC event generators. The idea  
is to produce a routine for every relevant analysis from any  
experiment that can be interfaced with any MC output and plot   
distributions that can be directly compared with the experimental  
result. More information is here:
http://projects.hepforge.org/rivet/

We are currently including many Tevatron Run I and Run II  
measurements. One important example is the Z pT measurement, which CDF  
has a Run I result for and D0 a Run I and Run II result. All the  
results have the feature that they use theoretical models to correct  
the data. While this is unavoidable at some level it is really  
preferable to keep this "contamination" of the data to an absolute  
minimum.

For example, all of these results have used Photos to correct the di- 
lepton pT so that the measurement gives the true Z pT. This is of  
course only true if Photos gets QED 100% correct, and it means that  
for event generators that include QED we have to somehow un-include it  
when we write the routine (by e.g. clustering back photons and  
leptons, for both electrons and muons). While it may be desirable for  
other reasons to correct for some theoretical effects, it would be  
really helpful if all such publications also included a measurement of  
what was actually observed. This is not only essential for us but also  
any other theorists who want to make comparisons to the data.

Since, for electrons, collinear photons are usually indistinguishable  
in the calorimeter, this would probably mean publishing the dilepton  
pT where an electron is a cluster of a certain size.

The same point is also true when correcting for acceptance cuts, it  
would actually be preferable to publish the dilepton pT for electrons  
with |eta| < X etc, rather than relying on models of the lepton decay  
distributions to correct for these cuts.

We are keen to make sure that any future Tevatron results take this  
approach to make using them for MC validation more robust. In general  
data that has been interpreted using the best theoretical model at the  
time age as the theory develops, but experimental results, corrected  
only for detector effects, last forever.

Please let us know your thoughts on this issue and whether you are  
willing to make this change to future publications.

Best regards,

Emily (for the Rivet team).

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.hepforge.org/lists-archive/rivet/attachments/20090423/f3a00830/attachment.htm