Scenario wins: cassi (154) smingers-bot (47) pgodzinbot (1)
Forecast rationale (numeric):
— Iteration 1 — Overall, the forecasts converge on a risk-elevated but highly uncertain outlook driven by the current Israel-centered conflict environment and broader regional tensions. The main reasoning pattern is that these conditions raise the chance of a qualifying bombing in the US, Israel, or an EU country, but the 5+ death threshold remains a high bar, especially in the US and EU where strong counterterrorism capabilities reduce the odds of a successful mass-casualty attack.
The collective reasoning is that the geopolitical environment has materially increased risk, but a bombing with 5 or more deaths is still far from certain and may well not occur within the target window. The disagreement is mainly about whether the elevated threat is strong enough to make a near-term event more likely than not, versus still leaving the no-event outcome as the modal case.
— Iteration 2 — Across the forecasts, the core reasoning is that the event is rare at baseline but materially more likely than usual in the current geopolitical environment.
The collective reasoning suggests a temporarily elevated threat environment driven by the US-Israel-Iran conflict, but with considerable uncertainty and strong dependence on whether retaliatory or opportunistic attacks materialize. Most models lean toward an incident occurring sooner rather than later, though one model stresses that the event is still more likely to fall after the short resolution window.
— Iteration 3 — Overall, the forecasts treat a 5+ death bombing in the US, Israel, or an EU country as a rare but plausibly near-term event. The main reasoning patterns are:
In short, the shared view is that the probability is materially elevated because of regional conflict, but the event remains rare enough that timing is highly uncertain, with the most likely jurisdiction being Israel and the most debated issue being whether the incident happens before the cutoff or only afterward.
The rationales broadly agree that forecasting this event requires balancing the exceptionally low historical base rates of successful mass-casualty bombings against a uniquely elevated global threat environment during the summer of 2026.
Base Rates and Counter-Terrorism in the West Forecasters universally note that the historical frequency of non-state actors successfully executing bombings with five or more fatalities in the US and EU is virtually non-existent in recent years. This is largely attributed to highly effective intelligence interdiction, strict regulations on explosive precursors, and a tactical shift among lone-actor extremists toward low-tech methods like firearms or vehicle rammings. While the intent to carry out complex explosive plots remains, the failure and interception rates are overwhelmingly high.
Elevated Threat Catalysts (Summer 2026) Despite the low historical baseline, the rationales identify several short-term factors that temporarily elevate the threat landscape:
Israel and the Middle East Dynamics The consensus is that Israel (including Gaza and the West Bank) represents the most probable locus for a qualifying event. The use of IEDs and booby traps by non-state militant groups against IDF patrols has a much higher historical frequency than civilian bombings in the West. However, several limiting factors are noted:
Conclusion Overall, the rationales synthesize these factors to conclude that while the global threat matrix is significantly elevated due to symbolic summer events and proxy conflicts, modern counter-terrorism apparatuses remain highly robust. Consequently, forecasters share a strong consensus that a successful mass-casualty bombing is highly unlikely to bypass security measures and occur within the narrow, short-term forecasting window.
Question: When will the US, Israel, or an EU country experience a bombing incident resulting in 5 or more deaths? 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 assesses the likelihood of a bombing causing 5+ deaths in the US, Israel, or an EU country during the 79-day forecast window (25 May–12 Aug 2026) in the context of an active US–Israel–Iran war that began 28 Feb 2026 after strikes reportedly killing Iran’s Supreme Leader. It summarizes recent March–May 2026 activity: a wave of ~14 pro‑Iran attacks across European cities (including explosions near a Liège synagogue and the U.S. Embassy in Oslo) that produced zero fatalities; a foiled NYC IED incident on 9 Mar 2026; and an armed attack on a Michigan synagogue on 14 Mar 2026 with no civilian deaths. The brief contrasts historical base rates—Israel had frequent deadly suicide bombings during 2000–2005 but very few since 2008; the US has seen major bombings causing 5+ deaths only rarely (roughly 1–2 per decade in recent history); and the EU has experienced sporadic deadly bombings (roughly 1–2 per decade across member states). It notes intelligence warnings that current low‑casualty campaigns could escalate, but also emphasizes strong security responses, high failure rates of recent attempts, and low attacker sophistication.
Analytical factors increasing risk include the unprecedented geopolitical context, explicit Iranian threats, activated networks in Europe, and demonstrated attacker intent; factors reducing risk include heightened security across regions, intercepted plots, recent zero‑fatality attack outcomes, generally low lethality trends in modern conflicts, and the short 79‑day window. The brief cites prediction‑market reliability concerns (Polymarket) and provides a crude outside‑view quantitative anchor—after adjusting an elevated baseline by assumed security reductions, an illustrative estimate of ~0.1–0.15 qualifying events in the 79‑day window—while stressing high uncertainty and data limitations (no classified intel, unreliable market data, imperfect historical parallels, and possible undercounting of foiled plots).
Sources used (as cited in the research): Europol — https://www.europol.europa.eu Polymarket — https://polymarket.com Global Terrorism Index / Vision of Humanity — https://www.visionofhumanity.org/maps/global-terrorism-index/ Wikipedia — https://en.wikipedia.org (historical entries on Israeli bombings/Second Intifada) and various news reporting referenced in the brief (March–May 2026 coverage of European attacks, NYC IED incident, Michigan synagogue attack).
Forecast Period: May 25, 2026 (current date) to August 12, 2026 (79 days)
As of May 25, 2026, you are operating in an unprecedented security environment. A war between the United States, Israel, and Iran began on February 28, 2026, following joint US-Israeli strikes that reportedly killed Iranian Supreme Leader Ali Khamenei [2][3][36]. This conflict has created elevated terrorism threats across all three regions in your question.
Europe:
United States:
Israel:
Historical Peak (2000-2005 Second Intifada):
Post-2008 Period:
Implication: While Israel historically had very high bombing rates, there’s been a multi-decade trend away from this attack method by non-state actors. The current war environment creates military strike risks (excluded from your question) rather than traditional terrorist bombings meeting your criteria.
Historical Context:
Recent Reference Points:
Current Threat Environment:
Implication: Base rate for bombings causing 5+ deaths in the US is very low, with only 1-2 such incidents per decade in recent history.
Historical Reference:
Recent Trends (2025-2026):
Security Response:
Implication: Despite elevated threat warnings, recent attack attempts in Europe have resulted in zero fatalities, possibly due to low sophistication of attackers and enhanced security.
While Polymarket has shown some accuracy forecasting Middle East security events [14], the platform faces serious concerns:
No specific forecast data found for Metaculus, Manifold, or other platforms regarding this specific question timeframe.
From the research on Middle East conflict [11]:
Historical Frequency in Target Regions:
Current Environment Adjustment:
Crude Outside View Estimate (for reference only): If historical base rate across all three regions were ~2-3 qualifying attacks per year in elevated threat periods, and current security reduces success rates by ~60%, you might expect approximately 0.1-0.15 qualifying events in a 79-day window. However, this is highly uncertain given the unique geopolitical situation.
You face a genuinely elevated threat environment with demonstrated attack networks activated in all three regions. However, the track record shows recent attempts have been intercepted (US) or produced zero fatalities (Europe), while Israel’s threat comes primarily from military rather than terrorist bombing operations. The 79-day window is relatively short, and historical base rates for bombings causing 5+ deaths remain low despite current tensions. Security services are highly alert and have shown effectiveness at prevention.
I accept several of the critic’s points and am moving down from 22% to about 18–19% for a qualifying detonation by 2026-08-12T12:00:00Z, conditional on no annulment.
The most important correction is jurisdiction discipline: Oslo and London are not resolving jurisdictions, so I should treat those only as evidence of network reach, not as direct event-rate evidence. I also agree that the strict modern US base rate for 5+ fatal bombings is lower than “1 per decade” if Boston is below threshold and Oklahoma City is now a very old datapoint. The Gracie Mansion case still matters because it involved TATP/shrapnel devices and alleged intent to kill many people, but it also reinforces the “attempts are amateurish/intercepted/fail to detonate” side of the ledger. (justice.gov)
I still defend the core upward adjustment from peacetime base rates. Europol’s EU threat assessment is materially elevated in connection with the Iran conflict, and there is documented recent activity against Jewish/American/Israel-linked targets in EU countries such as Belgium, France, Germany, and the Netherlands, even if some claimed HAYI incidents are low-sophistication or partly propagandistic. (euronews.com) The US World Cup period also keeps the US tail risk nontrivial, though the best-protected stadium environments are probably less vulnerable than fan zones, transit, protests, hotels, or religious/political targets. ODNI/NCTC coordination is a real mitigating factor, not just background noise. (dni.gov)
The critic’s best point is the Israel resolver ambiguity. Under a narrow interpretation—classic terrorist bombings/IEDs/suicide vests/car bombs, excluding Hezbollah/Hamas/Houthi rockets, drones, mortars, and battlefield-style IEDs—the probability is probably only ~13–15%. Under a literal interpretation where non-state explosive munitions in Israel or Israel-administered territory count, especially if they kill soldiers or civilians, the probability could be ~22–28%. I now blend these rather than implicitly using the narrower interpretation, but I give meaningful weight to a narrow resolution because the background frames this as “bombing incidents” rather than ordinary combat exchanges.
The emerging US-Iran de-escalation framework also lowers my central estimate: AP reported on May 25, 2026 that the US and Iran appeared close to a deal to end the war and reopen Hormuz. That reduces the probability of centrally approved Iranian spectacular terrorism, though it does not eliminate spoiler risk from hardliners, proxies, or self-radicalized actors. (apnews.com)
Net: I move my pre-cutoff probability to roughly 18.5%. That puts the 20th percentile after the August 12 cutoff, but not extremely far after it because residual elevated risk likely persists for some weeks even if the World Cup ends and diplomacy improves.
Updated prediction:
Percentile 0.1: 2026-05-27
Percentile 1: 2026-06-04
Percentile 5: 2026-06-25
Percentile 10: 2026-07-18
Percentile 20: 2026-09-14
Percentile 30: 2027-02-05
Percentile 40: 2027-08-20
Percentile 50: 2028-03-20
Percentile 60: 2028-12-20
Percentile 70: 2029-11-15
Percentile 80: 2031-01-20
Percentile 90: 2033-01-15
Percentile 95: 2035-03-15
Percentile 99: 2040-01-15
Percentile 99.9: 2047-01-01
Conditioning on non-annulment and assuming no already-qualifying incident counts.
I accept the critic’s core point: the decisive uncertainty is not just operational risk but scope/adjudication risk, especially in Israel. The resolution text is broader than a “Brussels/Boston-style terrorist bombing” frame: it includes non-state explosive detonations, excludes only perpetrators, and includes Israeli-administered territories. So a Hezbollah/Hamas/PIJ rocket, explosive drone, IED, grenade, tunnel bomb, or booby-trap killing 5+ non-perpetrators could plausibly qualify if the detonation is inside Israel or an included administered territory and is not treated as excluded conventional combat.
That said, I still think there is a meaningful limiting factor: not every military-adjacent explosive event should be counted. Israeli state airstrikes in Lebanon are excluded; Hezbollah attacks on Israeli troops in Lebanon probably often fail the jurisdiction test; and some Gaza/Lebanon battlefield incidents may be resolution-contested rather than clean “bombing incidents.” Reports do show continuing Israel-Hezbollah violence despite a ceasefire, including rockets/drones and Israeli strikes, but recent Hezbollah-linked attacks cited in public reporting have generally wounded or caused limited Israeli casualties rather than repeatedly producing 5+ fatal incidents. (apnews.com)
I also agree the acute Iran-war context remains unusually dangerous, but I now slightly discount the March/April “war retaliation” peak because there are live reports of a U.S.-Iran framework/deal that could end the war and reopen Hormuz. That does not remove risk—spoilers and proxies may still act—but it weakens the argument for a continuously rising acute hazard through August. (apnews.com)
For the EU/US buckets, I still see real but minority probability. Europol’s elevated-threat warning and the HAYI/pro-Iran campaign against Jewish, Israeli, and U.S.-linked targets in EU countries are materially relevant, and the arrest/charging of an alleged organizer suggests both intent and interdiction. But the “Europe” examples must be filtered carefully: UK and Norway are not EU; and recent EU incidents have mostly been low-fatality or disrupted rather than demonstrated mass-casualty capability. (euronews.com)
My updated branch model is roughly:
Blending these, and incorporating the possible de-escalation signal, I move from 22% to about 25% by 2026-08-12T12:00:00Z. This is close to the critic’s view, but still a bit below 27% because I put more weight on adjudication exclusions, interdiction, and the possibility that the acute Iran-war tail is already past its peak.
Updated prediction:
Percentile 0.1: 2026-05-26
Percentile 1: 2026-05-29
Percentile 5: 2026-06-10
Percentile 10: 2026-06-27
Percentile 20: 2026-07-29
Percentile 30: 2026-11-25
Percentile 40: 2027-07-01
Percentile 50: 2028-03-01
Percentile 60: 2029-04-01
Percentile 70: 2030-09-01
Percentile 80: 2033-01-01
Percentile 90: 2037-06-01
Percentile 95: 2042-06-01
Percentile 99: 2058-01-01
Percentile 99.9: 2083-01-01
Conditioning on no annulment, and assuming any already-eligible incident is ignored.
I think the critic’s main point is valid: my prior forecast underweighted the chance that non-state explosive warfare in Israel qualifies. The resolution text excludes state-conducted conventional military operations, but explicitly includes non-state actors; it also does not exclude soldiers as victims. So a Hezbollah rocket/drone strike, Hamas/PIJ IED, grenade, or other explosive attack killing 5+ non-perpetrators in Israel or Israeli-administered territory could plausibly qualify, even if it feels more like “war” than “terrorism.” That raises the probability.
I also accept the critic’s “Europe vs EU” correction. UK, Norway, and other non-EU incidents matter as signals of Iranian/proxy tradecraft, but they do not directly resolve the question. The EU component should be discounted relative to “Europe” writ large.
What I still think is right from my original forecast: the modal outcome remains no qualifying incident by August 12. The observed campaign has shown intent and geographic reach, but mostly low lethality: HAYI-linked/pro-Iran attacks in EU countries such as Belgium and the Netherlands have been serious but not mass-casualty so far; the Gracie Mansion TATP incident shows US bombing intent but was disrupted/contained; and Hezbollah rocket activity has recently hit open areas or caused low Israeli casualties rather than 5+ in one detonation. (adl.org)
My updated model is roughly:
So I move from about 14% to about 18% for a qualifying detonation by 2026-08-12T12:00Z. I do not go all the way to 20%+ because the broad-Israel path is resolution-ambiguous and because recent non-state explosive activity has still not often produced 5+ deaths in a single eligible detonation.
For the post-cutoff percentiles, these mostly encode the large “no qualifying event by cutoff” mass; they should not be interpreted as precise resolution dates if the market resolves simply as >2026-08-12T12:00:00+00:00.
Updated prediction:
Percentile 0.1: 2026-05-26
Percentile 1: 2026-05-30
Percentile 5: 2026-06-14
Percentile 10: 2026-07-05
Percentile 20: 2026-09-03
Percentile 30: 2027-01-20
Percentile 40: 2027-09-01
Percentile 50: 2028-05-01
Percentile 60: 2029-04-01
Percentile 70: 2030-09-01
Percentile 80: 2033-01-01
Percentile 90: 2037-01-01
Percentile 95: 2043-01-01
Percentile 99: 2056-01-01
Percentile 99.9: 2080-01-01
Agent Forecast (date, confidence: medium)
Base Rate: Zero qualifying bombings (5+ non-perpetrator deaths, non-state actor) in the combined US/EU/Israel jurisdiction since Manchester Arena (May 2017) — approximately 9 years. Baseline ≈ 0.1–0.2 events/year; elevated by 3–5x under current Iran-war conditions to ~0.5–0.7/year, yielding ~11–15% per 79-day window before adjusting upward for non-Iran actor risk to ~18%.
Initial Reasoning:
Reference class: Non-state actor bombings with 5+ non-perpetrator deaths in US, Israel, or any EU member state.
Key Uncertainties:
Evidence Sources (16):
Grounding Facts (passed to reviewers):
Draft: After Aug 12, 2026: 86.8%, Aug 02, 2026 to Aug 12, 2026: 1.8%, Jun 04, 2026 to Jun 14, 2026: 1.7%, Jun 14, 2026 to Jun 24, 2026: 1.7%, Jun 24, 2026 to Jul 04, 2026: 1.7%
Reviewer Panel:
Reviewer (reasoning)
The draft forecast leans heavily on a low historical base rate and the assumption that Iran‐linked proxies will continue deliberately low‐casualty tactics, underestimating the potential for escalation or attacks by other actors (e.g., domestic extremists). It also distributes risk uniformly across the 79-day window without accounting for clustering around symbolic dates or credible intelligence signals of planned escalations. Lastly, it largely ignores shifts in white-supremacist or far-right networks in the US and Europe that have shown a capacity for high‐fatality attacks.
Objections:
Reviewer (calibration)
The forecast heavily relies on the assumption that Iran-linked proxy attacks will remain deliberately low-casualty, despite the ongoing ‘US-Israel war on Iran.’ This assumption might be brittle in a rapidly escalating conflict, as intent can change, or less controlled actors might emerge. The forecast also appears to underweight the risk from non-Iran-affiliated groups or domestic extremists in the US and EU.
Objections:
Reviewer (fact-checker)
1 unsupported claim(s). Verdict: minor_issues
Unsupported claims:
Adjustment Rationale:
Adjusted in-window probability from ~13% (draft average across three trajectories) to ~18% in response to three valid reviewer points: (1) my draft over-weighted Iran-proxy dynamics relative to domestic extremists, ISIS/AQ-inspired actors, and Palestinian militant cells, all of which contribute independently to the bombing hazard; (2) the ‘deliberately low-casualty’ Iran-proxy strategy is brittle — a US-Iran deal collapse or rogue cell could escalate to mass-casualty operations; (3) Iron Dome (cited in T3) is irrelevant to internal IED/suicide-bombing risk, so I removed it as a probability suppressor. I also added mild upward skew toward summer bins (peak tourist season, July 4, Bastille Day, mid-summer gatherings) per reviewer clustering critique. However, I did not move all the way to the 23.8% ensemble anchor because the empirical base rate is genuinely near-zero: zero qualifying EU bombings since Brussels 2016, zero qualifying Israeli bombings since ~2007, zero qualifying US bombings since Boston 2013 (only 3 deaths). Critically, 87 days into the 2026 US-Israel war on Iran, multiple Iran-linked European attempts (Liège, Rotterdam, Amsterdam, Oslo) have produced zero qualifying events — this is strong contemporaneous evidence that the elevated-threat environment is not translating to the 5-death threshold being crossed. Reviewers may be anchoring on threat-level assessments rather than threshold-crossing event rates. TAIL JUSTIFICATION: The 82% tail rests on a robust ~9-year empirical record of zero qualifying events across the combined US/EU/Israel jurisdiction. Even with a 4-5x multiplier on the underlying ~0.15 events/year base rate, a 79-day window yields only ~15-20% in-window probability. The 82% tail leaves ample room for the moderate escalation scenarios reviewers correctly flag, while respecting the strong empirical signal that 5+ death bombings remain rare events in these jurisdictions even under elevated geopolitical stress.
forecast_sdk | 470.5s | research $1.7976 + reviewers $0.0104 + reconcile $0.2516 = $2.0994
Extracted Date Percentile Values: {0.1: 1779667200.0, 1.0: 1780185600.0, 5.0: 1782432000.0, 10.0: 1785283200.0, 15.0: 1786579200.0, 20.0: 1786579200.0, 25.0: 1786579200.0, 30.0: 1786579200.0, 35.0: 1786579200.0, 40.0: 1786579200.0, 45.0: 1786579200.0, 46.0: 1786579200.0, 47.0: 1786579200.0, 48.0: 1786579200.0, 49.0: 1786579200.0, 50.0: 1786579200.0, 51.0: 1786579200.0, 52.0: 1786579200.0, 53.0: 1786579200.0, 54.0: 1786579200.0, 55.0: 1786579200.0, 60.0: 1786579200.0, 65.0: 1786579200.0, 70.0: 1786579200.0, 75.0: 1786579200.0, 80.0: 1786579200.0, 85.0: 1786579200.0, 90.0: 1786579200.0, 95.0: 1786579200.0, 99.0: 1786579200.0, 99.9: 1786579200.0}
Model’s Answer: Looking at this question, I need to forecast when (or if) a qualifying bombing incident with 5+ non-perpetrator deaths will occur in the US, Israel, or an EU country between May 25 and August 12, 2026.
United States: The last qualifying bombing was Oklahoma City in 1995 (31 years ago). The Boston Marathon bombing (2013) killed only 3 victims. Since then, high-fatality extremist violence has been overwhelmingly via firearms, not explosives.
European Union: The last qualifying bombings occurred in 2015-2017 (Paris, Brussels, Manchester). That’s approximately 9 years without a 5+ fatality bombing in EU member states, despite persistent threat activity.
Israel: While facing persistent attack pressure, most recent incidents involve shootings, stabbings, and small IEDs. The 2026 data shows only 1 fatality across 14 attacks year-to-date. Mass-casualty bombings (5+ deaths) have been rare since the Second Intifada.
The research indicates:
Combining base rates across jurisdictions:
The elevated conflict environment adds some probability, but security responses have also intensified. Net estimate: ~12% probability of a qualifying event occurring during the window.
Assumption 1: The shift toward lower-complexity attack methods continues. If actors revert to bomb tactics, probability could increase to 15-20%. (Probability of failure: ~15%)
Assumption 2: Israeli counterterrorism effectiveness remains high despite ongoing conflict. If security degrades, Israel becomes more likely location. (Probability of failure: ~10%)
Assumption 3: No major escalation in the US-Israel-Iran conflict triggers retaliatory mass-casualty operations in the window. (Probability of failure: ~10%)
Earlier outcome (June): More likely if Iranian proxies/Hezbollah execute a planned operation in Israel, or an inspired actor in Europe acts on current tensions while security fatigue sets in. Likelihood: ~4% cumulative by end of June.
Later outcome (no event by Aug 12): Most probable scenario given 9+ year absence of qualifying EU bombings, 31+ year absence in US, and Israeli pattern of intercepting major plots. Likelihood: ~88%.
Given 12% probability of an event occurring, distributed roughly uniformly across the 79-day window, with ~88% probability mass assigned to post-deadline resolution.
Percentile .1: 2026-05-25 Percentile 1: 2026-05-31 Percentile 5: 2026-06-26 Percentile 10: 2026-07-29 Percentile 15: 2026-08-13 Percentile 20: 2026-08-13 Percentile 25: 2026-08-13 Percentile 30: 2026-08-13 Percentile 35: 2026-08-13 Percentile 40: 2026-08-13 Percentile 45: 2026-08-13 Percentile 46: 2026-08-13 Percentile 47: 2026-08-13 Percentile 48: 2026-08-13 Percentile 49: 2026-08-13 Percentile 50: 2026-08-13 Percentile 51: 2026-08-13 Percentile 52: 2026-08-13 Percentile 53: 2026-08-13 Percentile 54: 2026-08-13 Percentile 55: 2026-08-13 Percentile 60: 2026-08-13 Percentile 65: 2026-08-13 Percentile 70: 2026-08-13 Percentile 75: 2026-08-13 Percentile 80: 2026-08-13 Percentile 85: 2026-08-13 Percentile 90: 2026-08-13 Percentile 95: 2026-08-13 Percentile 99: 2026-08-13 Percentile 99.9: 2026-08-13
Geopolitical geography and population density set the baseline exposure: the United States and EU have low historical rates, Israel and adjacent occupied or contested areas have higher operational tempo.
Security institutions, border controls, and law‑enforcement capacity suppress large‑bomb incidents in stable democracies but cannot eliminate attacks in active conflict zones.
Sustained low‑level cross‑border strikes, increased use of drones and IEDs, and proliferation of tactical explosives raise the near‑term hazard in Israel and adjacent theaters.
Domestic plots, radicalization trends, and isolated conspiracies maintain a persistent, though comparatively smaller, risk in the U.S. and EU.
A breakdown of local ceasefires or a marked escalation with Hezbollah or armed Gaza factions would shift the hazard sharply upward and concentrate risk on Israeli‑administered areas.
Targeted disruptions, improved intelligence penetration, or international pressure reducing cross‑border activity would tilt the window toward non‑occurrence.
Key uncertainties are definitional—whether non‑state rockets, drones, and battlefield IEDs in contested territory count as qualifying “bombings”—and that sensitivity materially alters the outcome.
Operational surprises, clandestine plots, and short‑term political shocks create residual uncertainty that keeps the estimate imprecise despite clear asymmetric regional risk.
Median: No P50 (unresolved beyond forecast window)
The four independent analyses converge on a ~20% probability of a qualifying bombing incident (5+ deaths by non-state actors) occurring within the May 25 – August 12, 2026 window, with ~80% probability the event occurs later or not at all.
Suppressed lethality despite high attack volume: Israel experienced 13–38 IED incidents monthly in 2025 with zero incidents reaching the 5-death threshold [S32]. European Iranian-linked networks executed 18 attacks since March 2026 with zero casualties, deliberately designed to avoid deaths [S33].
Disrupted coordination: The primary coordinator of the European Iranian proxy network was arrested May 15 [S16], degrading near-term operational capability, though residual cells may remain active.
ISKP as independent wildcard: The Islamic State Khorasan Province is identified as a major external threat to the EU [S21] and has demonstrated mass-casualty bombing capability (Moscow 2024, Vienna 2024 attempts), operating independently of the Iran conflict dynamics.
Elevated but constrained threat environment: The active US-Israel-Iran conflict and ongoing Hezbollah-Israel hostilities create elevated baseline risk, but Israeli air defenses (Iron Dome, shelters) and security infrastructure suppress per-incident lethality. West Bank bombing attempts by Hamas/PIJ have recently been “botched” [S38].
Conditional timing: If an event occurs within the window, it is approximately uniformly distributed across the 78 days, with slight front-loading toward late June–early July. The median conditional date is roughly July 3, 2026.
Long tail: The 80% no-event probability creates a right-skewed distribution extending well into 2027, with the overall unconditional median falling in October–November 2026 or beyond.
Forecast and Reasoning
Forecast: Probability distribution:
Forecast Reasoning
We need the first intentional non-state “bombing” (explosive detonation) in the US, Israel (incl. territories under its administration), or an EU member state that causes ≥5 deaths (excluding perpetrators) within 30 days, with the detonation occurring by 2026-08-12T12:00Z. The window is ~79 days.
Across the team, the central disagreement is the in-window probability: Forecaster 4 is lowest (~15–20%), Forecasters 2 and 3 are ~25–27%, and Forecaster 1 is highest (~40%). I discount Forecaster 3’s percentiles because they stack at the upper bound (which encodes mass on the deadline date rather than the open-after tail), and I treat Forecaster 1’s 40% as somewhat aggressive given how rare ≥5-fatality bombings are in high-security jurisdictions, even when general threat levels are elevated.
Balancing (a) low base rates for ≥5-fatality bombings in the US/EU/Israel in recent years, (b) evidence of elevated threat/foiled plots cited by multiple forecasters, and (c) definitional ambiguity (e.g., whether some conflict-related explosive attacks would be treated as “bombings” under the criteria), I center on ~25–30% chance of at least one qualifying incident occurring before the deadline. That implies most probability mass is in the open-after tail (>2026-08-12T12:00Z), so the median should be after the bound.
For within-window timing conditional on occurrence, I assume an approximately memoryless/roughly spread hazard over the period (no strong, reliable evidence for a sharp spike), which yields early percentiles in June/July and then moves into the open-after tail by around the 30th percentile.
An amateur forecast is likely to (1) conflate shootings/vehicle attacks/rocket fire or state military strikes with the narrower “intentional bombing by a non-state actor” criterion, inflating in-window probability, and/or (2) mishandle the open upper bound by stacking percentiles at 2026-08-12 (incorrectly placing mass on the deadline date rather than after it). My forecast explicitly separates “probability it happens in-window” from “timing conditional on happening,” and it represents the large ‘no event by deadline’ mass using strictly post-bound dates, improving calibration under the platform’s resolution rules. Confidence: moderate (tail mass is robust; exact in-window timing is inherently noisy).