{"id":"a28bc3cf-ee49-40a8-8057-444140867e53","arxiv_id":"1908.01117","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Five-planet system lifetimes drop as initial eccentricity grows, but aligned eccentric orbits survive almost as long as circular ones, so relative eccentricity, not absolute eccentricity, controls stability.","lead":"This paper uses computer simulations to measure how long idealized systems of five Earth-mass planets remain stable when some planets start on eccentric orbits. It finds that slightly eccentric planets shorten lifetimes, but the alignment of the eccentric orbits matters as much as the eccentricity value.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Exclusion of first-synodic-period collisions from the fits is the main quantitative soft spot; a rerun with random phases would show whether the reported e1=0.05 slope and the 'one eccentric planet shortens lifetimes mainly at wide spacing' claim are robust.","rationale":"The reader's weakest_assumption correctly identifies the first-synodic-period exclusion and the phase representativeness as the main concern. My independent review confirms that this is load-bearing: the fitted slopes for e1 = 0.03 and e1 = 0.05 are computed only over beta ranges starting at 4.19 and 5.32, respectively, so the reported steep drop in slope with eccentricity is not measured at small beta where the excluded collisions occur. The paper's own text calls these collisions 'direct consequences of our initial longitudes,' which is an honest acknowledgment, but it does not quantify how the fits would change if they were included or if phases were randomized. A phase-average test would settle whether the exclusion is benign. The secondary concern about missing error bars on the multiplicative factors in Table 2/Figure 10 is real but less central; it affects the precision of the comparative claims (e.g., aligned runs survive 'far longer,' intermediate planets 'about half as long-lived') but not the main qualitative conclusions. I do not see an internal inconsistency in the derivation or simulation setup; the concern is about representativeness and post-hoc exclusion, which maps to CONDITIONAL rather than REJECT. The aligned-eccentric result is well supported by the direct comparison in Figure 11 and does not depend on the contested fits. I agree with the reader that CONDITIONAL is the right verdict, with the phase-exclusion check as the key condition.","tokens_in":14972,"tokens_out":2304,"duration_ms":21243,"concrete_test":"Re-run the e1 = 0.05 batch (and ideally e1 = 0.02, 0.03) with the same beta grid but many random initial longitudes/periapse angles per beta (e.g., 10-20 realizations at each beta), then fit Eq. 8 over beta >= 3.47 (including the first-synodic-period collision points). If the fitted slope and intercept for e1 = 0.05 remain within their standard errors of the values in Table 3 (b' = 0.785, c' = 1.02), the exclusion is harmless. If the slope changes substantially (e.g., by more than 2 sigma) or the log-linear form breaks down below beta = 5.32, then the paper's central claim about the persistence of the exponential trend for eccentric systems needs qualification, and the abstract's 'generally leads to shorter system lifetime' claim would need to be rephrased to emphasize the extrapolated nature of the small-separation behavior.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper's central quantitative claims rest on log-linear fits (Eq. 6/8, Table 3) to lifetimes versus beta. For the highest-eccentricity single-planet runs (e.g., e1 = 0.05), 22 systems suffered close encounters in under ten years, all between planets 1 and 2 during their first synodic period. Section 4 explicitly says these are 'direct consequences of our initial longitudes' and the fits start 'just after the widest separation system that has a collision during the first synodic period.' This post-hoc exclusion is load-bearing: the abstract and Section 9 claim that larger eccentricity 'generally leads to shorter system lifetime' and that the log-linear trend persists, but the fitted slopes (b' = 0.785 for e1 = 0.05 versus 1.080 for circular) are computed only over the beta range above the last excluded collision. The manuscript provides no evidence that the excluded collisions are purely phase artifacts rather than physically meaningful outcomes for the specified initial conditions. Because all runs use the same golden-ratio longitudes with omega = 0, the first-passage geometry that causes these collisions is not sampled; a different phase choice would place the particle near periapse at first conjunction and could remove the collisions, but it could also change the distribution of lifetimes at larger beta. The central claim 'for a given initial orbital separation, larger initial eccentricity generally leads to shorter system lifetime' is partly an artifact of this exclusion: at small beta, the fitted lines for e1 = 0.05 are extrapolated from beta > 5.32, so the claim that eccentric systems are shorter-lived at all separations is not directly measured where the effect is largest. A related but secondary issue is that Table 2's average multiplicative factors and Figure 10's ratios are quoted without error bars or scatter estimates, despite the paper itself noting up to two orders of magnitude scatter around the regression lines (Section 3).","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper uses N-body integrations (REBOUND/WHFast) to study the dynamical stability of idealized systems of five equal-mass (1 M_Earth) planets orbiting a solar-mass star, uniformly spaced in mutual Hill radii (separation parameter β). It extends prior work on circular orbits by giving one planet an initial eccentricity up to e=0.05, by examining which planet is eccentric, by including AMD-matched outer-planet runs, and by considering all-planet eccentric configurations with aligned or randomly distributed periapses. The main findings are that (i) for fixed β, larger eccentricity generally shortens the close-encounter time; (ii) the log-linear (exponential) scaling of lifetime with β persists for eccentric systems but with shallower slopes; (iii) there is little systematic dependence on which planet is initially eccentric; and (iv) systems with all planets at e=0.05 and aligned periapses are much longer-lived than single-eccentric systems but somewhat shorter-lived than circular systems. Quantitative results are presented as log-linear fit coefficients in Table 3 and average lifetime ratios in Table 2.","tokens_in":15278,"tokens_out":6567,"duration_ms":65140,"significance":"If the results hold, the paper provides a useful quantitative mapping of how initial eccentricity and periapse alignment affect the stability of tightly packed multi-planet systems. It successfully reproduces the circular-orbit fits of Smith and Lissauer (2009) and Obertas et al. (2017) with small uncertainties, giving confidence in the methodology. The conclusion that relative eccentricity (alignment) matters more than absolute eccentricity is directly relevant to interpreting Kepler multi-planet systems and to TTV-based eccentricity constraints. The AMD-matched comparison is a clever control and the randomized-periapse series brackets the phase dependence for the all-eccentric case. The simulation setup is standard and reproducible using the open-source REBOUND code.","major_comments":[{"comment":"The exclusion of first-synodic-period collisions from the log-linear fits is not justified by any control experiment. For the e1 = 0.05 batch, 22 systems with lifetimes under ten years are discarded because they are described as 'direct consequences of our initial longitudes', and the fit begins 'just after the widest separation system that has a collision during the first synodic period' (Figure 3 caption, Section 4). The fitted slope b' = 0.785 for e1 = 0.05 in Table 3, and the claim in Section 9 that the exponential trend persists for eccentric systems, therefore depend on a post-hoc choice of β-range. Please add at least one batch with randomly drawn initial longitudes (or an alternative deterministic phase) to show that those collisions are phase artifacts rather than physical outcomes for the stated initial conditions, or present the fits including those systems and discuss how the conclusions change.","section":"Section 4, Eq. (6), Table 3"},{"comment":"The stopping criteria (10^10-year cap and termination of a batch after five systems exceed 2×10^9 years) censor the lifetime distribution at the upper end, and the fits in Table 3 use lifetime = 10^10 years for surviving systems. Because the fit range for each batch is set by these censoring rules, the reported slopes b' and intercepts c' are range-dependent; the paper notes this but does not quantify it. A sensitivity analysis (e.g., refitting over a common β-range for all batches, or excluding censored systems) is needed to show that the conclusions—slopes decreasing with eccentricity, and aligned-eccentric systems being longer-lived than single-eccentric systems—are not driven by the censoring treatment.","section":"Section 2.2.2, Table 3"},{"comment":"All single-eccentric-planet batches use a single initial-phase prescription (θ_i = 2πiλ with the golden ratio λ, and ω = 0 for all planets). The claim of 'little systematic dependence of which planet is initially on an eccentric orbit' and the ordering of lifetimes for e1, e2, e3, e5 eccentricities are based on this one geometry. At e = 0.05, the radial excursion is roughly 5% of the semi-major axis, so the phase at first conjunction can materially affect the initial encounter. The manuscript should test at least one alternative phase choice (or a small set of random phases) for the e = 0.05 batches to confirm that the qualitative ordering and the log-linear slopes are not artifacts of the golden-ratio phasing.","section":"Section 2.2.1, Sections 4-9"}],"minor_comments":[{"comment":"The text reads 'we also set the argument of peripasisω = 0'; this should be 'periapsis' with a space before the equation.","section":"Section 2.2.1"},{"comment":"The sentence '(eq. 7. Table 3 also gives the values of σb′ and σc′...' is missing a closing parenthesis after 'eq. 7'.","section":"Section 8"},{"comment":"The notation 'e5 = eamd 1 |0.05' is not defined in the table caption; please clarify how the AMD-matched eccentricity values are denoted and computed.","section":"Table 3"},{"comment":"The axis label 'logte' appears to be a typo for 'log t_e'.","section":"Figure 13"},{"comment":"The paper reports R² values only for three batches; consider tabulating R² for all fits in Table 3 to allow a more direct comparison of the quality of the log-linear approximation.","section":"Section 7"},{"comment":"The 'average multiplicative factor' values are given without uncertainties; since they are computed from a finite range of β, providing standard errors or bootstrap intervals would strengthen the quantitative comparisons.","section":"Table 2"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid empirical study of an important question, and the central trends are likely robust. The main issue is the post-hoc exclusion of first-synodic-period collisions combined with a single initial-phase choice; this is fixable with a modest number of additional simulations or a careful sensitivity analysis. If the authors address this and the censoring issue, the paper would be suitable for Icarus. I also note that the simulation data are not made publicly available, which would be helpful for reproducibility but is not strictly required."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know up front. First, the main new result is that five-planet systems with all planets at e=0.05 and aligned periapses survive much longer than systems with a single planet at e=0.05, and almost as long as circular systems. That is a clean, useful finding for anyone modeling Kepler's multi-planet systems. Second, the paper is transparent about its biggest quantitative vulnerability: it excludes systems that collide during the first synodic period from its log-linear fits, and the fits for e1=0.05 start at beta > 5.32. If those excluded collisions are taken as physical rather than phase artifacts, the reported slope and the 'shorter lifetimes at all separations' claim would need revision.\n\nWhat is actually new: Obertas et al. did the circular case, Smith and Lissauer had a few eccentric runs. This paper systematically scans e=0.01-0.05 for each planet slot, adds AMD-matched comparisons, and quantifies the aligned-eccentric longevity effect. The fits reproduce the circular coefficients from prior work with standard errors, and the text explicitly discusses the scatter and the 10^10-year censoring. That counts for a lot.\n\nWhere the soft spots are: the exclusion issue is real and load-bearing, as I said. The paper calls the excluded collisions 'direct consequences of our initial longitudes,' which is plausible, but there is no test of that claim. A rerun with randomized initial phases for the e1=0.05 batch, or at least a sensitivity analysis, would settle it. Without that, the quantitative slopes in Table 3 for the e=0.05 runs carry an asterisk. The second issue is minor: the average multiplicative factors in Table 2 and Figure 10 come without error bars, even though the scatter around the fits is up to two orders of magnitude. That is a presentation flaw, not a fatal one.\n\nOverall the central argument holds. The qualitative conclusion that relative eccentricity matters more than absolute eccentricity is supported by the aligned versus single-eccentric comparison, and it matches Pu and Wu's earlier work. The citation pattern is appropriate; the self-citations are to the benchmarks being extended, not inflation.\n\nWho this is for: dynamicists working on Kepler multi-planet stability and TTV modeling. It deserves serious peer review. I would send it out with a request that the authors address the phase-artifact exclusion, either with a robustness run or by softening the slope claims for the highest eccentricities. As it stands I would not desk-reject it; I would expect a conditional accept after revision.","headline":"A transparent numerical extension of the circular-orbit stability results to eccentric systems; the main quantitative caveat is the exclusion of first-synodic-period collisions from the fits, which needs a robustness check.","tokens_in":15895,"tokens_out":3084,"would_cite":true,"duration_ms":32474,"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":"Relative eccentricity, not individual eccentricity, controls the lifetime of closely packed five-planet systems; aligned eccentric orbits are nearly as stable as circular ones.","keywords":["exoplanets","orbital stability","eccentricity","Hill radius","N-body simulations","mean motion resonance","angular momentum deficit","Kepler multi-planet systems"],"falsifier":"Run the same batches of simulations (e.g., e1=0.05) with hundreds of randomly chosen initial longitudes instead of the golden-ratio set. If close-encounter times under ten years still occur at the same rate for small β, the excluded runs are real and the fitted slopes for eccentric systems would steepen; if they disappear, the paper's exclusion is justified.","tokens_in":14775,"feed_emoji":"🪐","tokens_out":8268,"duration_ms":73905,"temperature":0.7,"pith_summary":"This paper asks how closely five Earth-mass planets can be packed around a Sun-like star and still remain dynamically stable for billions of years, when the planets start on mildly eccentric orbits. Using N-body integrations, it finds that raising one planet's eccentricity to 0.05 shortens the time to a close encounter by about two orders of magnitude, and that which planet is eccentric barely matters. The exponential 'wider spacing means much longer lifetime' trend known from circular systems persists, with a shallower slope. The central discovery is that the relative eccentricity between neighboring planets, not the absolute eccentricity, sets the lifetime: systems where all five planets have e=0.05 with aligned periapses survive almost as long as circular systems, while randomizing periapses makes them far shorter-lived.","feed_headline":"Aligned eccentric orbits keep planet systems stable far longer","feed_subtitle":"Simulations of five Earth-mass planets show relative eccentricity, not individual e, sets packing limits.","key_machinery":"The machinery is a set of N-body integrations using a symplectic Wisdom-Holman integrator. Systems of five equal-mass (one Earth-mass) planets around a one-solar-mass star are initialized coplanar with semi-major axes spaced by a fixed multiple β of their mutual Hill radius, with one planet given an initial eccentricity e up to 0.05 or with all planets eccentric. Lifetime is defined as the time until any two objects come within 0.01 AU, and the resulting close-encounter times are fit to the log-linear relation log T = b'β' + c', where β' = β - 2√3, the two-planet Hill stability limit. A second tool is the angular momentum deficit (AMD), the amount by which each orbit's angular momentum falls short of a circular orbit of the same semi-major axis; comparing batches with equal AMD isolates whether total AMD or the location of the eccentricity controls stability.","core_discovery":"The central claim is that in these idealized planar five-planet systems, the quantity controlling dynamical lifetime is the difference between neighboring planets' eccentricity vectors—their relative eccentricity—rather than the individual eccentricity values. A single planet at e=0.05 reduces the system's close-encounter time by roughly a factor of 75 at a given spacing, with intermediate planets causing slightly more damage than inner or outer ones. But when all planets start at e=0.05 with aligned periapses, lifetimes are far longer than the single-eccentric-planet case and close to the circular baseline, whereas randomly oriented periapses over 360 degrees cut lifetimes dramatically and flatten the exponential fit. The interpretation is that aligned eccentric orbits reduce the minimum initial separation between neighboring orbits less than random phases do, matching the idea that minimum initial orbital separation is the key stability parameter. The paper also shows that lifetimes do not scale with total angular momentum deficit, arguing that mean motion resonances, not secular effects, dominate the destabilization.","pith_inferences":["If relative eccentricity is the true control parameter, then a system's stability boundary may be determined by the distribution of periapsis separations across adjacent pairs; a direct test would be to run five-planet systems with the same minimum periapsis gap but different absolute eccentricities and check that lifetimes match.","The paper excludes very short lifetimes that occur during the first synodic period as artifacts of the chosen initial longitudes; an ensemble with randomized initial longitudes would quantify how often such prompt instabilities actually occur and whether the fitted slopes for e=0.05 change.","Because aligned periapses are strongly stabilizing, any physical process that aligns orbits—such as disk damping or tidal circularization—could permit more tightly packed planet systems than circular-orbit stability maps suggest; this is a testable prediction for observed multi-planet systems with moderate eccentricities."],"forward_implications":["At a fixed initial spacing, raising one planet's eccentricity from 0 to 0.05 shortens the system's lifetime by roughly two orders of magnitude, averaged over the β range studied.","The exponential relationship between initial spacing and close-encounter time holds for all eccentric setups tested, with slopes that decrease as eccentricity rises, so a single fitted line per configuration still predicts stability.","Systems with all five planets at e=0.05 and aligned periapses have lifetimes close to those of initially circular systems with the same spacing, so moderate aligned eccentricity does not force a system to be more spread out.","Mean motion resonances destabilize eccentric systems too, but by smaller factors than circular ones, making the resonance dips in the lifetime-versus-spacing curves less pronounced."],"supporting_citations":[{"why":"Provides the circular-orbit baseline lifetimes and the exponential fitting approach that this paper extends to eccentric initial conditions.","marker":"Obertas et al. (2017)"},{"why":"Earlier study of closely-spaced systems with some eccentric/inclined cases that this paper extends at ten times the resolution in initial separation.","marker":"Smith and Lissauer (2009)"},{"why":"Shows that minimum initial orbital separation (effectively relative eccentricity) sculpts the spacing of Kepler planets; the paper's aligned-versus-random results support this.","marker":"Pu and Wu (2015)"},{"why":"Gives the two-planet Hill stability criterion and the critical separation 2√3 used as the shift in the fit variable β'.","marker":"Gladman (1993)"},{"why":"Justifies starting near the two-planet critical separation for small eccentricities via Hill stability in the AMD framework.","marker":"Petit et al. (2018)"},{"why":"Supplies the open-source N-body code used for all integrations.","marker":"Rein and Liu (2012)"},{"why":"Supplies the symplectic Wisdom-Holman integrator used for the long-term integrations.","marker":"Rein and Tamayo (2015)"},{"why":"Provides the analytically derived AMD-stability criterion that frames the paper's AMD comparisons.","marker":"Laskar and Petit (2017)"}],"fun_headline_variants":["Relative eccentricity sets packing limit for five-planet systems","Aligned eccentric orbits mimic circular stability in packed systems","Eccentricity alignment, not eccentricity, decides planet stability","Random planet eccentricities shrink five-planet system lifetimes","Relative eccentricity governs close-packed planet lifetimes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The fixed initial longitudes (golden-ratio phases and zero periapsis angles) are assumed representative, and the very short lifetimes seen in high-eccentricity runs at close spacing are judged to be phase artifacts and excluded from the fits; if those runs are physically meaningful, the reported slopes and the conclusion that eccentricity mainly shortens lifetimes at wide spacings would need revision.","fun_headline_variants_meta":{"raw":{"variants":["Relative eccentricity sets packing limit for five-planet systems","Aligned eccentric orbits mimic circular stability in packed systems","Eccentricity alignment, not eccentricity, decides planet stability","Random planet eccentricities shrink five-planet system lifetimes","Relative eccentricity governs close-packed planet lifetimes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000643,"raw_usage":{"total_tokens":2997,"prompt_tokens":1023,"completion_tokens":1974,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":639,"completion_tokens_details":{"reasoning_tokens":1896}},"tokens_in":639,"tokens_out":1974,"duration_ms":13495,"temperature":1.0,"reasoning_tokens":1896,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:22:35.114172+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same batches of simulations (e.g., e1=0.05) with hundreds of randomly chosen initial longitudes instead of the golden-ratio set. If close-encounter times under ten years still occur at the same rate for small β, the excluded runs are real and the fitted slopes for eccentric systems would steepen; if they disappear, the paper's exclusion is justified.","supporting_citations":[{"cited_title":"and Petit, A","cited_arxiv_id":null,"evidence_quote":"Provides the analytically derived AMD-stability criterion that frames the paper's AMD comparisons."}],"review_version":1}