Published April 24, 2026 | mideastintel.com
In the intelligence profession, humility is both a virtue and an occupational hazard. Analysts are trained to acknowledge uncertainty, qualify their assessments, and resist the kind of confident prediction that makes for clean newspaper headlines but poor tradecraft. Yet over the past eighteen months, a class of actors conspicuously lacking formal intelligence training — the traders who populate cryptocurrency-denominated prediction markets — have repeatedly placed early, accurate, and well-calibrated bets on outcomes that professional analysts publicly doubted or quietly understated. The Iran crisis of 2025–2026 has become, in retrospect, a stress test for both methodologies. The results are uncomfortable for the intelligence community to absorb and genuinely instructive for anyone trying to understand the epistemology of geopolitical forecasting.
This is not an argument that financial speculation should replace intelligence analysis. It is an argument that probability aggregation from diverse, financially motivated participants captures certain forms of information — particularly regarding the timing and likelihood of discrete, observable events — with a precision that classified bureaucracies often cannot match, for structural reasons that deserve honest examination.
The Academic Foundation: Tetlock, IARPA, and the Superforecasting Tradition
The intellectual basis for taking prediction markets seriously as forecasting instruments predates the current crisis by over a decade. Philip Tetlock's seminal work, culminating in his collaboration with IARPA — the Intelligence Community's research arm — on the Good Judgment Project, produced findings that were striking and, within intelligence circles, quietly embarrassing. After four years, 500 geopolitical questions, and over one million individual forecasts, Tetlock's civilian superforecasters outperformed intelligence officers with access to classified information by approximately 30 percent. The Good Judgment Project emerged as the undisputed winner of the IARPA tournament.
Tetlock identified four drivers of forecasting accuracy: recruitment of numerically calibrated thinkers, structured cognitive-debiasing training, collaborative teamwork in prediction markets, and superior statistical aggregation of the wisdom of the crowd — the last factor alone accounting for a 35 percent improvement over unweighted averaging. Critically, about 70 percent of superforecasters maintained their elite ranking from year to year, ruling out the hypothesis that accurate forecasting was simply luck. What prediction markets do, at their best, is industrialize the Tetlockian insight: they convert the distributed, partially informed judgments of thousands of participants into a continuously updated, financially backed probability estimate.
This is the theoretical case. The empirical case, as the Iran crisis demonstrates, is more nuanced.
What the Markets Got Right: US Strike Probability
The June 2025 US airstrikes on Iranian nuclear facilities are the clearest case where prediction market signals outran the public expert consensus by a wide and telling margin. In the weeks preceding the operation, mainstream analysts and think-tank commentary largely dismissed the prospect of imminent US military action, citing diplomatic constraints, allied reservations, and the administration's stated preference for negotiated outcomes. The markets disagreed emphatically.
When President Trump publicly mentioned contemplating strikes on Iran's nuclear program, Middle East experts responded with characteristic skepticism. Prediction markets simultaneously assigned a 58 percent probability to US strikes occurring within the week — a figure that proved prescient when seven B-2 stealth bombers entered Iranian airspace carrying bunker-buster munitions. In the months running up to the February 2026 coordinated US-Israeli operation, Polymarket's "US military action against Iran before July" contract was pricing the probability at 71 percent as early as January 2026, a full month before the strikes. One senior energy analyst, citing those market signals, publicly estimated a 75 percent probability of near-term US action — a figure treated skeptically by regional experts at the time and vindicated by events.
The volume of capital behind these assessments is not trivial. Over $500 million traded in a single multi-outcome market on the timing of the US-Iran strikes. That liquidity matters: thin markets can be moved by a single large position, and their prices are correspondingly less reliable. Deep, liquid markets aggregate more genuine information. Platforms such as PolyMarket Predictions provide real-time probability assessments on geopolitical events that intelligence professionals increasingly monitor alongside classified estimates — and the volume in Iran-related markets gave those probabilities genuine evidential weight.
The Hormuz Signal: Calibrated Uncertainty in Real Time
The Strait of Hormuz — the 33-kilometer chokepoint through which approximately 20 percent of global oil supply passes — became a central prediction market category in 2025. The markets' handling of Hormuz risk demonstrates both the strengths and the inherent limits of the methodology.
Through the second half of 2025, as Israel-Iran tensions escalated through the tit-for-tat exchange of strikes that preceded the full-scale February operation, Polymarket's Hormuz closure contract fluctuated between 14 and 22 percent probability for a formal Iranian-imposed closure in 2025. That range — relatively modest by crisis standards — proved accurate in its conservatism: Iran did not formally close the Strait in 2025, preferring interdiction and harassment operations that fell short of full closure while substantially degrading transit volumes.
The markets updated sharply in early 2026. By February, traders had repriced normalization expectations dramatically downward: the probability of Hormuz traffic returning to normal by April 30, 2026 collapsed to approximately 6.5 percent, reflecting the blockade conditions and continued conflict disruption. The subsequent contracts — May 31 at 61 percent, June 30 at approximately 70 percent — encode a market expectation of gradual, not rapid, normalization that aligns closely with the analyst consensus on negotiating timelines. Here the markets and the professional forecasting community converged, which is itself a useful data point: convergence between market prices and expert estimates increases the credibility of both.
Intelligence analysts and strategic planners who tracked these probabilities through resources like PolyMarket Predictions had access to a continuously updating, financially backed probability distribution over Hormuz scenarios — a complement to, rather than a substitute for, classified source reporting on Iranian military intentions.
Where the Markets Struggled: Timing, Insider Risk, and the Nuclear Escalation Problem
The case for prediction markets as a forecasting tool must be made honestly, which requires acknowledging the methodological failures alongside the successes. Three categories of limitation are particularly relevant to the Iran crisis.
The first is timing calibration. Markets are generally better at assessing whether an event will occur than precisely when. In the run-up to the February 2026 operation, Polymarket priced US military action at high probability across multiple time horizons simultaneously — before January, before March, before July. Several traders who held the correct directional bet (yes, there will be strikes) nonetheless lost money because they had also bet on a specific window that passed before the operation occurred. The markets correctly identified the basin of attraction; they were less reliable at the specific precipitation point.
The second limitation is more troubling: the insider trading problem. Investigators in Israel have indicted two individuals, including a military reservist, for allegedly using classified information to place Polymarket bets during the Iran conflict. CNN reporting identified a trader who earned nearly $1 million from dozens of well-timed bets that correctly anticipated US and Israeli military operations, with positions entered hours before strikes occurred. The question this raises for the forensic use of prediction market data is serious: when a market moves sharply and correctly in the hours before a covert military operation, is that movement revealing genuine crowd wisdom — or reflecting the positions of someone with nonpublic access? The epistemological value of prediction markets as uncontaminated aggregators of public information is compromised if their prices are being moved by participants exploiting classified sources.
The third limitation is the nuclear escalation case. Polymarket offered markets on the probability of a nuclear detonation by year-end, which at one point reached approximately 22 percent. Those contracts were subsequently archived following public criticism. The problem with very-low-probability, catastrophic-consequence events is that prediction markets are systematically less reliable at pricing them: the liquidity is thin, the reference class of prior events is small or nonexistent, and the market-maker problem is acute (who takes the other side of a contract that resolves YES in a world where financial markets may not be functioning?). For scenarios at the tail of the probability distribution — the events that intelligence communities most urgently need to forecast — prediction markets offer the least reliable guidance.
The Intelligence Community's Structural Disadvantages
To understand why prediction markets outperformed professional analysts on specific Iran questions, it helps to understand the structural constraints under which intelligence assessments are produced. The intelligence community's internal prediction market — which ran on a classified network with over 4,000 participants and 190,000 trades — was quietly shuttered, in part because its results conflicted with official assessments in ways that created institutional discomfort. Bureaucracies reward consensus and punish dissent; prediction markets reward accuracy and punish overconfidence regardless of institutional position.
A National Intelligence Estimate on Iranian intentions must satisfy multiple stakeholders, survive interagency coordination, and be written in language that protects the sources and methods that underpin it. The result is often hedged to the point of uselessness for decision-makers who need to know not that Iran "could potentially" pursue a certain course of action but what probability they should assign to that course of action within a given time window. Prediction markets answer exactly that question, in real time, with financial skin in the game.
The Military Intelligence Professional Bulletin published an analysis in late 2025 — aptly titled "The Market Knows Best" — arguing that prediction market data should be integrated into standard intelligence products as a systematic complement to classified source reporting. The Council on Foreign Relations similarly acknowledged the institutional shift underway, noting that institutional investors, policy analysts, and intelligence professionals now routinely reference Polymarket and Kalshi odds alongside traditional forecasting. This is not a marginal development; it represents a quiet recalibration of how geopolitical probability is produced and consumed at the professional level.
The Integration Imperative: A Practitioner's Assessment
The most intellectually honest position for an analyst to hold, having examined the Iran crisis as a forecasting case study, is that prediction markets and intelligence estimates are not competitors — they are complements with different failure modes. Intelligence agencies have privileged access to signal intelligence, human reporting, and technical collection that no public market can replicate. They are irreplaceable for understanding adversary intentions, internal deliberations, and the specific operational details that shape outcomes. But they are structurally disadvantaged at aggregating the distributed, probabilistic judgment of large numbers of participants who have financial incentives for accuracy and no institutional incentive to shade their estimates toward the politically safe consensus.
The practical implication is integration. An intelligence professional tracking the Iran nuclear file in early 2026 who was also monitoring the Polymarket contract on US military action was receiving a consistent signal — one that the market was pricing at 58 to 71 percent when public expert commentary was still broadly skeptical — that warranted serious analytical weight. The question is not whether to trust the markets blindly but whether to ignore them, which is itself an analytical choice with consequences. For analysts and policy professionals seeking real-time probability assessments on geopolitical events, resources such as PolyMarket Predictions have become part of the standard toolkit — not because they replace judgment but because they encode, in a continuously updated numerical form, the aggregate judgment of a large and financially motivated forecasting community.
The Iran crisis will generate considerable post-mortem analysis in the years ahead. Some of that analysis will focus on the policy failures and the military decisions. Some will examine the diplomatic breakdowns and the ceasefire's fragility. A smaller but genuinely important strand of that analysis should examine the forecasting record — because understanding where the collective intelligence of open markets converged with, or diverged from, the classified assessments of professional analysts is not merely an academic exercise. It is a guide to improving the epistemic infrastructure of national security decision-making before the next crisis arrives.
Dr. Karim Nassar is a senior analyst at mideastintel.com specializing in Iranian strategic doctrine, Gulf security architecture, and geopolitical forecasting methodologies. Views expressed are those of the author.