Electron efficiency in LHC Run-2 with the ATLAS experiment

Otilia Ducu

Electron efficiency in LHC Run-2 with the ATLAS experiment

T0 review · 0 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper reports that ATLAS electron reconstruction efficiency exceeds 97% in LHC Run-2, with data-to-simulation ratios close to unity and sub-0.1% uncertainties.

desk verdict A clean, honest conference-proceedings summary of already-published ATLAS electron efficiency results; no new measurement, but a serviceable overview with minor internal inconsistencies. read the letter →

arxiv 2412.19323 v1 pith:U6IEHFMQ submitted 2024-12-26 hep-ex

Otilia Ducu This is my paper · ORCID

classification hep-ex PACS 06.30.-k14.60.-z

keywords electronefficiencyreconstructionidentificationisolationpromptelectronsfake/non-promptTag-and-ProbeLHCRun-2

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper reports how well the ATLAS detector finds and identifies electrons in 13 TeV proton-proton collisions from LHC Run-2, and how well the detector simulation reproduces that performance. The headline numbers are a reconstruction efficiency above 97%, identification efficiencies that range from roughly 60% at low transverse momentum to above 80% at high momentum, and correction factors close to one for all three identification and five isolation working points. The paper argues that these results, obtained with the Tag-and-Probe method in Z→ee and J/ψ→ee events, show that ATLAS Monte Carlo simulation describes electron performance well enough that the small correction factors can be applied to simulation with confidence. The practical payoff is that precision measurements of the Higgs boson and Standard Model processes, which rely on electrons, can trust their electron-related systematic uncertainties.

What carries the argument

The central machinery is the Tag-and-Probe method applied to Z→ee and J/ψ→ee events: one electron (the tag) triggers the event, while the other (the probe) tests the selection efficiency. Efficiency is extracted from fits to the invariant-mass distribution (Z-mass) or to the isolation-cone transverse energy (Z-isolation), and at low E_T from the pseudo-proper-time variable in J/ψ→ee events. Reconstruction itself rests on superclusters, which are variable-size calorimeter clusters that capture bremsstrahlung energy, plus track matching with a Gaussian-sum-filter refit and a likelihood discriminant built from inner-detector and calorimeter shower-shape variables. The isolation working points combine corrected calorimeter isolation, with pile-up subtraction from the ambient energy density, and track isolation with fixed or variable cone sizes.

What would settle it

Compare the Tag-and-Probe efficiency from Z→ee events with an independent measurement based on W→eν events in the same E_T and η bins; a disagreement beyond the quoted uncertainties would show that the background model is biased.

Watch

Extended reading notes

Core claim

Using the full 139 $fb^{-1}$ Run-2 dataset, the electron reconstruction efficiency in ATLAS is above 97%, and the data-to-Monte Carlo reconstruction efficiency ratio is close to unity with uncertainties generally below 0.1%. The identification efficiency depends strongly on transverse momentum: the Loose, Medium, and Tight likelihood working points give efficiencies that can be as low as about 60% for Tight at low E_T and rise to above 80-90% at high E_T, with data-to-simulation correction factors within 5% for all three working points. The five isolation working points—HighPtCaloOnly, TightTrackOnly VarRad, TightTrackOnly FixedRad, Tight VarRad, and Loose VarRad—have efficiencies that depend on the identification working point and on pile-up, but their correction factors also stay close to one, within 5-7%. An improved background-estimation method reduces the identification efficiency uncertainty by 30-50% in the low-E_T region.

Load-bearing premise

The whole measurement rests on the assumption that the background estimates in the Z→ee and J/ψ→ee Tag-and-Probe selections are accurate enough that the extracted probe efficiency equals the true electron efficiency.

Editorial extensions

If this is right

If these numbers are correct, ATLAS analyses can apply reconstruction correction factors of order unity with systematic uncertainties below 0.1% for electrons above 30 GeV.
Identification correction factors within 5% mean that the Loose, Medium, and Tight working points can be used in precision measurements without large efficiency penalties.
The isolation working points, especially Tight VarRad below 60 GeV and HighPtCaloOnly above 80 GeV, provide a practical recipe for rejecting fake and non-prompt electrons while keeping high prompt-electron efficiency.
The improved background estimation that cuts identification uncertainty by 30-50% at low E_T makes low-mass electron measurements, such as J/ψ and Drell-Yan processes, more precise.
The paper's comparison with the CMS experiment implies that electron performance is comparably good in both general-purpose detectors.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A corollary the paper leaves implicit is that if the data-to-simulation ratios are this close to unity, electron-channel systematic uncertainties in ATLAS Run-2 measurements are likely dominated by energy scale and resolution rather than by efficiency correction.
The paper's suggestion that particle-flow-based isolation will replace traditional isolation implies that Run-3 analyses should re-measure the same working points with particle-flow inputs to verify that the claimed efficiency gains persist at higher pile-up.
One could test the sub-0.1% reconstruction-ratio uncertainty by comparing the Tag-and-Probe result with a completely independent method, such as using W→eν events, in the same E_T and η bins.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Bottom line: this is a proceedings write-up, not a research paper. It compiles previously published ATLAS electron reconstruction/identification/isolation efficiencies (Refs 15, 16, 19) and explicitly reuses their figures. So the correct question is whether it is a fair and accurate review, not whether it breaks new ground. On that question, it largely succeeds. The text is clear, the selection of results is representative, and the Tag-and-Probe premise is honestly flagged as external – the paper points to Refs 15–17 for the method and notes that background estimation is the hardest part. Good.

What it adds beyond the cited papers: a short ATLAS-vs-CMS comparison and a recommendation of five isolation working points, both drawn from the same ATLAS performance documents. There is no new measurement, no new correction factor, no new derivation. That is fine for a conference proceedings, but it means the arXiv listing as hep-ex might mislead someone expecting a new result.

Soft spots are minor but real. The abstract says reconstruction data/MC uncertainties are 'generally smaller than 0.1%', while Section 3 says less than 0.5% below 30 GeV and 0.1% above. Section 4 says Tight identification can be as low as 60% below 15 GeV; the conclusions say as low as 70% below 50 GeV. These are presentation-level discrepancies, not load-bearing, because the underlying numbers come from the referenced papers and the figures are unchanged. Still, if someone reads only the abstract they get the wrong impression.

The paper leans on collaboration-internal references (15, 16), which is natural here – those are the primary sources. I would not flag self-citation as a problem.

Who it's for: a student or colleague who wants a quick orienting summary of ATLAS electron efficiency results without reading the full JINST papers. As a refereed contribution to a proceedings volume, it deserves light peer review focused on consistency with the cited sources. As an arXiv preprint, it is not a substantive new research claim. My recommendation: treat it as a review document; if the venue is a proceedings, send it for a light check; if it is meant as a research paper, tell the authors to reframe it as a review. Either way, no serious referee time is needed.

Referee Report

0 major / 5 minor

Summary. This manuscript is a conference-proceedings-style review of the ATLAS Run-2 electron reconstruction, identification, and isolation efficiency measurements, based on a presentation at ICNFP 2022. It describes the supercluster reconstruction algorithm, the Tag-and-Probe measurements using Z->ee and J/psi->ee events, the three likelihood-based identification working points, the five recommended isolation working points, and the data-to-Monte-Carlo correction factors, with comparisons to CMS performance. The paper presents no new measurement or derivation; its numerical results are taken from ATLAS publications and public plots, chiefly Refs. 15 and 16. The central claims are that reconstruction efficiency exceeds 97%, data/MC ratios are close to unity with uncertainties below roughly 0.1-0.5%, and identification and isolation correction factors are within a few percent of unity.

Significance. The document is a useful and generally clear summary of an important set of efficiency measurements that underpin ATLAS precision physics. Its significance is derivative: it does not add new experimental information, but it consolidates the Run-2 electron performance results, explicitly identifies background estimation in the Tag-and-Probe method as the most challenging ingredient (Section 4), and gives an accessible comparison with CMS. I find no internal circularity or omitted derivation, because the manuscript does not attempt to re-derive the measurement; the correctness of the reported efficiencies rests on the underlying ATLAS analyses, in particular Refs. 15-17, which is appropriate for a review. The presentation-level inconsistencies noted below do not put the physical conclusions in doubt but should be corrected before publication.

minor comments (5)

[Abstract vs Section 3] The abstract states that the data/MC reconstruction efficiency ratio has 'associated uncertainties generally smaller than 0.1%', whereas Section 3 states the systematic uncertainties are 'generally less than 0.5% (0.1%) when ET < 30 GeV (ET > 30 GeV)'. Please harmonize these statements, since the abstract alone overstates the precision at low ET.
[Section 4 vs Section 7] Section 4 says the Tight identification efficiency is 'as low at 60% below 15 GeV', while the Conclusions say that below 50 GeV the Tight working point can be 'as low as 70%'. The two numbers describe different ET ranges, but the wording invites misreading; please state the ET range explicitly in both places and correct the typo 'as low at'.
[Section 5] The sentence 'Figure 6, left, shows the results are shown as a function of the average number of interactions per bunch crossing' should refer to the right-hand panels of Figure 6; the left panels show ET dependence.
[Section 7 vs Section 3] The conclusion that reconstruction efficiency is 'better than 95%' for ET > 10 GeV is weaker than the Section 3 statement that the electron reconstruction efficiency is 'above 97%'; clarify that the former refers to the simulated stepwise efficiencies in Figure 4 (left) while the latter refers to the data efficiency in the Z->ee selection.
[Section 5, Eq. (1) and Table 1] Please fix the typo 'rspectively' in the Table 1 caption and make the cone-size notation consistent between Eq. (1) (EconeXX_T) and the text (Econe20_T), since the current mixture of XX and numeric subscripts is confusing.

Circularity Check

0 steps flagged · score 0.0 of 10

Honest non-finding: the paper is an explicit review of published ATLAS electron-efficiency measurements and contains no derivation that could reduce to its inputs.

full rationale

The document is a conference-proceedings review (Sections 1 and 7, plus the statement 'Document based on a presentation at ICNFP 2022'). It reports efficiencies and correction factors that are reused, with permission, from Refs. 15, 16, and 19 (ATLAS electron/photon performance papers and an ATLAS public plot). The only load-bearing premise is the Tag-and-Probe background estimation in the underlying ATLAS analyses; the paper explicitly attributes that method to earlier work ('the measurement is performed using the Tag&Probe method presented in Ref. 17' in Section 3; 'The most challenging task, for all methods, is the precise estimation of the background' in Section 4). Those cited measurements are data-driven, externally falsifiable analyses, not assumptions whose conclusion is this review. No equation in the paper is used to derive a result from a fitted parameter, and no 'prediction' is manufactured from the review's own inputs. The only internal issues are presentation-level discrepancies (abstract's 'uncertainties generally smaller than 0.1%' vs Section 3's 'less than 0.5% (0.1%) when ET<30 GeV (ET>30 GeV)', and Section 4's 60% low-ET Tight efficiency vs the conclusion's 70% below 50 GeV), which are accuracy/consistency concerns, not circularity. Therefore no circular step is present and the score is 0.

Assumptions & free parameters 0 free parameters · 2 assumptions · 0 invented entities

The paper introduces no new free parameters, invented entities, or ad hoc assumptions. It relies on standard assumptions of the underlying ATLAS analyses: that Tag-and-Probe selections provide unbiased efficiencies and that the MC simulation models the detector well enough for correction factors to be valid.

assumptions (2)

domain assumption The Tag-and-Probe method using Z to ee and J/psi to ee events yields an unbiased estimate of the true electron efficiency.
The paper relies on this in Sections 3 to 5 to claim measured efficiencies without presenting the full background estimation or closure tests.
domain assumption The ATLAS MC simulation models the detector response well enough that the data/MC correction factors are approximately unity and can be applied to MC.
The correction factors are derived from the ratio of data to MC efficiency; if the MC simulation were biased in a way that the Tag-and-Probe does not catch, the factors would not be correct. This is a standard assumption in the cited refs.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Electron efficiency in LHC Run-2 with the ATLAS experiment." pith.science (2026). https://pith.science/paper/U6IEHFMQ

@misc{pith2026241219323,
  author       = {Pith},
  title        = {Pith review of: Electron efficiency in LHC Run-2 with the ATLAS experiment},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/U6IEHFMQ}},
  note         = {Machine review of arXiv:2412.19323}
}

abstract

The document presents a general overview of the electron reconstruction, identification and isolation performance in the ATLAS experiment. The results are obtained using 13 TeV proton-proton collision data collected during the LHC Run-2. The electron reconstruction efficiency is higher than 97%, and the ratio of data to Monte Carlo simulation efficiency is close to unity, with associated uncertainties generally smaller than 0.1%. The electron identification is shown for three working points, and depending on the electron $E_T$, it can be as low as 60%, increasing to more than 80% above 50 GeV. The correction factors are close to one, generally within 5%. Five isolation working points are recommended in the ATLAS experiment, to successfully reject fake/non-prompt electrons. Their dependency on the electron identification working points is shown and discussed, as well as their pile-up dependency, and their performance versus electron $E_T$ and $\eta$. Document based on a presentation at the XI International Conference on New Frontiers in Physics (ICNFP 2022). keywords; prompt electrons, reconstruction, identification, isolation, fake/non-prompt electrons

Figures

Figures reproduced from arXiv: 2412.19323 by the authors.

**Figure 2.** Illustration of the electron path through the ATLAS sub-detectors. The red, hashed line [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Summary of various steps performed for the electron and photon reconstruction in AT [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: Electron reconstruction efficiency. Reused with permission from Refs. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

**Figure 5.** Figure 5: Electron identification efficiency. Reused with permission from Refs. [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

**Figure 6.** Figure 6: Electron isolation efficiency, for Loose (top), Medium (middle) and Tight (bottom) elec [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 8 canonical work pages

[1]

ATLAS Collaboration, Observation of a new particle in the search for the Standard Model Higgs boson with the ATLAS detector at the LHC , Phys. Lett. B 716 (2012) 1, [arXiv: 1207.7214 [hep-ex]]

arXiv 2012
[2]

CMS Collaboration, Observation of a new boson with mass near 125 GeV in pp col- lisions at √s = 7 and 8 TeV , JHEP 06 (2013) 081, [arXiv: 1303.4571 [hep-ex]]

arXiv 2013
[3]

131 (2023) 251802, [arXiv: 2308.04775 [hep-ex]]

ATLAS Collaboration,Combined measurement of the Higgs boson mass from the H → γγ and H → ZZ ∗ → 4ℓ decay channels with the ATLAS detector using √s = 7, 8 and 13 TeV pp collision data , Phys.Rev.Lett. 131 (2023) 251802, [arXiv: 2308.04775 [hep-ex]]

arXiv 2023
[4]

ATLAS Collaboration, Evidence of off-shell Higgs boson production from ZZ leptonic decay channels and constraints on its total width with the ATLAS detector , Phys. Lett. B 846 (2023) 138223, [arXiv: 2304.01532 [hep-ex]]

arXiv 2023
[5]

ATLAS Collaboration, Differential cross-section measurements of the production of four charged leptons in association with two jets using the ATLAS detector , CERN- EP-2023-152, [arXiv: 2308.12324 [hep-ex]]

arXiv 2023
[6]

ATLAS Collaboration,Computer generated image of the whole ATLAS detector, [CDS: https://cds.cern.ch/record/1095924]

work page arXiv
[7]

ATLAS Collaboration, Electron reconstruction and identification in the ATLAS ex- periment using the 2015 and 2016 LHC proton-proton collision data at √s=13 TeV, Eur. Phys. J. C 79 (2019) 639, [arXiv: 1902.04655 [hep-ex]]

arXiv 2019
[8]

Evans and P

L. Evans and P. Bryant,LHC Machine , JINST, vol. 3, p. S08001

work page

Show all 23 references

[9]

Bruning, P

O. Bruning, P. Collier, et al.,LHC Design Report , CERN-2004-003-V-1 Electron efficiency in LHC Run-2 with the ATLAS experiment 14 Otilia Ducu

2004
[10]

ATLAS Collaboration, The ATLAS Experiment at the CERN Large Hadron Collider , JINST, vol. 3, p. S08003, 2008

2008
[11]

ATLAS Collaboration, Expected Performance of the ATLAS Experiment - Detector, Trigger and Physics , 2009, [arXiv: 0901.0512 [hep-ex]]

2009 arXiv
[12]

CMS Collaboration,The CMS Experiment at the CERN LHC, JINST 3 (2008) S08004, [CDS: https://cds.cern.ch/record/1129810]

2008
[13]

ATLAS Collaboration,Performance of the ATLAS Trigger System in 2015, Eur. Phys. J. C 77 (2017) no.5, 317, [arXiv:1611.09661 [hep-ex]]

2017 arXiv
[14]

ATLAS Collaboration, The ATLAS Collaboration Software and Firmware , ATL- SOFT-PUB-2021-001, https://cds.cern.ch/record/2767187

2021
[15]

ATLAS Collaboration,Electron and photon performance measurements with the AT- LAS detector using the 2015-2017 LHC proton-proton collision data , JINST 14 (2019) P12006, [arXiv: 1908.00005 [hep-ex]]

2019 arXiv
[16]

ATLAS Collaboration,Electron and photon efficiencies in LHC Run 2 with the ATLAS experiment, CERN-EP-2023-182, [arXiv: 2308.13362 [hep-ex]]

2023 arXiv
[17]

ATLAS Collaboration, Electron efficiency measurements with the ATLAS detector using 2012 LHC proton–proton collision data , Eur. Phys. J. C 77 (2017) 195, [arXiv: 1612.01456 [hep-ex]]

2017 arXiv
[18]

CMS Collaboration, Electron and photon reconstruction and identification with the CMS experiment at the CERN LHC , JINST 16 (2021) P05014, [arXiv: 2012.06888 [hep-ex]]

2021 arXiv
[19]

ATLAS Collaboration, Electron efficiency with full Run2 , https://atlas.web.cern.ch/Atlas/GROUPS/PHYSICS/PLOTS/EGAM-2022-02/

2022
[20]

ATLAS Collaboration, Tools for estimating fake/non-prompt lepton backgrounds with the ATLAS detector at the LHC , JINST 18 (2023) 11, T11004, [arXiv:2211.16178 [hep-ex]]

2023 arXiv
[21]

ATLAS Collaboration, Higgs boson production cross-section measurements and their EFT interpretation in the 4 ℓ decay channel at √s = 13 TeV with the ATLAS detector, Eur. Phys. J. C 80 (2020) 957, [arXiv:2004.03447 [hep-ex]]

2020 arXiv
[22]

ATLAS Collaboration,Test of CP-invariance of the Higgs boson in vector-boson fusion production and its decay into four leptons , CERN-EP-2023-030, [arXiv:2304.09612 [hep-ex]]

2023 arXiv
[23]

Seth Moortgat, Lepton Isolation Using Particle Flow Objects for the ATLAS Detector , CERN-STUDENTS-Note-2014-208, https://cds.cern.ch/record/1756841

2014

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

ATLAS Collaboration, Observation of a new particle in the search for the Standard Model Higgs boson with the ATLAS detector at the LHC , Phys. Lett. B 716 (2012) 1, [arXiv: 1207.7214 [hep-ex]]

arXiv 2012

[2] [2]

CMS Collaboration, Observation of a new boson with mass near 125 GeV in pp col- lisions at √s = 7 and 8 TeV , JHEP 06 (2013) 081, [arXiv: 1303.4571 [hep-ex]]

arXiv 2013

[3] [3]

131 (2023) 251802, [arXiv: 2308.04775 [hep-ex]]

ATLAS Collaboration,Combined measurement of the Higgs boson mass from the H → γγ and H → ZZ ∗ → 4ℓ decay channels with the ATLAS detector using √s = 7, 8 and 13 TeV pp collision data , Phys.Rev.Lett. 131 (2023) 251802, [arXiv: 2308.04775 [hep-ex]]

arXiv 2023

[4] [4]

ATLAS Collaboration, Evidence of off-shell Higgs boson production from ZZ leptonic decay channels and constraints on its total width with the ATLAS detector , Phys. Lett. B 846 (2023) 138223, [arXiv: 2304.01532 [hep-ex]]

arXiv 2023

[5] [5]

ATLAS Collaboration, Differential cross-section measurements of the production of four charged leptons in association with two jets using the ATLAS detector , CERN- EP-2023-152, [arXiv: 2308.12324 [hep-ex]]

arXiv 2023

[6] [6]

ATLAS Collaboration,Computer generated image of the whole ATLAS detector, [CDS: https://cds.cern.ch/record/1095924]

work page arXiv

[7] [7]

ATLAS Collaboration, Electron reconstruction and identification in the ATLAS ex- periment using the 2015 and 2016 LHC proton-proton collision data at √s=13 TeV, Eur. Phys. J. C 79 (2019) 639, [arXiv: 1902.04655 [hep-ex]]

arXiv 2019

[8] [8]

Evans and P

L. Evans and P. Bryant,LHC Machine , JINST, vol. 3, p. S08001

work page