Watch the data in Real-time. Become your #1 broker.
A sharper view of Bitcoin.
Refreshing reconnects market feeds. Saved history stays on this device.
The tell
—Current Kalshi target and expiry
Loading the model projection.
Model & live data details ▾
Collecting trades.
Each trade has both sides. These figures compare which side initiated it.
UP/DOWN compares the shared projected close with the contract target at its exact expiry. Every valid projection has an UP or DOWN call; a small margin is shown as context. If updates stop, the last call stays visible with its original time and expiry. EXIT REVIEW uses two opposing completed-minute forecasts; it is not a tested sell rule. This is not an instruction to buy or a proven advantage. Kalshi can settle on a different index or averaging window.
Bitcoin price chart
Waiting for exchange data
Receiving exchange trades.
No contract selected.
Read the selected candle ▾
Each candle summarizes one minute. Green means the price went up; red means it went down. The thin wick shows the highest and lowest prices.
Dashed purple = experimental future estimate. Shaded area = model range, not a guarantee.
USD flow bars match each candle’s time window: green = buyer-started value; red = seller-started value. Faded bars are partial. A grey dash means the buy/sell split was not collected, not zero trading.
The transparent line is the average of five candles, not a price forecast. Your practice entry appears when you hold virtual BTC on this exchange.
Saved history is local to this browser. Live collection runs while this page is open.
What the data says
Waiting for exchange trades
Listening to the market
Live buying and selling pressure, explained using received trades.
Each bar = 10 seconds. Grey = incomplete collection. Bar height shows the USD difference; the latest bar is forming.
Waiting for both exchanges.How to read this
Every trade has a buyer and a seller. This compares the USD value started by buyers versus sellers on the selected exchange. “More” means at least 55% for that side; 45–55% is called closely balanced. The price response is the change between the first and latest received trades inside the same rolling minute. A pressure mismatch describes these observations; it does not prove hidden orders or predict the next move.
The five-minute pace compares this minute’s USD trading value with the average per minute in the prior four minutes. Missing connections and buffer limits disable a reliable reading. The two exchanges are checked separately, never added together. No collection happens while this page sleeps or closes.
Prices waiting right now
BEST BID / ASKThe highest displayed buyer offer and lowest seller ask. These quotes can change before an order fills.
Sources & meaning
Top price level only, not full order-book depth. Displayed sizes are offers, not completed trades. Coinbase publishes ticker updates on trades; Kraken is subscribed to best-price changes. Quote freshness expires after 10 seconds and is separate from the trade feed. Quotes are context only; no orders are placed.
The big trades stay here
Completed BTC trades from both exchanges, held for 30 minutes.
Waiting for large trades.
Each row is one received trade, not a whole order or an identified whale. Saved locally on this browser; collection stops when this page sleeps or closes. Hold the list, or scroll to hold it automatically. Older saved events expire after 30 minutes.
Detailed trade reader ▾
Buying & selling activity
Completed exchange trades
Waiting for trades.
| Time window | Buyer value | Seller value | Difference | Trade count | Largest trade |
|---|
Every trade has a buyer and seller. Buy/sell here means the side that started the trade. Highlighted rows contain a single trade worth at least $100,000. Large trades do not identify a whale or a wallet; orders can be split. This is completed trading, not a list of waiting orders. Scroll to pause the table automatically.
The ghost’s closing call.
Exchange candles anchor this estimate. Shared flow and contract quotes are tested as bounded learned adjustments.
Type a target from any market, or your own price. This estimates ending above it, not touching it at any time.
Waiting for history and fresh prices.
Green = past candle closes. Purple dashed = future model median. Shading = a model-based range; actual prices can fall outside it. No real trade or guaranteed entry is implied.
How the estimate updates & learns ▾
Each completed minute updates recent price variation and a damped trend from the last 60 candle-to-candle changes. A simple log-return model extends that trend and widens uncertainty with time. It assumes returns behave more smoothly than real markets sometimes do. Candle shapes are shown for context, not assigned invented success rates.
Verified shared market ghosts adjust the common model’s price bias and range. This device’s custom checks stay separate. Learning history compares price errors with simply keeping the starting price; more samples can improve or worsen results.
Live estimates refresh from each shared completed-minute snapshot. Every eligible current-contract minute is recorded on the shared server with a frozen base-model comparison; separate custom checks remain local. Missed or late observations are excluded. This device stops collecting while asleep or closed. Shared ghosts are checked when any client requests the server; there is no always-on background scheduler. Fees are not included in target probabilities.
Kalshi base targets & 15-minute contracts ▾
Read-only public data. Your exchange spot price can differ from the contract’s settlement index and averaging window.
No Kalshi request made yet.
Contract details arrive with each shared minute; the browser checks for a new snapshot every 15 seconds while awake. The same shared snapshot drives The Tell and the current-market estimate.
Kalshi public API source ↗Learn the candle shape
FORMINGReal shapes from your selected chart.
Explore the pattern library
Waiting for two fresh feeds
Asynchronous last trades, not an executable spread.
Feed details & saved candles ＋
Freshness expires after 10 seconds without a fresh trade. Heartbeats alone cannot make prices live. Delay includes device clock differences. Missing intervals remain empty.
| Start | Start price | Highest | Lowest | Latest / end | Data coverage |
|---|
Evidence before conviction.
Current estimates share one snapshot. Original and closing calls stay fixed. Official outcomes and price errors are scored separately.
Is the model getting more accurate?
Earlier local calls & shared training records ▾
Current shared prediction loading…The live estimate above is the same calculation as The Tell. Rows below are original saved predictions, frozen at their save time; the export also retains the shared input timestamp. Tracking history is local to this browser and may differ between devices.
Saved original predictions
Verified shared ghosts ▾
One original ghost per source and market, saved before expiry. A published exchange closing candle resolves its price error; Kalshi’s official result is scored separately. Every verified ghost contributes to shared calibration.
| Recorded | Closed | Original forecast | Exchange close | Price error |
|---|
Waiting for live market data.
Waiting for a published target and fresh inputs.No official outcomes yet.
| Recorded | Contract closes | Original call | Spot estimate | Official index value | Official outcome |
|---|
The price estimate uses exchange candles. The contract may settle on a different index and average. Official YES/NO outcomes come from Kalshi, including after you reopen the page. Calls are recorded once before expiry; missed markets are not invented. More calls alone do not prove better accuracy or profit.
Spot-price forecast checks & countdowns ▾
Does the ghost match what happens?
FORWARD OBSERVATIONSThese averages use saved checks on this device. Shared market ghosts train a separate common calibration, shown above. Lower error is better; compare with the unchanged-price baseline. No improvement is claimed until the evidence supports it.
| Saved at | Ends | Source | Horizon | Middle estimate | Status | Actual price | Price error |
|---|
What is the market pricing?
Prices people are paying for an outcome. These are separate from your Bitcoin chart.
Source: Polymarket public markets ↗ · Order-book quotes ↗
15-minute candles summarize price over 15 minutes. A 15-minute market is a separate contract with its own 15-minute outcome window and resolution rules. The shared model uses completed 1-minute candles.
Read-only public data. No wallet or betting account connection. Regional/network restrictions may block access. Retrieval age and provider timestamps are distinct. Book quotes are not guaranteed executable and omit provider fees.
Workspace preferences & limits ＋
Alerts appear inside this open page with a 60-second cooldown. No notification permissions are requested.
Shared completed-candle predictions, 12 pattern heuristics, live observed trade feeds and local saved outcome history. Device sleep pauses local collection; reopening requests a current shared snapshot.
Legacy records are backed up locally before the active account key is retired. No simulator runs in this version.
Patterns are experimental definitions, not a complete taxonomy or a proven advantage. No blockchain whale monitor or multi-asset trading is connected.
Trading words, in plain English
- Candle
- One block of time showing where price started, the highest and lowest it reached, and where it ended or is now.
- Wick and body
- The thin wick marks the high and low. The thick body connects the start and end prices.
- Trade vs waiting order
- A trade already happened. An order is an offer to buy or sell that may still be waiting. This activity table shows completed trades.
- Spread and slippage
- Spread is the gap between buying and selling prices. Slippage means the fill price differs from the price you expected.
- Buying share
- The share of received trade value or BTC amount started by buyers. Every completed trade also has a seller.
- Liquidity and market odds
- Liquidity describes available trading interest. Prediction-market prices reflect what people pay for an outcome, not a guaranteed chance of winning.