Read the statistics behind each round.

Shares and traded value are different totals

Share volume adds the number of outcome shares in recorded trades. Traded value adds price multiplied by share size, expressed in the quote currency. Ten shares at 20¢ represent $2 of traded value; ten shares at 80¢ represent $8. Equal share counts therefore need not mean equal traded value.

A larger traded-value share on Up describes more recorded turnover on that token in the selected round. It does not identify net buying, prove that more traders support Up, or establish a probability of winning: every matched trade has a buyer and a seller.

Traded value = Σ(price × shares)

VWAP: the average price paid per share

VWAP means volume-weighted average price. Weighting each execution by its share size makes a large trade count more than a small one. In Omens, the terminal label “Avg. traded price” presents this quantity in cents.

VWAP = total traded value ÷ total traded shares

Why the round window matters

The round summary uses available executions across the round window, independently of the chart’s zoom. A chart showing only the final minute should not redefine the full-round total. The platform’s event lifetime volume, a 24-hour total and Omens recorded round value can differ because they cover different periods and sources.

Comparisons require compatible sources and units. Missing or incomplete data should remain identifiable rather than becoming a zero. A blank value and a measured zero have different meanings.

Ask a precise question about past outcomes

For example: among eligible five-minute Bitcoin rounds, how often did Up settle when its sampled price was in the 70–80¢ band with one minute remaining? Fix the asset, round duration, settlement source, checkpoint and price band before looking at the answer.

Omens round research groups supported historical samples at 120, 60 or 30 seconds before close. Its current checkpoint method uses the latest qualifying chain settlement in the preceding five-second window. Chain settlement time and websocket execution time are different clocks, so these samples should not be treated as identical to a live last-trade quote.

A band displays a rate after at least 20 eligible rounds. That is a display threshold, not a guarantee of reliability. Inspect the count, period and exclusions alongside the percentage.

Historical settlement frequency = matching settled outcomes ÷ eligible sampled rounds

Inspect price and final outcome samples

Check what was left out

Pending results, unscanned windows, unavailable source history and checkpoints without a recent qualifying settlement are excluded. The sampled rounds can therefore differ from the full market population. More active markets may be easier to sample.

Each round can appear once for each side, but Up and Down from the same round are not independent observations. Repeated markets can also share the same news or price move. A larger row count does not automatically mean proportionally more independent evidence.

Historical frequency is not a prediction for the next round, and it is not a trading return. Fees, spread, execution timing and liquidity are separate questions. Use the probability guide to understand calibration and statistical uncertainty.

Probability, calibration and sample uncertainty

Polymarket candlestick charts, built from trades.

Explore Polymarket OHLCV charts with recorded volume, side-by-side outcomes and settlement references. Understand what each candle actually measures.

What does a market price say about probability?

Learn how to compare prediction market prices with settled outcomes using calibration, sample sizes, confidence intervals and Brier scores.