Sunday's discussion

We would like to devise a strategy to interpret the data among the large number of theoretical approaches. This is true for all our measurements we want to perform. At the moment, we use each calculation available and published (also a preprint is OK), once we have the corresponding programs to run the calculation. 

We would like to discuss how we can choose whether to adopt a calculation (e.g. is having a publication enough?).

Once/if we adopt more than a calculation, what can we quote as final result?  At the moment we present all the results from all the adopted calculations. Can we just use one calculation instead of others to quote the final results? Can we make an average of the different calculations and e.g. use the maximum difference as theoretical error?

Similarly, also from the experimental point of view we do have different analyses for each final state. In the HFAG now we are averaging all the measurements for the same final state (only in the inclusive Vub we show also the results for each kind of analysis).  Does it make sense to combine the results from all the measurements (of the same final states) from the theoretical point of view? Moreover, for sa same analisys (eg B-recoil) we present distributions in more than a kinematical variable and accordingly different values of Vub. Should we choose among them? Eg the one with the smallest total error?