{"id":"c25420e7-ef2d-470e-a16d-106ea792a4e9","arxiv_id":"2507.11225","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"The apparent excess of hot Jupiters in phase-space overdensities is driven by younger, more massive, more metal-rich host stars, not by an intrinsic environmental formation channel.","lead":"Hot Jupiters appear more common around stars in dense stellar neighborhoods, but this paper finds the effect largely disappears once host star age, mass, and metallicity are taken into account. The apparent clustering signal is better explained by demographics than by star clusters directly forging these extreme planets.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"ZGR23 residual after demographic correction is dismissed without quantitative support; isochronal age uncertainties may explain or mask the residual, so the central null claim is not yet established.","rationale":"The paper is a careful re-analysis that improves on earlier work by using homogeneous stellar parameter estimates and two complementary samples. The SWEET-Cat sample, which has more precise ages (20 per cent median relative uncertainty), shows no significant residual after detrending or matching, supporting the demographic interpretation. However, the ZGR23 sample—larger, more complete for transit planets, but with much larger age uncertainties (55 per cent)—retains a residual excess that is statistically significant after detrending and remains borderline after matching. The authors explicitly acknowledge this residual and conjecture that it stems from age uncertainties, but they do not demonstrate that the observed residual has the magnitude expected from such uncertainties. Because the central claim is a null result, the burden is on showing that the residual is consistent with the instrumental and model-related noise in the age estimates. A Monte Carlo simulation that resamples ages from the posterior distributions and repeats the full correction pipeline would settle whether the ZGR23 residual is within the noise floor. This is precisely the weakest assumption identified by the reader, and the reader's conditional verdict remains appropriate. No new concern requiring a different verdict has been identified; the analysis is internally coherent and the authors are transparent about limitations, but the abstract's phrasing overstates the strength of the null result relative to the ZGR23 findings.","tokens_in":17293,"tokens_out":9325,"duration_ms":120944,"concrete_test":"Run a Monte Carlo noise test: for each host in the ZGR23 sample, draw ages repeatedly from the isoclassify posteriors (or perturb the nominal ages with the reported 55 per cent relative uncertainties), redo the detrending and the like-for-like matching exactly as in the paper, and compute the resulting HJ vs non-HJ KS p-value and the Poisson p-value after matching. Repeat 1000 times. Compare the observed values (p_KS ≈ 4.1e-3 and median p ≈ 0.063) with the simulated distribution. If the observed residual is typical of the age-noise-induced residuals, then the age-uncertainty explanation is supported; if it is a strong outlier, the residual is unlikely to be explained by age errors and the central claim would need to be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that the hot Jupiter overdensity excess disappears after correcting for mass, metallicity, and age—relies on isochronal ages with median relative uncertainties of 20 per cent (SWEET-Cat) and 55 per cent (ZGR23). In the ZGR23 sample, the larger and more transit-dominated of the two, a residual difference between HJ and non-HJ hosts survives detrending (p_KS ≈ 4.1e-3; after a 1–5 Gyr cut, p_KS ≈ 0.042), and after like-for-like matching the median Poisson p is 0.063 with 39.7 per cent of 1000 resampled matches giving p < 0.05. The authors attribute this residual to age uncertainties and/or detection biases, but they do not quantitatively test that attribution. If isochronal ages are noisy, the linear detrending slope is attenuated, leaving a residual that mimics the observed one; if ages are systematically biased with phase space density (for example through age–metallicity correlations or activity-driven parameter errors), the correction could be either incomplete or artificially complete. Without a quantitative check, the claim that demographics fully explain the HJ excess is an assumption rather than a demonstrated result.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper re-examines the reported excess of hot Jupiters (HJs) in phase-space overdensities, testing whether the excess survives after controlling for host-star demographics. The authors construct homogeneous stellar parameters for FGK main-sequence planet hosts from SWEET-Cat and the Zhang et al. (2023, ZGR23) Gaia XP catalogue, derive isochronal ages with isoclassify, and reproduce the Winter et al. (2020) Mahalanobis phase-space density decomposition. They find that overdensity hosts are younger, more massive, and more metal-rich, and that after detrending phase-space density against age or matching overdensity and underdensity hosts in age, mass, and metallicity, the HJ excess is strongly diminished. The paper concludes that the previously reported environmental correlation is driven by host-star demographics rather than an intrinsic clustering effect.","tokens_in":17592,"tokens_out":3339,"duration_ms":44313,"significance":"If the null result were fully established, this would be an important resolution of an active debate about whether hot Jupiters form preferentially in clustered environments. The paper has real strengths: it uses two independent homogeneous stellar-parameter samples, propagates uncertainties through resampling, constructs size-controlled comparison samples, and makes its data publicly available. However, the central claim of 'no significant differences' is undercut by the authors' own ZGR23 results, where a residual difference survives detrending and the matching exercise produces a substantial fraction of significant draws. Because the paper's headline conclusion is stronger than the evidence it presents, the result is not yet fully supported.","major_comments":[{"comment":"The abstract's statement that 'we find no significant differences in the HJ populations between over- and underdense regions' is not supported by the ZGR23 results. In the full ZGR23 sample the residual phase-space density distributions of HJ and non-HJ hosts differ at p_KS = 4.1e-3 even after detrending, and the difference remains marginally significant (p_KS = 0.042) after the 1-5 Gyr age cut. These are the paper's own numbers, so the conclusion must either be restricted to the SWEET-Cat sample or the ZGR23 residual must be quantitatively explained before a general null claim can appear in the abstract.","section":"Abstract and Section 3.1, Figure 4"},{"comment":"The paper reports that hosts in the bottom half of the residual density distribution are significantly alpha-enhanced in the ZGR23 sample (p_KS = 1.7e-5), and notes that [alpha/Fe] is commonly used as an age proxy. This is direct evidence that the detrending has not removed all age-related information: the residual phase-space density still correlates with an independent age indicator. The manuscript mentions this only as a possibility and does not test whether the residual difference in HJ occurrence can be explained by this residual age correlation. A quantitative test, such as including [alpha/Fe] in the detrending or matching variables, is needed before attributing the residual to 'inadequate detrending or isochronal age uncertainties'.","section":"Section 4, alpha-enhancement analysis"},{"comment":"The matching results for ZGR23 are reported as a median Poisson p-value of 0.063 with 39.7% of 1000 resampled matching draws giving p < 0.05. The authors argue that the high fraction of significant draws cannot be interpreted as grounds for rejecting the null because the draws use resampled rather than independent data. While the p-values are indeed not independent, under a true null one would still expect roughly 5% of draws to fall below 0.05, not 39.7%. The observed fraction is therefore itself a sign of a residual difference. The manuscript should either provide a proper permutation test that accounts for the resampling design or explicitly acknowledge that the matching analysis does not establish a null result for ZGR23.","section":"Section 3.2, Figure 5"},{"comment":"The central demographic correction depends heavily on isochronal ages, and the authors report median relative age uncertainties of 55% in the ZGR23 sample. Detrending against a very noisy independent variable attenuates the regression slope, so a residual difference of the kind seen in Figure 4 is exactly what one would expect even if the underlying effect were entirely demographic. The manuscript claims that the residual 'could be due to inadequate detrending or isochronal age uncertainties' but does not test this. A simulation that injects realistic age noise into a demographics-only null model and compares the resulting residual p-value distribution to the observed one would make the argument quantitative and is necessary to support the paper's central conclusion.","section":"Section 2.2 and Section 4, age uncertainties"}],"minor_comments":[{"comment":"The captions state 'log10 pKS = 10.21' and similar values, which is ambiguous. It should be written as p_KS = 10^{-10.21} or the text should explicitly say that the displayed number is the base-10 logarithm of the p-value.","section":"Figure 3 and Figure 4 captions"},{"comment":"The conclusion that the absence of a significant RV/transit subsample difference is 'not purely a sample size effect' is based on randomly drawing subsets of the full sample down to the subsample size. This is a useful sanity check, but it is not a formal power calculation, and the wording should be softened to say that the test is suggestive rather than definitive.","section":"Section 3.3, detection-method subsamples"},{"comment":"The paper removes sources with normalized phase-space density greater than 50 as outliers, but then assigns hosts with rho > 50 to overdensities regardless of P_high. Please clarify how these two statements are reconciled in the actual sample construction.","section":"Section 2.3, outlier removal"},{"comment":"The numbers in Table 1 are informative, but the text does not explain why the number of hosts with valid ages (690 in SWEET-Cat, 1392 in ZGR23) is much smaller than the number of hosts with atmospheric parameters (711 and 1826). A brief explanation of the cuts leading from one to the other would help the reader.","section":"Section 2.1 and Table 1"},{"comment":"The matching threshold of 0.25 in min-max normalized feature space is described as producing 'generally a good match', but no sensitivity analysis is shown for this choice. Given that the matching result is used to support the null claim, a small exploration of threshold values would strengthen the robustness argument.","section":"Section 3.2, matching threshold"}],"recommendation":"major_revision","confidential_remarks":"The paper is a careful and valuable contribution, but the abstract overstates the null result. The ZGR23 residual and the matching statistics need to be addressed quantitatively, and the authors should either soften the central claim or provide a simulation that shows the residual is consistent with a demographics-only model. The use of the Winter et al. (2020) phase-space metric, co-authored by one of the present authors, is transparent and the test is not circular, but the manuscript would benefit from explicitly acknowledging this lineage in the discussion of the metric's interpretation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my read. The paper is a serious re-analysis of the Winter et al. 2020 claim that hot Jupiters are overabundant in phase-space overdensities. The new thing is scale and uniformity: two large homogeneous samples (SWEET-Cat and ZGR23), homogeneous isochronal ages from isoclassify, and like-for-like matching with resampling and size-controlled samples. That is exactly what the earlier Adibekyan/Mustill suggestions lacked. The analysis is careful, the limitations section is honest, and the data are public. Credit where due.\n\nThe soft spots, in proportion. The abstract says 'no significant differences' after correction, but in the larger, transit-dominated ZGR23 sample a residual KS difference survives detrending (p≈4e-3, or 0.042 after a 1–5 Gyr cut), and 39.7% of matched resamples still give p<0.05. That is not a clean null. The authors attribute the residual to isochronal age noise and detection biases, but they never quantitatively test that attribution. With median relative age uncertainties of 55% in ZGR23, attenuation of the detrending slope is a real concern, and their own residual metallicity and [alpha/Fe] differences suggest incomplete correction. So the headline conclusion is plausible but not established.\n\nThat said, I don't think this is fatal. Both detrending and matching move in the same direction, the SWEET-Cat sample is clean, and the residual, if real, is not specific to hot Jupiters—HJ and CJ distributions are indistinguishable in ZGR23, and the effect is driven by giant vs non-giant planets, likely a metallicity/age tag. The paper does not misrepresent its own body; it just over-compresses in the abstract.\n\nBottom line: this deserves a serious referee. The referee should ask the authors to either provide a quantitative test of whether age noise explains the ZGR23 residual, or soften the abstract to 'at most marginally significant' for that sample. The paper is a step forward even if the final word needs asteroseismic ages.","headline":"A careful, largely convincing demographics explanation for the hot Jupiter phase-space overdensity, but the abstract overstates the null result — the ZGR23 residual is left unresolved.","tokens_in":18098,"tokens_out":2439,"would_cite":true,"duration_ms":30779,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The apparent excess of hot Jupiters in dense regions of phase space disappears once host star age, mass, and metallicity are taken into account.","keywords":["hot Jupiters","planet occurrence","phase space density","stellar clustering","host star demographics","isochronal ages","stellar kinematics","planetary systems"],"falsifier":"Take a sample of planet hosts with precise asteroseismic ages, which are far more accurate than isochrone ages for main-sequence stars, and repeat the detrending and like-for-like matching. If a significant excess of hot Jupiters in phase-space overdensities persists after matching on those ages plus mass and metallicity, the paper's central conclusion is wrong. A cheaper check is to split the overdensity and underdensity samples into narrow $[\\mathrm{Fe}/\\mathrm{H}]$ bins and ask whether the hot-Jupiter fraction still differs within bins where the age distributions overlap.","tokens_in":17117,"feed_emoji":"🪐","tokens_out":7918,"duration_ms":88028,"temperature":0.7,"pith_summary":"Hot Jupiters are gas giants on orbits shorter than about ten days, and earlier work reported that they orbit stars in dense regions of six-dimensional phase space more often than expected, hinting that crowded birth environments helped form them. This paper asks whether that signal is real or an illusion created by the ages, masses, and metallicities of the host stars. Using a parent sample of 2265 confirmed exoplanet hosts and two homogeneously derived sets of stellar parameters, it confirms the raw excess: hot-Jupiter hosts do sit at higher phase-space density than other hosts. But stars in overdensities are also younger, more massive, and more metal-rich, and these are exactly the stars known to host hot Jupiters more often. After detrending the density against stellar age or matching over- and underdense hosts like-for-like, the paper finds the excess largely disappears or falls to at most marginal significance, and concludes that the clustering signal is demographic rather than environmental.","feed_headline":"Crowded-space hot-Jupiter excess vanishes after age control","feed_subtitle":"After like-for-like matching, crowded and sparse regions host equal hot-Jupiter rates.","key_machinery":"The central object is the Mahalanobis phase-space density $\\tilde{\\rho}_{M,20}$, a local measure of how many stellar neighbours a host has in six-dimensional phase space (position plus velocity), computed by inverting the 20th-nearest-neighbour Mahalanobis distance after normalising by the covariance of the local neighbourhood. Hosts are split into over- and underdensities with a two-component Gaussian mixture model on $\\log_{10} \\tilde{\\rho}_{M,20}$, with hosts above $\\tilde{\\rho}_{M,20} > 50$ treated as bound clusters. The argument is carried by two correction schemes applied to this split. First, ordinary least squares detrending of $\\log_{10} \\tilde{\\rho}_{M,20}$ against the logarithm of isochronal age, where ages are computed with isoclassify and MIST isochrones from homogeneous SWEET-Cat and Gaia XP parameter sets; this removes the age channel linking density to planet occurrence. Second, like-for-like matching of over- and underdensity hosts in a min-max rescaled space of age, mass, and metallicity, repeated 1000 times with values resampled from isoclassify posteriors, which controls the full multivariate host distribution. Significance is assessed with a Poisson means test (E-test) for hot-Jupiter host counts and Kolmogorov-Smirnov tests on density residuals.","core_discovery":"The paper's central claim is that the previously reported preference of hot Jupiters for phase-space overdensities does not survive correction for host star properties. In both homogeneous samples the raw difference in hot-Jupiter host fraction between over- and underdense regions is significant, with Poisson p-values around $10^{-3}$ to $10^{-4}$, but the overdensities are dominated by younger, more massive, and more metal-rich stars. Removing the age correlation by linear detrending of log phase-space density against log age erases the difference in one sample and reduces it to marginal in the other, and the residual difference there is between giant and non-giant planets rather than between hot and cold Jupiters. Like-for-like matching on rescaled age, mass, and metallicity likewise shifts the Poisson p-values to 0.28 and 0.08 in the two samples, values that random sampling of the same size does not reproduce. The paper therefore states that phase-space density is largely a proxy for stellar kinematics and age, and that no intrinsic environmental enhancement of hot-Jupiter formation is required.","pith_inferences":["If this conclusion holds, earlier interpretations of phase-space clustering as evidence for environment-driven hot-Jupiter formation are overturned; the more direct test would be to look for a clustering signal among wide-orbit planets, which are the planets that stellar encounters actually perturb.","The paper's residual $[\\alpha/\\mathrm{Fe}]$ difference in the ZGR23 sample is a hint that age is not fully removed by detrending; a sharp test would be to repeat the analysis with asteroseismic ages, which are far more accurate than isochrones for main-sequence field stars.","We would also read the results as predicting that any genuine environmental effect on planets should appear preferentially around older, kinematically heated hosts with wide-orbit planets, rather than around the young overdensity stars that dominate the current sample."],"forward_implications":["If the central claim is right, the phase-space-density correlation cannot be used as evidence that stellar flybys in clustered environments create hot Jupiters via high-eccentricity migration.","The Mahalanobis phase-space density should be treated as a kinematic-age proxy, so studies using it must control for host age, mass, and metallicity before attributing planetary differences to the environment.","The small residual difference in the transit-dominated sample is between giant and non-giant planets, not hot and cold Jupiters, so even the residual does not point specifically to hot-Jupiter formation.","Splitting the homogeneous sample into RV and transit hosts separately removes the raw excess, implying the apparent signal is inflated by the different detection mixes in dense and sparse regions.","Future larger samples should match hosts on multivariate property distributions rather than individual parameters, because single-parameter cuts can leave spurious environmental signals."],"supporting_citations":[{"why":"The original claim that hot-Jupiter hosts preferentially reside in phase-space overdensities, which this paper re-tests.","marker":"Winter et al. (2020)"},{"why":"Questioned that result by showing phase-space density correlates with peculiar velocity and stellar age, motivating the detrending approach.","marker":"Mustill et al. (2022)"},{"why":"Provided homogeneous isochronal ages and found underdensity hosts older, but with a small radial-velocity-only sample.","marker":"Adibekyan et al. (2021)"},{"why":"Source of the homogeneous SWEET-Cat stellar parameters for FGK hosts used in the first sample.","marker":"Santos et al. (2013)"},{"why":"Update of SWEET-Cat that supplies the atmospheric parameters and uncertainties used for age fitting.","marker":"Sousa et al. (2021)"},{"why":"Gaia XP spectroscopic parameters that form the second homogeneous sample, ZGR23.","marker":"Zhang et al. (2023)"},{"why":"Showed that hot and cold Jupiters detected by the same method have similar phase-space-density distributions, suggesting the raw signal is method-driven.","marker":"Chen et al. (2023)"},{"why":"Independent work attributing the hot-Jupiter phase-space-density variation to metallicity rather than environment.","marker":"Rampalli et al. (2025)"},{"why":"The isoclassify code used to derive the isochronal ages that power the detrending and matching.","marker":"Huber et al. (2017)"}],"fun_headline_variants":["Hot-Jupiter excess in clusters is just an age effect","Crowded regions lose hot-Jupiter advantage after stellar-age control","Host age, not environment, explains hot-Jupiter clustering","Dense-space hot-Jupiter counts matched after correcting for star age"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the isochronal ages of main-sequence FGK field stars are accurate enough that, after detrending or matching, any residual correlation between age and phase-space density is small; if those ages are systematically biased in a way that tracks phase-space density, the corrections would over- or under-shoot, and the disappearing hot-Jupiter excess could be an artefact of the age model rather than a real demographic effect.","fun_headline_variants_meta":{"raw":{"variants":["Hot-Jupiter excess in clusters is just an age effect","Crowded regions lose hot-Jupiter advantage after stellar-age control","Host age, not environment, explains hot-Jupiter clustering","Dense-space hot-Jupiter counts matched after correcting for star age"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000239,"raw_usage":{"total_tokens":1572,"prompt_tokens":1057,"completion_tokens":515,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":673,"completion_tokens_details":{"reasoning_tokens":444}},"tokens_in":673,"tokens_out":515,"duration_ms":6903,"temperature":1.0,"reasoning_tokens":444,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T17:13:18.318612+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a sample of planet hosts with precise asteroseismic ages, which are far more accurate than isochrone ages for main-sequence stars, and repeat the detrending and like-for-like matching. If a significant excess of hot Jupiters in phase-space overdensities persists after matching on those ages plus mass and metallicity, the paper's central conclusion is wrong. A cheaper check is to split the overdensity and underdensity samples into narrow $[\\mathrm{Fe}/\\mathrm{H}]$ bins and ask whether the hot-Jupiter fraction still differs within bins where the age distributions overlap.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provided homogeneous isochronal ages and found underdensity hosts older, but with a small radial-velocity-only sample."}],"review_version":1}