{"id":"86847a53-44cd-4a91-9964-ff2c2640fe9e","arxiv_id":"2508.03809","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"Simulated effective-spin distributions match LVK O1-O3 data for isolated binaries only when black hole spin axes are tossed at formation, or if roughly 72% of mergers have a dynamical origin.","lead":"This paper compares computer simulations of binary black hole mergers with 83 observed LIGO-Virgo-KAGRA events to test whether black hole spin axes get randomly tossed when the black holes form. It finds that, if most mergers come from isolated binary stars, observations favor spin-axis tossing unless about 72% of mergers come from dynamical channels with random spin directions.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No-tossing failure hinges on perfect alignment of both BH spins with the pre-SN orbit; the paper flags but does not test the Baibhav & Kalogera (2024) counter-argument, so the tossing inference remains conditional.","rationale":"The reader identified the same load-bearing assumption: perfect alignment of both BH spins with the pre-SN orbital angular momentum. My analysis agrees that this is the weakest link in the central argument. The paper's geometric result—that no-tossing yields only positive χeff—is valid only under that idealization. The authors themselves cite Baibhav & Kalogera (2024) as a potential caveat, but they do not test how their conclusion changes if first-born BH alignment is incomplete. A single targeted simulation with a first-born spin-tilt distribution drawn from that counter-argument could settle whether the no-tossing scenario can be rescued. Until such a test is done, the evidence for spin-axis tossing is conditional. The reader's CONDITIONAL verdict already reflects this, so no change is warranted. I do not see an additional concern that would shift the verdict to REJECT, because the paper is transparent about the assumption, provides falsifiable forecasts, and its geometric argument is internally consistent under the stated idealization. The proposed test would either confirm the robustness of the no-tossing failure or reveal that the central contrast is an artifact of an overly strong alignment assumption.","tokens_in":30171,"tokens_out":9475,"duration_ms":112098,"concrete_test":"Re-run the no-tossing Monte Carlo simulations of Section 4.4.1 with the first-born BH spin tilt Θ1 drawn from a distribution that includes a fraction f of isotropically oriented spins, following the prescription of Baibhav & Kalogera (2024) or a simplified model (e.g., Θ1 uniform in cosΘ1 for a fraction f of systems), while keeping Θ2 = δ and all other parameters unchanged. Vary f from 0 to 0.5 and recompute the KS, CvM, and AD p-values against the 83 LVK O1–O3 χeff values. If p > 0.05 for any f consistent with observational constraints on BH spin misalignment (e.g., from X-ray binaries such as MAXI J1820+070), then the central claim that no-tossing cannot explain the data—and hence the inference of spin-axis tossing—is not robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (Section 4.6) that no-tossing isolated binaries cannot explain observed negative χeff values rests on the assumption, stated in the introduction to Section 4, that the first-born BH spin is fully aligned with the pre-SN orbital angular momentum L0 by accretion (so Θ1 = δ), and that the second-born helium star is tidally locked so Θ2 = δ in the no-tossing case. Under these assumptions, χeff = cosδ (M1χ1 + M2χ2)/MT > 0 for small δ, making negative χeff impossible without tossing. However, the paper itself flags a known counter-argument in a footnote to Section 4.4: \"The first-formed BH is also expected to experience spin-axis tossing, but its spin axis will later align due to tidal interactions and mass accretion ... (but see Baibhav and Kalogera, 2024).\" If accretion-alignment is incomplete, the first-born BH can retain a misaligned spin component, so its tilt relative to the post-SN orbital angular momentum can exceed 90° even when the kick-induced δ is small, producing negative χeff without any tossing of the second-born BH. The same applies if tidal locking of the helium star is imperfect, so Θ2 need not equal δ. The no-tossing p-values below 0.001 are therefore not a test of \"no spin-axis tossing\" but of \"no tossing plus perfect spin-orbit alignment of both BHs.\" The contrast between tossing and no-tossing could collapse if a realistic misalignment distribution for the first-born BH is adopted, and the paper does not quantify this sensitivity.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes the LVK O1-O3 effective-spin (chi_eff) measurements of 83 binary black hole (BH+BH) mergers and compares the empirical distribution with Monte Carlo simulations of the second supernova in isolated binary evolution, with and without spin-axis tossing at BH formation. The authors use kernel density estimation, functional boxplots, and three two-sample tests (KS, CvM, AD). Their main finding is that isolated-binary simulations without spin-axis tossing always give p-values below 0.001, whereas simulations with tossing can match the data with p-values up to 0.882. They also infer a mass-reversal fraction near 29% and, in the no-tossing case, a dynamical-origin fraction of about 72±8%. The central claim is that the data support the spin-axis tossing hypothesis if isolated binaries dominate the BH+BH merger channel.","tokens_in":30496,"tokens_out":5032,"duration_ms":56538,"significance":"If the central contrast between tossing and no-tossing simulations is robust, the paper would provide population-level support for Tauris's (2022) spin-axis tossing hypothesis and a transparent way to constrain isolated versus dynamical formation fractions. The strengths of the paper are its explicit Monte Carlo setup, the use of three complementary statistical tests, the careful treatment of chi_eff measurement uncertainty through functional boxplots, and the fact that the predictions can be checked with O4+O5 and 3G detector data. The main risk is that the central no-tossing rejection depends on an alignment assumption that the paper itself flags as contested, and several model inputs are fitted to the same empirical data used for the final p-values.","major_comments":[{"comment":"The statement that no-tossing isolated binaries cannot produce the observed negative chi_eff tail is load-bearing, and it depends on two assumptions stated in the text: the first-born BH spin is fully aligned with the pre-SN orbital angular momentum so that Θ1=δ, and tidal locking forces Θ2=δ in the no-tossing case. Footnote 2 cites Baibhav & Kalogera (2024) as challenging exactly this alignment story, but the manuscript does not quantify the sensitivity. If the first-born BH retains a misaligned spin component, or if tidal locking is incomplete, Θ1 or Θ2 can exceed 90° even with small kicks, producing negative chi_eff without any tossing. I request a quantitative test: rerun the no-tossing simulations with a misalignment distribution for the first-born BH (or a residual tilt for the helium star) and report the resulting p-values and the fraction of systems with chi_eff<0.","section":"Section 4 (opening items i–ii), Section 4.4.1, Fig. 16"},{"comment":"The chi1 and chi2 beta distributions are obtained by fitting the simulated chi_eff curve to the empirical functional-boxplot LSCV curve; these same empirical data are then used as the reference for all later p-values. This is a circular step that inflates the reported agreement: the model is scored against the same dataset used to set its spin inputs. The later use of the GWTC-3 chiA/chiB credibility intervals (Section 4.6) partially mitigates this, but the Section 4.2 spin inputs remain data-fitted. Please quantify how much of the tossing/no-tossing separation survives when chi1 and chi2 are instead drawn from independent population estimates or from the full prior range allowed by Fig. 18.","section":"Sections 4.2.1–4.2.2 and Fig. 8"},{"comment":"The 'fitted tossing' distribution P(Φ2) = 1/2[β(1.36,2.51)+β(7.9,5.3)] is fitted to minimize RMSE to the same empirical LSCV curve, so the subsequent comparison (Fig. 15C and Section 4.6) uses a distribution that has already been tuned to the data. This makes the high p-values for the tossing scenario partly a measure of the flexibility of the two-beta mixture rather than evidence for isotropic tossing. The paper should state this explicitly and provide out-of-sample validation, for example by using only O1-O2 data for fitting and O3 for testing, or by forecasting the chi_eff distribution for O4.","section":"Section 4.4, Eq. (9)"},{"comment":"The final high p-value (0.882) is obtained from panel A, which the text itself calls an extreme fit that maximizes chiA and minimizes chiB within the 90% credibility intervals, combined with the 29% mass-reversal fraction that was itself selected to maximize p-values in Section 4.5. As stated, the analysis selects the best case among several alternatives and then reports that best case as the headline. The choice of panel A needs to be presented as a selection effect: report the p-values for all panels (A-D) and, ideally, the distribution of p-values over randomly drawn spin-component curves inside the credibility region, so the reader can see how often the tossing/no-tossing contrast is achieved.","section":"Section 4.6, Fig. 18"}],"minor_comments":[{"comment":"The definition q ≡ M2/M1 ≤ 1 is used in Eq. (1), but Section 4.5 explicitly discusses q > 1 after mass reversal; the definition should allow q > 1 or the mass-reversal text should refer to the re-labelled mass ratio.","section":"Eq. (1)"},{"comment":"The symbol Φ2 is used both for the tossing angle and for the post-SN spin tilt angle Θ2; please unify the notation to avoid confusion.","section":"Fig. 14 and Eq. (8)"},{"comment":"The AD p-values are capped at 0.250, but the main text does not explain this cap until the table note; please state this where the p-values are first quoted.","section":"Table 1 and Section 4.5"},{"comment":"The detection-bias discussion dismisses the effect on the basis of Vitale et al. (2022), but a sentence explaining the sign and expected size of the bias would help the reader assess the sensitivity of the negative-tail inference.","section":"Section 3.2.1"},{"comment":"The phrase 'we cannot completely rule out this scenario' for Fdyn=1 is stronger than the preceding statistical discussion warrants; consider rewording to 'not strongly excluded' or similar.","section":"Section 4.7"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for New Astronomy and the methodology is largely transparent and reproducible. My main concern is that the central no-tossing rejection is conditional on an alignment assumption that the authors themselves flag as contested, and that several fitted inputs (chi1/chi2 beta parameters, the tossing distribution, the mass-reversal fraction, and the selected spin-component panel) precede the reported p-values. I therefore recommend major revision rather than rejection: the claims can be made robust with targeted sensitivity tests and a more honest treatment of selection effects."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this paper is a useful, transparent quantitative comparison, but the central claim is more conditional than the abstract lets on. The no-tossing result is solid under the paper's alignment assumptions, but the tossing result is partly built by fitting to the same LVK data it is then compared against.\n\nThe new contribution is real: a Monte Carlo campaign over a 9-dimensional parameter space, 100,000 simulated merging systems per setup, three two-sample tests (KS, CvM, AD), functional boxplots to encode measurement uncertainty, and quantified fitted parameters: ~30% mass reversal, ~72±8% dynamical fraction needed to rescue no-tossing. The core no-tossing failure is geometrically robust: with small natal kicks and both spins aligned to the pre-SN orbit, chi_eff stays positive, while O1–O3 contains negative events. That is the paper's strongest result.\n\nWhat it does well: the authors are unusually explicit about assumptions, including which distributions are fitted, and they make falsifiable forecasts for O4/O5 and 3G detectors. They also acknowledge the Baibhav & Kalogera counter-argument in a footnote, which is more than many papers do.\n\nThe soft spots are proportional. First, the high p-values for the tossing scenario (up to 0.882) are weakened by the fact that chi1/chi2 beta parameters, the tossing-angle distribution, the mass-reversal fraction, and the fast/slow spin-component panel are all tuned against the same LVK chi_eff and spin-component data used in the test. High p-values here are partly a measure of fitting flexibility, not independent confirmation. Second, the no-tossing conclusion rests on perfect alignment of the first-born BH spin with the orbital angular momentum and tidal locking of the helium star. If the first-born BH retains a misaligned spin component, negative chi_eff can arise without tossing. The paper flags this in a footnote but does not test it; the stress-test note lands on that point correctly. Third, detection bias in chi_eff is acknowledged and set aside; that is defensible for now but worth quantifying for O4+. No code is shipped, so full reproducibility is untested.\n\nThe 72±8% dynamical mixture is also a fitted solution, not a prediction. The paper's own language is more careful than the abstract: they say support for tossing \"if isolated binaries dominate.\" I would carry that conditional through the abstract too.\n\nWho this is for: gravitational-wave population folks and binary-evolution modelers working on formation channels. It deserves a serious referee; a good referee should push for sensitivity to first-born spin misalignment and selection effects, and for the code. I would cite it, and bring it to reading group.","headline":"A transparent, useful Monte Carlo comparison whose no-tossing result is robust under strict alignment assumptions, but whose tossing evidence is partly fitted to the same data — so the conclusion stays conditional.","tokens_in":31115,"tokens_out":2646,"would_cite":true,"duration_ms":30447,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["04.25.dg","95.85.Sz","97.80.-d","97.60.Bw"],"model":"deepseek-v4-flash","headline":"Simulations with random spin-axis tossing at black hole formation reproduce the observed binary black hole spins, while simulations without it fail.","keywords":["binary black hole mergers","effective spin","spin-axis tossing","black hole natal kicks","isolated binary evolution","dynamical formation","mass reversal","gravitational wave observations"],"falsifier":"Look for a single isolated-binary merger with a confidently negative effective spin and a small inferred natal kick: the no-tossing model predicts that such a system always has positive $\\chi_{\\rm eff}$ because the maximum tilt is only $6.42^\\circ$, so one clean counterexample would falsify the no-tossing branch. Conversely, a large future sample of field mergers with only positive effective spins would remove the statistical need for tossing.","tokens_in":29930,"feed_emoji":"🕳️","tokens_out":10274,"duration_ms":108062,"temperature":0.7,"pith_summary":"Gravitational-wave observatories have recorded 83 binary black hole mergers whose effective spin values are asymmetric around zero and include a notable number of negative entries. This paper asks whether that pattern can arise if the mergers are produced by isolated binary stars, and concludes that it can only if the collapsing star that forms the second black hole 'tosses' its spin axis in a random direction during core collapse. Simulations with such tossing reproduce the observed distribution, with two-sample test p-values up to 0.882, whereas simulations without tossing yield p-values below 0.001 except in an extreme wide-orbit, high-kick corner. The paper also finds that a universe without tossing would need about 72±8% of detected mergers to come from dynamical environments with random spin directions, and that the best isolated-binary fits include mass reversal in about 30% of progenitor binaries.","feed_headline":"Black hole spin data favor spin-axis tossing at birth","feed_subtitle":"If isolated binaries dominate, only tossing reproduces the observed spins; otherwise ~72% must be dynamical.","key_machinery":"The argument is carried by a Monte Carlo model of the final core collapse in a synthetic population of isolated binaries. The simulation draws nine parameters: the two black hole masses and spin magnitudes, the pre-supernova orbital separation, the kick velocity and two isotropic direction angles, and the tossing angle $\\Phi_2$ of the second-born black hole's spin axis. The tilt assignments do the decisive work: accretion alignment gives $\\Theta_1=\\delta$, tidal locking gives $\\Theta_2=\\delta$ in the no-tossing case, and isotropic tossing replaces $\\Theta_2$ with draws from $P(\\Theta_2)=\\frac12\\sin\\Theta_2$. Simulated $\\chi_{\\rm eff}$ distributions are compared with the 83 observed mergers using kernel density estimates, functional boxplots, and the Kolmogorov-Smirnov, Cramér-von Mises, and Anderson-Darling tests, with each observed value varied inside its 90% credibility interval.","core_discovery":"On its own terms, the paper's discovery is that the measured effective spin distribution of binary black hole mergers is a readable record of the second collapse. Because accretion torques align the first-born black hole's spin with the pre-supernova orbit, its tilt angle $\\Theta_1$ is just the kick-induced misalignment $\\delta$; in the widely used no-tossing picture, tidal locking gives the second-born black hole the same tilt, $\\Theta_2=\\delta$, so all merger spins stay aligned with the orbit and only positive $\\chi_{\\rm eff}$ values result (maximum tilt $6.42^\\circ$). Against the 83 observed events this scenario is statistically excluded, with p-values below 0.001. If instead the second-born black hole's spin axis is drawn from an isotropic tossing distribution, $P(\\Theta_2)=\\frac12\\sin\\Theta_2$ on $[0,180^\\circ]$, the simulated $\\chi_{\\rm eff}$ distribution matches the data with p-values as high as 0.882 (KS), 0.742 (CvM), and 0.250 (AD, capped). The authors therefore conclude that the observations support spin-axis tossing if isolated binaries dominate the merger channel.","pith_inferences":["If tossing is real, the toss angle should not be universal: the direction and magnitude of the toss likely depend on the pre-supernova structure, mass loss, and orbital period, so future data could search for a conditional, non-isotropic $\\Phi_2$ distribution.","The two surviving scenarios—tossing-dominated isolated binaries versus roughly 72% dynamical mergers—make different predictions for merger eccentricities, host environments, and rates, so combining $\\chi_{\\rm eff}$ with those observables could break the degeneracy.","The ~30% mass-reversal preference implies that which black hole is spinning faster already encodes which one formed second; more precise individual-spin measurements could turn this into a direct formation-order test."],"forward_implications":["Population synthesis of isolated binary black hole mergers should include spin-axis tossing of the second-formed black hole; the paper suggests a fully isotropic distribution as a simple default.","A no-tossing interpretation is not dead, but it requires roughly 72±8% of detected mergers to come from dynamical channels with random spin directions.","The best isolated-binary fits prefer mass reversal in about 30% of progenitor systems, aligning with independent estimates from spin data.","The tossing scenario's high p-values mean it cannot be rejected with current data; future observing runs and third-generation detectors should sharpen the $\\chi_{\\rm eff}$ comparison and probe whether tossing is isotropic or depends on progenitor properties."],"supporting_citations":[{"why":"Introduces the spin-axis tossing hypothesis and the Monte Carlo recipe that this paper extends to observed masses and spins.","marker":"Tauris (2022)"},{"why":"Supplies the empirical population of 83 mergers and the 90% credibility intervals for the fast and slow component spins used in the final simulations.","marker":"Abbott et al. (2023b)"},{"why":"Provides the up-to-32% mass-reversal estimate that the paper's ~30% best-fit fraction tracks.","marker":"Mould et al. (2022)"},{"why":"Establishes the Lense-Thirring mechanism by which the first-born black hole's spin aligns with the pre-supernova orbit.","marker":"Bardeen and Petterson (1975)"},{"why":"Supports accretion-disk alignment of black hole spin with the orbital angular momentum.","marker":"King et al. (2005)"},{"why":"Gives the gravitational-wave merger time used to retain only post-supernova systems that merge within a Hubble time.","marker":"Peters (1964)"},{"why":"Motivates the mixed isolated-plus-dynamical channel analysis by arguing no single formation channel dominates.","marker":"Zevin et al. (2021)"}],"fun_headline_variants":["Spin-axis tossing at black hole birth preferred by LVK data","Black hole merger spins favor spin-axis tossing in isolated binaries","No tossing means 72% of mergers must be dynamical, study finds","Mass reversal in 30% of progenitors if isolated binaries dominate","Spin data reject no-tossing scenario for isolated binary black holes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument stands on one premise: accretion fully aligns the first-born black hole's spin with the pre-supernova orbit, and tides lock the collapsing helium star to that same orbit, so without tossing both black hole tilts equal the kick angle.","fun_headline_variants_meta":{"raw":{"variants":["Spin-axis tossing at black hole birth preferred by LVK data","Black hole merger spins favor spin-axis tossing in isolated binaries","No tossing means 72% of mergers must be dynamical, study finds","Mass reversal in 30% of progenitors if isolated binaries dominate","Spin data reject no-tossing scenario for isolated binary black holes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000405,"raw_usage":{"total_tokens":2201,"prompt_tokens":1136,"completion_tokens":1065,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":752,"completion_tokens_details":{"reasoning_tokens":977}},"tokens_in":752,"tokens_out":1065,"duration_ms":10141,"temperature":1.0,"reasoning_tokens":977,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T04:13:35.663692+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Look for a single isolated-binary merger with a confidently negative effective spin and a small inferred natal kick: the no-tossing model predicts that such a system always has positive $\\chi_{\\rm eff}$ because the maximum tilt is only $6.42^\\circ$, so one clean counterexample would falsify the no-tossing branch. Conversely, a large future sample of field mergers with only positive effective spins would remove the statistical need for tossing.","supporting_citations":[],"review_version":1}