How We Evaluate Prediction Market Platforms
Answer first
We score prediction market platforms with seven continuous factors. Contract resolution quality and execution and liquidity each receive twenty percent. Eligibility, regulatory, and geographic fit and fees and total cost each receive fifteen percent. Market breadth and relevance, funding, custody, and settlement, and tools and decision support each receive ten percent.
The category-relative v2 method is contract-aware and venue-aware. Each factor can take any evidence-backed value from zero to one hundred. Missing evidence is stored as unknown and blocks publication of the overall category score; it never becomes numeric product weakness. Evidence confidence and current legal, geographic, account, and contract eligibility remain separate from product quality. Product claims stay dated August 22, while this score calibration was reviewed August 25, 2026.
Decision table
| Decision | Practical guidance | What to verify |
|---|---|---|
| Is the user and contract eligible? | Fail the recommendation when official current evidence does not support access. | Check the exact venue, jurisdiction, account type, age, residency, and contract restriction. [8] [9] [6] |
| Can the order be executed? | Score spread, depth, quote quality, and exit conditions at a defined size. | Use current order-book evidence rather than a platform-wide volume claim. |
| Can the rules be explained? | Score definitions, source hierarchy, deadlines, corrections, disputes, and settlement clarity. | Read the exact contract and governing rules, not only a help-page summary. [1] [3] [10] |
| Can total cost be priced? | Score the actual fee model and in-app disclosure for the order. | Check category, probability, expected earnings, exchange, account program, spread, funding, and exit cost. [11] [5] [7] |
Best for and avoid if
Best for
- Readers who want to see which evidence and trade-offs sit behind a platform comparison.
- Editors updating fees, access, and contract mechanics without merging different venues.
- Users who compare one executable contract at a time instead of ranking brands by headline volume.
Avoid if
- You want a platform score to decide whether an event outcome will occur.
- You expect one legal or fee statement to cover every user and contract.
- You want recent volume, promotions, or affiliate economics to determine the score.
Source and venue hierarchy
The preferred evidence is the governing regulator record, exchange rule or filing, current fee page, official eligibility page, contract rule, and current product screen. Broker documentation can support how a client reaches an exchange, but it does not replace the exchange's contract rules. Current reporting can explain adoption and distribution, but it does not settle legal access or contract mechanics.
Every fact is bound to a venue. International Polymarket protocol documents apply to that protocol. A Polymarket US filing applies to the named US contract family. Kalshi help and filings apply to Kalshi products under their current scope. Interactive Brokers pages describe eligible client access and published schedules that can vary by program. Unclear required evidence is stored as unknown and blocks publication of the overall category score; it is never converted into a low capability value.
Eligibility, regulatory, and geographic access: separate hard gate
Eligibility is not part of the product-quality score. Before using any numeric score, check the platform entity, exchange designation where relevant, user country or state, age, residency, identity requirements, account affiliate, funding eligibility, and contract-specific restrictions. A platform can be lawful in a framework while a particular user or market remains unavailable.
A failed or unresolved eligibility check blocks a recommendation for that route without adding or subtracting product-quality points. The article does not suggest workarounds, location masking, or use of a different venue's documentation. Polymarket's geographic-restrictions page, Kalshi's international help, and IBKR's eligibility notes show why current access must be verified independently of the score.
Execution and liquidity: 25 percent
Liquidity is measured at the contract and intended order size. We review best bid and offer, spread, depth at several levels, recent executed activity, likely market impact, and whether an exit can be placed before settlement. Platform-wide volume receives little credit because it can be concentrated in unrelated markets.
A market can show a probability-like midpoint without an executable price there. Limit orders can reduce price uncertainty but add non-fill risk. A wide spread is a direct cost. The score uses a repeatable size scenario and records when the comparison was taken, since event-market depth can change sharply around news.
Contract resolution quality: 25 percent
Strong rules define the outcome, exclusions, measurement period, time zone, end date, resolution source, source hierarchy, correction treatment, void conditions, settlement timing, and dispute process. The method asks whether a reasonable user can distinguish a winning real-world thesis from a winning contract interpretation.
Polymarket international documents predefined rules and an optimistic-oracle dispute path. A reviewed Polymarket US filing describes official source hierarchies for one contract family. Kalshi product filings tie settlement to exchange rules and specific contract terms. ForecastEx education centers published outcomes. Each mechanism is scored on its own documentation.
Fees and total cost: 15 percent
We calculate cost from the order, not from a promotional table. The review includes exchange or protocol fees, broker commissions, spread, market impact, funding or payment costs, crypto conversion or network costs, withdrawal or intermediary costs, and a second spread if the user exits early. A winning one-dollar settlement is not the same as one dollar of profit.
Polymarket documents category and probability-dependent taker fees for certain markets. Kalshi documents expected-earnings-based fees and possible maker fees. IBKR lists per-contract charges by exchange and account context. Since all can change, current primary evidence and the in-app preview control. Unsupported precision fails.
Market breadth and relevance: 10 percent
Breadth means relevant, well-defined, tradable contracts rather than a large page count. We review topic coverage, contract cadence, duration, repeated listings, clarity, and whether the markets a target user needs have usable order books. A venue with fewer clear and liquid contracts can outrank a broad but thin catalog.
We also separate direct listings from distributed access. An app can display contracts from another exchange. That may improve convenience without adding an independent order book or rule set. The page records the underlying venue so apparent breadth is not counted twice.
Sources for this section: [6]
Funding, custody, and settlement: 10 percent
This category identifies the collateral currency or asset, funding methods, holder of funds, withdrawal process, settlement credit, recovery, and operational dependencies. Crypto-native wallet flows and brokerage cash accounts have different risks. Neither receives automatic preference. The score rewards clear current documentation and a usable withdrawal path.
Polymarket international says it does not charge for USDC deposits or withdrawals while intermediaries may charge. US DCM filings can describe full at-risk collateral. IBKR provides contracts inside a brokerage account. The method keeps these models separate and never infers one from another venue with a related brand.
Tools and decision support: 15 percent
We review limit and marketable order behavior, fee-inclusive quotes, contract search, rule visibility, watchlists, alerts, charts, data export, API access, and tools for closing or scaling. A tool earns points when it improves an informed decision. Faster order placement without clear rules or cost can reduce the practical score.
IBKR says its prediction-market app can identify favorable fee-inclusive prices for identical contracts across exchanges. Its documentation also discusses ScaleTrader for ForecastEx. Polymarket publishes developer documentation for markets, orders, and redemption. Tool availability still needs a current account and contract check.
Scoring, tie breaks, and refresh
The six weighted product-quality factors sum to one hundred percent. Eligibility is a separate hard gate with no numeric weight. Each scored factor uses the full continuous zero-to-one-hundred range, then the audited weight is applied. ROUND_HALF_UP is applied after weighted arithmetic. If required evidence is unknown, the factor has no numeric value and publication of the overall score fails closed. We do not add points for affiliate payouts, sign-up offers, audience size, or unverified accuracy marketing.
A refresh rechecks the official entity, eligibility, fee schedule, contract examples, rules, funding, and product tools. High-drift facts are checked again at release. If a current primary source is missing or conflicts with another venue's documentation, the factor is stored as unknown and the overall category score is withheld. The method does not average incompatible claims or assign a numeric product penalty for missing evidence.
Decision checklist
- Bind the venue: Record exchange, distributor, product, rule set, and jurisdiction for every factual claim.
- Test eligibility: Fail closed on user or contract access that current official evidence does not support.
- Capture the book: Use a dated spread and depth snapshot at the defined comparison size.
- Read the contract: Extract outcome, exclusions, source, deadline, corrections, void terms, and settlement process.
- Price all costs: Include fee, commission, spread, impact, funding, conversion, network, withdrawal, and early exit.
- Score all factors: Keep the evidence, continuous factor value, audited weight, and rationale for each of the six product-quality factors; record eligibility separately as a hard gate.
- Rebuild before release: Re-run deterministic link, source, schema, depth, visual, and action checks.
Risk review
| Risk | Why it matters | Control |
|---|---|---|
| Scope leakage | A rule for one venue or contract family can be applied to another. | Name the venue and scope beside every legal, fee, or resolution claim. |
| Non-executable pricing | A midpoint or last trade can be mistaken for an available fill. | Score spread and depth at a defined size and time. |
| Incomplete rules | A title can hide definitions and source hierarchy that decide settlement. | Read and record the complete contract before scoring. [1] |
| Stale fees | Fee structures can vary and change by market or account. | Use the current primary schedule and in-app preview. [11] [5] [7] |
| Commercial bias | Promotions and affiliate payouts can distort rankings. | Exclude them from scoring and include no referral URLs. |
These controls reduce avoidable errors. They do not remove market, contract, custody, or legal risk.
How this page was evaluated
The scorecard uses the category-relative v2 continuous 0-100 model, calibrated August 25, 2026. Each required factor receives an evidence-backed score across the full range, then its audited vertical weight is applied. Missing evidence is stored as unknown and blocks publication of the overall category score; it never becomes numeric product weakness. Score confidence and current user eligibility are reported separately. This page contains the full rubric. A category score is a decision aid, not universal user suitability or a promise of profit or platform safety. We recheck volatile product, fee, access, and rule facts against the cited pages before release.
| Criterion | Weight |
|---|---|
| Contract Resolution Quality | 25 |
| Execution Liquidity | 25 |
| Fees Total Cost | 15 |
| Market Breadth Relevance | 10 |
| Funding Custody Settlement | 10 |
| Tools Decision Support | 15 |
Frequently asked questions
How are prediction market platforms scored?
The continuous product-quality model gives twenty-five percent each to contract resolution quality and execution and liquidity, fifteen percent each to fees and total cost and tools and decision support, and ten percent each to market breadth and relevance and funding, custody, and settlement. Eligibility, regulatory, and geographic access is a separate hard gate with no numeric weight.
Can an ineligible platform still rank first?
No for that user and contract. Eligibility is a blocking gate, so a high product score cannot override current legal, geographic, age, residency, or account restrictions.
What happens when prediction-market evidence is missing?
The affected factor is marked unknown with no numeric value, and the overall TM category score is withheld until the required evidence is available. Missing documentation never lowers the product-quality score, and current user eligibility still fails closed.
Does platform volume determine liquidity score?
No. Liquidity is measured at the exact contract and intended size. Platform-wide volume can be concentrated elsewhere and receives no category-quality pass by itself.
Sources checked
- Polymarket ResolutionScope: Polymarket's international protocol documentation says markets use predefined resolution rules and UMA's optimistic oracle, with a two-hour challenge period and a dispute path that can take four to six days. Caveat: These mechanics apply to the international protocol documentation. Do not apply them to Polymarket US DCM contracts without contract-specific proof.
- CFTC Trading Organizations: Designated Contract MarketsScope: The CFTC lists QCX LLC doing business as Polymarket US as a designated contract market, designated July 9, 2025. Caveat: The designation does not make every international Polymarket market a CFTC-regulated US contract or prove universal user access.
- Polymarket US Athletic Tie Contract FilingScope: The filing describes one Polymarket US athletic tie contract family with $1 contracts, full at-risk collateral, official source hierarchies, settlement procedures, and participant restrictions. Caveat: This filing concerns a specific contract family. Do not generalize its source, restriction, or settlement language to every Polymarket US product.
- How Is Kalshi Regulated?Scope: Kalshi says it is a CFTC-regulated designated contract market, and the CFTC DCM list independently confirms Kalshi's designation. Caveat: DCM status does not remove contract-specific, state, age, eligibility, litigation, liquidity, or loss risk.
- Kalshi FeesScope: Kalshi says transaction fees depend on expected earnings and that some markets can also have maker fees, with the applicable fee displayed in the product and current schedule. Caveat: Do not reduce Kalshi fees to one universal percentage. Verify the exact market, side, size, account, and current in-app quote.
- IBKR Prediction MarketsScope: Interactive Brokers provides one interface for event contracts from Kalshi, ForecastEx, and CME, says eligible clients must be at least 21, and limits ForecastEx election contracts to eligible US residents. Caveat: Eligibility varies by affiliate, country, account, and contract. A broker interface does not make every listed contract a ForecastEx product.
- Interactive Brokers Prediction Market CommissionsScope: IBKR's direct-client schedule lists no IBKR commission plus a one-cent exchange fee for ForecastEx contracts, and one-cent IBKR commission plus a one-cent exchange fee for Kalshi and CME contracts. Caveat: Rates can change and introduced or advised accounts may pay more. Verify the exact fee-inclusive in-app quote.
- Polymarket Geographic RestrictionsScope: Polymarket publishes geographic restrictions for access to its international product and instructs users to follow applicable location rules. Caveat: Restrictions can change and public market data does not prove trading eligibility. Verify actual current location and product access without using a workaround.
- Can I Trade on Kalshi from Outside the United States?Scope: Kalshi says it is available in many countries subject to its member agreement, identity verification, residency, local restrictions, and differing payment options. Caveat: Broad availability language does not prove that every country, payment rail, account, or contract is eligible. Verify current access in-app.
- Kalshi CONTROL Product CertificationScope: The product certification illustrates that Kalshi contracts use contract-specific definitions, source language, trading conditions, and settlement procedures. Caveat: This filing is evidence for its own contract family, not a universal description of every Kalshi market. Read the exact contract rules.
- Polymarket FeesScope: Polymarket says makers are not charged fees, taker fees vary by category and probability, and geopolitics or world-events markets are fee-free. Deposit or withdrawal intermediaries may still charge. Caveat: Fee parameters can change by market. Keep international protocol fees separate from Polymarket US products and verify the current order quote.
- Economic Forecast Contracts from ForecastExScope: IBKR course material describes ForecastEx contracts designed around published economic, financial, and climate outcomes and their settlement research workflow. Caveat: Course material is educational and contracts vary. Read the current named source, period, threshold, correction rule, and settlement terms.
- IBKR Prediction Markets AppScope: IBKR says its prediction-market app can identify a favorable fee-inclusive available price for identical event contracts offered by different exchanges. Caveat: The routing claim applies only after contracts are proven identical. Similar topics with different sources, deadlines, or settlement terms are not interchangeable.
- Using ScaleTrader for ForecastEx ContractsScope: Interactive Brokers documents ScaleTrader as a way for eligible users to stage ForecastEx contract orders across price levels. Caveat: Scaled orders can create partial exposure, non-fill, and a changing average price. The page's cost language does not replace the current fee-inclusive quote.
