Bloomberg analysis: longshot bets that lose 98% dominate Kalshi and Polymarket

Bloomberg analysis published on 7 October 2026 finds users of Kalshi and Polymarket mostly trade "longshot" contracts that lose approximately 98% of the time. The trading pattern on both platforms shows a sustained preference for low‑probability outcomes, turning much of the volume into high‑loss bets rather than clean market signals.

## The key finding: longshot bets and the 98% loss rate

Bloomberg examined trading behaviour on Kalshi and Polymarket and documented that activity is dominated by contracts paying out only if an unlikely event occurs. Those contracts end up "not paying" roughly 98% of the time. The analysis does not attribute responsibility to a single user group; the preference for longshots appears consistently across participants on both platforms.

"The volumes reflect more of a betting dynamic than a price‑discovery process," Bloomberg reported about the consolidated data from both platforms.

This matters because Kalshi and Polymarket publicly position these markets as information aggregators or "truth machines" that, according to promotional material, improve forecasting by harnessing collective intelligence. If most volume comes from arguably irrational wagers on very unlikely outcomes, the emerging price signal would not be an efficient probability estimate but noise.

## How this reshapes risk assessment for operators and regulators

Operators and regulators will need to revisit risk profiles associated with retail‑accessible prediction markets. The fact that low‑probability contracts lose 98% of the time provides a technical argument to treat these platforms as products behaving like casino gambling rather than market instruments designed to discover probabilities.

Regulators considering retail access to event contracts could cite the pattern as evidence that the product's practical function is betting: volume driven by the pursuit of large payouts at low probability. Operators that provide the interface, product portfolios and user experience must decide whether their model seeks to attract sharper flow—clients who trade on information—or to monetise the entertainment flow that currently generates the repeated outcomes.

## Practical implications for Kalshi, Polymarket and their products

Both platforms face a product and marketing dilemma. Maintaining an information‑market narrative limits legal defensibility if data show an overwhelming bias toward losing longshots. Alternatively, acknowledging an entertainment function could require adjustments in advertising, product categorisation and customer‑protection safeguards.

Concrete options include:

• 1. segmenting the product between contracts with high informational value and contracts positioned explicitly as entertainment;

• 2. changing price presentation and adding warnings for retail users;

• 3. implementing limit policies and risk‑management tools designed to reduce repeated exposure to longshots.

If Kalshi and Polymarket choose to attract sharper flow, they will need to show incentive mechanisms and products that reward informed trading. If they exploit entertainment demand, regulators may require compensating consumer‑protection measures.

## Why traders seeking market signals are disadvantaged

When most volume comes from bets that fail 98% of the time, the price set of probabilities becomes contaminated. A trader using quotes as a probability estimator faces a crowd that systematically seeks unlikely outcomes and pays a premium for them. That crowd does not correct market errors; it reinforces them.

The practical result is that prices may reflect preference for payout rather than statistical plausibility. Anyone treating Kalshi or Polymarket as a source to improve a forecast encounters a structural headwind: aggregation does not turn noise into signal when most noise responds to betting biases.

## What operators, vendors and compliance officers should watch

For operators and technology providers the warning is clear: risk profile and product architecture should align with actual user behaviour. Areas to monitor include:

• • traceability of the origin of volume flow;

• • segmentation and labelling of contracts by purpose (informational vs recreational);

• • customer‑protection measures in products with a high probability of repeated loss.

Compliance teams preparing documentation for auditors or regulators will need to present granular user metrics that show how much volume corresponds to longshots, the frequency of retail customer participation and the economic consequences for those players.

## What regulators will point to and what to expect in the market

Regulators debating retail access to prediction markets could use Bloomberg's analysis to justify stricter consumer‑protection requirements or even product limits. The argument will be technical: if the product operates in practice as gambling, its licensing, marketing and controls should reflect that reality.

Commercially, Kalshi’s and Polymarket’s responses will shape competitive futures. Their product and marketing choices will reveal whether they try to attract sharper flow or accept positioning as entertainment platforms that generate revenue from user demand for longshots.

For operators and institutional clients, the immediate operational recommendation is to review flow‑quality metrics before relying on market prices as probability signals. For regulatory teams, the analysis provides an empirical sample that can be used in supervision and rule‑design.