{"id":"c8aa2ef0-fbfb-441c-99f0-ad754052a4db","arxiv_id":"2504.16669","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A Bayesian framework, SideKicks.jl, infers black-hole natal kicks and mass loss from inert black-hole binaries using full 3D velocity and orbital data, and tightens VFTS 243's kick upper limit to about 27 km/s.","lead":"Astronomers built a new Bayesian tool to measure the kick a black hole receives at birth by combining the full three-dimensional motion and orbit of black-hole binaries with a normal star. Applied to the system VFTS 243, it constrains the natal kick to below about 27 km/s and the mass loss to below about 2.9 solar masses.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"VFTS 243 birth-velocity prior is specified as sigma=(12,12,12) km/s in Sec. 2.5 but sigma=(42,40,11) in Table 5; the headline 90% kick/mass-loss limits depend on this width and are not currently reproducible.","rationale":"Reader's weakest_assumption is the circular pre-collapse orbit prior. I agree that assumption is fragile: the paper's own Fig. 5c shows a misspecified circular prior on an e_i = 0.3 system returns tight but biased posteriors, and Fig. C.2 shows the VFTS 243 vk 90% upper limit rises from about 26.5 km/s to about 76 km/s under a flat eccentricity prior. However, this fragility is acknowledged and discussed in Sec. 5.1, and the abstract conditions the recovery on circularity. The more readily checkable threat to the exact headline numbers is the internal inconsistency in the birth-velocity prior width. The observable Delta_v is the difference of two velocities, so the prior on v_i is load-bearing; a factor of about 3.5 in its width can absorb or expose the observed offset. This is not a disagreement with consensus; it is a reproducibility issue in the paper's own specifications. The method's strengths (explicit forward model, open code, injection tests that validate the machinery when the priors are correct) are real, so the concern does not warrant rejection. The reader's CONDITIONAL verdict should stand, with the additional condition that the Venv prior values be reconciled and a sensitivity test reported.","tokens_in":33792,"tokens_out":11941,"duration_ms":112581,"concrete_test":"Open the archived VFTS 243 SideKicks.jl configuration (Zenodo 10.5281/zenodo.15196531, injection scripts 10.5281/zenodo.15267344) and extract the exact Normal priors used for v_alpha,i, v_delta,i, and v_r,i. Re-run the SV3 circular-prior model with the Sec. 2.5 values (sigma = 12,12,12) and with the Table 5 values (sigma = 42,40,11). If the 90% upper limits on vk and Delta_m2 differ by more than about 20%, the paper should quote the limits from the actual prior and add a sensitivity statement; if they do not differ, the inconsistency is cosmetic and can be fixed by a correction.","verdict_should_be":"UNCHANGED","load_bearing_attack":"VFTS 243 is the demonstration of the central claim, and its quoted limits (vk <= 27 km/s, Delta_m2 <= 2.9 Msun) come from the SV3 model. In that model the kick information is carried by Delta_v = v_f - v_i, with v_i drawn from the birth-environment prior Venv. Section 2.5 sets Venv,alpha = N(393,12), Venv,delta = N(143,12), Venv,r = N(271,12), while Table 5, the prior table used for the VFTS 243 analysis, lists Venv,alpha = N(393,42), Venv,delta = N(146,40), Venv,r = N(270,11). The tangential widths differ by a factor of roughly 3.5. If the wider prior is the one actually used, the observed v_f can be attributed more easily to the birth velocity and the inferred kick/mass-loss upper limits may be substantially larger; if the narrow prior is correct, Table 5 is wrong. Either way, the headline numbers are not reproducible until the discrepancy is resolved. This is independent of the acknowledged e_i = delta(e_i) assumption, which the paper's own Fig. 5c shows can also produce precise but wrong values, and Sec. 2.5 itself warns that the environment sample choice is subjective.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents SideKicks.jl, a Bayesian inference framework for constraining the natal kick and mass loss of the BH progenitor in inert black hole binaries (BHBs) from combined spectroscopic and astrometric observations, including the orientation of the systemic velocity relative to the orbit. The forward model maps pre-core-collapse orbital elements and collapse parameters (kick magnitude and direction, mass loss, explosion true anomaly) to post-collapse observables, and the posterior is explored with the No-U-Turn Sampler. Injection tests on a circular mock binary show precise recovery when a circular pre-collapse prior is used and the environment velocity dispersion is small; using an agnostic eccentricity prior substantially broadens the constraints, and applying the circular prior to an eccentric system yields precise but wrong values. Applying the SV3 model (spectroscopy plus all three velocity components, no resolved orbit) to VFTS 243 under the circular prior yields 90% upper limits vk < 26.5 km/s and Delta_m2 < 2.9 Msun. The paper also argues that BH formation channels can be distinguished without a resolved orbit. The code and data are publicly archived.","tokens_in":34092,"tokens_out":10705,"duration_ms":97922,"significance":"If the results are robust, the framework will be a valuable community tool for the BLOeM survey and Gaia DR4, and the VFTS 243 constraints would strengthen the case for weak kicks and small mass loss in BH formation. The paper's strengths are its complete forward model, the explicit treatment of the velocity-vector orientation, the open-source implementation, and the injection tests that expose the role of the eccentricity prior. However, the headline VFTS 243 numbers are conditional on the circular-prior assumption, and the birth-environment velocity prior is specified inconsistently between Section 2.5 and Table 5, so the central quantitative claims are not currently reproducible as written.","major_comments":[{"comment":"The birth-environment velocity priors used for the VFTS 243 analysis are stated inconsistently. Section 2.5 sets Venv,alpha = N(393,12), Venv,delta = N(143,12), Venv,r = N(271,12), while Table 5 lists Venv,alpha = N(393,42), Venv,delta = N(146,40), Venv,r = N(270,11). Because the inferred Delta_v = vf - vi, and hence the 90% upper limits on vk and Delta_m2, depend directly on these widths, the quoted headline constraints (vk < 26.5 km/s, Delta_m2 < 2.9 Msun) are not reproducible without resolving this discrepancy. The authors should state which prior was actually used, correct the other occurrence, and if the wider prior is the correct one, rerun the analysis and report the resulting limits.","section":"Section 2.5 and Table 5"},{"comment":"The 90% credible limits quoted for VFTS 243 are obtained exclusively with the circular pre-collapse prior pi(ei)=delta(ei) (Table 5, Section 4.1). The paper's own injection test in Fig. 5c shows that the same prior applied to a system with true ei=0.3 produces posteriors that are precise but far from the truth. Since the pre-collapse eccentricity of VFTS 243 is not directly measured, the headline constraints are assumption-driven. I request that the abstract and conclusions explicitly state that the limits are conditional on a circular pre-collapse orbit, and that the corresponding 90% limits under the agnostic prior pi(ei)=U(0,1) (already computed in Fig. C.2) be reported numerically in the text so that the sensitivity is transparent.","section":"Sections 3.3 and 4.1; Fig. 5c; Abstract"},{"comment":"The injection tests in Section 3.2 show that the recovery degrades markedly with increasing birth-environment velocity dispersion: at sigma_env = 10 km/s the true values fall outside the 2D 90% credible region. The dispersion inferred for the Tarantula environment and used for VFTS 243 is 11-42 km/s (Table 5), i.e., larger than or comparable to the largest tested value, and Section 2.5 acknowledges that the SB2 sample used to derive it is small and subjectively chosen. This means the reliability of the VFTS 243 kick and mass-loss limits is not established by the current injection tests. The authors should run injection-style robustness checks at the actual VFTS 243 environment widths, or marginalize over the hyperparameters of Venv as they suggest, before the headline numbers are used.","section":"Section 3.2 (Fig. 4) and Section 2.5"},{"comment":"The prior on the pre-collapse inclination is given as pi(cos ii) = U(0,1) (Section 2.3.1 and Table 5), which restricts ii to [0,pi/2]. However, Appendix A.5 states that the inclination is defined on the range [0,pi] to account for all geometries in which the star recedes at the ascending node, and Eq. (A.44) computes if = arccos(Lhat_f dot Ohat), which can exceed pi/2. This is an internal inconsistency. If the sampling prior is literally U(0,1), then retrograde configurations are not represented and the orientation sampling is incomplete, which could bias the inferred pre-collapse parameters and the derived kick posteriors. The authors should clarify the convention and, if needed, use pi(cos ii) = U(-1,1).","section":"Section 2.3.1 and Appendix A.5"}],"minor_comments":[{"comment":"The citation 'Homan & Gelman 2014' for the No-U-Turn Sampler is a misspelling; the correct author is Hoffman, so this should read 'Hoffman & Gelman 2014' in both the text and the reference list.","section":"References (Sec. 2.3)"},{"comment":"There are several typographical errors, including 'the the orbital inclination' in Section 2.1 and 'a ≈ 25 M_sun O-type star' in Section 4 (should be 'an'). A careful proofreading pass is needed.","section":"Sec. 2.1 and Sec. 4 opening"},{"comment":"Some figure captions contain unresolved LaTeX artifacts, such as '/g4(ei) = /g5(ei)' in Figs. 3, 5, and B.1, which will confuse readers; these should be corrected in the production version.","section":"Figure captions"},{"comment":"The references listed for the Venv priors (Gaia Collaboration et al. 2023 and Almeida et al. 2017) point to data sources, but the specific values (means and dispersions) are derived in Section 2.5; cross-referencing that section would help avoid the appearance of a discrepancy and would clarify the provenance of the numbers.","section":"Table 5"}],"recommendation":"major_revision","confidential_remarks":"The inconsistency between Section 2.5 and Table 5 is the most serious technical issue; if the wider prior was actually used, the headline VFTS 243 limits may be substantially weaker. I recommend that the editor ask the authors to provide the exact prior table used for the VFTS 243 run and to make the injection tests at realistic environment dispersions. The paper is otherwise a solid methods contribution with a publicly available implementation and clear injection tests."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Rein, this paper is worth your time. The new thing here is a Bayesian forward model for inferring natal kicks and mass loss from inert black hole binaries that uses full 3D systemic velocity and its orientation relative to the orbital plane, and that handles eccentric pre-collapse orbits and partial observing scenarios. The implementation, SideKicks.jl, is open source, and the injection tests in the idealized circular case recover the input parameters precisely. The framework also allows distinguishing BH formation channels without a resolved orbit, which is useful for the BLOeM and Gaia DR4 samples.\n\nThe paper is honest about its main fragility: the VFTS 243 results assume a circular pre-collapse orbit via a delta prior at ei=0, and their own Fig. 5c shows that if the real orbit was eccentric (ei=0.3), the same analysis gives precise but wrong answers. This is a real modeling vulnerability, and the abstract presents the 27 km/s and 2.9 Msun limits without that caveat. The derivation of the birth-environment velocity dispersion is also acknowledged as subjective, based on a small SB2 sample that could include runaways.\n\nThere is also a concrete reproducibility issue. Section 2.5 sets the environment velocity priors for VFTS 243 as N(393,12), N(143,12), N(271,12) in the three components. Table 5, the actual prior table for the VFTS 243 analysis, lists N(393,42), N(146,40), N(270,11). The tangential sigmas differ by a factor of about 3.5. Since the inference uses the width of that prior to decide how much of the observed velocity can be attributed to birth rather than kick, the reported upper limits change with the choice. I don't know which numbers were used in the MCMC, but the mismatch makes the headline results non-reproducible until resolved. That's a fixable error, not a flaw in the method.\n\nAll of this said, the core method is sound. The forward model is standard, the likelihoods are Gaussian, and the paper's own sensitivity tests expose the circular-prior problem rather than hiding it. For anyone working with inert BHBs, the framework will be useful. I'd cite it. Send it to a serious referee, but require that the velocity-prior inconsistency be fixed and the circular-prior caveat be stated in the abstract. It's a solid contribution with a messy corner that needs cleaning up.","headline":"A genuinely useful Bayesian framework for natal-kick inference from inert BHBs, but the VFTS 243 headline numbers are not reproducible as written because the birth-velocity prior widths in Sec. 2.5 and Table 5 disagree.","tokens_in":34687,"tokens_out":3591,"would_cite":true,"duration_ms":29759,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that inert black hole binaries can tightly constrain the natal kick and mass loss of black-hole-forming core collapses, and that for VFTS 243 this yields 90% upper limits of about 27 km/s and 2.9 solar masses.","keywords":["natal kicks","black hole formation","inert black hole binaries","VFTS 243","Bayesian inference","core-collapse mass loss","binary eccentricity","BLOeM survey"],"falsifier":"The paper's own Fig. 5c is the falsifier: a system that was actually eccentric ($e_i=0.3$) but analyzed with the circular prior yields 90% intervals that exclude the true kick and mass loss by a wide margin. Therefore, an independent determination that VFTS 243's pre-collapse orbit was non-circular—for example, a measured eccentricity distribution of similar short-period post-mass-transfer binaries with a substantial tail above $e\\approx 0.1$—would invalidate the reported 27 km/s and 2.9 solar-mass bounds.","tokens_in":33604,"feed_emoji":"🕳️","tokens_out":8178,"duration_ms":74511,"temperature":0.7,"pith_summary":"This paper tries to establish that inert black hole binaries—wide binaries with a stellar-mass black hole, a massive main-sequence star, and no X-ray emission—can serve as precision probes of how black holes are born. It presents a Bayesian inference framework that combines spectroscopic orbital elements, astrometric proper motions, and the three-dimensional systemic velocity of the binary relative to the velocity dispersion of its birth association, in order to separate the natal kick (the recoil imparted at core collapse) from symmetric mass loss. Applied to the benchmark system VFTS 243, the framework yields a natal kick below $27\\ \\mathrm{km\\,s^{-1}}$ and a mass loss below $2.9\\,M_\\odot$ at 90% credibility, with both quantities peaking at zero. The broader claim is that even when the orbit cannot be resolved, the full velocity information distinguishes black hole formation channels, and that a circular pre-collapse orbit assumption makes the inference tight.","feed_headline":"Quiet black hole binaries reveal their birth kicks","feed_subtitle":"A Bayesian reading of VFTS 243 caps its natal kick at 27 km/s and mass loss at 2.9 solar masses.","key_machinery":"The load-bearing object is a forward model of core collapse in a binary: twelve free parameters describing the pre-collapse orbit, orientation, and collapse (initial period, eccentricity, both masses, three orientation angles, mass loss, kick magnitude and direction, and true anomaly at explosion) are mapped through the two-body impulse equations onto all six post-collapse orbital elements plus the center-of-mass velocity vector in the observer's frame. The decisive step is to treat the pre-collapse systemic velocity as drawn from the measured velocity dispersion of the birth association, so the difference between the observed velocity and the birth velocity becomes a direct constraint on the kick. When the pre-collapse orbit is assumed circular, the true anomaly drops out, the number of free parameters matches the number of constraints, and the collapse parameters can be recovered to essentially arbitrary precision, limited only by measurement noise.","core_discovery":"The central claim is that the recoil and mass loss of a black hole-forming core collapse can be read off an inert black hole binary when all available information—the orbital period, eccentricity, radial-velocity semiamplitude, companion mass, orientation angles, and the full three-dimensional velocity of the system relative to its birth environment—are modelled jointly. The paper demonstrates with injection tests that, in the best-observed scenario with a circular pre-collapse orbit, the mapping from observables to collapse parameters is effectively one-to-one and the natal kick and mass loss are recovered to measurement precision. For VFTS 243, using the circular prior, the inference gives a natal kick peaking at $0$ with a 90% upper limit of $26.5\\ \\mathrm{km\\,s^{-1}}$ (quoted as $<27\\ \\mathrm{km\\,s^{-1}}$) and a mass loss peaking at $0$ with a 90% upper limit of $2.9\\,M_\\odot$, with 68% limits of about $13.5\\ \\mathrm{km\\,s^{-1}}$ and $1.34\\,M_\\odot$. The paper further claims this framework can distinguish formation channels, such as direct collapse versus a kicked supernova, even without a resolved astrometric orbit, which is exactly the situation expected for distant SMC systems in the upcoming survey.","pith_inferences":["The tight VFTS 243 numbers are conditional on a delta-function circular prior: the paper's own injection test shows that a circular prior applied to a system that was actually eccentric ($e_i=0.3$) returns precise but incorrect values, so the quoted bounds should be read as upper limits only if the short-period mass-transfer circularization assumption holds.","Because the inference weakens quickly as the assumed birth-velocity dispersion grows (the paper's Fig. 4 shows the true values leaving the 90% region by $10\\ \\mathrm{km\\,s^{-1}}$), population applications will depend on building clean, uncontaminated host-association samples.","The same machinery could be inverted: with a sample of dozens of inert binaries, the inferred joint kick and mass-loss distribution could calibrate supernova explosion models, since the framework treats the explosion mechanism agnostically.","One could extend the method to nearby astrometric black-hole binaries by adding the same velocity-dispersion treatment, though their older, lower-mass companions make the no-perturbation assumption harder to justify."],"forward_implications":["For distant inert black hole binaries with no resolved orbit, the full three-dimensional velocity information is enough to separate direct-collapse formation (zero kick, near-zero mass loss) from formation with a substantial kick.","If the pre-collapse orbit is circular, the inference becomes tightly constrained, so determining the eccentricity distribution of pre-collapse binaries—observationally and theoretically—directly controls how precise kick inferences can be.","For VFTS 243, the framework rules out natal kicks above about $27\\ \\mathrm{km\\,s^{-1}}$ and core-collapse mass loss above about $2.9\\,M_\\odot$ at 90% credibility, consistent with a negligibly kicked, low-mass-loss birth.","The same pipeline applies to inert neutron-star binaries, offering a way to probe the weak-kick end of the neutron-star kick distribution.","For the upcoming survey of massive stars in the Small Magellanic Cloud, the method provides a ready analysis path for the many inert black hole binaries that survey is expected to find."],"supporting_citations":[{"why":"Provides the discovery, orbital parameters, and companion mass estimate of VFTS 243 that anchor the analysis.","marker":"Shenar et al. 2022b"},{"why":"Supplies the parallax and proper-motion measurements used to derive VFTS 243's transverse velocities and the Tarantula velocity dispersion.","marker":"Gaia Collaboration et al. 2023"},{"why":"Gives the radial-velocity orbit and systemic radial velocity of VFTS 243 plus the SB2 sample used for the birth-velocity prior.","marker":"Almeida et al. 2017"},{"why":"Population-synthesis comparison giving an earlier natal-kick limit of about 33 km/s for VFTS 243.","marker":"Stevance et al. 2022"},{"why":"Earlier inference on VFTS 243 that the present framework tightens by including the transverse velocity information.","marker":"Banagiri et al. 2023"},{"why":"Previous kick and mass-loss inference for VFTS 243 whose 68% limits are compared directly in the present work.","marker":"Vigna-Gómez et al. 2024"},{"why":"Establishes the symmetric mass-loss (Blaauw) kick that the framework separates from the natal kick.","marker":"Blaauw 1961"},{"why":"Derives post-explosion orbital elements for eccentric binaries, providing the basis of the forward model extended here.","marker":"Pfahl et al. 2002"}],"fun_headline_variants":["Black hole birth kicks capped at 27 km/s","Inert binaries reveal quiet black hole births","VFTS 243: natal kick under 27 km/s, mass loss 2.9 Msun","Bayesian method reads kicks from silent black hole binaries"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The reported VFTS 243 constraints assume the binary's orbit was exactly circular before core collapse, so the observed near-zero eccentricity is attributed entirely to a weak kick and small mass loss; if the orbit was even moderately eccentric, the same observation would be produced by different kick and mass-loss values.","fun_headline_variants_meta":{"raw":{"variants":["Black hole birth kicks capped at 27 km/s","Inert binaries reveal quiet black hole births","VFTS 243: natal kick under 27 km/s, mass loss 2.9 Msun","Bayesian method reads kicks from silent black hole binaries"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00032,"raw_usage":{"total_tokens":1929,"prompt_tokens":1195,"completion_tokens":734,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":811,"completion_tokens_details":{"reasoning_tokens":662}},"tokens_in":811,"tokens_out":734,"duration_ms":6848,"temperature":1.0,"reasoning_tokens":662,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T10:57:51.901358+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"The paper's own Fig. 5c is the falsifier: a system that was actually eccentric ($e_i=0.3$) but analyzed with the circular prior yields 90% intervals that exclude the true kick and mass loss by a wide margin. Therefore, an independent determination that VFTS 243's pre-collapse orbit was non-circular—for example, a measured eccentricity distribution of similar short-period post-mass-transfer binaries with a substantial tail above $e\\approx 0.1$—would invalidate the reported 27 km/s and 2.9 solar-mass bounds.","supporting_citations":[{"cited_title":"F., Ghodla, S., Richards, S., et al","cited_arxiv_id":null,"evidence_quote":"Population-synthesis comparison giving an earlier natal-kick limit of about 33 km/s for VFTS 243."},{"cited_title":"2002, The Astrophysical Journal, 573, 283 Pietrzy´nski, G., Graczyk, D., Gallenne, A., et al","cited_arxiv_id":null,"evidence_quote":"Derives post-explosion orbital elements for eccentric binaries, providing the basis of the forward model extended here."}],"review_version":1}