Yi's questions/comments

Nice paper....though it will be good to elaborate a bit more on details. Whether to include these in the paper or not is another question, but I would like to know a few things as follows.

1. Line 26. Why positive sign?
2. Line 54. Cone size of 0.7 is large, especially since the eta range is only 1.4 units or so for each region. Is there any particular reason smaller cone sizes are not used/compared with? Were there supporting studies using other jet algorithms or same algorithm with different parameters?
3. Line 80-83. Elaborate a bit more on the mass-dependent correction? The reference didn’t say much either.
4. Line 92. How is the 20% determined?
5. Figure 1. The choice of mass division points is perplexing...
6. Figure 1. How much systematics is expected on pythia prediction? How wide would the band be?
7. Line 100. What is rho_j exactly?
8. Line 108-109. Is the overall shift in units of ratio? Could the choice of constant shift be justified?
9. Line 129-134. Where does the 10x amplification come from?

EXO-10-002: Search for New Physics in the Dijet Ratio. Caltech Comments.

Links

CERN Wiki of the paper

Dorian's comments

1. In general I like it. It is a straightforward measurement with limit settings and discussion of uncertainties.

2. The JES uncertainties can have a very large impact on the very signal that we would like to detect. As the authors point out, new physics will have a rapidly varying shape vs the dijet mass and it will have very large uncertainty coming from JES. Since we do not only want to set limits and since the main problem is an experimental one, I would have expected a more detailed study of this. One obvious thing is to try different jet algorithms. They use anti-kt which is supposed to be less sensitive ot soft-energy depositions (but their problem is not UE or PU, yet) and has a more regular jet area. The regular area might make it easier to calculate energy corrections, but it is possible that those corrections are larger due to the rigidity of the jet boundary. In fact theoriticians suggest in general to use more than 1 algorithm. I have noticed that the present ATLAS papers on the subject also use anti-kt.

3. There is no real explanation in the paper why Pythia with NLO corrections does so much worse than the NLO calculations, especially since we are looking at dijet events. The funny thing is also that the non-perturbative corrections applied to the NLO calculations are derived from Pythia itself which puzzles me even more but maybe I have to think more carefully about this in order to understand it.

4. I find the conclusions of the paper not so optimistic. All they can conclude is that at 4pb-1 we could set the same exclusion limits as Tevatron, but what about discovery? Btw., things might have changed but I remember even more pessimistic numbers here. I remember that we were claiming before that we could reach the Tevatron exclusion limits with 10pb-1. I wonder what has changed.

5. Finally, I do not like the 'centrality' in the title since it is unclear, but this is a subjective matter.

-- Main.dkcira - 2010-10-04