How Odds Aggregation Works on AllTop
When you look at a market on AllTop, you're not seeing one platform's price. You're seeing the same real-world question priced by multiple prediction market venues, pulled into a single view, with the best available price flagged for each outcome.
This article explains where that data comes from and how it gets onto the page.
How does AllTop combine venues into one market?
Prediction market platforms list contracts independently. Kalshi and Polymarket might both run a market on the same Fed rate decision, but they list it under different titles, different contract structures, and different prices.
AllTop's backend collects each venue's version, then clusters provider markets covering the same real-world event into one canonical event. That single event is what renders on the page: one question, with a per-venue price table underneath it showing each platform's current odds.
The matching happens three ways. Title patterns identify obvious matches automatically, a scoring matcher handles less obvious ones, and our editorial team curates links directly where automation isn't confident. The result is that a market page reflects a real-world event, not a single platform's listing.
Once a market is matched to an event, we align it with the news covering that same event, so the price and the story it belongs to sit together. How that connection is made is covered in how a news story links to a market.
Which venues does AllTop cover?
AllTop currently ingests live market data from Kalshi, Polymarket, and Polymarket US. Other platforms appear on AllTop in editorial coverage, including our platform reviews and the regulatory table on the markets hub, but their odds are not part of the live data feed. When you see a best price on AllTop, it means the best price among the venues we ingest, and we say so on the page.
How current is the data?
Two systems run side by side. The backend ingests each venue's API feed in near real time, cleans the data, and sorts it by venue. Scheduled jobs handle the slower-moving work: refreshing each venue's full market catalog every few minutes, pulling price history, and rolling up volume totals hourly. The page you're looking at also refreshes itself, every 15 seconds on single-market pages and every 30 seconds on game pages.
Every market page carries a data status label, such as "LIVE · updated 5m ago," so you can see exactly how current the numbers are. If a venue's feed degrades or goes stale, the label says so instead of pretending otherwise.
Why compare venues at all?
Collecting the same market from multiple venues does a few things a single platform can't.
First, it surfaces the best available price per outcome, since venues frequently price the same event differently.
Second, it shows you the spread between venues, which is itself useful information: a wide spread often means the market is moving on news and the venues haven't converged yet.
It provides a holistic view of how the market is pricing a real-world event to happen (or not).
The aggregation doesn't factor trading fees into the best-price comparison. Fees vary by venue and are shown as separate disclosure on the page, so check them before treating the highlighted best price as the best net payout.