BTC market overview
Public Binance ticker and one-minute candles. This is a reference feed, not a forecast.
Recent BTC candles
Last 60 one-minute candles fetched from Binance. Candle times are UTC.
| UTC | Open | High | Low | Close | Volume | Return bps |
|---|---|---|---|---|---|---|
| No candles loaded yet. | ||||||
Import BTC history
Fetch public one-minute candles by date range or import a historical CSV file.
Automatic public-data import
Start with a short date range to verify API access.
Import CSV history
Common columns supported: timestamp, open, high, low, close, volume. ISO timestamps and Unix seconds/milliseconds are supported. Binance kline CSV exports can also be used.
History quality report
Validation is run after import and can be repeated.
Compare predictorv37 vs Candle Intelligence
Enter or import prediction records from both sites. The comparison uses only rows you provide; it does not invent V37 predictions.
Bundled prediction-history files
This V38 build includes the available original scorecard and V28 learning-record CSVs. These are model prediction/outcome histories, not a full raw minute-by-minute OHLCV archive. Load either into Site A for review or comparison.
Site A — predictorv37 / original scorecard
Import the scorecard CSV exported from your current site. Supported fields include predicted_direction, actual_direction, result, timestamps and spot/forecast prices.
Site B — Candle Intelligence
Use the same time period and settlement definition. Site B baseline predictions can be generated from imported candles, then compared with outcomes.
Comparison details
Accuracy alone is not profitability. Keep settlement source, contract timing, and fees consistent.
| Contract / time | Site A prediction | Site A outcome | Site B prediction | Site B outcome | Agreement |
|---|---|---|---|---|---|
| Load records from both sites to compare. | |||||
Correlation & baseline research
Calculates market patterns from available candle history. No results are shown until you run the analysis.
Return correlation plot
Run research after importing candles.
Average return by minute in 15m cycle
Method & limitations
- Uses non-overlapping UTC-aligned 15-minute windows.
- First five-minute return is compared with the following ten-minute return.
- Missing candles are skipped; the app does not synthesize prices.
- Historical BTC return tests are not the same as Kalshi contract profitability.
- Historical candles cannot reconstruct V37's forecasts unless its past predictions are supplied.
Export & settings
Data stays in this browser and on this site origin unless you export it.
Download data
Restore / clear
This only clears this new site. It does not change predictorv37. Export before resetting.