Scenario wins: Panshul42 (71) lewinke-thinking-bot* (5)
Question: how many distinct FederalRegister.gov documents listing the NRC will appear from 2026-06-08 through 2026-08-12 (observed 2026-08-15)? Forecasters base their view on publication-day capacity (~46 Federal Register publication days in the 66-calendar-day window), historical same-window counts (an opaque 2000–2025 range of 66–167 and a closest 2025 analogue around 82–84), and stable June–August seasonality (2024–25 full summer totals ~110). Recent indicators push the expectation upward: Justia shows 215 NRC documents year-to-date through 2026-06-05 and FederalRegister.gov listed two public-inspection items for 2026-06-08, implying an active, lumpy stream of routine notices and some rulemaking follow-ups that could fall inside the window. All teams agree the most likely outcome lies above the platform’s 75.5 top-coded bin (one team put ~85% mass in the open upper bin), though they differ on the central point estimate (mid/high-80s vs. 85–100 or low-90s). Remaining uncertainty comes from publication lumpiness, possible indexing timing, and the platform’s open upper bin — forecasters therefore retain modest lower-tail probability for an unusually quiet or disrupted summer.
Forecast: Probability distribution:
Key quantity is the count of Federal Register documents tagging the Nuclear Regulatory Commission (NRC) as an agency with publication dates 2026-06-08 through 2026-08-12 inclusive (a 66-day window).
Evidence/base rates from the prompt:
Main downside risks that could pull the window down toward/below 75: summer seasonality, docket batching/administrative bottlenecks, and governance/process disruptions (e.g., commission/quorum or other operational slowdowns). But to land at 75 or below would require a sustained rate ≤ about 1.14/day across the whole window, which is meaningfully below the 2023–2025 analog pace and well below 2026 YTD.
Synthesis: I place substantial mass in the open-above bin (count > 75) while retaining a non-trivial ~10–20% chance of ≤75 outcomes driven by a sharp summer slowdown (with the 2022 analog of 66 anchoring the low tail). Percentiles above 75 are expressed as values strictly past 75 to correctly represent open-upper probability mass under the question mechanics (not as a claim that the true count is capped at 94).
An amateur forecast often (a) anchors on the 0–75 visible range and incorrectly compresses the distribution near 75, and/or (b) ignores that most recent same-window analogs and 2026 YTD rates imply the outcome is very likely above 75. My forecast explicitly targets the main scoring-relevant uncertainty—P(count > 75)—using the historical same-window base rate plus 2026’s higher run-rate, while still keeping a realistic downside tail for summer/operational slowdowns. Confidence in improvement: moderate; the dominant uncertainty is how persistent the 2026 elevated pace is through mid-summer.
Forecast rationale (numeric):
— Iteration 1 — Across the forecasts, the dominant approach is to anchor on 2026 year-to-date NRC publication pace and then adjust for recent rolling counts, historical summer-window patterns, and policy-driven volatility.
The collective view is that the NRC will likely publish more than the recent historical summer-window norm, but probably not far above the current 2026 pace-based projection. The most plausible outcome is a high-80s to mid-90s count, with meaningful uncertainty extending into the low 80s on the downside and roughly around 110 on the upside.
— Iteration 2 — The forecasts mostly reason from three inputs: historical same-window counts, 2026 year-to-date strength, and the most recent issuance pace. The historical analogs for this June-to-August period tend to cluster in the low-to-mid 80s, but all of the rationales note that 2026 has been running above prior years, with recent 35- to 65-day windows implying a materially higher run rate than the historical average.
The central disagreement is over how much to weight that elevated 2026 pace versus seasonal and operational drag. More bullish estimates push the count into the high 90s to low 100s, citing strong recent throughput and possible acceleration from licensing activity or executive-order-driven nuclear actions. More conservative reasoning pulls the estimate down toward the 70s, either because of explicit bounding constraints or because summer seasonality, administrative slowdown, or other disruptions could dampen output.
Overall, the shared view is that the Federal Register count should likely finish above the historical same-window norm, but with a wide uncertainty band. Downside risks include shutdown or quorum-related delays; upside risks come from sustained elevated filing volume and policy-driven acceleration.
— Iteration 3 — Across the rationales, the core approach is to anchor the forecast in recent publication pace and then adjust for seasonality and institutional factors.
Recent pace extrapolation is the starting point.
The forecasts lean heavily on year-to-date publication rates and recent rolling-window counts, which imply something near the mid-to-high 90s for the 66-day window. Recent windows were cited as especially informative because they capture current NRC/Federal Register throughput better than older history.
Historical analogs pull the estimate downward somewhat.
Older same-window periods, especially from 2023–2025, tend to cluster in the low-to-mid 80s, suggesting that a simple extrapolation may overstate volume if current activity normalizes.
Seasonality and administrative friction are expected to reduce output.
Several rationales explicitly discount the raw pace because the period falls in summer, when processing can slow. Some also mention possible operational drag from litigation, quorum issues, or broader administrative slowdown.
Policy and project activity support an elevated baseline.
Even with caution, the forecasts treat current NRC-related activity as above average, citing active drivers such as nuclear executive actions, SMR licensing, and major reactor restart efforts. These factors justify staying above older historical norms.
The shared logic is: use current pace as the baseline, discount it for summer/operational slowdowns, then widen uncertainty because NRC publication volume can be lumpy and sensitive to policy or administrative shocks.
Summary of Forecasting Rationales
Calendar and Historical Baselines The forecasting window spans exactly 66 calendar days, which translates to 46 federal business publishing days after accounting for weekends and federal holidays (Juneteenth and Independence Day). Historically, the NRC’s publication output during this exact summer window has been remarkably consistent, clustering tightly in the 80s over the past three years.
Elevated 2026 Pacing and Executive Mandates Forecasters universally note that 2026 is operating at a significantly elevated regulatory baseline. This surge is largely driven by Executive Order 14300, which mandated a wholesale rewrite of NRC regulations for advanced reactors and Small Modular Reactors (SMRs). This led to a massive “spring blitz” of major proposed rules in early 2026. However, the most recent data leading up to the target window indicates that this blistering spring pace has mathematically cooled, reverting toward a more normalized—but still historically elevated—baseline of routine notices and license amendments.
Catalysts for Continued Volume Several specific actions will guarantee a steady flow of Federal Register traffic throughout the summer:
Constraints and Downward Pressures Despite the high volume of regulatory action, several structural and administrative counter-currents are expected to cap explosive growth in the distinct document count:
Conclusion The rationales synthesize these factors by projecting that the recent “cooled but elevated” post-spring baseline will hold steady. The final outcome is expected to land somewhat above historical summer clusters due to E.O. deadlines and site-specific restarts, but it will be constrained by procedural bundling, seasonal meeting gaps, and bureaucratic bandwidth limits.
Question: How many Nuclear Regulatory Commission documents will the Federal Register publish between 6/8/2026 and 8/12/2026? Final Prediction: Probability distribution:
Total Cost: extra_metadata_in_explanation is disabled Time Spent: extra_metadata_in_explanation is disabled LLMs: extra_metadata_in_explanation is disabled Bot Name: extra_metadata_in_explanation is disabled
Forecaster 1: Probability distribution:
Forecaster 2: Probability distribution:
Forecaster 3: Probability distribution:
The research compiles historical Federal Register counts for NRC documents during the June 8–August 12 window (66 days): 66 (2022), 88 (2023), 88 (2024), and 82 (2025), giving a four‑year mean of 81 and a recent three‑year cluster of 82–88 (mostly in the 76+ upper‑tail bin). For 2026 through June 4 the NRC was running at an elevated year‑to‑date pace of 228 documents across 155 days (1.47/day), with recent shorter windows at 1.63/day (Apr 1–Jun 4, 65 days) and 1.40/day (May 1–Jun 4, 35 days). Applying different pace assumptions to the 66‑day forecast window yields point projections of roughly 81 (historical 4‑yr mean), 92–97 (recent paces), and 108 (recent 65‑day pace), with a recent 3‑yr cluster mean of 85.
The research also catalogs institutional drivers likely to push volumes up (May 2025 executive orders accelerating nuclear development, multiple active rulemakings and licensing actions in 2026 including byproduct rule proposals, fusion framework, microreactor rules, and high‑profile license/restart proceedings) and moderating forces (summer seasonality, NRC administrative bandwidth, post‑Loper Bright regulatory uncertainty, and early‑2026 congressional funding/staffing disruptions). Key analytical observations are that 2026 shows materially elevated pace versus historical same‑window rates, three of four historical analogs fall in the 76+ bin, momentum is mixed across recent windows, and the modal analytical central tendency concentrates in the upper‑tail (roughly 85–100 documents) though uncertainty remains.
Sources used (as cited in the research):
You’re forecasting the number of Nuclear Regulatory Commission (NRC) Federal Register documents published between June 8 - August 12, 2026 (a 66-day window). The forecast bins range from 0-75 (discrete integers) with an open upper-tail bin for 76+.
Historical Same-Window Performance (June 8 - August 12):
2026 Year-to-Date Trends (as of June 4, 2026):
The 2026 YTD pace (1.47/day) is 19% above the historical same-window rate (1.23/day average), and recent momentum shows even higher activity.
Upward Pressures on NRC Publication Volume:
Executive Policy Acceleration [2]: President Trump signed four executive orders on May 23, 2025, directing acceleration of nuclear deployment, NRC procedural reform, and fast-tracking of small modular reactor (SMR) licensing. The NRC’s mission statement shifted to “enabling benefits” and introduced 18-month generic milestone schedules.
Major Active Rulemakings [1][3]:
Downward/Moderating Pressures:
Summer Seasonality: The background information notes that summer months historically show lower regulatory hearing scheduling than spring/fall periods.
Administrative Constraints: NRC publication throughput is structurally constrained by Office of the Secretary calendar bandwidth and is not infinitely elastic regardless of executive-order pressure.
Regulatory Uncertainty [9]: The 2024 Supreme Court decision in Loper Bright Enterprises v. Raimondo stripped NRC of Chevron deference authority, creating litigation uncertainty that may complicate NRC actions.
Government Funding Challenges [16][19][22][24]: Congress experienced two shutdowns in early 2026, including the longest partial DHS shutdown in history (55+ days through April). However, no evidence suggests imminent shutdown affecting NRC specifically in the June-August window. The question’s fine print notes that shutdown doesn’t change the resolution count.
General Federal Register activity reached an all-time high of 106,109 pages in 2024 (19% increase over 2023), with a trend toward fewer but longer and more complex rules [11]. This macro environment supports elevated agency publication activity.
No specific prediction markets found for this NRC Federal Register document count question. I searched Kalshi, Polymarket, Metaculus, and Manifold Markets [12][13][14][15], but found no active markets on this specific outcome. The prediction market ecosystem focuses primarily on political, economic, and geopolitical events rather than granular regulatory metrics.
Based on different pace assumptions for the 66-day forecast window:
Three of four historical analogs (82, 88, 88) fall in the 76+ upper-tail bin, with only the 2022 deep-low of 66 falling within discrete bins.
2026 exhibits materially elevated pace: The YTD rate of 1.47 docs/day represents sustained acceleration above all recent historical periods.
Momentum indicators are mixed: The 65-day window shows very high activity (1.63/day), but the most recent 35-day window shows some moderation (1.40/day), possibly indicating a peak in spring with summer deceleration ahead.
Institutional pressures are net positive: Executive orders, major project pipelines, and the NRC’s “enabling benefits” mission shift create throughput requirements that outweigh seasonal and administrative constraints.
Modal outcome likely 85-100 range: Given elevated 2026 pace and strong project pipeline, probability mass should concentrate in the upper-tail bin (76+), with central tendency around 85-100 documents.
I agree with the critic’s main adjustment: my first forecast leaned a bit too hard on 2026 YTD/recent-window extrapolation. The recent NRC Federal Register flow is visibly lumpy, with multi-document batch days such as the May 1 and May 18 clusters, so a 35- or 65-day pace can overstate the sustainable summer rate if it captured backlog clearing rather than a new steady-state intensity. (nrc.gov)
That said, I still think the critic is only a modest correction, not a reversal. Conditioning on non-annulment, the strongest evidence remains: the same June 8–August 12 window was 82, 88, 88 in the last three years, and 2026 is running hotter than the recent historical baseline. Even if policy acceleration does not translate linearly into more distinct document numbers, NRC Federal Register notices are often legally required and are tied to routine licensing, meetings, information collections, and docket actions, so there is a substantial irreducible publication flow. (nrc.gov)
The query currently returns no documents for the future window, so this remains a fully forward-looking forecast rather than one with observed in-window data. (federalregister.gov) I also give some weight to the critic’s point that the 2022 value of 66 is not ignorable: a low-burst summer plus batching just outside the window can plausibly put the result in the 60s or low 70s without any dramatic institutional disruption.
My revised view: 76+ remains clearly favored, but I reduce my center from about 95 to about 92, and reduce probability above 75 from the mid/high-80s to roughly 83%. I keep a fairly long upper tail because a few large batch days can still push the count well above 110.
Updated prediction:
Percentile 0.1: 34.1
Percentile 1: 48.1
Percentile 5: 61.1
Percentile 10: 69.1
Percentile 20: 78.1
Percentile 30: 84.1
Percentile 40: 88.1
Percentile 50: 92.1
Percentile 60: 96.1
Percentile 70: 101.1
Percentile 80: 107.1
Percentile 90: 116.1
Percentile 95: 125.1
Percentile 99: 144.1
Percentile 99.9: 166.1
I accept the critic’s main correction: my earlier distribution was too tight around clearing 75. The strongest point is that the April–early-June burst may be partly front-loaded. The Federal Register agency page’s significant 2026 NRC items are indeed concentrated in March–April, and the EO-driven rulemaking timeline plausibly pulled some activity forward rather than establishing a fully persistent summer rate. The critic is also right that the relevant cadence is about 46 Federal Register publication days in this window, since the Federal Register is published every federal working day, not daily. (federalregister.gov)
I still think the core bullish call is correct. The target search currently has no counted documents only because the window has not begun, while the NRC page already shows two NRC documents scheduled for June 8, 2026, the first day of the window. (federalregister.gov) More importantly, the same-window base rate is not “creator thought likely under 75”; it is 66, 88, 88, 82 from the prompt, with three of four prior years and all three recent years above 75. Combined with the prompt’s 2026 YTD pace and recent-window pace, that still makes 76+ the dominant side.
Where I update downward is in the mapping from “nuclear-policy acceleration” to “distinct Federal Register document count.” EO 14300 pushes toward faster NRC licensing and regulatory revision, but it also explicitly includes reorganization, reductions in force in some functions, streamlined hearings, standardized approvals, and general/high-volume licensing pathways. Those features can increase nuclear deployment work while not necessarily increasing the number of separate Federal Register documents in this exact summer window. (federalregister.gov)
My revised view: central tendency modestly down from 94 to about 92, with a fatter lower tail. I now put roughly 16–18% probability at 75 or below, rather than about 10%. I still condition this forecast on the question not being annulled.
Updated prediction:
Percentile 0.1: 34.4
Percentile 1: 48.4
Percentile 5: 60.4
Percentile 10: 69.4
Percentile 20: 79.4
Percentile 30: 84.0
Percentile 40: 88.0
Percentile 50: 92.0
Percentile 60: 96.4
Percentile 70: 101.4
Percentile 80: 107.8
Percentile 90: 116.8
Percentile 95: 125.8
Percentile 99: 145.8
Percentile 99.9: 168.8
Conditioning on non-annulment, I think the Round 2 critic makes a good correction: my prior revision probably gave too much weight to front-loading / summer-lull risk. I still think that risk is real, but the same-window historical base rate already includes summer seasonality. The strongest fact remains that the last three comparable windows were 82, 88, and 88, all above the 75 cap.
I especially accept the critic’s publication-day framing. The Federal Register is published every Federal working day, not every calendar day, so the 6/8/2026–8/12/2026 window has roughly 46 publication days after weekends and the Juneteenth / Independence Day observed holidays. That means 76 documents is only about 1.65 NRC documents per publication day, while a 90-ish outcome is about 2 per publication day—quite consistent with recent history and the current first-day public-inspection signal. NARA confirms the Federal Register is published every Federal working day. (archives.gov)
I also verified the first-day signal: the NRC agency page shows two pending NRC documents scheduled for publication on June 8, 2026, including one on the Crane Clean Energy Center and one on mandatory-hearing policy. The same page shows several NRC documents recently published on June 4–5, which supports an active, ordinary-to-high pipeline rather than an immediate drop-off. (federalregister.gov)
Where I still disagree with the critic is on becoming too confident above 75. EO 14300 can generate documents, but it can also disrupt workflow: it calls for NRC reorganization and reductions in force while also ordering a dedicated rulemaking team, proposed-rule notices within 9 months, and final rules/guidance within 18 months. That makes a spring surge followed by some summer digestion still plausible. (federalregister.gov) So I do not want to push P(>75) into the high 80s.
My updated forecast is a modest move back upward: median around 90–91, P(>75) about 80%, with meaningful but not dominant mass on a 2022-like low outcome. The modal scoring bin remains clearly 76+, but I retain a roughly one-in-five chance of landing at or below 75 because administrative batching and a post-spring pause could plausibly put the count in the high 60s or low 70s.
Updated prediction:
Percentile 0.1: 42.5
Percentile 1: 51.5
Percentile 5: 61.5
Percentile 10: 67.8
Percentile 20: 76.1
Percentile 30: 82.4
Percentile 40: 86.9
Percentile 50: 90.7
Percentile 60: 95.1
Percentile 70: 100.5
Percentile 80: 107.4
Percentile 90: 118.4
Percentile 95: 129.4
Percentile 99: 153.5
Percentile 99.9: 181.5
Modal: Above 75 (68.0%) • partial_shift • 6m35s
Reading: strict
The resolution criteria are highly specific and literal: a named search URL, a precise observation timestamp (12:00 UTC 2026-08-15), a strict publication-date window, an explicit source hierarchy, and a snapshot-at-observation-time finality rule — leaving almost no room for interpretive discretion. The resolver counts exactly what the query returns at the specified moment, with no charitable broadening for late-indexed or corrected documents.
Total evidence registered (shared pool): ?
| Variant | Perspective | Model | Turns | Tools | Status |
|---|---|---|---|---|---|
| 0 | inside_view (inside_view_v1) | openai/gpt-5-mini | 15 | 14 | OK |
| 1 | outside_view (outside_view_v1) | anthropic/claude-sonnet-4-6 | 21 | 32 | OK |
| 2 | contrarian (contrarian_v1) | openai/gpt-5 | 7 | 7 | OK |
Evidence confidence: high
high evidence]2026 YTD publication pace is 58% above 2025 for the same period; the 67-day April–June 2026 window produced 110 documents; EO 14300 rulemaking schedule lists at least 8 specific rules targeted for publication in the Jun–Aug window; arrival-process means of 96–110 place the distribution center well above 75; all three most recent prior-year analogs (82, 88, 88) also fall above 75.
Summer seasonality historically runs below the full-year rate; staffing losses of 500+ may constrain throughput; Office of the Secretary bandwidth is finite; a government shutdown could reduce document issuance; overdispersion is high, so the distribution has a fat left tail.
medium evidence]Summer seasonality dampens pace; the 2025 analog was 82 but 2025 YTD pace was materially lower; staffing constraints could slow EO 14300 processing; the answer-range grid upper bin at 75 captures a plausible scenario if 2026 pace reverts partway toward historical norms. The 35-day May 2026 run rate (49 docs, ~1.4/day) projected over 67 days gives ~94 — but with high overdispersion, outcomes in 66–75 remain plausible.
No prior-year analog in the 66–75 range except 2022’s 66; the 2026 YTD acceleration is the strongest in the four-year set; EO 14300 pipeline is front-loaded in Jun–Aug; rate projections center well above 75.
low evidence]A government shutdown during the window, or a significant operational disruption from the NRC reorganization effective June 2026 (two new offices created), could sharply curtail publication throughput. Commissioner vacancy and quorum issues could also block Commission-level documents.
No prior-year analog below 66 in the four-year data; the resolution rules specify counts are drawn from documents actually published even during a shutdown; 2026 YTD pace is at a multi-year high; EO 14300 imposes statutory-equivalent deadlines.
medium evidence]If the April–June 2026 rate of ~1.64 docs/day holds linearly over the 67-day window, the expected count is ~110. The dense EO 14300 schedule, the NRC reorganization notices, and the Palisades/TMI restart docketing all add incremental volume on top of baseline license amendment notices.
Summer seasonality historically draws the rate below the spring pace; 2025 summer was ~1.18/day vs. 2025 YTD ~1.29/day; overdispersion is high (p5 = 16–19), reflecting genuine regime uncertainty. No prior-year analog has reached 100 in this window.
inside_view_v1)Inside-view causal chain analysis:
Current state → mechanism → outcome:
The causal chain is: 2026 YTD NRC publication rate of ~1.48–1.64 docs/day (per Federal Register JSON API, authoritative source) × 67-day window = projected 96–110 documents. This is driven by:
- EO 14300 rulemaking deadlines forcing at least 8 proposed/final rules into the Jun–Aug window (NRC Chairman Ho K. Nieh confirmed 21 rules through July 2026)
- Structural acceleration: 2026 YTD is 58% above 2025 YTD (232 vs. 147 docs through June 6, per authoritative API)
- Active licensing proceedings: Palisades restart, TMI Unit 1, AI data-center co-location arrangements
Scenario weighting by evidence strength:
>75 (bin_8): HIGH evidence strength — All three most recent prior-year analogs (82, 88, 88) fall in this bin. The 2026 run rate projects a mean of 96–110 for the 67-day window (arrival forecast, per computations from authoritative API data). EO 14300 pipeline adds incremental volume above baseline. The “open upper-tail outcome” scenario carries
relative_evidence_strength: highand deserves dominant mass (~55%).66–75 (bin_7): MEDIUM evidence strength — The “in-bin moderate outcome” scenario is tagged
medium. Summer seasonality dampens pace (~1.18/day historical summer rate). Staffing losses (500+ per Commissioner Crowell’s testimony). High overdispersion means fat left tail. The 35-day May rate (1.4/day) is slightly below the 67-day spring rate. This scenario gets moderate mass (~15%).56–66 (bin_6): Lower probability — Possible via summer deceleration + staffing constraints together, but overdispersion distributes some probability here. ~8%
bins 0-5 (below 56): LOW evidence strength disruption scenario — Only plausible under government shutdown or major operational disruption. No prior-year analog below 66. Combined ~12% across bins 0-5.
Within bins 0-6 distribution: The low-evidence “deep deceleration” scenario gets modest tail mass; bins 1-5 each get small but non-trivial probability given overdispersion (p5=16–19 from arrival forecast).
Key calibration note: The >75 (bin_8) scenario has
highevidence strength per the brief’s tagging, while the “extreme upper tail >100” scenario ismedium. Given the overdispersion model suggests broad variance, but the central tendency strongly points above 75, I weight bin_8 most heavily while maintaining meaningful tail mass below 75 for scenarios involving summer seasonality reversion, staffing constraints, and overdispersion downside.Distribution:
75 (bin_8): ~0.55 — high evidence strength; all recent analogs; 2026 acceleration
- 66-75 (bin_7): ~0.15 — medium evidence; summer seasonality, partial reversion
- 56-66 (bin_6): ~0.08 — includes the 2022 analog lower bound territory
- 47-56 (bin_5): ~0.06
- 38-47 (bin_4): ~0.05
- 28-38 (bin_3): ~0.04
- 19-28 (bin_2): ~0.03
- 9-19 (bin_1): ~0.02
- 0-9 (bin_0): ~0.02 Sum = 1.00
outside_view_v1)Reference class: Federal Register NRC-document counts in the identical 6/8–8/12 calendar window for 2022–2025 (values: 66, 88, 88, 82) — 3 of 4 prior years exceeded 75 and the four-year mean ≈ 81 (per the brief / Federal Register API, sources 4–5). Base rate: P(>75) ≈ 0.75 (3/4 of prior-year analogs). Adjustment: I shift only modestly from that base rate to reflect strong 2026-specific evidence that pulls mass toward the open upper tail (the 2026 run-rate: 110 docs in Apr–Jun 2026 → ~1.64/day and YTD 2026 = 232 documents, +58% YoY; per sources 8–9) and the dense EO 14300 rulemaking pipeline scheduled across June–July (per sources 22–24). Countervailing factors that reduce certainty: summer seasonality (summer 2025 ~1.18/day), staffing losses and reorganizational frictions (Crowell testimony and NRC reorg; sources 12, 25–26), and genuine overdispersion in arrival timing (source 10). Because the brief explicitly tags the open-upper-tail scenario as the high-evidence scenario, I give it materially more mass than the other scenarios while still allocating non-trivial probability to the in-bin historical upper tail and a small left tail for disruption/shutdown risk (per the brief’s scenario strength tags). Final (outside-view calibrated): P(bin_8, Above 75) = 0.72; P(bin_7, 66–75) = 0.20; P(bin_6, 56–66) = 0.06; small residual (bin_5) = 0.02 for low-probability operational-disruption outcomes. Rationale for shape and numbers: (a) historical analogs put central tendency >75 (3/4 years >75; mean ≈81) so the outside-view anchor is high (per sources 4–5); (b) 2026 empirical run-rates and arrival-process projections place the distribution center even further above 75 (per sources 8, 10–11), justifying the heavy weight on bin_8; (c) summer-seasonality and staffing/reorg/shutdown risks justify retaining ~0.28 total mass below the open-tail (sources 12, 25–26 and the resolution rules’ shutdown caveat); (d) confidence in the evidence is high per the brief, so the distribution is relatively concentrated rather than extremely wide. Reading: strict — the forecast assumes the resolver will use the named FederalRegister search/API snapshot at the observation time and counts documents exactly as indexed there (per the resolution criteria and interpretation).
contrarian_v2)Adversarial reading
The bulk of the evidence in the brief pulls toward the open upper-tail bin (>75 documents). The high-evidence scenario (relative_evidence_strength: high) is the open upper-tail outcome driven by the 58% YoY acceleration in 2026 publication pace, the 110-document 67-day spring window at ~1.64 docs/day, the EO 14300 pipeline of at least 8 scheduled rules, and the three most recent historical analogs (82, 88, 88). All rate-based arrival projections center well above 75.
Counter-evidence found (or absence)
Counter-evidence exists but is weaker: summer 2025 rate of ~1.18 docs/day (source 12), 500+ net staffing losses (source 26), reorganization risk effective June 2026 (source 25), and high overdispersion producing a fat left tail (p5=16–19 in arrival forecasts). These support the medium- and low-strength scenarios (in-bin 66–75 or ≤65). The 2022 analog of 66 is the sole prior-year value inside the discrete bins.
Distribution implications
Because the high-evidence upper-tail scenario materially outweighs the medium/low counter-scenarios, bin_8 receives the largest single share. However, the adversarial reading of seasonality, staffing constraints, and overdispersion justifies shifting more mass into 66–75 (bin_7) and 56–66 (bin_6) than a naive consensus would assign, while keeping negligible mass in the lowest bins. The resulting distribution reflects the evidence-strength weighting required by the brief.
Strategy: logit_mean • Drafts: 3/3
anchoring_v1The draft rationale states the “inside view” places ~55% on bin_8 and the “outside view” places 72% on bin_8, with the logit-mean blending yielding ~61%. The anchoring concern is that the forecaster treated the CURRENT 2026 YTD run-rate (~1.64 docs/day from the April–June spring window) as the p50 projection for the June–August summer window, rather than the modal expectation. The brief itself explicitly notes that “summer-month seasonality historically shows modestly lower daily rates; 2025 summer (June 1–August 31) averaged ~1.18 docs/day,” and the 35-day May 2026 rate was ~1.40/day — both below the spring peak. A flat-prior reading of the seasonal adjustment implies the modal summer rate is closer to 1.2–1.4 docs/day, projecting 80–94 documents — still above 75 but right on the boundary. The forecaster qualitatively notes seasonality as a “counter-current” but anchors quantitatively on the optimistic spring peak rate (1.64/day) rather than the seasonally-adjusted median rate (~1.3/day). This artificially pushes the p50 into the high-110 range rather than the more conservative 80–90 range, correctly placing most mass in bin_8 but underweighting the 66–75 bin and overconcentrating on the extreme upper tail.
ceiling_v1The structural ceiling concern here is not a ceiling that constrains outcomes downward — it is the bin-grid boundary at 75 that creates an artificial floor on the probability assigned to the open upper-tail bin. The brief explicitly states: “forecasters who project the outcome above 75 allocate the corresponding probability mass to the open upper-tail bin (this includes all of the three most recent historical analogs 82, 88, and 88).” The draft rationale correctly identifies this but the structural math confirms it is even more binding than acknowledged: ALL FOUR prior-year analogs fall at or above 66, THREE fall strictly above 75 (82, 88, 88), and the rate-based arrival projections produce medians of 80–92 and means of 96–110 — all landing in bin_8. The structural constraint implied is: given that the historical base rate P(>75 | same calendar window) = 3/4 = 0.75, and given that the 2026 YTD pace is 58% above the 2025 YTD from which the 2025 analog of 82 was drawn, the conditioned probability P(>75 | 2026 pace) is materially higher than 0.75. The draft assigns bin_8 = 0.61, which is actually BELOW the unconditioned historical base rate of 0.75, let alone the conditioned 2026 rate. This is a ceiling-floor mismatch: the draft is applying mass to bins 0–7 (specifically bins 0–6 get ~0.19 combined) in excess of what the structural facts support — there is no prior-year analog in bins 0–6 except the 2022 value of 66 (in bin_6), and the current 2026 acceleration makes even that scenario low-probability.
math_v1The forecast’s probability distribution sums to 1.0 and is mathematically coherent, but exhibits internal inconsistency between the stated rationale and the assigned probabilities. The rationale emphasizes the “high evidence strength” of the open upper-tail scenario (bin_8, >75), citing the four-year mean of ~81, all three recent prior-year analogs (82, 88, 88) falling above 75, and arrival-process projections of mean 96–110. Yet bin_8 receives only 0.612 (61.2%), while bins 6–7 (56–75) combined receive 0.279 (27.9%) — nearly half as much mass as the high-evidence scenario. This is mathematically inconsistent with the rationale’s statement that bin_8 “carries dominant mass (~55%)” and should be weighted “most heavily,” and it contradicts the outside-view calibration statement (“P(bin_8) = 0.72”). The actual assignment of 0.279 to the medium/low-evidence scenarios (66–75 and 56–66) overstates the tail risk from summer seasonality and staffing constraints relative to the explicit quantitative evidence: if the 1.48–1.64 docs/day rate is correct, Poisson/negative-binomial projections place p25 at 46–52, meaning ≥25% of probability should lie below 52 — yet the forecast assigns only ~0.07 to bins below 56. The distribution appears to have underweighted bins 0–5 (disruption/shutdown scenarios) and overweighted the in-range bins 6–7, compared to both the stated high/medium/low evidence strengths and the quantitative arrival-process projections provided in the brief.
Reviews consumed: 3
All three reviewers converged on shifting mass into bin_8 and trimming the low-bin tail. The ceiling reviewer (high confidence) flagged that bin_8 = 0.61 sits below the unconditioned historical base rate of 0.75 and that bins 0–5 are overweighted given no prior-year analog exists below 66. The math reviewer (medium) independently flagged the same inconsistency: the rationale claims bin_8 deserves dominant mass (~72% outside view) yet the draft assigned only 61%, and the lower bins 0–5 carry implausible mass given no disruption analog. The anchoring reviewer flagged a mild over-anchoring on the spring peak rate (1.64/day) vs. summer-seasonality-adjusted ~1.3/day, suggesting slightly more weight on bin_7 — I partially accepted this by keeping bin_7 meaningful (~0.18) rather than collapsing it. Net: I shifted bin_8 upward to ~0.68 (between draft 0.61 and ceiling’s 0.72), trimmed bins 0–5 substantially per ceiling/math, and kept bin_7 at ~0.18 per anchoring concern. This is a partial_shift.
| Bin | Probability |
|---|---|
| 0-9 (bin_0) | 0.6% |
| 9-19 (bin_1) | 0.7% |
| 19-28 (bin_2) | 0.9% |
| 28-38 (bin_3) | 1.1% |
| 38-47 (bin_4) | 1.3% |
| 47-56 (bin_5) | 2.2% |
| 56-66 (bin_6) | 7.2% |
| 66-75 (bin_7) | 18.0% |
| Above 75 (bin_8) | 68.0% |
Due to aggressive automated scraping of FederalRegister.gov and eCFR.gov, programmatic access to these sites is limited to access to our extensive developer APIs. Please visit FederalRegister.gov API documentation or eCFR.gov API documentation to learn more about how to access the API.
Your request has been flagged as potentially automated. If you are human user receiving this message, please complete the CAPTCHA (bot test) below and click “Request Access”. You may occassionally be asked to complete the CAPTCHA again, this is normal and pa
wooley_rhino • run 5cd5ce61 • $2.99 • 6m35s
Extracted Percentile Values: {0.1: 50, 1.0: 58, 5.0: 68, 10.0: 74, 15.0: 78, 20.0: 81, 25.0: 83, 30.0: 85, 35.0: 87, 40.0: 89, 45.0: 91, 46.0: 91, 47.0: 92, 48.0: 92, 49.0: 93, 50.0: 93, 51.0: 94, 52.0: 94, 53.0: 95, 54.0: 95, 55.0: 96, 60.0: 98, 65.0: 100, 70.0: 103, 75.0: 106, 80.0: 109, 85.0: 113, 90.0: 118, 95.0: 126, 99.0: 140, 99.9: 155}
Model’s Answer:
I need to forecast the count of NRC Federal Register documents published from June 8 to August 12, 2026 (66 calendar days). The resolution counts distinct documents with NRC listed as an agency and publication dates in the window.
Same window (6/8-8/12) in prior years:
2026 Pace Assessment:
Upward pressures:
Downward pressures:
Starting from the 2023-2025 cluster average of 86:
Cross-validation with daily rates:
Critical Assumption 1: No government shutdown during window
Critical Assumption 2: Executive order implementation proceeds smoothly
Critical Assumption 3: Summer seasonality effects similar to historical patterns
A major administrative disruption, severe commissioner quorum issues, or appropriations lapse could compress output toward 2022-like levels. Summer slowdown could be more severe than typical. Probability: ~5%
Full executive order implementation, accelerated restarts, SMR licensing surge, and minimal summer slowdown could push output significantly above historical norms. Probability: ~5%
The central estimate centers around 93-95 documents, reflecting the 2026 elevated pace moderated by summer seasonality and administrative constraints. The distribution is right-skewed because there are more pathways to exceed historical highs (executive pressure, restart proceedings, SMR surge) than to fall significantly below the 2023-2025 cluster. The 2022 low of 66 was an outlier caused by specific judicial proceedings, unlikely to recur without similar disruption. I’m setting relatively wide confidence intervals given institutional uncertainty around executive order implementation and commissioner dynamics.
Percentile .1: 50 Percentile 1: 58 Percentile 5: 68 Percentile 10: 74 Percentile 15: 78 Percentile 20: 81 Percentile 25: 83 Percentile 30: 85 Percentile 35: 87 Percentile 40: 89 Percentile 45: 91 Percentile 46: 91 Percentile 47: 92 Percentile 48: 92 Percentile 49: 93 Percentile 50: 93 Percentile 51: 94 Percentile 52: 94 Percentile 53: 95 Percentile 54: 95 Percentile 55: 96 Percentile 60: 98 Percentile 65: 100 Percentile 70: 103 Percentile 75: 106 Percentile 80: 109 Percentile 85: 113 Percentile 90: 118 Percentile 95: 126 Percentile 99: 140 Percentile 99.9: 155
Forecast: median = N/A (no P50)
How many Nuclear Regulatory Commission documents will the Federal Register publish between 6/8/2026 and 8/12/2026?
Key figures
Historical context
Historically, the count of NRC documents in the June 8 to August 12 window has shown a remarkably tight cluster in recent years. In 2023 and 2024, the Federal Register published exactly 88 documents in each window. In 2025, the count was 82. This three-year average of 86 stands in contrast to 2022, which was a historical low at 66 documents. The ‘regime change’ of 2026 is visible in the YTD numbers: through early June 2026, the NRC published 228 documents, a 59% increase over the 143 documents published through the same date in 2025, and an 18% increase over 2024. This historical shift is largely attributed to the implementation of the 2024 ADVANCE Act and the aggressive nuclear-focused executive orders issued in May 2025.
Tailwinds
Headwinds
Detailed reasoning
My prediction is primarily driven by the observed 2026 year-to-date (YTD) publication rate, which represents a structural break from the previous four years. As of June 4, 2026, the NRC had published 228 documents in 155 days. This average of 1.47 documents per calendar day provides a strong statistical anchor for the 66-day window (June 8 to August 12). Applying this daily rate yields a projection of approximately 97 documents. Furthermore, the most recent 65-day window leading into the forecast period (April 1 to June 4) produced 106 documents, or 1.63 per day, which would scale to nearly 108 documents over the forecast window if sustained.
However, I have tempered these mechanical extrapolations by accounting for historical ‘summer seasonality’ and institutional bottlenecks. Historically, the June-to-August window is slightly less active than the spring or fall periods due to administrative scheduling and the summer-month regulatory hearing lull. While the 2023-2025 period saw a consistent ‘cluster’ of 82 to 88 documents, 2026 is uniquely impacted by a flurry of executive mandates. President Trump’s Executive Order 14300 and the ADVANCE Act of 2024 have essentially forced a ‘throughput requirement’ on the NRC. Key projects like the Crane Clean Energy Center (Three Mile Island Unit 1) restart and the Palisades Nuclear Plant restart are generating high volumes of license amendments, environmental assessments, and hearing notices.
I weighted the current elevated 2026 pace at 85% of the forecast, centering the most likely outcome at 90 documents. This is intentionally conservative relative to the 97-108 document run-rate to account for potential friction from the NRC’s mid-summer reorganization and the ‘Trump v. Slaughter’ litigation, which could temporarily impact commission-quorum requirements for contested matters. A smaller 15% probability weight was assigned to a ‘downside’ scenario where administrative slowdowns or a reversion to 2025 levels (82 documents) occur. Because three of the four recent historical analogs (82, 88, and 88) already sit above the question’s 75.5 bin threshold, my analysis places the vast majority of the probability mass in the open upper-tail region, indicating that an ‘in-range’ outcome (0-75) is highly unlikely.
Key uncertainties
Conclusion