{"claim": "The Deep Space Network is structurally oversubscribed by construction: supply (antenna aperture and band coverage) is fixed over any operational planning horizon while demand grows continuously, so in recurring windows requested antenna-time exceeds available antenna-time and the schedule must drop or reduce a subset of requests. This establishes that the contention problem the dissertation studies is real, not hypothetical.", "evidence": [{"source": "Goh, Johnston et al., Deep Space Network Scheduling via Mixed-Integer Linear Programming (the DSN scheduling problem is explicitly oversubscribed: only a subset of requested activities can be scheduled and operators must decide which to exclude), IEEE Access (2021)", "doi_or_url": "https://doi.org/10.1109/access.2021.3064928", "grade": "A"}], "facet": "empirics", "chapter": "ch1_introduction", "subclaim": "real"}
{"claim": "DSN demand is growing along several axes at once (a larger deep-space spacecraft population, rising per-mission data return, and a thirty-year traffic horizon projecting unparalleled growth in robotic spacecraft and the entry of tracking-intensive human exploration missions), so the oversubscription that motivates the science-throughput-penalty question is worsening rather than transient. This grounds the materiality of the problem.", "evidence": [{"source": "Abraham, MacNeal, Heckman, Traffic Modeling for the Deep Space Network in the Human Exploration Era, SpaceOps (2016)", "doi_or_url": "https://doi.org/10.2514/6.2016-2523", "grade": "B"}], "facet": "empirics", "chapter": "ch1_introduction", "subclaim": "material"}
{"claim": "The crewed-exploration era is a fielded, geometry-constrained reality for the DSN, not a forecast abstraction: NASA's DSN already supported the Artemis I crewed-exploration-era mission as a fixed-complex, geometry-constrained network supporting the emerging high-rate, time-critical, deferral-intolerant human-exploration traffic class. This grounds that the load mix the penalty is indexed to is materially shifting.", "evidence": [{"source": "Harmon et al., Pre-launch lessons learned from NASA's Deep Space Network support for the Artemis I mission to the Moon, Acta Astronautica Vol. 210 (2023)", "doi_or_url": "https://doi.org/10.1016/j.actaastro.2023.05.016", "grade": "A"}], "facet": "empirics", "chapter": "ch3_literature_review", "subclaim": "material"}
{"claim": "The DSN schedule of record is produced by a documented multi-stage, peer-to-peer negotiation among the user community via the Service Scheduling Software: the DSN Scheduling Engine interprets user requirements, elaborates them into tracking passes, and resolves conflicts within a negotiated process extended to long-range planning. This grounds that the request, negotiated schedule, and realtime log are distinct archived objects, so a catalog of record and a multi-stage data structure genuinely exist to be specified.", "evidence": [{"source": "Johnston, Tran, Arroyo, Sorensen, Tay et al., Automated Scheduling for NASA's Deep Space Network, AI Magazine 35(4) (2014)", "doi_or_url": "https://doi.org/10.1609/aimag.v35i4.2552", "grade": "A"}], "facet": "measurement", "chapter": "ch4_data_and_measurement", "subclaim": "material"}
{"claim": "The DSN antenna-satellite scheduling problem is formally an oversubscribed mixed-integer linear program coupling many missions through shared complex-band-epoch capacity, with incremental (Delta-MILP) re-solve methods established. This grounds that an optimal-allocation counterfactual and a dual/shadow-price benchmark are well-posed and constructible on the same data, which the descriptive design currently lacks.", "evidence": [{"source": "Claudet, Alimo, Goh, Johnston, Delta-MILP: Deep Space Network Scheduling via Mixed-Integer Linear Programming, IEEE Access (2022)", "doi_or_url": "https://doi.org/10.1109/access.2022.3164213", "grade": "A"}], "facet": "identification", "chapter": "ch3_literature_review", "subclaim": "alternatives"}
{"claim": "Without a certified optimality gap, a scheduling result reported only against the deployed heuristic's realized output is an assertion, not a benchmark: by LP/MILP duality the irreducible-versus-avoidable loss split is the gap between the optimal and realized objective, and the shadow price on a binding capacity constraint is the marginal value of that resource at the optimum, which distinguishes a physically oversubscribed window from one the incumbent discipline merely deprioritized. The candidate benchmarks concentration only against a ten-percent uniformity baseline and a Gini coefficient, so the structural-tail-versus-scheduler-suboptimality question is unresolved.", "evidence": [{"source": "Dantzig & Wolfe, Decomposition principle for linear programs, Operations Research 8(1) (1960) (dual pricing of subproblems coupled by shared linking resources)", "doi_or_url": "https://doi.org/10.1287/opre.8.1.101", "grade": "A"}], "facet": "identification", "chapter": "ch6_analysis_plan", "subclaim": "residual_risk"}
{"claim": "A convex lost-downlink-versus-load curve is exactly what a near-optimal allocator also produces as it approaches a binding capacity frontier (the optimal value function is convex in the binding right-hand-side capacity and the shadow price is its increasing slope), so the candidate's convex load-response is observationally confounded with ordinary increasing marginal cost at a capacity limit and is not by itself a fat-tail/fragility signature. Dropping the extreme load decile does not address this confound; fitting the curve separately to the deployed output and to a re-solved near-optimal allocation on the same demand would.", "evidence": [{"source": "Dantzig & Wolfe, Decomposition principle for linear programs, Operations Research 8(1) (1960)", "doi_or_url": "https://doi.org/10.1287/opre.8.1.101", "grade": "A"}], "facet": "empirics", "chapter": "ch6_analysis_plan", "subclaim": "alternatives"}
{"claim": "System state lives in stocks (accumulations such as backlog and unmet downlink) that change only through flows, and a stock's behavior over time cannot be read off a snapshot of its flows: the accumulation must be modeled as a level over time. The candidate's estimators (Cox hazard, tail fit, cross-sectional logistic/GLM) carry no accumulating state variable and the measurement codebook contains no time-integrated backlog level, so the design tests the loop's signature, not the stock-and-flow structure it invokes.", "evidence": [{"source": "Forrester, Industrial Dynamics / Dynamic Models of Economic Systems and Industrial Organizations (System Dynamics Review reprint)", "doi_or_url": "https://doi.org/10.1002/sdr.284", "grade": "A"}, {"source": "Sterman, Bathtub Dynamics: initial results of a systems thinking inventory, System Dynamics Review", "doi_or_url": "https://doi.org/10.1002/sdr.198", "grade": "A"}], "facet": "mechanism", "chapter": "ch2_theoretical_framework", "subclaim": "mechanism"}
{"claim": "The decisive system-dynamics test of the reinforcing-loop story is endogeneity: reproduce the observed concentration from the system's own internal feedback with no exogenous shock. The candidate's cross-sectional design holds mission and epoch fixed and records padding/early-submission only as a static hedging bias, never as a response regressed on a mission's own recent backlog, so it cannot close the feedback edge and an exogenous bursty arrival process would produce the same convex signature.", "evidence": [{"source": "Forrester, Counterintuitive Behavior of Social Systems (the endogeneity test: trouble must reproduce from internal feedback alone)", "doi_or_url": "https://doi.org/10.1007/bf00148991", "grade": "A"}, {"source": "Sterman, Bathtub Dynamics: initial results of a systems thinking inventory", "doi_or_url": "https://doi.org/10.1002/sdr.198", "grade": "A"}], "facet": "identification", "chapter": "ch2_theoretical_framework", "subclaim": "mechanism"}
{"claim": "The dominant request-to-allocation loop delay is asserted to exceed the window but is never measured; the concentration estimators (Gini, top-decile share over geometry-phase cells) carry no lag structure, so the design cannot separate delay-generated concentration from concentration that is the fixed celestial-mechanics geometry of overlapping viewing windows (which the candidate itself promotes to a behavior-independent instrument and which carries no feedback at all). A measured loop-delay-versus-window estimate and a geometry-only placebo would discriminate them.", "evidence": [{"source": "Forrester, Industrial Dynamics / Dynamic Models of Economic Systems and Industrial Organizations (System Dynamics Review reprint)", "doi_or_url": "https://doi.org/10.1002/sdr.284", "grade": "A"}, {"source": "Sterman, Bathtub Dynamics: initial results of a systems thinking inventory", "doi_or_url": "https://doi.org/10.1002/sdr.198", "grade": "A"}], "facet": "empirics", "chapter": "ch2_theoretical_framework", "subclaim": "alternatives"}
{"claim": "An exogenous-demand null and an endogenous deferral-hedge loop predict the same static top-decile concentration but opposite lag structure: in a within-mission fixed-effects panel regressing this cycle's padding/early-submission on prior-cycle realized pass-loss at the same geometry-phase window, the exogenous null gives a lag-1 coefficient near zero or negative once geometry is absorbed, while the endogenous loop gives a positive, geometrically decaying lag profile. The discriminant is the sign-and-decay of the lag, not the cross-sectional concentration both hypotheses reproduce, so the static GLM has potentially reversed the causal arrow.", "evidence": [{"source": "Forrester, Industrial Dynamics / Dynamic Models of Economic Systems and Industrial Organizations (System Dynamics Review)", "doi_or_url": "https://doi.org/10.1002/sdr.284", "grade": "A"}, {"source": "Forrester, Counterintuitive Behavior of Social Systems (1971)", "doi_or_url": "https://doi.org/10.1007/bf00148991", "grade": "A"}, {"source": "D'Ambrosio et al., Novel Source-Sink Model for Space Environment Evolution with Orbit Capacity Assessment (the temporal profile of an inflow, not only its total, sets stock-system stability), Journal of Spacecraft and Rockets (2023)", "doi_or_url": "https://doi.org/10.2514/1.A35579", "grade": "A"}], "facet": "identification", "chapter": "ch6_analysis_plan", "subclaim": "alternatives"}
{"claim": "A closed-loop deferral controller and an open-loop scarcity story predict the same static concentration but different post-intervention dynamics: after a window is relieved (added aperture, arraying, logged policy change), open-loop scarcity gives a rebound fraction near zero with requested load decaying toward the new capacity, while closed-loop set-point restoration gives a rebound fraction near one (loop gain near unity) with the carrying window reappearing or migrating to the next-binding slot. Forrester policy resistance plus the Meadows leverage hierarchy (parameters are weak, loop gain is strong) imply targeted aperture durably recovers science only if the measured rebound fraction is near zero.", "evidence": [{"source": "Forrester, Counterintuitive Behavior of Social Systems (policy resistance; relieved constraints rebound)", "doi_or_url": "https://doi.org/10.1007/bf00148991", "grade": "A"}, {"source": "Meadows, Leverage Points: Places to Intervene in a System (parameter interventions #12 are weak; balancing-loop strength #8 and reinforcing-loop gain #7 are strong)", "doi_or_url": "https://doi.org/10.4324/9781849773386-15", "grade": "A"}, {"source": "Forrester, Urban Dynamics (canonical capacity-intervention rebound / set-point-restoration case)", "doi_or_url": "https://doi.org/10.2307/214050", "grade": "A"}], "facet": "rival", "chapter": "ch7_discussion", "subclaim": "alternatives"}
{"claim": "A closed-loop deferral controller manufactures convexity for free: as backlog rises the hedge gain rises (more padding and escalation), so measured lost-downlink can accelerate against measured load even if physical service capacity is perfectly linear. Conditioning the load-response on the executed-versus-requested ratio (a proxy for instantaneous hedge gain) decomposes the convex curve into a physical service-saturation component and an endogenous hedge-amplification component; if the convexity vanishes when the ratio is held fixed, the fragility is in the loop, not the antennas, and the static cross-sectional GLM cannot perform this decomposition.", "evidence": [{"source": "Forrester, Industrial Dynamics / Dynamic Models of Economic Systems and Industrial Organizations (System Dynamics Review)", "doi_or_url": "https://doi.org/10.1002/sdr.284", "grade": "A"}, {"source": "Meadows, Leverage Points: Places to Intervene in a System (loop gain vs physical parameters)", "doi_or_url": "https://doi.org/10.4324/9781849773386-15", "grade": "A"}, {"source": "Forrester, Counterintuitive Behavior of Social Systems (1971)", "doi_or_url": "https://doi.org/10.1007/bf00148991", "grade": "A"}], "facet": "mechanism", "chapter": "ch6_analysis_plan", "subclaim": "alternatives"}
{"claim": "Power-law / heavy-tail fits to empirical data are over-prone to false positives, which is why the Clauset-Shalizi-Newman maximum-likelihood-plus-goodness-of-fit-plus-likelihood-ratio procedure is the disciplined standard; the fitted tail rests on the few extreme order statistics above the threshold and is sensitive to them. The candidate correctly imports this procedure and a generalized-Pareto shape parameter as the gating test, which is the right tool for the wait-time-tail clause of H1.", "evidence": [{"source": "Clauset, Shalizi & Newman, Power-Law Distributions in Empirical Data, SIAM Review (2009)", "doi_or_url": "https://doi.org/10.1137/070710111", "grade": "A"}], "facet": "identification", "chapter": "ch6_analysis_plan", "subclaim": "mechanism"}
{"claim": "Most empirically claimed power laws do not survive scrutiny once alternative heavy- and light-tailed forms are tested; an apparent scaling can be produced by a handful of influential points or a limited range rather than a structural mechanism, and incorrect likelihood-based inference has produced spurious heavy-tail conclusions in published work. The heavy-tail verdict therefore requires a leave-one-out / top-k-exceedance-deletion robustness check and a calibrated false-positive guard against a light-tailed-plus-drift null, neither of which the design currently runs.", "evidence": [{"source": "Stumpf & Porter, Critical Truths About Power Laws, Science (2012)", "doi_or_url": "https://doi.org/10.1126/science.1216142", "grade": "A"}, {"source": "Edwards et al., Incorrect Likelihood Methods Were Used to Infer Scaling Laws of Marine Predator Search Behaviour, PLoS ONE (2012)", "doi_or_url": "https://doi.org/10.1371/journal.pone.0045174", "grade": "A"}], "facet": "rival", "chapter": "ch6_analysis_plan", "subclaim": "residual_risk"}
{"claim": "Deadline-expired DSN requests are exactly the deferred subclass the contention mechanism predicts, so treating their censoring as non-informative biases the wait-time distribution; competing-risks cause-specific and cumulative-incidence methods are the established correction, and the naive Kaplan-Meier that treats a competing event as ordinary censoring is known to be biased. The candidate adopts the competing-risks treatment, which is the correct handling of expiry-as-abandonment.", "evidence": [{"source": "Andersen et al., Competing risks in epidemiology: possibilities and pitfalls, International Journal of Epidemiology (cited as design ref [64])", "doi_or_url": "https://doi.org/10.1093/ije/dyr213", "grade": "A"}], "facet": "measurement", "chapter": "ch5_research_design", "subclaim": "mechanism"}
{"claim": "Fragility is detectable as a convex (second-order) response of a system's loss to a dose of a stressor, a method meant to substitute for forecasting where the event cannot be extrapolated; using an in-sample convex load-response fitted on today's mission mix to price a crewed-exploration concurrency beyond all recorded load is the extrapolation the fragility framework was built to avoid. A within-support convex fit says nothing about behavior past an unreached saturation or instability point, so an out-of-sample backtest (fit on low-load cycles, predict held-out high-load cycles) is required before the second-order response can be claimed to forecast crewed-era exposure.", "evidence": [{"source": "Taleb & Douady, Mathematical definition, mapping, and detection of (anti)fragility, Quantitative Finance (2013)", "doi_or_url": "https://doi.org/10.1080/14697688.2013.829244", "grade": "B"}, {"source": "Flyvbjerg et al., regression to the tail (megaproject cost convexity as an in-sample, not out-of-regime, characterization), Environmental Science & Policy (cited as design ref [61])", "doi_or_url": "https://doi.org/10.1016/j.envsci.2021.06.013", "grade": "B"}], "facet": "empirics", "chapter": "ch7_discussion", "subclaim": "residual_risk"}
{"claim": "In Ostrom's Institutional Analysis and Development framework an allocation outcome is generated jointly by biophysical conditions, community attributes, AND the rules-in-use; modeling only geometry, band, and load with fixed effects holds the priority-and-deferral rule constant by omission, so attributing a heavy tail to physics while the allocation rule is unmodeled is an identification error. The rule-in-use is a first-class explanatory variable that should enter as a covariate.", "evidence": [{"source": "Ostrom et al., A general framework for analyzing sustainability of social-ecological systems, Science 325 (2009)", "doi_or_url": "https://doi.org/10.1126/science.1172133", "grade": "A"}, {"source": "Claps et al., Xi-MILP: Deep Space Network Scheduling via Mixed-Integer Linear Programming, IEEE Access (2022) (DSN is a demand-oversubscribed scheduled resource resolved by a constrained allocation, so an allocation rule exists to model)", "doi_or_url": "https://doi.org/10.1109/access.2022.3164213", "grade": "B"}], "facet": "identification", "chapter": "ch2_theoretical_framework", "subclaim": "alternatives"}
{"claim": "Cheap, mutual, accountable monitoring is the keystone that makes graduated sanctions and trust self-reinforcing in a common-pool resource; without it a CPR collapses into the open-access loss dynamic independent of load. Strategic padding of a requested quantity is a monitoring-and-enforcement failure in this vocabulary, so 'loss concentrates where bargaining discipline breaks down' is a distinct generative mechanism competing with the geometry-window hypothesis, with a request-versus-execution gap as its natural observable. A contended space resource is formally a CPR for which this framing applies.", "evidence": [{"source": "Ostrom, Governing the Commons: The Evolution of Institutions for Collective Action, Cambridge Univ. Press (1990)", "doi_or_url": "https://doi.org/10.1017/CBO9780511807763", "grade": "A"}, {"source": "Weeden & Chow, Taking a common-pool resources approach to space sustainability: A framework and potential policies, Space Policy 28(3) (2012)", "doi_or_url": "https://doi.org/10.1016/j.spacepol.2012.06.004", "grade": "A"}], "facet": "mechanism", "chapter": "ch2_theoretical_framework", "subclaim": "alternatives"}
{"claim": "Ostrom's panacea critique holds that no single institutional lever is a universal remedy and that prescription must follow a diagnostic decomposition of the specific situation; because the deferral discipline is a collective-choice variable, not a biophysical constant, a fraction of concentrated loss is institutional and curable by re-ruling rather than by capital. Concentrated loss must therefore be decomposed into a physical-oversubscription component and a deferral-rule-artifact component before aperture is recommended, or capital risks being spent to fix a collective-choice problem.", "evidence": [{"source": "Ostrom, A diagnostic approach for going beyond panaceas, PNAS 104(39):15181-15187 (2007)", "doi_or_url": "https://doi.org/10.1073/pnas.0702288104", "grade": "A"}, {"source": "Moving beyond panaceas in fisheries governance, PNAS (2018)", "doi_or_url": "https://doi.org/10.1073/pnas.1716545115", "grade": "A"}], "facet": "governance", "chapter": "ch7_discussion", "subclaim": "residual_risk"}
{"claim": "The DSN schedule is built through a documented multi-stage peer-to-peer negotiation among roughly 37 user projects via the Service Scheduling Software, which maintains versioned schedule workspaces and a revision history across distinct planning, mid-range negotiation, and near-real-time phases. The logs can therefore in principle be partitioned into the rules-on-paper stage (published priority discipline) versus the rules-in-use stage (realized inter-mission de-conflict trades), which would diagnose whether concentration reflects a fragile rule or a fragile institution; absorbing inflated requests as exogenous noise discards exactly the collective-choice action situation that should be the unit of analysis.", "evidence": [{"source": "Johnston, Tran, Arroyo, Page, Automating Mid- and Long-Range Scheduling for NASA's Deep Space Network, SpaceOps/AIAA (2012)", "doi_or_url": "https://doi.org/10.2514/6.2012-1296235", "grade": "B"}, {"source": "Johnston et al., Automated Scheduling for NASA's Deep Space Network, AI Magazine 35(4) (2014)", "doi_or_url": "https://doi.org/10.1609/aimag.v35i4.2552", "grade": "A"}], "facet": "governance", "chapter": "ch4_data_and_measurement", "subclaim": "alternatives"}
{"claim": "The DSN is not a single redesignable central allocator but a polycentric, bounded peer-to-peer negotiation among a defined user community (roughly 37 projects including international partners and ground-based science/calibration users), each able to edit shared schedule workspaces, and the problem is oversubscribed so realized loss falls on a subset of appropriators. The candidate's three remedy levers presuppose a single allocator the architecture contradicts; the binding diagnostic is whether loss concentrates on appropriators weakly represented at the negotiation table (a representation problem solvable only by changing who governs the window) versus a contention problem solvable by aperture.", "evidence": [{"source": "Goh et al., Deep Space Network Scheduling via Mixed-Integer Linear Programming, IEEE Access (2021) (oversubscribed: only a subset of requests can be scheduled, so loss falls on a subset of appropriators)", "doi_or_url": "https://doi.org/10.1109/access.2021.3064928", "grade": "A"}, {"source": "Johnston, Tran et al., Automating Mid- and Long-Range Scheduling for NASA's Deep Space Network, SpaceOps (2012) (distributed peer-to-peer negotiation among all users)", "doi_or_url": "https://doi.org/10.2514/6.2012-1296235", "grade": "B"}], "facet": "governance", "chapter": "ch7_discussion", "subclaim": "residual_risk"}
{"claim": "A causal estimate is licensed by the assignment mechanism, not the outcome model: the candidate must state the exclusion restriction that target viewing-window geometry enters pass-loss only through concurrent contention and is conditionally independent of the pass's own declared science value given covariates. Because the same encounter/occultation/conjunction geometry that compresses a window also raises the pass's science value, geometry plausibly co-determines both treatment and outcome (a confounder), and propensity/balancing adjustment removes bias only from observed covariates while mission fixed effects remove only the time-invariant level, so relabeling geometry an instrument does not rescue identification.", "evidence": [{"source": "Rubin, Bayesian Inference for Causal Effects: The Role of Randomization, Annals of Statistics (1978) (the assignment mechanism is the central licensing object)", "doi_or_url": "https://doi.org/10.1214/aos/1176344064", "grade": "A"}, {"source": "Rosenbaum & Rubin, The central role of the propensity score in observational studies for causal effects, Biometrika (1983) (adjustment removes bias only from observed confounders)", "doi_or_url": "https://doi.org/10.1093/biomet/70.1.41", "grade": "A"}], "facet": "identification", "chapter": "ch5_research_design", "subclaim": "residual_risk"}
{"claim": "Contention is constructed as the count and aggregate requested antenna-time of OTHER missions, so the treatment one mission receives is constituted by other missions' treatments: this is interference by construction, violating SUTVA's no-interference limb, which makes a unit-level estimand ill-posed rather than merely biased. Mission fixed effects cannot fix a contemporaneous spillover; the coherent target is an interference-aware exposure-mapping estimand, and strategic padding additionally threatens the no-hidden-versions limb because a 'requested hour' that is sometimes genuine need and sometimes hedge is not a single well-defined treatment.", "evidence": [{"source": "Rubin, Bayesian Inference for Causal Effects: The Role of Randomization, Annals of Statistics (1978) (SUTVA; restate the estimand to respect interference)", "doi_or_url": "https://doi.org/10.1214/aos/1176344064", "grade": "A"}, {"source": "Imbens & Rubin, Causal Inference for Statistics, Social, and Biomedical Sciences, Cambridge Univ. Press (2015) (SUTVA's two limbs; violations make the estimand incoherent)", "doi_or_url": "https://doi.org/10.1017/cbo9781139025751", "grade": "A"}], "facet": "mechanism", "chapter": "ch5_research_design", "subclaim": "residual_risk"}
{"claim": "The policy-relevant causal estimand is the potential-outcomes contrast Y_i(low-load) minus Y_i(high-load) on a fixed tracking-pass-request unit (not Y_i(served) versus Y_i(lost), since served/lost is the outcome), and a treatment effect is estimable for a unit only where comparable units exist under both load arms. Where the high-value critical-phase windows (encounter, EDL) never occur under low load there are no donor units and any effect there is extrapolation past support, so the design must report the no-common-support covariate cells (declination x band x phase x complex) before fitting and narrow the estimand to the sub-population where overlap holds.", "evidence": [{"source": "Rubin, Estimating causal effects of treatments in randomized and nonrandomized studies, Journal of Educational Psychology (1974) (the unit-level contrast Y_i(1) - Y_i(0))", "doi_or_url": "https://doi.org/10.1037/h0037350", "grade": "A"}, {"source": "Rubin, For objective causal inference, design trumps analysis, Annals of Applied Statistics (2008) (overlap/positivity; demonstrate common support before examining outcomes)", "doi_or_url": "https://doi.org/10.1214/08-aoas187", "grade": "A"}], "facet": "empirics", "chapter": "ch5_research_design", "subclaim": "residual_risk"}
{"claim": "Concentration on the carrying windows is observationally equivalent to a survivorship/selection artifact unless the propensity for a request to enter and be retained as a logged loss is shown flat across geometry windows: the logged-loss indicator is itself an assignment-into-the-sample whose mechanism must be reconstructed. The candidate's own design firewalls the concentration claim to the descriptive layer ('stands or falls on the completeness of the log, not on an identification argument'), and the out-of-stream withdrawal/entry trail needed to break the rival is not in the panel (the only independent load cross-check is the NTRS aggregate, and mission criticality records are the least-sourced dataset), so the survivorship rival is not ruled out.", "evidence": [{"source": "Rosenbaum & Rubin, The central role of the propensity score in observational studies for causal effects, Biometrika (1983) (the assignment-into-sample propensity is the central object to model)", "doi_or_url": "https://doi.org/10.1093/biomet/70.1.41", "grade": "A"}, {"source": "Imbens & Rubin, Causal Inference for Statistics, Social, and Biomedical Sciences, Cambridge Univ. Press (2015) (logged outcome under non-random assignment is a missing-data problem)", "doi_or_url": "https://doi.org/10.1017/cbo9781139025751", "grade": "A"}], "facet": "identification", "chapter": "ch5_research_design", "subclaim": "residual_risk"}
{"claim": "The geometry-criticality confound (per-request anticipated criticality drives both whether a mission fights-for-and-logs a pass and the tight-geometry window it falls in, since encounters and EDL are simultaneously highest-criticality and tightest-geometry) varies request-to-request within a mission-epoch and so cannot be absorbed by mission or epoch fixed effects; it requires an explicitly measured, outcome-independent request-level criticality covariate built from mission-declared critical-event windows (not requested duration, which is strategically inflated). The candidate defines exactly this covariate but concedes the mission/criticality records are its least-sourced dataset, so the confound is acknowledged but unresolved.", "evidence": [{"source": "Rosenbaum & Rubin, The central role of the propensity score in observational studies for causal effects, Biometrika (1983) (a within-unit confounder unabsorbed by fixed effects must enter as a measured covariate)", "doi_or_url": "https://doi.org/10.1093/biomet/70.1.41", "grade": "A"}, {"source": "Imbens & Rubin, Causal Inference for Statistics, Social, and Biomedical Sciences, Cambridge Univ. Press (2015) (design with covariates fixed while blind to outcomes)", "doi_or_url": "https://doi.org/10.1017/cbo9781139025751", "grade": "A"}], "facet": "rival", "chapter": "ch5_research_design", "subclaim": "residual_risk"}
{"claim": "Strategy is the dialectic of two reciprocally adapting wills, so a concentration result that holds for a single cycle proves planning against a passive opponent, not a structural invariant against an adapting one; and every measure that increases one's freedom of action must be checked for the recoil it grants the adversary, so an indirect scheduling-discipline lever whose recoil (requesters re-optimizing padding against the new rule) equals its intended effect is a redistribution, not a recovery. The candidate concedes first-order wait redistribution and offers post-intervention re-running but models no second-order re-gaming, leaving cross-cycle stability and the freedom-of-action ledger untested.", "evidence": [{"source": "Beaufre, An Introduction to Strategy (1965), via the dialectic-of-two-wills reading (Crossref record)", "doi_or_url": "https://doi.org/10.2307/2147679", "grade": "A"}, {"source": "Croson, Donohue, Katok, Sterman, Order Stability in Supply Chains, Production and Operations Management (2013) (self-interested hedging is the bullwhip coordination-risk mechanism that re-amplifies against a changed discipline; design ref [63])", "doi_or_url": "https://doi.org/10.1111/j.1937-5956.2012.01422.x", "grade": "A"}], "facet": "mechanism", "chapter": "ch7_discussion", "subclaim": "residual_risk"}
{"claim": "Establishing that bit-loss tracks lost science return rather than merely asserting the mapping requires a genuine qualitative strand connected to the quantitative concentration result at the sampling point (the 'connecting through sampling' integration operation) and reported in a joint display, with a pre-committed obligation to report whether the strands confirm, expand, or contradict. Re-weighting bits by mission-declared criticality is a within-strand quantitative transformation that performs no integration, so on the mixed-methods framework it cannot close the proxy-to-construct gap; a divergence (heavy-bit-loss windows scientifically cheap, low-bit-loss windows catastrophic) would invert the aperture-aiming conclusion and is invisible to a single-strand design.", "evidence": [{"source": "Fetters, Curry & Creswell, Achieving Integration in Mixed Methods Designs: Principles and Practices, Health Services Research (2013)", "doi_or_url": "https://doi.org/10.1111/1475-6773.12117", "grade": "A"}, {"source": "Guetterman, Fetters & Creswell, Integrating Quantitative and Qualitative Results in Health Science Mixed Methods Research Through Joint Displays, Annals of Family Medicine (2015)", "doi_or_url": "https://doi.org/10.1370/afm.1865", "grade": "A"}], "facet": "measurement", "chapter": "ch4_data_and_measurement", "subclaim": "residual_risk"}
{"claim": "The dissertation's socio-technical science-throughput-penalty contribution requires naming the mixed-methods design in the typology and choosing the integration point (explanatory-sequential: quantitative concentration first, then interview affected mission teams; or exploratory-sequential: interview first to surface dimensions of scientific cost, then build the criticality-weighting instrument), each fixed by interaction, priority, timing, and mixing point; the deletion test (if removing the qualitative strand changes nothing, it is decorative) is the auditing standard. Because the payoff is explicitly distributional (whose deferred passes are protected), the design also owes an explicit paradigm declaration (pragmatist or transformative) and a priority-class stratification of where loss concentrates.", "evidence": [{"source": "Ivankova, Creswell & Stick, Using Mixed-Methods Sequential Explanatory Design: From Theory to Practice, Field Methods (2006)", "doi_or_url": "https://doi.org/10.1177/1525822X05282260", "grade": "A"}, {"source": "Fetters et al., A Comprehensive Taxonomy of Research Designs... Achieving Design Naming Conventions in Mixed Methods Research, Journal of Mixed Methods Research (2023)", "doi_or_url": "https://doi.org/10.1177/15586898221131238", "grade": "A"}], "facet": "governance", "chapter": "ch5_research_design", "subclaim": "residual_risk"}
{"claim": "A pooled non-stationary DSN catalog can manufacture an apparent power law from a mixture of regimes, so record completeness and definitional stability must be established before any derived tail statistic: McDowell's catalog-anchored, trend-not-snapshot discipline obliges fitting the wait-time and loss tails within fixed antenna-complement epochs. The candidate's epoch fixed effects address level shifts in a mean regression but do not test stability of a Clauset-Shalizi-Newman / generalized-Pareto tail exponent or the top-decile share within a fixed-configuration regime, which a power-law-from-regime-mixture artifact requires.", "evidence": [{"source": "McDowell, The Low Earth Orbit Satellite Population and Impacts of the SpaceX Starlink Constellation, Astrophysical Journal Letters 892 L36 (2020) (exemplar of catalog-anchored trend accounting)", "doi_or_url": "https://doi.org/10.3847/2041-8213/ab8016", "grade": "A"}, {"source": "Clauset, Shalizi & Newman, Power-Law Distributions in Empirical Data, SIAM Review (2009) (tail-fit discipline a mean-regression epoch dummy does not satisfy)", "doi_or_url": "https://doi.org/10.1137/070710111", "grade": "A"}], "facet": "rival", "chapter": "ch6_analysis_plan", "subclaim": "residual_risk"}
{"claim": "An independent second record exists in principle to reconcile the candidate's request-derived counts: the DSN scheduling problem is a published, mathematically specified resource-allocation problem (mixed-integer LP against DSN requirements) with conflict resolution, and rescheduling operates on a baseline revised for equipment outages and weather, so executed-pass and station-downtime records are produced independently of the demand-request stream and can serve as the reconciliation source. A concentration claim resting on one arrangement-filtered stream that supplies both demand and outcome is, until reconciled window-by-window, unverifiable on McDowell's two-source discipline.", "evidence": [{"source": "Claudet, Alimo, Goh, Johnston, Delta-MILP: Deep Space Network Scheduling via Mixed-Integer Linear Programming, IEEE Access (2022)", "doi_or_url": "https://doi.org/10.1109/access.2022.3164213", "grade": "A"}, {"source": "Johnston, Tran, Arroyo, Page, Automating Mid- and Long-Range Scheduling for NASA's Deep Space Network (the DSN schedule of record and its negotiated workspaces), NASA NTRS 20130009147", "doi_or_url": "https://ntrs.nasa.gov/citations/20130009147", "grade": "C"}], "facet": "identification", "chapter": "ch4_data_and_measurement", "subclaim": "residual_risk"}