We’re opening your workspace.
If the connection takes too long, you can return home without submitting an operation.
If the connection takes too long, you can return home without submitting an operation.
QUÉ BOLÁASERE AI · ORACLEEspañol ↗Crypto, indices, futures and metals. Compare experimental 5-minute, 1-hour, 4-hour and daily forecasts, and see how their trends align.
Coinbase Exchange · USD
Data through — ET · 0/60 candles
Waiting for actual provider candles.
Each row has its own forecast. Compare the smaller timeframe’s direction with the next larger one.
Loading this market’s timeframes.
The current models are preserved. Additional horizons start at the next 5-minute block and may span session closures. Alignment uses completed candles; it is not a probability of success.
“Starting to align” compares the reading used by the forecast with the previous window of completed candles. WAIT does not count as alignment.
—
Waiting for this market’s data provider.
If you check at 3:10 p.m., the target is 4:00 p.m. Each hour starts a new forecast for BTC, ETH, SOL and XRP.
The target time is exact; the price is an experimental estimate and may differ from the market.
The next hour and the 23 earlier hours. Only forecasts issued before their target are shown.
Bitcoin · next hour and 23 earlier hours in this table
Coverage: observations within the original range / verified outcomes. Point error: mean |estimate / observation − 1| × 100. Reference: the same error using the last known price at issue time. Only actually issued and observed forecasts count; historical simulations are shown separately.
The percentage uses verified outcomes only. Pending and unscored issues are excluded. It is a historical result for this sample, not a future success probability.
Green: the observed price was inside the published range. Red: it was outside. This measures the range; it does not mean the point price estimate was exact.
| Target time | Issued | Estimated price | Indicative range | Observed | Absolute error | Status |
|---|---|---|---|---|---|---|
| Waiting for server time and this market’s records. | ||||||
The observed price is the close of the 5-minute candle ending at the target time, once Coinbase publishes that candle. It may differ from an instantaneous quote or another exchange’s price. Missing records are not filled with retrospective predictions.
Damped drift model: combines recent changes in closed 5-minute candles and reduces their projection with the remaining time to the target. This is a transparent rule, not a trained model or a promise of returns.
When comparable historical errors are insufficient, the range uses recent volatility. It is indicative and does not guarantee that the price will remain inside.
There are not enough comparable windows yet to display MAE and a historical baseline.
A team of models to anticipate movements and learn from observed outcomes.
Loading sources…
View of the selected issue. The walkthrough illustrates connections; it is not a live stream.
Checks timestamps, crypto continuity, sessions and freshness before calculation.
Estimated price in USD
Agreement measures model votes, not success probability. The point estimate and range can be wrong.
Targets at the next UTC 1-hour, 4-hour or midnight boundary. Times shown in New York.Without a current issue, readings are not presented as published predictions.
The last figure describes the bottom 10% of simulated scenarios. It is not a guaranteed maximum loss.
Checks do not eliminate uncertainty.
Minimum observed outcomes per active analyst for this market, timeframe and regime.
Price models start with equal weights; context and flow receive smaller weights. After 12 outcomes per analyst, weights adjust to error against the last price. Persistent underperformance reduces influence while evaluation continues.
Regime for this issue: —. Latest observed target: —.
Blocks of 3 hourly returns are resampled into 1,000 paths with a fixed seed and a volatility floor. Scenarios depend on the recent sample; they do not represent every possible crisis or calibrated probabilities.
Scenarios use closed hourly candles. Traditional-market session breaks can make those observations span more calendar time.
Alternative.me supplies the general crypto Fear & Greed index; it is not asset-specific sentiment. Headlines use a simple word rule with a small weight. Yahoo Finance is a public source with possible delays and contract rolls. Gold and silver are COMEX futures.
This context may be newer than the saved issue. Its original sources and values remain frozen in that issue.
No recent relevant headlines available for this check.
Existing directional models and the hourly table retain their own rules. This council is an additional model, with separately measured outcomes.
Bitcoin · 1h · last 24 issues within the 7-day retention window
Coverage: observations within the original range / verified outcomes. Point error: mean |estimate / observation − 1| × 100. Reference: the same error using the last known price at issue time. Only actually issued and observed forecasts count; historical simulations are shown separately.
The percentage uses verified outcomes only. Pending and unscored issues are excluded. It is a historical result for this sample, not a future success probability.
| Target | Estimated | Published range | Observed | Point error | Outcome |
|---|---|---|---|---|---|
| No issues recorded for this selection yet. | |||||
Watchlist, trend matrix, alignment and observed history.
Choose markets and consult the existing readings at 5 min, 1 h, 4 h and 1 day. Refresh is manual.
| Market | 5m | 1h | 4h | 1 day |
|---|---|---|---|---|
| BitcoinBTC-USD | ||||
| EthereumETH-USD | ||||
| SolanaSOL-USD | ||||
| S&P 500^GSPC |
Choose a cell with data to inspect factors, age and conditions that weaken the reading.
One first-issued prediction per target; revisions are not independent trials. Revisions excluded from this table: 0.
| Market | Timeframe | Correct / incorrect | Pending | Unscored |
|---|
Up to 120 directional readings in the studio. Enable saving in My space to retain them between visits. An outcome is scored only when a later refresh contains every exact candle in the target window. Horizons without verifiable boundaries remain unscored. WAIT is excluded.
First-issued predictions in this studio · up to 120 stored readings
The percentage uses verified outcomes only. Pending and unscored issues are excluded. It is a historical result for this sample, not a future success probability.
| Market / horizon | Target | Reading | Observed outcome |
|---|
History begins with your first refresh. Earlier outcomes are not invented.
No validated success probability is available. Your studio history is a partial sample and does not represent all market activity. Sources: Coinbase Exchange and Yahoo Finance; some markets may be closed or delayed.