Comments on Bs->Jpsi pi+pi-
===========================

Questions
=========
General: The paper contains also less relevant information,
especially in figures that do not differ from each other
with the only purpose to show that they are not different.
Candidates for removal may be:
- Fig. 2.
- The additional resonances in Fig. 12, Fig. 13, Fig. 14, Table 5, and Table 6.
- Subpanels b, and d to h of Fig 18. 
  Just comment that we do not observe a significant structure 
  in the other angular moments in agreement with the 3R+NR model.

Line 112-114: 
It comes a bit late and not very prominent here as the main object of this
paper that we want to analyze this (obvious?) structure.

Line 160: 
Please provide some more explanation as to why the rho(770) 
has to be included: 
You used the sPlot technique to isolate the background and compared 
the found distribution with the wrong sign di-pion distribution. 
The observed discrepancy could be removed by adding background 
from the rho(770). Right?

Fig 10: 
Given the errors on the data points, is the fit not tweaked 
too much to the data? 
Cannot we use a smoother function? 
How does this specific choice influence the results of the full fit 
as illustrated in Fig. 13?

Lines 215-218:
Do we understand why we have two isolated maxima in the likelihood for 3R+NR?
It can be regarded as arbitrary to take the most likely solution
as 'alternate' and the slightly less likely one to be 'preferred'.
If for one or other reason, statistics, binning, correlations, or just coincidence,
we have different islands at a 1 sigma contour the solution might be to
look at the connectivity at 2 or 3 sigma just to avoid the ugly
result of a 'preferred' and an 'alternate' solution.
From the higher sigma contours we can construct a reasonable
value for a 1 sigma estimate, as well as a central value.

To use the accidental occurrence of the two maxima to define
a systematic uncertainty is even more arbitrary and results in
asymmetric uncertainties which can be avoided with the above
mentioned procedure.
What if we would have missed the second maximum?
Or if there is another one that we missed?

Fig. 12: 
The pull plot shows that we systematically underestimate the data 
around 1.5 GeV^2. 
How sure are we that this is due to a statistical fluctuation 
and not due to interference between the f0 and the f2?

No difference is visible in (b), (c), and (d) with respect
to (a). What is the aim of these three plots apart from showing
what is already written in Table 4?
The only difference of the signal can be found in (d) at low mass 
and is probably due to an error in the selection of the plot.

Line 180: 
Where does the kinematic upper limit of 1600 MeV come from?
And why is it 1550 MeV in line 247?
The naive limit lies at mBs-mJpsi = 2270 MeV and matches both 
the statements in the original analysis notes for Bs->Jspif0 
(mass window 300-2230 MeV) and the Dalitz plot (m^2 < 5.15 GeV^2). 
Now we seem to ignore half of the Dalitz plane.

Table 1: 
Why do we a priori not consider the other 0+(0++) 
and 0+(2++) resonances listed by the PDG: 
f0(1710), ... , f0(2200), f2(1430), ... , and f2(2150)? 
Their contribution will undoubtedly be negligible, but we should 
at least explain why we can ignore them.

Currently we give the impression that we are only looking 
for resonances formed by the Feynmann diagram in Fig. 1, 
whereas we are analyzing the pi+pi- final state, 
which in principle could also have contributions from other diagrams. 
[We don't have an example though] 
Again other contributions will undoubtedly be negligible, 
but we should at least explain why we can ignored them.

Line 196, Table 5 and further:
Helicity is given as |1|. This does not seem logic,
if the vertical bars mean the absolute value of the helicity.
The bars should then be around the helicity variable, |\lambda|=1.
If in an unknown definition of helicity a unit vector is meant, 
it should be indicated with a bold 1.

Section 6: 
How are the ratio of branching fractions measured?

Table 10: How does the specific choice of signal and background models 
and parameters affect the systematics? 
What uncertainty is assigned to varying these models and parameters?


Comments in detail
==================

- Since this is an untagged analysis, we do not distinguish between Bs0 
  and Bs0bar events. 
  Therefore, all 'Bs0bar' should be changed to 'Bs', starting
  from the title to the last sentence of the Conclusions.

- The first five pages should have lowercase roman page numbers.

Abstract:
- Remove :The decay ... violation".
  This is a very good sentence for the introduction where we motivate
  our paper, but not for the abstract where we shortly summarize the
  results.
- Remove "A detailed understanding of its substructure is imperative 
  in order to optimize its usefulness." 
  It is not made clear which 'substructure' is meant, 
  who is requiring in a 'imperative' way,
  and for what it should be 'useful'.

- Replace "> 97.7\%" by "larger than 97.7\%", 
  This should also be replaced in all other occasions where '<' is
  not part of a mathematical expression (lines 65, etc.). 

Line 59: 
- Replace "the primary vertex" by "a primary vertex".
  A collision event can have more than one PV.

Figure 2:
- Replace "The data has been fit with a double ... dashed line"
  by "A double ... dashed line was fitted to the data".

Line 78-79: 
- Remove "P-wave". 
  A fit to the P-wave contribution cannot have a fraction of 
  S-wave contribution.

Line 100-101:
- We should give the Bs mass window. 
- We should state that the background fraction is fixed to the amount 
  found from the invariant mass fit in Fig. 3, 
  as is done in the analysis note

Line 104:
- Suggestion to replace "therefore would not ... disentangling"
  by "might not add discriminating information in the analysis of".
  "Little structure" is also information and 'therefore' is not
  necessarily true.

Line 110: 
- Replace "To ensure that there are no visible ..." 
  by "To investigate if there are visible ...". 
  'Ensure' suggests that we force it not to have structure, 
  whereas we want to investigate it.

Line 116: 
- Replace "the daughter momenta of Bs meson" 
  by "the momenta of the daughter particles of the Bs meson".

Line 138: 
- Omit ',' after gKK.

Figure 7:
- Give the unit for the color scale: 'arbitrary units' or '%'.
- What is meant by 'relative' efficiency?
  Suggestion: "Parametrized detection efficiency as a function
  the invariant mass squared of the Dalitz pairs, m^2(pi+pi-)
  and m^2(J/psi pi+), of the Bs->J/psi pi+pi- decay."

Line 145: 
- Replace "Each axis ranges" by "Both s12 and s13 range".
- We suggest to re-order Lines 144-148 as the number 18.9 comes 
  out of the blue in Eq. (15).

Line 149:
- Add "of $\epsilon$(s12,s23)" after "polynomial function".

Line 160: 
- Replace "from rho(770)" by "combinatorial background formed by J/psi 
  and random rho(770)" for clarity.

Line 164: 
- A typo in "The are".

Figure 8: 
- Change the labels in the plots as this is simulation, not LHCb data.

Figure 10:
- Refer to the line in the figure caption.

Line 174b:
- Suggestion to remove 'the' in "listed in the Table 1".

Figure 11:
- Replace in the caption "background fit with the function" 
  by "background and the fitted function".

Line 183: 
- Change this sentence.

Line 185:
- Replace "try the addition of additional resonances"
  by "allow for additional resonances".

Table 2: 
- Change "non-resonance" to "non-resonant".

Line 203-208: 
- We suggest to merge these lines with 5.2 into one section.
  This is (partially) repeated in Section 5.2; 

Line 218:
- Replace "is improved noticeable adding components up"
  by "is only slightly changed with the addition of other resonances".

Line 238: 
- Remove first "In all cases".
  It appears twice.

Line 264: 
- Replace "alternative" by "alternate".

Line 299: 
- Replace "measured with in bins" 
  by "measured in bins" 
  and 'p_t' in 'p_T'.

Line 304b-d:

- Remove the ambiguity of "taking into account",
  which still can be read due to the long list of corrections 
  including the detection efficiency for the J/phi phi channel, 
  presumably applying to phi->K+K-.
  A suggestion might be: "multiplying by"
  or earlier "the J/psi K+K- yield divided by B(phi->K+K-)".

Line 306: 
- Remove extra '.'.

Line 323-324:
- Replace "the one found earlier" by "our first observation
  with 33 pb^-1 integrated luminosity [1]."
- Write "The quoted result for the decay specific
  ratio of the branching fractions 
  B(Bs->J/psi f0, f0->pi+pi-)/B(Bs->J/psi phi, phi->K+K-) 
  multiplied by the branching fraction B(phi->K+K-)=(48.9\pm 0.5)% 
  is 0.123+0.026/-0.022 to be compared with the value given above."

  Important is to avoid any possible suggestion that in this 
  or the other paper the numbers might be incorrectly presented.

Line 342:
- Remove 'is' in "uncertainty is assigned".

Line 364:
- Add "pi+pi-" before final states for clarity.
- Is 'individual' correct in "by individual interfering decay amplitudes"?

Line 373-374: 
- Rewrite the sentences.

Line 376b:
- Replace "The total rate in this final state is given by the ratio"
  by "The measured relative branching ratio for the total pi+pi-
  invariant mass range is".

Ref [13]: 
- Replace "Flatte" by "Flatt\'e".

Ref [14]: 
- Replace "Sjostrand" by "Sj��strand". 
  We suggest to use the LHCb template bib file.