Comments from Conor 
===================
These comments are based on the draft dated 25 September.

General comments:
Overall, I think the paper is in pretty good shape. Most of my
comments are fairly minor. However, I do have a suggestion to change
how the plots and tables containing data are presented. At present,
plots containing data are shown as part of the bkg estimate section,
and as part of the NN discriminant description. I think it would be
better if all plots and tables containing data were put into a new
Results section. I explain more in the detailed comments.

% thanks 

Abstract: Good abstract. Perhaps you would like to add how the
cross-section limits obtained compare to the SM Higgs expected values?

%  added "... pb or 7.5 to 101.9 times the SM expectation for ..."

Page 1, line 36: 'one more' -> another

% fixed 

Page 3, line 41: "In this paper, we present ..."
Perhaps you should say this overview line before explaining the
analysis improvements in lines 35-40?
The order could be:
 - discuss previous CDF paper
 - state that this paper is update to to 1.9 fb-1
 - brief description of analysis improvements
 - contents of the current paper

% rephased 

Page 3, line 62: " about 30 um of which is due to ..." -> I think "of which
about 30 um is due to ..."  reads a little better, but this is just a
personal preference, so you are free to ignore it if you disagree

% fixed 

Page 3, line 66: "E_T = E sin(theta)" 
You already defined E_T in line 53, no need to do it again

% fixed 

P4, L 71: "Plug" is a little bit too much CDF jargon. Better to use
'forward' I think. Or at least define "plug" = "forward" right away,
eg "in the forward calorimeter, known as the 'plug' "

% fixed 

P4, L 81-88: do you require muons to have minimum ionizing deposits
in the calorimeter? Might be worth adding that to the nice description
here.

% No, no minimum energy is required in the calorimeter.

P6, L 142: add 'non-overlapping' or 'exclusive' to make it clear, as
in "Three non-overlapping b-tagged event categories are considered"

% fixed with ' exclusive' 

P6, L 143: 'event' -> 'events'

% fixed 

P6, L 145: 'NN' -> I dont think that NN has been defined yet. Prefer
to use Neural Net, or make sure the abbreviation is defined already.

% fixed 

P6, L147: "With a SecVtx mistag rate of ..."
I agree with this statement, but I dont see it's relevance to this
particular section? It belongs in a later section on the algorithms,
not in the event selection section.

% removed 

P6, L154: "vertex" -> "Vertex"  All your other subsections have
capital letters for every word...

% fixed 

P7, L 158: "NN" -> again, I prefer 'Neural Net'

% fixed 

P7, L 160-163: "The b-quark has a relatively long lifetime..."  
This type of conceptual discussion belongs in the introduction to this
section, before you start discussing any of the individual algorithms

% fixed 

P9, L 225: "heavy-flavor jets from events with an electron candidate"
You should explain briefly why this sample is used for
validation. It's not clear to me. Is it a b-enriched sample?

% fixed 

P9, L 228: you should define the NN output variable first. Something
like: "The output of the neural net is a value ranging from 0 to 1,
denoting the probability that ... etc"

% fixed 

P9, L 232: "At these cut values .." 
I would put this sentence directly after the values chosen, before
talking about the scale factors.

% fixed 

P9, L 228-234: Clearly the selection of the cut value is a trade-off
between efficiency and purity. But its not clear from the current
description why you chose the 90% efficiency value. For example, if
you were willing to cut tighter and reduce the efficiency to 80%,
would it massively or only marginally improve the purity?

%  The mistag is small part of background, so, cut hard does not improve
   the search sensitivity. 

P10, L 236 onwards: It's a nice description, but I think it would help
the reader if you could more explicitly address how jet-prob tagging
differs from SECVTX?

% rephased 

P10, L 246: 'While' -> 'Since'

% fixed 

P10, L 251: '72 categories' -> the reader may be interested in how you
make 72 categories - 4 bins in eta, 9 in pt, 2 in quality?

% removed 72. 

P11, L 292: "due to" -> "of"

% fixed 

P11, L 292: "mimic" -> "mimicked"

% fixed 

P11, L 293: remove 'production', just leave "by other processes"

% fixed 

P12, L 299: 'with' -> 'by producing'

% fixed 

P12, L 302: 'be observed via' -> 'result from'

% fixed 

P12, L 303-304: 'due to some unknown reasons' - this may be true, but
I wouldnt say it like this!

% rephased as ' due to some limitations '

P12, L305: '(pretag)' -> '(known as the 'pretag' sample)'

P12, L 311: '4 sidebnad sectors' - well, one is actually the signal
region, not a sideband, so better just to say '4 sectors'

% rephased 

P14, L 358: 'due to MET trigger bias' - suggest to remind the reader
that the MET+PEM trigger is used for forward region, which causes the
MET trigger bias

% fixed 

P14, L 369: 'jets with at least 2 tracks well measured in the silicon
detector' - taggable jets were already defined on P13, L 327, so no
need to define them again

% ok as is since the previous definition is no longer exist.  

P14, L 371-378: suggest a slightly different ordering to the
information presented in this paragraph, to make it a bit
clearer. Perhaps something like:

"Negative mistags are defined as tags with unphysical...
The positive mistag rate can be obtained from the negative mistag rate
with an additional correction factor, reflecting an enhancement of
positive mistags due to light-flavor ...
The correction factor is measured in a control sample ...
The systematic uncertainty ..."

% rephased as suggested 

P14, L 378: 'control sample' - how is this control sample defined?

%  inclusive jet samples 

P15, L 389: 'used' -> 'derived and applied'

% fixed 

P15, L 394: 'programs' -> 'event simulators'?

% fixed

P15, L 399: 'programs' -> 'simulations'?

% fixed  

P15, L 403-406: the description of 1B,1C,2B,2C should go in the
caption for Table II, as is done already for Table III

% fixed 

P18, Section E: it seems to me that this section, which is described
as 'Summary of Background Estimate' is really the 'Results' section.
The figures and tables here are not just the background - they also
show the data!
I think a better idea would be to create a new section called
'Results' and put these figures and tables in it.

The ordering of the sections would now be:
  Sec V Background (subsections A-D; no need to include the summary E)
  Sec VI Signal Acceptance
  Sec VII NN Discriminant
  Sec VIII Results (encompassing all plots and tables that contain
                      data, including the NN discriminant plots)
  Sec IX Higgs Limit (or perhaps make this a subsection of Results?)

The idea behind this suggested reordering is that all plots and tables
that contain data are collected together in the same Results
section, and all analysis methods (bkg estimates, signal acceptance, NN
discriminant) are described before any data is presented.

% created a new result section as you suggested and made Higgs limit as subsection of Results 

P18, Fig 1: 'In this plots, the number of central and plug region are
merged' -> 'This plot combines the information from the central and
plug regions of the detector.' Although you say this in the text
anyway, so it may not be necessary to say it in the caption.

% fixed 

Also, the order of tables and figures should match the way it is
described in the text. One way that seems reasonable to me is to put
the figures first, since they show all jet bins, then show the tables,
since they have the details of the signal bin. Alternatively, you
could separate them according to b-tag category, and give the
appropriate fig and table for each category in order. It doesn't
matter too much, as long as there is some logic to it, and the text
description should match that.

% rephased as suggested 

Table V, VI, VII: 'as a function of jet multiplicity' should be removed from
the caption in all these tables, since these results are just for the
signal bin Njet=2.
Also, 'central' and 'plug' should probably start with uppper-case letters.

% fixed 

P19, L 442: 'programs' -> 'generators'? [I just dont like calling
these programs]

% fixed 

P19, L 442: 'The PYTHIA program was used ...' -> just 'PYTHIA is used
...' Note also the suggested change to present tense.

% fixed 

P19, L 447: 'where epsilon ...respectively ' -> I have a preference
for explaining the symbols as 'where A is ..., B is ... ' but this is
not a big deal, so the authors should do what they prefer here. I just
point out my preference.

% ok as is 

P20, L 453: 'kinematics' is a very jargon-y word. Do you mean the
'event selection criteria'?

% fixed 

P20, L 455-456: It may be worth a little more explanation of where the
e_z0, e_trigger and e_lepton are obtained from.

% rephased. 

P22, Fig 2: why are the plots from these two b-tag categories combined
into the same figure, while the SECVTX+NN gets a figure all of its
own? Unless there is a particular reason for this, I suggest you give
each b-tag category its own figure, so that the current Fig 2 gets
split into 2 separate figures.

% will combine two double tags together since S/B is similar. So, there will
  be one for single tag and one for double tags.

P22, Fig 3: 'various b-tag strategies' - this phrase occurs a lot, but
I don't really like it. I prefer 'categories' to 'strategies'. And
'various' sounds a bit too random ... how about 'the selected' or
something like that to reinforce the idea that you intentionally chose
these 3 categories?

% fixed 

P23, Table VIII, IX: In the tables themselves, replace 'One tag w/NN tag' with
'One SECVTX with NN tag'. That should be clear enough, so you can
remove the explanation of the tag categories from the captions. The
reader will be familiar with the 3 categories by this stage of the
paper.

% fixed 

P23, L481-482: this statement of the increase in acceptance gained
from the inclusion of plug electrons probably belongs somewhere
earlier in this section, rather than just right at the end...

% fixed 

P24, L 491: '...from a list of 76 possible choices considered from two
jets, MET, and lepton kinematics and correlation between them'
- this part of the sentence could be phrased better. Perhaps something
like: "... from a list of 76 possible variables, based on the
kinematic information of the two jets, lepton and MET in the events
(including correlations between these objects)."

% fixed 

P24, L 500: 'constant against the Higgs mass' -> 'constant as a
function of the Higgs mass'

% fixed 

P24, L 504: personally, I would replace the comma here with a colon
before the list of the 6 variables, and then in each item, I would
replace the current colon, with a comma. However, I'm sure the PRD
copyeditors will have their own suggestions on how to handle this.

% fixed 

P25, Fig4 & P26 Fig 5:  why not combine these two into 1 figure which
shows all six variables used in the NN?
At the same time, since they contain data, I think they should be
moved into the new Results section that I proposed earlier.

% plots are no longer needed. 

P27, Fig 7: similar comment to Fig 2 - why are these 2 categories
combined while the other category gets its own figure? Unless there is
some reason, I think each category should get its own figure.

% will combine two double tags together 

P27, Eq 13: this equation and its accompanying text seems unnecessary
to me. Eq 15, since it includes syst uncertainties, is the only one
you need to give and explain here.

% removed eq 15. 

P28, L 553: 'credibility' -> 'confidence'?

% ok as is 

P28, L 561: you could probably shorten this sentence to just say 'the
three 3 b-tag categories' if you wanted...

% fixed 

P28, L 573: 'of forward' -> 'of the forward' 
% fixed 

P29, Fig 8: the caption should state that this is not the
cross-section limit, but rather the limit divided by the SM
expectation.

% fixed 

P29, Conclusion: it might be nice to add here how the observed limit
corresponds to the SM expectation, ie the information in Fig 8.

% fixed