{"id":"5d1c1526-1981-4ee4-8cb8-25e1f3f98642","arxiv_id":"1908.01679","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"The period distribution of giant planets around metal-rich, sun-like single stars shows two peaks separated by a gap from roughly 494 to 924 days, a feature the author argues is very unlikely to arise by chance.","lead":"This paper reports a gap in the orbital period distribution of giant planets around metal-rich, sun-like single stars, splitting the usual period pileup into two peaks. The finding matters because such a sharp feature, if real, would constrain how giant planets form and migrate.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Selection effects on RV detectability, not modeled, could create the rSLSS-only gap; the binomial argument in §5.2.1 assumes period-independent rSLSS fraction.","rationale":"Reader's weakest assumption is the right one. The paper's defenses are real but incomplete: splitting by observer group addresses whole-program cutoffs, not per-star sensitivity; comparing LSG/pSLSS/SLBS counts addresses a gross failure to find rSLSS objects in one period range, not a smooth period-dependent completeness difference tied to metallicity and logg. The binomial calculation in §5.2.1 is the mathematical core of the argument and it explicitly assumes the rSLSS fraction is constant in period. That assumption is load-bearing and untested. The proposed injection-recovery test would settle whether the gap is a real feature. I do not think the paper deserves rejection; the feature may well be real, but the current analysis does not rule out the selection-effect explanation, so the conditional verdict stands.","tokens_in":16315,"tokens_out":4927,"duration_ms":54348,"concrete_test":"Run an injection-recovery test on the actual RV timeseries for the rSLSS and comparison stars (or, failing that, on the published observation baselines and cadences from exoplanets.org): inject synthetic planets with periods drawn from a smooth log-uniform distribution over 100–5000 days and masses drawn from the observed msini distribution, recover them with the same detection pipeline, and compare the recovered period distribution in the rSLSS subset with the other subsets. If the recovered rSLSS distribution still shows a gap at 653–924 days at significance above 5%, the concern is refuted; if the gap disappears, the central claim is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim depends on the gap being physical rather than a period-dependent selection effect. The paper's argument in §5.2.1 computes the probability of 33 consecutive non-rSLSS objects in the deep gap as 280(1−113/313)^33 ≈ 1.1e−4, which assumes the probability that an object in the gap is rSLSS is the same as the global ROI fraction 113/313. This is exactly the assumption that RV detectability is period-independent and identical across the rSLSS and comparison populations. Section 2.6 shows only that the major observing groups found planets on both sides of the gap; it does not model how detection probability varies with period, eccentricity, or stellar properties such as metallicity and logg. Since the rSLSS subset is defined by the very stellar properties that correlate with RV survey target selection and planet eccentricity, a period-dependent detection bias confined to this subset cannot be excluded by the count comparisons. If such a bias exists, the empty 653.2–923.8 day region and the surrounding peaks could be an artifact.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes radial-velocity planets from exoplanets.org and defines a subset of 113 'rSLSS' planets hosted by metal-rich, sun-like (log g >= 4), single stars with 4500 < Teff < 6500 K. Within a 100-5000 day region of interest it finds a bimodal pileup in the period distribution separated by a shallow gap at 493.7-923.8 days and a completely empty deep gap at 653.2-923.8 days. It argues via Monte Carlo spacing tests, a binomial run statistic, and a comparison of observing groups that this structure is neither random nor produced by observational effects, and that the gap is strongest for single-planet, roughly Jupiter-mass systems. The authors interpret the feature as evidence that planet formation is more uniform than previously assumed.","tokens_in":16557,"tokens_out":10595,"duration_ms":104511,"significance":"If the gap is physical, the claim is significant: it would be a sharp, large-population feature in the giant-planet period distribution, with strong implications for migration and formation models. The paper has the merit of using a public catalog, presenting detailed count tables, and attempting several independent arguments against a random or observational origin. However, the analysis does not currently establish the feature: the null model is unrealistic, the selection boundaries are post hoc, and no RV completeness model is included. The central result is therefore plausible but unproven; the paper would need a substantially more rigorous statistical treatment and preferably an independent-sample validation before this claim can be accepted.","major_comments":[{"comment":"The gap boundaries (493.7-923.8 d; deep gap 653.2-923.8 d) and the msini cut 0.30 < m sin i < 9.0 MJ are chosen after inspecting the observed distribution: §3.1 explicitly says the cut is 'made possible' by the two in-gap objects lying at the extremes of msini. The significance calculations do not include a multiplicity correction for the number of possible gap locations, gap widths, and selection cuts that could have been tried. Consequently the p-values in Table 3 and the strengthened tail-to-peak ratios for the rlJ subset cannot be quoted as false-alarm probabilities for the discovery. The authors should either pre-specify the cuts, or evaluate the null distribution of the full search procedure (allowing gap position, width, and msini bounds to vary), or validate the feature on an independent later sample.","section":"§2.3, Table 1, §3.1"},{"comment":"The Monte Carlo null draws N points uniformly in log period over a chosen ROI. This is not a realistic null for a period distribution that already contains a broad pileup at hundreds of days (Figure 1). A central gap will appear artificially unlikely when compared to a uniform distribution, because the uniform null does not reproduce the high-density wings that bracket the gap. The authors should generate nulls from a smooth single-peaked model fitted to the comparison populations or to the combined all-planet distribution, and should run the same gap-finding algorithm (including a search over positions and widths) on those nulls. In addition, the algorithm described in §5.1.2 (largest empty spacing) does not by itself produce the 'Full Gap' or 'Both-Together' entries in Table 3, so those calculations are not reproducible from the text.","section":"§5.1, Table 3"},{"comment":"The binomial calculation treats the 33 objects in the deep gap as independent draws with probability 113/313 of being rSLSS. This assumes that the rSLSS fraction is constant across the ROI, i.e., that RV detectability is independent of period and identical for rSLSS and comparison stars. That is precisely the assumption at issue. The rSLSS subset is defined by log g, [Fe/H], and Teff, which enter RV target selection, and planet eccentricity (which affects detection probability) also correlates with these stellar properties. Section 2.6 shows only that several observing groups found planets on both sides of the gap; it does not model detection probability versus period, eccentricity, or stellar parameters. Without a completeness model or a matched control sample, a period-dependent detection bias confined to this subset cannot be excluded, and the empty deep gap could be an artifact.","section":"§5.2.1"},{"comment":"The rlJ selection is constructed by choosing msini bounds that exclude the two objects remaining in the gap. The subsequent statement that the gap becomes 'even more distinct' in this selection is therefore guaranteed by construction and is not independent evidence. The abstract's claim that the feature is characteristic of planets with masses near Jupiter's rests on this circular step. The mass cut should be defined before inspecting the gap, or the rlJ test should be applied to an independent data set.","section":"§3.1, §3.2"}],"minor_comments":[{"comment":"The data download date is given as 2017 January 31 in §2.2 but as 2016 January 31 in the Table 1 caption; this should be reconciled, since it affects reproducibility.","section":"§2.2 vs Table 1 caption"},{"comment":"The abstract says 'nearly 40% of planets with periods past 200 days' and the ROI starts at 100 days, while the quoted fraction 113/313 is 36% of the 100-5000 day ROI; please use consistent period boundaries and percentages.","section":"Abstract, §2.2"},{"comment":"Figure 6 appears to be never referenced in the text; it should be discussed in Section 4 or 5, or removed.","section":"Figure 6"},{"comment":"The text says the metallicity dividing line is set at [Fe/H]=0 for simplicity but 'would be better drawn at -0.03'; Table 1 should state explicitly which boundary is used, and the counts should be recomputed for the alternative boundary if it changes the rSLSS membership.","section":"Section 4"},{"comment":"The Monte Carlo program is not provided; including code, a random seed, and the exact gap-detection statistic is necessary to reproduce the 1-in-10^4 values in Table 3.","section":"§5.1.2, Table 3"},{"comment":"There are numerous typos and incomplete placeholders: the abstract begins 'Thepileupofplanets', §2.6 has 'detemined', §2.1 has 'of of', and the citation block contains duplicated 'How cite this article' lines and a placeholder journal abbreviation.","section":"General"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The one thing to know: this paper reports a genuinely new empirical pattern — a gap splitting the period pileup of planets around metal-rich, sunlike, single stars (rSLSS) into two peaks — and it would be a useful constraint on giant-planet formation if it holds. The other thing to know: the statistical support is weaker than the paper claims, and the main hole is the unmodeled RV selection function.\n\nWhat is actually new: the specific bimodal structure in this population. The paper uses a public RV catalog, defines the rSLSS subset, and shows that the gap around 500–900 days is present in data from both major observing groups. That cross-group check is a good instinct. It also shows the gap is not present in the same period range for low-surface-gravity stars or metal-poor sunlike stars, which is a fair way to argue against a simple period-dependent selection effect.\n\nWhere it gets soft: the gap boundaries and the metallicity cut are chosen after looking at the data, so the Monte Carlo tests that count how often a gap \"anywhere\" appears are not fully post-hoc — the paper does allow the gap to move — but the null is a uniform random distribution. A uniform null is not realistic for a period distribution that has a broad physical pileup; testing against a smooth, single-peaked null would be more convincing, and would probably weaken the claimed 10^-4 to 10^-5 numbers. The deeper issue is in §5.2.1: the binomial probability of 33 consecutive non-rSLSS objects assumes each object in the gap region has the global 113/313 chance of being rSLSS. That is exactly the assumption that detection efficiency is period-independent and identical for the rSLSS and comparison populations. RV sensitivity depends on period, eccentricity, and stellar properties, and the rSLSS subset is defined by the very stellar properties that correlate with target selection. So the gap could be deepened by a period-dependent bias confined to this subset. The cross-group comparison helps, but it doesn't model detectability.\n\nThere are also minor issues: the \"no low msini\" and \"no high msini\" cuts are applied after seeing that the two remaining gap objects are at the extremes, which is a mild post-hoc selection, and the paper acknowledges some of this.\n\nNet: the empirical pattern is plausibly real and worth taking seriously, but the significance analysis is not robust enough to carry the strong conclusions. This paper deserves a serious referee — I would send it out, with the expectation of major revision. It is a good candidate for a reading group on selection effects and post-hoc inference, though I probably wouldn't cite it in my own work in the near term.","headline":"A genuinely new empirical claim about a bimodal period distribution in metal-rich sunlike single stars, but the significance analysis is too post-hoc and the detection-bias argument is not modeled; still worth refereeing.","tokens_in":17035,"tokens_out":2044,"would_cite":false,"duration_ms":19872,"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 period distribution of planets around metal-rich, Sun-like, single stars is bimodal: two peaks separated by a sharp, almost empty gap from 653 to 924 days.","keywords":["exoplanets","planet period distribution","planet-metallicity correlation","radial velocity","planet formation","bimodal period distribution","Sun-like stars"],"falsifier":"Track the detection completeness of the RV surveys as a function of period for rSLSS stars: if a period-dependent completeness model predicts a dip as deep as observed in exactly 653–924 days, or a blind RV survey with longer than five-year baselines finds a statistically significant number of such planets in that window, the claim of a physical gap is falsified.","tokens_in":16114,"feed_emoji":"🪐","tokens_out":12275,"duration_ms":107665,"temperature":0.7,"pith_summary":"This paper reports that the long-known 'pileup' of exoplanet orbital periods at roughly one year and beyond is not a single smooth bump for a major population of planets. For planets orbiting single stars with roughly solar surface gravity and metallicity at least solar, the pileup splits into two peaks separated by a gap from 493.7 to 923.8 days, with zero such planets found between 653.2 and 923.8 days. The author argues that the gap is neither a random fluctuation nor an observational artifact: Monte Carlo simulations put the chance of a gap of this size appearing at random below 1 in 10,000, and the same period range is densely populated among planets of lower-metallicity stars, binary stars, and evolved stars. The structure is sharpest for near-Jupiter-mass planets in single-planet systems, suggesting the feature belongs to a specific planet population. If correct, the discovery implies that planet formation and migration are more regular and uniform than the current stochastic picture assumes.","feed_headline":"Empty 270-day gap splits planet periods of metal-rich Sun-like stars","feed_subtitle":"Two sharp peaks around an empty 653-924 day gap; Monte Carlo odds under 1 in 10,000.","key_machinery":"The load-bearing quantity is the period distribution of the 'rSLSS' population: planets orbiting single stars with solar-like surface gravity (log g ≥ 4), effective temperature 4500–6500 K, and metallicity [Fe/H] ≥ 0. The paper bins the log periods of 113 such objects and defines a natural unit, the gap-width (0.272 in log period, 493.7–923.8 days), to compare densities across regions. Within this unit, the mean ROI density is 18.1 objects per gap-width; the gap contains only six objects (0.33 of mean), while the quarter-gap-width bins adjacent to the gap hold 10 and 13 objects respectively (2.2 and 2.9 times the mean). The unlikelihood argument runs on two independent tests: Monte Carlo simulations of random distributions over two bracketing ranges (one log-period wide and three-gap-width wide) yield deep-gap frequencies of roughly 1 in 5,000 to 1 in 9,000, full-gap frequencies of about 1 in 2,000 to 1 in 4,500, and both-together frequencies of 1 in 40,000 to 1 in 85,000; and a Bernoulli-style calculation shows a consecutive run of 33 non-rSLSS objects in the deep-gap region has probability about 1.1 × $10^{-4}$.","core_discovery":"Using 113 radial-velocity planets around 'rSLSS' stars—single stars with surface gravity log g ≥ 4, effective temperature 4500–6500 K, and metallicity [Fe/H] ≥ 0—the author finds that the period distribution in the 100–5000 day range consists of two distinct peaks, not one. The short-period peak (SPP) and long-period peak (LPP) are separated by a shallow gap from 493.7 to 923.8 days containing only six objects, and by a completely empty deep gap from 653.22 to 923.8 days. Density per log period in the bins immediately adjacent to the gap is 2–3 times the mean ROI density, while the gap sits at 0.33 of that mean. The gap survives in multiple-planet systems but is most prominent among single-planet systems with masses 0.3–9 Jupiter masses; applying that mass cut removes the last two gap objects. The author quantifies unlikelihood in two ways: Monte Carlo draws of random period distributions with the same counts produce comparable deep gaps with frequency below $10^{-4}$, and the probability that 33 consecutive objects in the deep-gap region all fall outside the rSLSS selection is about 1.1 × $10^{-4}$. Because the comparison populations (metal-poor sunlike, binary, and low-surface-gravity-hosted planets) have their own pileup across exactly that period range, the paper concludes the gap is a physical feature, not a selection effect.","pith_inferences":["A full detection-completeness model of the underlying RV surveys, treating period, stellar metallicity, and binarity as covariates, could either confirm or weaken the physical-gap conclusion, which this paper reaches by comparing group counts rather than by modeling detectability.","If the gap is set by a condensation front or resonance, its location should scale with stellar mass or disk temperature; observations of lower-mass stars could test whether the gap shifts in period.","A natural next check is whether eccentricity differs across the gap; the author's earlier eccentricity–metallicity correlation suggests that the two peaks may belong to dynamically distinct populations.","The persistence of the gap in multi-planet systems could be sharpened by testing whether its boundaries correlate with the presence of outer companions, which would point to dynamical sculpting rather than formation alone."],"forward_implications":["The period distribution of giant planets is not smooth: a population comprising nearly 40% of planets past 200 days shows preferred period bands and a nearly forbidden band between roughly 653 and 924 days.","Standard stochastic planet-formation models need a mechanism that suppresses or relocates planets in this window for metal-rich single stars, since random assembly would not preserve such a sharp gap.","New planet surveys that measure host-star metallicity, surface gravity, and binarity can test the claim directly; even period-only surveys should show a notch in the all-planet histogram where the gap sits.","The sharp edge of the gap at 923.8 days is a clock-like feature: its abruptness constrains migration timescales and disk properties regardless of the exact formation mechanism.","For binary-hosted planets of similar metallicity, the gap is partially filled, so stellar companions serve as a control population that can help isolate the mechanism."],"supporting_citations":[{"why":"Supplies the exoplanet orbit database from which every period, metallicity, and multiplicity value is drawn.","marker":"Han et al., 2014"},{"why":"Establishes the planet–metallicity correlation that justifies the metal-rich division of the sample.","marker":"Fischer & Valenti, 2005"},{"why":"Defines the short-period valley and the eccentricity–metallicity framework that sets the region of interest.","marker":"Dawson & Murray-Clay, 2013"},{"why":"Shows binary-hosted planets have higher eccentricities, motivating the restriction to single-star hosts.","marker":"Taylor, 2013"},{"why":"Found the short-period eccentricity–metallicity correlation that the paper extends to the period distribution.","marker":"Taylor, 2012"},{"why":"Reviews the metallicity–giant-planet connection used to frame the rSLSS population.","marker":"Udry & Santos, 2007"}],"fun_headline_variants":["Two peaks, one gap: planet orbits around metal-rich stars","Unexpected bimodal planet periods around metal-rich Sun-like stars","Metal-rich Sun-like stars show twin peaks in planet periods","Empty gap splits planet periods into two peaks around metal-rich stars"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes that radial-velocity surveys would detect planets in the 653–924 day period range around metal-rich, Sun-like, single stars as readily as around the comparison stars, so the empty gap reflects a real scarcity rather than a period-dependent selection effect tied to the very stellar properties used to define the subset.","fun_headline_variants_meta":{"raw":{"variants":["Two peaks, one gap: planet orbits around metal-rich stars","Unexpected bimodal planet periods around metal-rich Sun-like stars","Metal-rich Sun-like stars show twin peaks in planet periods","Empty gap splits planet periods into two peaks around metal-rich stars"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000869,"raw_usage":{"total_tokens":3814,"prompt_tokens":1048,"completion_tokens":2766,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":664,"completion_tokens_details":{"reasoning_tokens":2696}},"tokens_in":664,"tokens_out":2766,"duration_ms":18824,"temperature":1.0,"reasoning_tokens":2696,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:47:48.240374+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Track the detection completeness of the RV surveys as a function of period for rSLSS stars: if a period-dependent completeness model predicts a dip as deep as observed in exactly 653–924 days, or a blind RV survey with longer than five-year baselines finds a statistically significant number of such planets in that window, the claim of a physical gap is falsified.","supporting_citations":[{"cited_title":"APACrefauthors \\ 2013 05 , ArXiv e-prints","cited_arxiv_id":null,"evidence_quote":"Shows binary-hosted planets have higher eccentricities, motivating the restriction to single-star hosts."},{"cited_title":"APACrefauthors \\ 2012 11 , ArXiv e-prints","cited_arxiv_id":null,"evidence_quote":"Found the short-period eccentricity–metallicity correlation that the paper extends to the period distribution."}],"review_version":1}