{"q": "With 30-60 missions partitioned into competing-risk strata AND calendar-period (era) fixed effects, can the cohort exhibit a single era containing BOTH a meaningful count of first-of-kind active-sensor missions AND a meaningful count of passive-radiometer heritage missions, so the archetype contrast is identified WITHIN era rather than borrowed ACROSS eras? Report the realized archetype-by-era cross-tabulation before any modeling; if every diagonal cell is near-empty the era fixed effects are collinear with archetype and the dominance result is the launch-market longue duree wearing an archetype mask.", "facet": "identification", "raised_by": "braudel", "priority": "high", "query_terms": ["NASA Earth-observing mission archetype by launch era cross-tabulation", "first-of-kind active sensor vs passive radiometer heritage cohort composition", "within-era identification competing risks small sample"], "status": "open"}
{"q": "Launch-side covariates (vehicle class, provider, shared-vs-dedicated manifest, provider-in-development) are realizations of a launch world-economy whose centre shifted from an Old Space core to a NewSpace core across the sample, so 'shared manifest on a provider in its own development' means something categorically different in the EELV era than in the reusable-commercial era. Can you show from the data that these covariates carry comparable hazard meaning across the transition (test for proportional-subdistribution-hazard violation and covariate-by-era interaction on the launch event) rather than assuming one time-invariant launch coefficient? If the launch hazard is non-proportional across the provider transition, the launch CIF averages two regimes.", "facet": "measurement", "raised_by": "braudel", "priority": "high", "query_terms": ["proportional subdistribution hazard violation test launch covariate", "covariate-by-era interaction Fine-Gray non-proportionality", "Schoenfeld residual test competing risks time-varying"], "status": "open"}
{"q": "Sensor archetype and launch era may be the same gradient in the deep structure: heritage passive-radiometer continuity missions are the late, routinized material-civilization layer of an established lineage and first-of-kind active sensors are the frontier-capitalism layer, and the two layers were funded and flown in systematically different launch epochs. Can the cohort distinguish 'instrument physics dominates for active sensors' from 'active-sensor missions simply clustered in the epoch where launch slip was structurally suppressed'? Concretely: holding launch epoch fixed, does within-program heritage progression (an active first-of-kind followed by its own later passive-continuity rebuild on comparable launch infrastructure) reproduce the dominance reversal? If no such within-lineage pairs exist in the cohort, on what evidence is the archetype gradient separated from the longue-duree maturation gradient it rides on?", "facet": "rival", "raised_by": "braudel", "priority": "high", "query_terms": ["within-program heritage progression Landsat continuity launch infrastructure", "active first-of-kind followed by passive continuity rebuild matched pairs", "novelty versus program maturation confound spacecraft schedule"], "status": "open"}
{"q": "Is the launch-availability subdistribution hazard a fixed-effect era nuisance, or a conjoncture (a monotonic decade-long structural trend from EELV near-monopoly to competitive rideshare) that period dummies misspecify, so the correct launch covariate is a contemporaneous market-tightness index at each mission's KDP-B rather than a static shared-vs-dedicated flag?", "facet": "identification", "raised_by": "braudel", "priority": "high", "query_terms": ["launch market tightness index active providers manifest backlog months", "EELV monopoly to commercial rideshare structural trend", "monotonic launch-availability hazard trend NASA missions"], "status": "open"}
{"q": "Are launch-driven first slips independent per-mission competing-risks draws (transparent market economy), or do they co-move when one dominant launch program slips (concentrated capitalism / opaque top layer), so the per-mission subdistribution hazard understates launch risk for the heritage archetype precisely when the launch economy is concentrated? Can the candidate measure cross-mission clustering of launch-driven first slips by provider-and-year?", "facet": "mechanism", "raised_by": "braudel", "priority": "high", "query_terms": ["cross-mission clustering launch slip provider and year", "shared launch provider development slip propagation NASA missions", "launch provider concentration cascade mission schedule"], "status": "open"}
{"q": "Can the launch hazard be anchored to a measured physical scarcity scalar (manifest-congestion: slots demanded vs slots available in the assignment window) that predicts launch-driven slip onset, rather than to an administrative shared-vs-dedicated label assigned post hoc, per the geography-first test?", "facet": "measurement", "raised_by": "braudel", "priority": "high", "query_terms": ["manifest congestion scalar slots demanded versus available launch assignment window", "launch slot scarcity predicts schedule slip onset", "integration site capacity launch manifest constraint"], "status": "open"}
{"q": "At what per-stratum events-per-variable count does the archetype-by-TRL interaction in the ridge-penalized partial likelihood remain distinguishable from its shrunk-to-zero null: how many instrument-driven first-slip events fall in the first-of-kind active stratum versus the passive-radiometer heritage stratum, and does the cross-validated ridge penalty regularize away the heterogeneous interaction you most care about?", "facet": "empirics", "raised_by": "callaway_santanna", "priority": "high", "query_terms": ["events per variable subdistribution hazard competing risks small sample", "ridge penalty shrinkage interaction term cross-validation survival", "minimum events per variable Fine-Gray archetype stratum"], "status": "open"}
{"q": "State which aggregation your reserve-allocation decision requires (cause-specific subdistribution hazard ratio on TRL deficit, marginal CIF plateau at a fixed horizon, or a cohort-weighted average across adoption eras), then show from the CADRe-plus-GAO cohort that the reported archetype dominance is robust to that weighting choice, since a subdistribution-hazard dominance and a CIF-plateau dominance can disagree in finite samples.", "facet": "identification", "raised_by": "callaway_santanna", "priority": "high", "query_terms": ["aggregation robustness subdistribution hazard versus CIF plateau finite sample", "cumulative incidence function plateau dominance reserve allocation", "Callaway Sant'Anna transparent weighting aggregation competing risks"], "status": "open"}
{"q": "From the actual mission-by-year cohort, show the cross-tabulation of archetype against launch era and demonstrate the instrument-versus-launch dominance reversal survives estimation within era, not merely with an additive calendar fixed effect, or concede the dominance attributed to sensor archetype is partly an era effect the additive fixed effects cannot purge.", "facet": "rival", "raised_by": "callaway_santanna", "priority": "high", "query_terms": ["archetype by launch era cross-tabulation within-era re-estimation", "additive calendar fixed effects era confounding archetype", "era versus sensor archetype separability NASA cohort"], "status": "open"}
{"q": "When the within-stratum CIFs and the penalized interaction coefficient are collapsed into the single 'dominance' verdict that drives reserve allocation, what is the DECLARED weighting (CIF plateau difference at a fixed horizon? area between curves? penalized SHR point estimate?), at WHICH calendar horizon, and show by worked example that this declared aggregation, not the ridge penalty's cross-validated implicit shrinkage, determines whether instrument-slip is called dominant.", "facet": "empirics", "raised_by": "callaway_santanna", "priority": "high", "query_terms": ["CIF plateau evaluation horizon competing risks reserve decision", "ridge shrinkage versus declared aggregation worked example subdistribution", "fixed calendar horizon cumulative incidence dominance verdict"], "status": "open"}
{"q": "Your competing events are coded from narrative attribution reconciled across two sources; the bound on the whole result is the measurement-error rate in that coding, yet you treat it only as a recode-both-ways check. Will you estimate cause-coding reliability directly (the inter-source agreement rate between CADRe Part A and GAO on the dominant cause across codable spells) and propagate that disagreement rate as the explicit error bound on the cumulative incidence functions, so the reader sees how wide the CIF bands become at the observed coding-disagreement level rather than at assumed-perfect coding? What agreement rate do the two corpora yield, and at what disagreement level does archetype dominance cease to be distinguishable?", "facet": "measurement", "raised_by": "fogel", "priority": "high", "query_terms": ["inter-source agreement CADRe Part A versus GAO dominant slip cause", "cause-coding reliability propagated CIF confidence band", "narrative attribution measurement error bound cumulative incidence"], "status": "open"}
{"q": "Build the explicit counterfactual and put a number on the social saving: from the CADRe baseline-movement record, compute for the as-flown cohort the committed reserve a pooled-first-slip-CIF board would set versus the archetype-conditional-CIF reserve, and report the delta in reserve dollars and schedule-months per mission. If the delta is a low single-digit percent of mission reserve (railroad social-saving order of magnitude), separability is statistically real but an economic rounding error.", "facet": "economics", "raised_by": "fogel", "priority": "high", "query_terms": ["pooled versus archetype-conditional reserve delta dollars schedule months", "social saving reserve allocation NASA mission counterfactual", "committed reserve percentile cost growth probability NASA"], "status": "open"}
{"q": "Name the behavioral channel and show it binds: regress the historical committed instrument-vs-launch reserve split on the archetype variable in the CADRe confirmation baselines and show boards do NOT already price archetype by intuition. The model's marginal value is the gap between intuitive reserve and CIF-optimal reserve; if boards already price archetype the channel carries zero marginal social saving.", "facet": "economics", "raised_by": "fogel", "priority": "high", "query_terms": ["board reserve split instrument versus launch archetype intuition", "reserve setting confirmation baseline archetype pricing", "marginal information gap intuitive versus model reserve"], "status": "open"}
{"q": "Apply substitution-elasticity discipline: from the CADRe baseline-movement and reserve-draw-down records estimate the cost of in-flight reallocation of reserve between the instrument and launch pools. If that substitution cost is low, the elasticity is high and the value of getting the ex-ante archetype split right collapses toward zero, leaving the decision-relevance claim resting on an unmeasured near-zero substitution elasticity.", "facet": "economics", "raised_by": "fogel", "priority": "high", "query_terms": ["in-flight reserve reallocation cost instrument launch pool substitution", "reserve fungibility draw-down NASA mission execution", "substitution elasticity reserve allocation value"], "status": "open"}
{"q": "Specify schedule-and-cost reserve as an explicit shared stock with its draw-down flows in the CADRe baseline-movement record, and show the cohort cross-cause hazard correlation is consistent with two structurally separable processes rather than one shared reserve-depletion loop generating both the instrument and launch events. If the instrument-then-launch slip sequence cannot be reproduced without the shared loop, the subdistribution hazards are measuring loop dynamics, not separable owners.", "facet": "mechanism", "raised_by": "forrester", "priority": "high", "query_terms": ["shared reserve stock draw-down flows CADRe baseline movement", "cross-cause hazard correlation shared reserve depletion loop", "instrument-then-launch slip sequence feedback system dynamics"], "status": "open"}
{"q": "Descope/gating is a balancing loop, not a confounder to control out: a mission anticipating instrument slip trims TRL ambition, suppressing the instrument-slip flow being measured and freeing the loop to bite the launch side; measuring entry TRL before the slip window does not break the loop because the gating decision that set that TRL is the loop's actuator. Mark where the competing-risks estimator acts on the causal diagram, and predict whether reserve-steering toward instrument maturation merely re-routes slip onto the launch side, leaving aggregate slip unchanged. Can CADRe/GAO measure whether total first-slip incidence is invariant to the instrument-versus-launch mix?", "facet": "rival", "raised_by": "forrester", "priority": "high", "query_terms": ["total first-slip incidence invariance instrument versus launch mix", "policy resistance reroute slip balancing loop descope", "aggregate slip reserve steering reroute NASA"], "status": "open"}
{"q": "Competing-risks gives each mission's FIRST slip, but reserve, manifest congestion, and launch-market state are system-level stocks integrating across the whole cohort and across calendar time; calendar-period fixed effects treat the era as an exogenous shock, yet launch-side hazard is plausibly driven by an endogenous manifest stock in which every mission's slip re-manifests slots that change the launch hazard for every contemporaneous mission, a fallacy of composition in the aggregation. Can the CADRe/GAO record show that launch-driven first-slip events cluster in time in a way a per-mission independent hazard cannot generate, i.e., that one mission's launch hazard is a function of the contemporaneous slip stock of the others rather than of that mission's own covariates?", "facet": "identification", "raised_by": "forrester", "priority": "high", "query_terms": ["launch slip temporal clustering manifest stock cross-mission", "fallacy of composition aggregate launch hazard endogenous stock", "contemporaneous slip stock launch hazard composition"], "status": "open"}
{"q": "The two competing launch/instrument risks are modeled as per-mission independent hazards, but launch-driven slip propagates through a shared, unnamed stock: the launch-services manifest backlog. Mission A's instrument slip releases a slot, that slot accumulates as manifest congestion, and the congestion becomes mission B's launch-driven slip months later. Specify this as a stock-and-flow loop at the mission scale (backlog stock, slot-release inflow from other missions' instrument slips, assignment outflow, release-to-relaunch lag) and test from CADRe/GAO whether first-slip events are manifest-clustered (one mission's instrument slip Granger-precedes another's launch slip on the same vehicle family), which would break the Fine-Gray independence-of-the-at-risk-set assumption.", "facet": "mechanism", "raised_by": "forrester", "priority": "high", "query_terms": ["shared reserve stock draw-down flows CADRe baseline movement", "cross-cause hazard correlation shared reserve depletion loop", "instrument-then-launch slip sequence feedback system dynamics"], "status": "open"}
{"q": "Rival-4 concedes anticipatory descope suppresses the instrument-slip being measured and handles it as a covariate, but descope is a rate-triggered nonlinear balancing loop with delay, not a linear covariate shift: a program office watches reserve draw down, and when the draw-down RATE crosses a threshold it descopes, halting the instrument-slip flow before it registers. The first-of-kind active-sensor missions claimed to show dominant instrument-slip are exactly the missions with the most reserve and descope authority, so the loop bites hardest there and biases the dominant hazard downward in the dominant stratum. Can you show the instrument-slip hazard is independent of where each mission sits on its reserve draw-down curve, or is the linear archetype coefficient an artifact of which missions could afford to self-arrest?", "facet": "identification", "raised_by": "forrester", "priority": "high", "query_terms": ["total first-slip incidence invariance instrument versus launch mix", "policy resistance reroute slip balancing loop descope", "aggregate slip reserve steering reroute NASA"], "status": "open"}
{"q": "Callaway-Sant'Anna aggregation discipline is invoked to refuse a pooled coefficient, but disaggregation stops at archetype, not at the actor whose feedback generates the slip. The mechanism required for the archetype effect to persist is a within-mission reinforcing loop: a TRL deficit causes a test failure, the failure consumes schedule reserve, reserve loss raises pressure that causes a corner-cut that causes the next test failure. If real, the instrument-slip hazard is path-dependent on the ORDER of test failures, not a proportional function of entry-TRL, and proportional subdistribution hazards are misspecified. From CADRe milestone-by-milestone and TechPort TRL-progression records, does entry-TRL at KDP-B carry the instrument-slip signal, or is the slip driven by the rate of TRL stall AFTER KDP-B (the reinforcing-loop signature), which a KDP-B-frozen covariate cannot see?", "facet": "empirics", "raised_by": "forrester", "priority": "high", "query_terms": ["post-KDP-B TRL stall rate reinforcing loop instrument slip", "entry TRL snapshot versus TRL progression TechPort milestone", "reinforcing loop test failure reserve corner-cut path dependence"], "status": "open"}
{"q": "State the instrument: report the inter-coder reliability (e.g. Cohen's kappa) on the instrument-versus-launch cause assignment AND the inter-source agreement rate between CADRe Part A and GAO on the same mission-year, computed before any modeling. Without a kappa, the competing risks are a coding artifact.", "facet": "measurement", "raised_by": "glaser_strauss", "priority": "high", "query_terms": ["Cohen kappa inter-coder reliability instrument versus launch cause", "CADRe GAO inter-source agreement rate dominant cause", "content analysis coding reliability before modeling"], "status": "open"}
{"q": "What fraction of first slips in the 30-60 mission cohort are genuinely single-dominant-cause versus co-occurring, and does the binary survive specifically the missions where CADRe and GAO disagree, or are the disconfirming incidents being discarded?", "facet": "identification", "raised_by": "glaser_strauss", "priority": "high", "query_terms": ["single-dominant-cause versus co-occurring first slip fraction", "binary survives disputed-attribution CADRe GAO disagreement subset", "constant comparison disconfirming incidents category survival"], "status": "open"}
{"q": "Run the inverse test: open-code the proximate-cause language in the Part A and GAO narratives with no a priori bin. Do exactly two dominant categories emerge and saturate, or do the data generate a different set (funding/phasing, descope, partner-delivery, workforce) that the two-risk model forces into 'instrument' or 'launch' and thereby mis-specifies the competing-risks structure?", "facet": "rival", "raised_by": "glaser_strauss", "priority": "high", "query_terms": ["open coding proximate cause narrative saturation categories", "two competing risks versus richer cause set funding descope workforce", "grounded theory inverse test category emergence slip cause"], "status": "open"}
{"q": "Pull the full set of distinct CADRe Part A / GAO proximate-cause phrasings assigned to 'instrument-driven,' lay them in a constant-comparison table, and show how many DIMENSIONS the bin contains (detector-immaturity, calibration-budget non-closure, environmental-test failure, parts obsolescence, workmanship). Show the bin is dimensionally homogeneous or concede it is a forced container collapsing five physically distinct sub-causes with different reserve levers.", "facet": "measurement", "raised_by": "glaser_strauss", "priority": "high", "query_terms": ["Cohen kappa inter-coder reliability instrument versus launch cause", "CADRe GAO inter-source agreement rate dominant cause", "content analysis coding reliability before modeling"], "status": "open"}
{"q": "Your cohort is every NASA Earth mission reaching KDP-B from ~1990 (30-60 missions), fixed in advance for census-completeness, which is sampling for representativeness, the opposite of theoretical sampling. Name the point at which adding the next mission stopped yielding a NEW proximate-cause property. If you cannot identify that saturation point, concede category boundaries are warranted by cohort completeness rather than saturation, and say how a property first appearing in mission 55 would be detected rather than absorbed into a frozen two-bin scheme.", "facet": "identification", "raised_by": "glaser_strauss", "priority": "high", "query_terms": ["single-dominant-cause versus co-occurring first slip fraction", "binary survives disputed-attribution CADRe GAO disagreement subset", "constant comparison disconfirming incidents category survival"], "status": "open"}
{"q": "Strip the imported scaffolding. Your bins, competing-risks frame, and archetype contrast all arrive from the methodological literature (Fine-Gray, Fogel decomposition, Callaway-Sant'Anna) BEFORE any incident is read. Re-derive cause categories with that scaffolding removed: open-code the raw CADRe/GAO slip narratives with no a priori instrument-vs-launch dichotomy and no archetype expectation, and report the data's OWN dominant axis of variation. If the leading natural cleavage is 'estimate-realism vs execution' or 'internal vs external-to-program' rather than instrument-vs-launch, the two competing events were forced onto the data to fit the apparatus.", "facet": "rival", "raised_by": "glaser_strauss", "priority": "high", "query_terms": ["open coding proximate cause narrative saturation categories", "two competing risks versus richer cause set funding descope workforce", "grounded theory inverse test category emergence slip cause"], "status": "open"}
{"q": "The two-month net-launch-date-movement threshold is a round-number convention, the schedule analogue of the 100 km Karman line, and should be justified by where the objects sit, not by tradition. Plot the empirical distribution of net launch-date movements and state how many coded slip events fall within one month either side of the two-month line. If events cluster at the boundary, which missions count as 'slipped' is set by the cutoff and dominance could reorder under the pre-registered 1-to-4-month sweep. Show cause-specific dominance is stable through the dense band, or concede the result is boundary-drawn.", "facet": "identification", "raised_by": "mcdowell", "priority": "high", "query_terms": ["two-month net launch-date movement threshold boundary pile-up", "slip event distribution one-to-four-month sweep dominance stability", "round-number schedule slip cutoff sensitivity competing risks"], "status": "open"}
{"q": "Give the four-way join across CADRe, NICM, GAO, and TechPort before any hazard model. For the 30-to-60 mission KDP-B population, state the count carrying a non-imputed CADRe record AND a matchable GAO project-year AND a TechPort sensor-TRL entry AND a NICM instrument-taxonomy record (the true intersection), versus the count that survives only by imputing one or more sources. What is the effective sample size, the intersection rather than the union?", "facet": "measurement", "raised_by": "mcdowell", "priority": "high", "query_terms": ["four-way join CADRe NICM GAO TechPort intersection effective sample", "non-imputed match count versus imputation-padded union", "catalog reconciliation effective N competing risks"], "status": "open"}
{"q": "KDP-B is the spell origin and launch readiness date is the censoring boundary, but both are moving, outcome-contaminated catalog entries. Name the single catalog of record that fixes the KDP-B month per mission, and state whether the censoring launch date is the original committed date or the as-flown date. Show that neither origin nor censoring boundary is a function of the slip being measured.", "facet": "identification", "raised_by": "mcdowell", "priority": "high", "query_terms": ["KDP-B catalog of record spell origin censoring committed versus as-flown", "moving outcome-contaminated origin censoring boundary survival", "spell origin independent of slip launch readiness date"], "status": "open"}
{"q": "Publish the definitional-stability audit for the archetype variable that anchors the H1 dominance hinge. Across the cohort, what fraction of missions get an identical archetype label whether read from the NICM instrument-type field, the CADRe instrument description, or the GAO project narrative? Where the three disagree, which is authoritative by a pre-stated rule?", "facet": "measurement", "raised_by": "mcdowell", "priority": "high", "query_terms": ["archetype label agreement NICM CADRe GAO authoritative rule", "definitional stability archetype effect modifier three-source", "archetype source-selection artifact agreement rate"], "status": "open"}
{"q": "Draw the DAG (sensor archetype, entry TRL, descope/confirmation-delay decision, launch-manifest assignment) and name the minimal adjustment set that closes every back-door path to the instrument-slip subdistribution-hazard event. Because the descope decision is plausibly a common effect of anticipated instrument risk and of archetype, and the cohort conditions on missions that reached KDP-B and were manifested, is 'carry descope history as a covariate' not collider/selection conditioning that opens a non-causal archetype<->slip path? Is the effect identifiable from CADRe/TechPort variables or does an unobserved anticipated-risk node leave it non-identifiable?", "facet": "empirics", "raised_by": "pearl", "priority": "high", "query_terms": ["DAG conditional independence implications competing risks falsification", "d-separation testable implications small sample empty cell", "graph-implied independence test versus Gray's test estimator"], "status": "open"}
{"q": "Enumerate the full mediator sequence from archetype (first-of-kind active sensor) to first instrument-slip (entry-TRL deficit -> environmental-test failure or calibration non-closure -> detector rework -> descope-or-delay decision -> committed-baseline movement), and for each mediator state whether the CADRe Part A / GAO / TechPort record OBSERVES it as a coded variable or merely ASSUMES it; identify the one mediator whose absence from the measured set means the chain is no longer fully captured by observed intermediates.", "facet": "identification", "raised_by": "pearl", "priority": "high", "query_terms": ["front-door criterion fully mediating measured variable instrument slip", "standing-army review-board scrutiny bypass edge unmeasured mediator", "mediation decomposition identification competing risks reduced form"], "status": "open"}
{"q": "The micro-observable that would CONFIRM (vs merely co-witness) the mediating channel -- a logged within-mission TRL non-advancement or a named test/calibration failure event time-ordered BEFORE the committed-baseline movement, PRESENT in active-sensor missions that slipped instrument-first and ABSENT (clean TRL maturation, passed tests) in active-sensor missions that did not -- is NOT resolvable in this record at the mission level. Entry TRL is coded once at KDP-B (Appendix A: 'measured before the slip window opens'), so no within-spell TRL transition series exists to test non-advancement; test/calibration failure and detector rework have no dictionary entry at all; and cause attribution rests on reconciling the CADRe Part A and GAO NARRATIVES, which Ch 5.4.1 and Ch 8 admit is 'irreducibly subjective' narrative attribution, a proxy for the physical cause 'not the physics.' Because the slipped vs non-slipped active-sensor missions are therefore indistinguishable on any coded intermediate event log (only the endpoint cause-code separates them), the mediating pathway is UNCONFIRMED and the archetype->slip association could be transmitted by an unmeasured driver -- precisely the standing-army/scrutiny bypass of g1. The candidate flags the cognate residual itself: Ch 7.3.4 / R3 names undocumented anticipatory descope as 'the rival most likely to bias toward H0,' and the whole work is a design-stage dissertation whose numbers are 'explicitly illustrative and not executed on the full cohort,' so the contrast cannot be settled on this record as written.", "facet": "empirics", "raised_by": "pearl", "priority": "high", "query_terms": ["within-mission TRL non-advancement test failure event time-ordered", "mediating channel confirm slipped versus non-slipped active sensor", "logged calibration failure detector rework before baseline movement"], "status": "open"}
{"q": "Jackknife the 30-60 mission cohort (leave-one-out, leave-two-out) and additionally recode/delete the two missions with the largest Schoenfeld-style influence on the instrument-slip subdistribution coefficient; report whether the SIGN of the archetype dominance contrast survives every perturbation or flips. If one mis-coded marquee mission or one first-of-kind blow-up can reverse dominance, the point estimate is not the planning quantity and the cohort is an Extremistan sample wearing Mediocristan CIs.", "facet": "empirics", "raised_by": "taleb", "priority": "high", "query_terms": ["leave-one-out leave-two-out influence subdistribution coefficient sign", "Schoenfeld influence marquee mission recode dominance flip", "jackknife robustness small cohort competing risks sign survival"], "status": "open"}
{"q": "The reserve-allocation prescription is asymmetric in consequence but evaluated symmetrically. Re-cast the decision as bounding tail loss from a wrong reserve call (not maximizing dominant-hazard classification accuracy) and show, from the empirical CIFs, the maximum schedule overrun each archetype incurs WHEN the dominance call is wrong. If you cannot show steering reserve away from the undifferentiated pool reduces worst-case overrun rather than expected overrun, you have optimized the ensemble average of a non-ergodic, ruin-bearing process and left the tail uncovered.", "facet": "governance", "raised_by": "taleb", "priority": "high", "query_terms": ["bounded tail loss worst-case overrun wrong reserve call archetype", "ruin non-ergodic reserve allocation tail versus expected overrun", "asymmetric consequence reserve decision tail coverage"], "status": "open"}
{"q": "The prescription is a via-positiva move: it ADDS a differentiation rule that concentrates reserve on the predicted-dominant hazard and thins it on the other. Using the off-diagonal mass of the archetype-by-cause CIFs, estimate the rate of dual-cause/cross-cause first slips and what fraction of overrun magnitude lands in the off-diagonal cell a steered posture leaves unprotected; compare against a via-negativa single undifferentiated pool sized to the worst archetype. If off-diagonal mass is non-trivial and two-source narrative coding flags rather than codes dual-cause events (Sec 4.5.2), steering reserve adds a brittle classification dependency without removing the underlying uncertainty.", "facet": "rival", "raised_by": "taleb", "priority": "high", "query_terms": ["off-diagonal cross-cause CIF mass dual-cause first slip rate", "via-negativa undifferentiated reserve pool worst archetype", "transfer of fragility steered reserve residual tail exposure"], "status": "open"}
{"q": "Plot the empirical distribution of total cost overrun for the 30-60 Earth missions from the actual CADRe/GAO record and report its tail: the share of total cohort overrun dollars from the worst 3 missions and whether mean overrun exceeds median by more than 2x. If the dollar distribution is fat-tailed, the 'slips-first-for-instrument-reasons' archetype can be the cheap archetype while a rare launch-driven cascade owns the tail, inverting the reserve prescription.", "facet": "empirics", "raised_by": "taleb", "priority": "high", "query_terms": ["empirical dollar overrun distribution tail worst-3 missions share", "mean to median overrun ratio fat-tailed Earth mission cohort", "fat-tail cost overrun distribution NASA Earth-observing cohort"], "status": "open"}
{"q": "A reserve-steering rule that moves the shared stock toward each archetype's dominant first-slip hazard necessarily removes coverage from the non-dominant cause. From the off-diagonal of the archetype-by-cause CIFs, what fraction of an active-sensor archetype's realized overruns historically originated on the launch side it is now told to under-reserve, and what is the largest single launch-driven overrun an instrument-steered active-sensor mission absorbed? If steering toward the modal cause leaves the rare-but-large off-diagonal uncovered, you lower the expected miss while raising the tail miss: a transfer of fragility.", "facet": "rival", "raised_by": "taleb", "priority": "high", "query_terms": ["active-sensor overrun originating launch-side off-diagonal share", "largest single launch-driven overrun instrument-steered mission", "transfer of fragility expected miss versus tail miss reserve"], "status": "open"}
{"q": "The estimator's spell ends at first slip or at launch, so program cancellation / termination-level descope, the absorbing state that actually ends the game, is invisible: a cancelled mission never records a 'first slip of either cause' and drops out as if censored. From the cohort census, how many candidate Earth missions reaching KDP-B in the era were cancelled or terminated before launch, by which dominant cause, and are they in the at-risk set or silently excluded? If the worst instrument-driven outcomes self-select out by cancellation, the instrument-slip hazard is conditioned on survival and understates the very tail a reserve policy exists to guard.", "facet": "identification", "raised_by": "taleb", "priority": "high", "query_terms": ["cancelled terminated mission census before launch dominant cause", "survivorship bias instrument hazard conditioned on survival NICM", "absorbing termination at-risk set competing risks NASA"], "status": "open"}
{"q": "Needs external citation: the era/archetype partial-confounding concession (additive calendar fixed effects cannot fully purge an era-correlated archetype composition on a 30-60 mission cohort) is currently grounded only in the candidate's own dissertation text. Find a peer-reviewed external source on additive fixed effects failing to identify a within-stratum contrast under composition-correlated treatment, to externally warrant the residual-risk caveat.", "facet": "rival", "raised_by": "callaway_santanna", "priority": "high", "query_terms": ["additive fixed effects within-group identification composition confounding", "calendar period fixed effects cannot purge group-by-era confound", "two-way fixed effects heterogeneous composition bias"], "status": "partial"}
{"q": "Needs external citation: the reproducibility-asymmetry argument (public GAO arm reproducible without a data-use agreement, restricted CADRe/ONCE/NICM arm not) rests only on the candidate's Chapter 4 and an internal McDowell dossier pointer. Find an external methods source on reproducibility/auditability standards for analyses built on access-restricted administrative data to externally warrant the governance caveat.", "facet": "governance", "raised_by": "mcdowell", "priority": "high", "query_terms": ["reproducibility restricted administrative data data-use agreement", "auditability falsifiability access-controlled government records research", "open data versus restricted cost data reproducibility standard"], "status": "partial"}
{"q": "Needs external citation: the claim that the reserve prescription is incidence-weighted rather than dollar-tail-weighted and therefore mis-ranks causes under fat-tailed loss is grounded only in the candidate's text plus an internal Taleb dossier pointer. Find an external peer-reviewed source on fat-tailed cost-overrun loss distributions in aerospace/megaprojects to externally warrant the tail-mis-ranking caveat.", "facet": "empirics", "raised_by": "taleb", "priority": "high", "query_terms": ["fat-tailed cost overrun distribution aerospace megaproject", "tail risk reserve allocation non-normal overrun loss", "extreme value cost growth NASA defense project"], "status": "partial"}
{"q": "MODERATOR MISSING ANGLE: No panelist tested external decision-utility / adoption, whether a NASA or JPL reserve board would actually change its KDP-B reserve posture on an archetype-conditional CIF rather than its current undifferentiated pool, and what decision threshold (CIF-plateau gap, dollar delta, or shadow-priced time delta) would move the call. The panel spanned identification, measurement, mechanism, rival, economics-as-social-saving, and tail risk, but never the implementation/governance question of whether the prescription is actionable inside existing NASA reserve policy and acquisition gates. Phase 1 must source the reserve-policy and KDP-B governance literature to frame this.", "facet": "governance", "raised_by": "moderator", "priority": "high", "query_terms": ["NASA schedule reserve policy Key Decision Point B confirmation", "joint cost confidence level reserve unallocated future expense NASA", "program reserve allocation decision threshold archetype mission", "NPR 7120.5 reserve posture acquisition gate adoption"], "status": "open"}