Kalshi · current contracts
15-minute markets compare the final reference average to a target. Hourly contracts ask whether the price finishes above a specific strike.
Each winning contract pays $1. A sixty-cent ask needs a 60% win rate to break even before costs. YES and NO asks can add to more than $1 because they are separate offers. The bars show quoted prices.
Finding the current market…
Waiting for visible depth
Waiting for a snapshot
Finding the current market…
Waiting for visible depth
Waiting for a snapshot
Finding the current market…
Waiting for visible depth
Waiting for a snapshot
CF Benchmarks reference feeds via Kalshi. Settlement follows each contract’s final-minute average rules. Mini charts show observations from this visit, with a two-minute, automatically scaled view.
The line follows the halfway price between the best YES bid and ask. The shaded band shows the space between them. Switch to Spread to see how wide that space was. The tape keeps the latest observed quote per receipt second during this visit. Gaps over 12 seconds break the line. Each contract starts a new tape.
Kalshi / 15-minute crypto
Follow a contract from its quoted price to its official outcome. Choose a price band to see what happened.
Each bar groups contracts by the closing YES ask of the exact one-minute candle ending at the selected minute. Bar height is the share that officially settled YES. The diamond shows the break-even win rate for buying one contract at each observed ask and collecting $1 on each YES, before fees. A 50¢ average ask gives a 50% break-even rate. The displayed gap subtracts that rate from the observed YES share, in percentage points. Each price band includes its lower bound and excludes its upper bound. Missing, duplicated, invalid, 0¢ and 100¢ quotes are excluded.
These candles were archived after settlement. The percentages describe these dates only. Quote size, fills and fees are unknown, and contracts in nearby windows can move together.
CoinMarketCap
Aggregate spot prices, returns and rolling volume across the wider market.
Loading shared market context…
Global context checked hourly
Waiting for source data
Waiting for source data
Waiting for source data
Changes and volume describe broader periods, not fifteen-minute order flow. Recorded samples preserve gaps; sentiment and past performance do not predict the next contract.
Explore the break-even map, reference price and live order book
Two or three assets · a shared expiry · joined history · public combo books where quoted
Waiting for market & reference feed
Available during the final 60 seconds
For the observed 60-second mean to reach the YES target. Requires every elapsed second. Uses our latest observation per second; official sampling and rounding may differ.(60 × target − sum of completed observed seconds) ÷ remaining seconds. Missing elapsed seconds suppress the calculation. Each contract’s settlement rules govern the official outcome.
Server-observed average, not the official settlement value. We use the latest received index value in each completed source second of the final minute, weight those seconds equally and omit gaps. Coverage starts when this server receives data; restarts and feed interruptions can leave gaps. Kalshi’s market rules and official determination govern settlement. How settlement works ↗
Coverage appears as source seconds complete.
TradingView Lightweight Charts™ · Copyright © 2025 TradingView, Inc. · TradingView · License
Waiting for the opening book.
Relative size at each side’s best bid.
A winning contract pays $1. At a 54¢ YES ask with 2¢ of assumed costs, YES needs to win more than 56% of the time for positive expected value. A 49¢ NO ask under the same assumption needs YES to win less than 49% of the time. The middle band includes break-even ties.
Bid-size balance equals 100 × (YES size − NO size) / total best-bid size. Positive values mean more resting size on YES. Orders can disappear. This measures displayed liquidity and has no validated win rate. The cost input is your scenario assumption. Actual fees and execution prices may differ.
One square. Fifteen minutes. An official result.
Past results describe this window; they do not predict the next contract. Hindsight quote analysis excludes fees and does not establish fill availability.
These 15-minute markets compare the CF index average from the minute before closing with the average from the minute before opening. The target above is supplied by Kalshi. The current index tick is not the final settlement value.
The individual market rules and contract terms govern the outcome. Gaps in our observations prevent us from reconstructing an exact settlement average. Read Kalshi’s explanation ↗
Connecting to Kalshi…
Two or three picks. One shared finish.
Choose UP or DOWN. Percentages are individual ask prices as a share of $1, before fees.
UP is YES at or above the target; DOWN is NO below it. Every selected leg must win for a $1 combo payout. Picks must share the same 15-minute window. Asset prices are never multiplied into a combo price.
Choose directions to inspect shared history.
BTC, ETH and SOL can move together. Adding picks does not increase the chance of winning. Historical matches do not price a new combo or predict the next window.
A shared research canvas for BTC, ETH and SOL. Inspect official outcomes, compare recorded market context, and see exactly where the evidence ends.
Loading recorded observations…
Loading recorded spot prices…
Each point is a recorded aggregate spot price. The band is a 20-sample moving mean plus or minus two population standard deviations. RSI uses simple sums of gains and losses over 14 observed changes. Realized volatility is the square root of the sum of squared log returns, in percent over those changes. Sampling intervals vary. A gap over 11 minutes resets the calculations. The scrubber uses only observations through the selected time. Kalshi settlement uses its specified reference index.
Brier error is the mean squared distance between each YES probability and the result (1 for YES, 0 for NO). Lower is better. All three methods use the same contracts. Several contracts share a closing time.
| Band | Contracts | Forecast | Settled YES |
|---|
Lower Brier error means probabilities were closer to official outcomes. Assessing returns also requires executable prices, fees, liquidity and a fixed selection rule. The live test saves forecasts before results and compares methods on identical scored contracts. Missing forecasts and pending outcomes remain visible in coverage. Daily reports are refreshed by the operator.
Open a private daily paper report to inspect each decision at the opening, follow the later book and compare outcomes. The file stays in this tab.
Each column closes at the same time across all three assets. Select a cell to inspect the source records and open “Before this outcome.” Missing results stay in place.
Each bar shows the share of next windows that settled YES after a YES or a NO. Only adjacent windows with both official results count. The difference describes this selected history and can swing sharply with few pairs.
Select a completed outcome cell above, then choose a time within that window. Later intervals are excluded. Historical downloads are labeled separately from saved collection timestamps.
Select an outcome cell to begin.
Up to 24 hours of stored CMC aggregate prices, rebased to 0% at the first shared sample. This is market context, separate from the CF settlement index.
Correlation uses consecutive synchronized sample returns, never price levels. Gaps and unchanged source timestamps are excluded. +1 means moving together, −1 means opposite. A short sample describes this period; it does not predict the next contract.
Blockworks aggregate five-minute candles provide a separate market perspective. These are neither the CF settlement reference nor Kalshi contract prices.
Source: Blockworks · Powered by Blockworks · Saved history pilot; no live collection.
Blockworks history is not connected.
Coverage belongs beside every chart. An absent or stale source is a data condition, not a zero.
| Source | Available in this view | What it measures | Source / retrieval time |
|---|
Read-only research. Alignment frequencies are historical descriptions, not probabilities for a new window. Export contains only the displayed Kalshi archive summaries and CMC response; it is not a complete tick archive or a point-in-time training dataset. All source gaps are preserved.