Skip to content
All legal documents

API Disclaimer

Accuracy, methodology, and Kalshi-integration disclaimers for the SplitWinner API.

Version 1 · Effective June 13, 2026

SplitWinner API Disclaimer

Last updated: June 2026

This disclaimer applies to the SplitWinner API at api.splitwinner.com. It is in addition to the API Terms, not in place of them. Where this disclaimer and the API Terms address the same subject, the API Terms control.

1. No guarantee of accuracy

SplitWinner predictions are probabilistic forecasts generated by machine-learning models. They are not guarantees, not investment advice, not betting advice, not financial advice, and not legal advice. Past model performance is not indicative of future results. Sports outcomes are inherently uncertain.

We provide the API AS IS and AS AVAILABLE, without warranty of any kind, express or implied, including warranties of merchantability, fitness for a particular purpose, accuracy, non-infringement, uptime, or that the API will be uninterrupted or error-free.

The API may return outputs that are incorrect, late, incomplete, or inconsistent. You are responsible for any decision you make using the API and for any losses — including financial losses from sports betting or related activity — that result from those decisions.

2. Model methodology

SplitWinner uses statistical and machine-learning methods to estimate probabilities for sporting events. You should understand:

  • Models are statistical tools that estimate probabilities — they are not guarantees of outcomes.
  • Model accuracy varies by sport, season, sample size, data quality, and market conditions.
  • Predictions may be based on incomplete, delayed, or inaccurate third-party data.
  • No prediction, regardless of confidence level, should be treated as a certainty.
  • Models are retrained periodically and may produce different results over time.

The substantive methodology — what we do, what the public audit-trail proves, and what it does not — is documented at splitwinner.com/verification. The public ledger is available at github.com/SplitWinner/audit_trail.

3. Performance data

Any performance statistics returned by the API or displayed in SplitWinner documentation (win rate, ROI, units, accuracy, calibration, and similar metrics) reflect historical outcomes. They are provided for transparency and do not predict or guarantee future results.

CFTC Rule 4.41 — Hypothetical performance disclosure

Hypothetical or simulated performance results have certain limitations. Unlike an actual performance record, simulated results do not represent actual trading. Also, since the trades have not been executed, the results may have under-or-over compensated for the impact, if any, of certain market factors, such as lack of liquidity. Simulated trading programs in general are also subject to the fact that they are designed with the benefit of hindsight. No representation is being made that any account will or is likely to achieve profit or losses similar to those shown.

4. Sandbox is synthetic

The free sandbox tier returns mocked data. Sandbox responses match the response shape, HTTP semantics, and value formats of the paid API, but:

  • The values themselves are synthetic. Identifiers, timestamps, scores, odds, confidence numbers, audit hashes, and aggregates are generated for integration testing only — they are not predictions.
  • Documented integration coverage is approximately 90%. Edge cases (rate limits under load, slate-day boundary edges, upstream-error pass-throughs, conformal-coverage tails) may differ from production.

Sandbox responses must not be presented to any third party as actual SplitWinner predictions, redistributed or sold, or used to make real betting, investment, or operational decisions.

5. Kalshi integration — what we do and do not do

SplitWinner provides prediction and market-match data via /v1/kalshi/match and /v1/kalshi/markets. For enterprise B2B callers, those endpoints return matched tickers and team mapping only — price, volume, and similar live-market fields are stripped server-side.

We do not:

  • Sign Kalshi API requests on your behalf
  • Proxy live Kalshi prices to enterprise callers
  • Place, route, or settle any Kalshi order
  • Hold your Kalshi credentials, including your Kalshi RSA private key, on our servers

You do:

  • Maintain your own Kalshi account in compliance with Kalshi's Terms of Service and applicable rules.
  • Fetch live market prices from api.elections.kalshi.com using your own credentials on your own infrastructure.
  • Place every order yourself, from your own systems. Your Kalshi audit trail stays on Kalshi.com.

SplitWinner is not registered with the U.S. Commodity Futures Trading Commission (CFTC) and does not act as a designated contract market, futures commission merchant, introducing broker, commodity trading advisor, or commodity pool operator. Kalshi, Inc. is a CFTC-regulated designated contract market; your relationship with Kalshi is governed by Kalshi.

6. Third-party data

SplitWinner's predictions are derived from publicly available third-party data sources. We do not guarantee:

  • The accuracy, completeness, or timeliness of any third-party data
  • That data sources will remain available or consistent
  • That statistical information has not been altered, delayed, or omitted by the original source

7. No professional relationship

Use of the SplitWinner API does not create a fiduciary, advisory, broker-client, or any other professional relationship between you and SplitWinner. Nothing the API returns is personalized advice tailored to your specific financial situation.

8. Downstream responsibilities

If you redistribute or surface SplitWinner outputs in a customer-facing product, you are responsible for conveying these disclaimers to your end users in a clear and prominent manner. The redistribution and resale restrictions in the API Terms §3 apply in full.

Contact