I did some more thinking about the L/E analysis. I dismissed it in my
earlier note because it did not have enough statistics to show a clear
oscillation signature (as opposed to a general suppression) in the L/E
plot for the preferred oscillation parameters of Delta m^2 = 2.5E-3.
However, on further reflection I think it may still be worthwhile to
pursue the L/E analysis, because it may give us a better rejection of
smaller Delta m^2 than the angular distribution alone. I have updated
my note, still at
http://www.cithep.caltech.edu/~nolty/LOverE.ps
with these thoughts; see pages 12 and 14 (all other pages unchanged).
I am not sure whether a similar analysis of the Spurio events could
add information; we can use pathlength through absorber to estimate E
but there is an enormous ambiguity in L. Still, I think it bears
looking into.
I don't know if we want to pursue this prior to publishing the current
two letters, or save this for a global fit paper later.
Bob