Institutial Review of CMS paper SUS-16-033

Instructions

We kindly ask you to submit a review before the deadline of Fri, Mar 31, 2017. Please see the detailed record of the paper at http://cms.cern.ch/iCMS/analysisadmin/cadi?ancode=SUS-16-033

Please acknowledge this mail and confirm that you will provide a review in due time by clicking http://cms.cern.ch/iCMS/analysisadmin/sirpaccept?id=649

A guideline for the type of comments expected can be found at the end of this message. We very much look forward to receiving your review.

Best regards

cheers, Bill CMS Publication Committee

=============================================

Guidelines for the Institutional Comments:

Spirit and intent of the Collaboration-Wide Review:

As a general guideline, we ask that individuals and groups reviewing a CMS publication draft concentrate on substantive matters of science and questions of clarity. For the sake of streamlining the process, comments to a CMS publication draft are divided in two sets:

A. English/Style/Formatting (including figures)

B. Everything else (e.g. strategy, paper structure, emphasis, additions/subtractions, etc).

Comments of type B are the responsibility of the authors, who have to furnish clear answers to all such comments on CDS before the Final Reading can be scheduled. Comments of type A are sent to the Language Editor responsible for [assigned to?] this paper, (S)he will decide on any actions arising from such comments. In general, responses to type A comments will not be posted on CDS, except in cases where an issue of language or style significantly affects the publication in question.

How to post comments during the CWR:

The review should be posted as a 'comment' to the paper, indicating clearly that this was a CMS Institutional Review, along with the name of your institution. All comments received from the reviewing institutions are answered by the authors and ARC on CDS. To post a comment, you visit the paper link (http://cms.cern.ch/cds/SUS-16-033) and click on 'Detailed Record'. This gives you access to the full paper and one of the links at the bottom says 'Submit a comment'.

Comments

Dominick

Type A

- figs 7,8,(9): errors on the ratio may be easier to see using a light solid gray band instead of a hatched band.

- fig 9: some of these may be easier to distinguish signal from bkg in the tails by using a log plot, especially the middle row plots.

- app B tables: should state somewhere that errors are first stat then syst.

Type B

- l122-126: a bit confusing to introduce charged tracks here without any motivation and describe them as not being an isolated electron or muon. Then the isolation criteria is described and it talks about identifying them as PF electrons or muons. Consider rephrasing to first state why isolated tracks are used then just defining them.

- l132-135: give pt and eta cuts for b jets

- sec 6: personal preference, but I find it a bit strange to describe signal systematics before the background methods

- l240: non-isolated -> isolated track or lepton?

- fig 6: why not show the R&S errors on the ratio?

- fig 9: why show HTmiss for T2bb and NJet for T2qq?

Giovanni

Type A

Type B

Mark

Type A

Type B

Mario

Type A

- General: missing line numbering before every equation, to correct.

- Abstract: would add "of supersymmetry" after "simplified models", since you only use simplified models of supersymmetry.

- L.15: would replace "elements" with "objects"

- L.33: would invert phrasing to introduce QCD multijet background: "... to evaluate the background from SM events with multijet final states ..., referred to in the following as QCD background"

- L.72: would replace "reaction" with "process"

- L.148: why not make the "i" counter explicit? j1 for the first, j2 for the second, and so on and so forth.

- L.156: would add "not identified" + "as such".

- L.158: would specify, dPhi_{MHT, ji} (with i = 1,2,3,4).

- L.167: mismeasurementS

- Fig. 2 (caption): I believe equation to infinity is mathematically undefined. Should write ">2100", and ">1000" respectively.

- L.256: would add "potential" signal contamination

- L.313: repetition of "missing"

- Eq.1: suggest making dependencies of each term on binning variables.

- L.388 + Fig. 4: why 20% and not 40%, as stated in L. 386?

- L.405: would replace "misreconstruction" with "mismeasurement"

- after L.414: "rebalance,". Correct typo

- Fig.5: lines in bottom panel stop at 2, despite range goes beyond 2

- Fig.5: why no band in lower panel for error?

- Fig.5: Should show errors for points at 0, both in top and bottom panels.

- L.453: Some points seem to be significantly away from 1, so I find this statement not true (if all uncertainties are displayed in the ratio). Would rephrase to "reduce" the statement, or explain why this is considered satisfactory.

- L.477: full sets of points are (very) far from closing. Is it understood why this is happening?

- L.479: this statement seems too strong. The two estimates do not agree by a significant amount for many bins. It looks like R&S is systematically providing larger estimate. Should explain why this is considered satisfactory, and why no additional uncertainty is accounted for to cover for this?

- Fig.6: why ratio goes negative? Not clear at all...

- Fig.6: what is the error on the ratio? It includes errors from both predictions? Find very confusing that different plots have different ways to show the error on the ratio, some with bands, others without. Need to fix it and make it homogeneous all throughout the paper.

- L.491: should add quantitative statement. How much worse these regions do wrt. the full search bins?

- Fig.7: error lines for zero points do not start at 0.

- Fig.7,8,9: zero points are not shown in ratio panel. Should fix this.

- Fig.7,8,9: it is singular to show the relative difference in the bottom panel. Why? Why not a simple ratio? Would be more interesting to see a simple ratio.

- L.520-522: does this mean you are including the CRs into the fit? The way it is phrased reads like your BG prediction is model dependent. Would rephrase to make it clearer.

- Fig.10,11: why not drawing line along diagonal. This is not in line with all other CMS SUSY analyses. Should fix.

- L.548: how do these limits compare to previous searches? I understand these models were not included in Ref.16, but there are results from different searches in CMS. Does this analysis extend existing results?

- L.565: Why specifying pT and not eta?

- L.568: remove "the" before data.

- L.571: add "of supersymmetry" after "simplified models".

- App. B: Pre-fit vs prefit (title)? Would keep the first and make it homogeneous.

- App. B: pre-fit background was never defined. Should define it.

- App. B: replace "+" with ">", and/or ">=" for bin thresholds.

Type B

-- MarioMasciovecchio - 2017/03/17