?
Shows the current 0-100 score for each signal family. Higher means that area looks more like previous Bitcoin bottom conditions.
Component Scores
Cycle Structure57.89
Market Capitulation63.16
Publication Quality81.33
Social Capitulation63.40
Overview
Overview
Current signal state, score ceiling, confidence, and alert blockers.
?
The headline live state. It combines the total bottom score, confidence, alert decision, and the reason the monitor is or is not sending an alert.
BTC Live Price$64,2032026-08-18
1D: -0.47%7D: +1.02%ATH DD: -48.54%
Static close shown · browser live starts after load
Last checked18 Aug 2026
Bottom Likely
Not bottom_likely: validated reclaim, PIT replay, cross-provider checks, and negative-control evidence still block promotion.
Bottom Score
66.09
ConfidenceMedium
Should alertNo
Alert reasonstate-machine bottom_likely gate is missing; reclaim/PIT/stabilisation validation is required before live bottom_likely alerts
?
Shows supporting details for Why this is not bottom likely. Use it as context for the surrounding score, diagnostic, or research summary.
Why This Is Not Bottom Likely
Current User RegimeBottom Likely
Internal Research StateStabilisation
Bottom LikelyNo
Reclaim GateBlocked Candidate Reclaim Not Passed
PIT ReplayBlocked Pit Snapshots Missing
Negative-Control StatusBlocked Negative Control Noise Detected
Reliability Upgrade AllowedNo
Why This Is Not Bottom Likely
- Live user regime is Bottom Likely, while the internal research state is Stabilisation; exhaustion is not a bottom-likely alert.
- Validated aggregate and STH cost-basis reclaim inputs are still missing or blocked, so the reclaim gate is not passed.
- Candidate reclaim gate status is Blocked Candidate Reclaim Not Passed; candidate-only values are not signal inputs.
- Live bottom_likely alerts and the canonical bottom_likely flag are now bound to the guarded state machine; the unguarded engine regime cannot override missing reclaim, PIT, or stabilisation evidence.
- Cross-provider validation status is Blocked Cross Provider Incomplete and PIT replay status is Blocked Pit Snapshots Missing.
- Negative controls show 1075 family-weighted alert days outside true bottom windows, so family evidence alone is noisy.
- Reliability is capped; validation infrastructure and explanatory audits do not raise forward predictive evidence.
What Would Invalidate The Current Read
- A new lower low would reset the lower-low failure clock and weaken stabilisation evidence.
- Provider-definition or cross-provider disagreement would keep NUPL/STH-NUPL-derived reclaim candidates blocked.
- Any PIT replay known-at leakage, stale snapshot, or export hash mismatch keeps reclaim and reliability blocked.
- Any candidate-only metric entering score, state machine, reliability, or alerting would invalidate the current safety lock.
- Negative-control noise in shifted, random, or bull-market guard windows prevents any threshold/reliability upgrade.
Plain-language uncertainty summary only. It does not promote candidate reclaim, change bottom_likely, loosen reclaim, or raise reliability.
?
Checks that every page is rendered from one canonical state object and explains the difference between the user alert regime and internal state-machine state.
Canonical State Consistency
Canonical Render ConsistencyPASS
User Alert RegimeBottom Likely
Internal ChartInspect StateStabilisation
Bottom LikelyNo
Lower Low Failure Days49
Active Reliability Cap7.50
State Hash2936aff77a650ef0
| Check | Status | Detail |
|---|---|---|
| Watch Vs Exhaustion Explained | Pass | User alert regime=bottom_likely; internal ChartInspect state=stabilisation |
| Lower Low Canonicalized | Pass | canonical=49; monitor=49 |
| Bottom Likely Consistent | Pass | bottom_likely remains false/no |
| Candidate Metrics Not Promoted | Pass | promoted candidates=[] |
| Reliability Cap Present | Pass | active cap=7.5 |
All pages are rendered from one dashboard data object and share the canonical_state block. Watch is the user alert regime; Exhaustion is the internal ChartInspect state-machine state.
?
Compact recent BTC-USD daily close chart from the same live market collector used by the scoring engine. It gives visual context without changing the bottom score.
BTC Price Chart
Latest Close$64,203
1D Change-0.47%
7D Change+1.02%
Drawdown From ATH-48.54%
Chart Range2026-04-21 → 2026-08-18
Auto-updates every 60s while this page is open using a browser-side Binance BTCUSDT feed; the static collector close remains the fallback.
?
Rates the signal itself, not today's BTC score. It combines backtest quality, false-positive control, strategy usefulness, independent confirmation, live validation, and benchmark coverage. It is capped below 10 because market signals cannot be guaranteed.
Signal Reliability Scorecard
Research Framework Reliability7.2 / 10
Forward Predictive Evidence7.0 / 10
Active Cap7.5 / 10
GradeB
Rating LabelResearch Grade
Max Honest Rating9.0 / 10
| Dimension | Score | Status | Evidence |
|---|---|---|---|
| Event Definition Integrity | 6.80 | Broad Window Caveated | Historical fit uses only a few BTC bottom windows plus a broad 180d lead / 120d lag tolerance; dashboard must caveat 4/4 claims. |
| Point In Time Data Integrity | 6.00 | Manifest Added Audit Pending | Known-at delay manifest is present, but provider snapshot/vintage replay is still required before raising the cap. |
| Backtest Robustness | 7.50 | Robust | Threshold perturbation verdict robust; metric dropout survivors 2/6. |
| Continuous Non Event Control | 6.20 | All History Alert Density Review | Outside-bottom alert days 63; outside alert density 1.35%. |
| False Positive Control | 8.80 | Guard Clean | Guard alert days 0; false-positive guard windows 0. |
| Practical Timing Usefulness | 7.00 | Useful But Drawdown Risk | Detected 4/4 bottom windows; worst 365D post-alert drawdown -44.3%. |
| Independent Confirmation | 6.20 | State Machine Shadow Added | State-machine latest state stabilisation; shadow stack clean runs 0 / 30; auto-promotion remains disabled. |
| Live Validation | 3.40 | Early Live Observation | 0/30 clean shadow runs; ledger has 46 live rows and 28 false-setup candidates; no future live capitulation event has occurred since build-out. |
| Ops Data Reliability | 7.00 | Daily Static Dashboard And Live Ledger | Cron/dashboard/data.json are reproducible; live data freshness remains monitored by clean-run checks. |
| Dashboard Honesty | 8.00 | Uncertainty Visible | Reliability cap, small-sample caveat, PIT audit status, and live-proof limits are shown directly on the dashboard. |
Reliable as a conservative BTC bottom-zone warning system; not reliable as an exact-entry or autonomous trading signal.
?
Tracks what cannot be proven by backtests: clean live observation and future real-world false-setup or capitulation behavior after the model was built.
Live Proof Tracker
StatusAwaiting Live Regime Proof
Clean Shadow Runs0 / 30
Ledger Runs46
False Setup Candidates28
Needs Future Capitulation EventYes
Cannot Be Completed InstantlyYes
Phase 5 cannot be completed instantly: real reliability upgrades require future live false-setup and/or capitulation evidence after the model was built.
?
Daily audit trail of live monitor runs. It records the score, alert decisions, shadow state, clean-run status, and false-setup candidates without changing reliability automatically.
Live Proof Ledger
Total Runs46
Clean Runs1
Latest Clean Streak0
False Setup Candidates28
Pending Follow-Up28
Confirmed Correct Rejections0
Primary / Shadow Alerts0 / 0
| Date | Score | Regime | Primary Alert | Shadow Alert | False Setup Status | Clean Run |
|---|---|---|---|---|---|---|
| 2026 08 18 | 66.09 | Bottom Likely | No | No | Not A False Setup Candidate | No |
| 2026 08 17 | 70.22 | Bottom Likely | No | No | Not A False Setup Candidate | No |
| 2026 08 16 | 75.89 | Bottom Likely | No | No | Not A False Setup Candidate | No |
| 2026 08 15 | 73.73 | Bottom Likely | No | No | Not A False Setup Candidate | No |
| 2026 08 14 | 69.31 | Bottom Likely | No | No | Not A False Setup Candidate | No |
| 2026 08 13 | 65.98 | Accumulation Watch | No | No | Not A False Setup Candidate | No |
| 2026 08 12 | 66.82 | Accumulation Watch | No | No | Not A False Setup Candidate | No |
| 2026 08 11 | 66.87 | Accumulation Watch | No | No | Not A False Setup Candidate | No |
| 2026 08 10 | 72.62 | Accumulation Watch | No | No | Not A False Setup Candidate | No |
| 2026 08 09 | 66.31 | Accumulation Watch | No | No | Not A False Setup Candidate | No |
| 2026 08 08 | 64.91 | Watch | No | No | Pending Followup | No |
| 2026 08 07 | 65.49 | Accumulation Watch | No | No | Not A False Setup Candidate | No |
| 2026 08 06 | 67.09 | Accumulation Watch | No | No | Not A False Setup Candidate | No |
| 2026 08 05 | 67.09 | Accumulation Watch | No | No | Not A False Setup Candidate | No |
Ledger rows are live observations only. They can support later reliability upgrades after future market follow-up, but they do not auto-promote or auto-raise the rating.
?
Explains why the reliability rating did or did not move after live evidence. Most days should say no rating change until future market follow-up proves the setup.
Reliability Change Log
- Reliability stayed at 7.2/10 because no future live capitulation proof has completed yet.
- 1 clean live observations are recorded in the ledger.
- 28 watch-level false-setup candidates are pending follow-up; unresolved setups do not improve reliability.
- Auto-promotion remains disabled; shadow evidence still requires manual review.
?
Shows whether each component has enough fresh, diverse data to trust the reading. Low confidence can mean missing or stale inputs, not just a weak signal.
Component Confidence
| Component | Displayed Confidence | Confidence Type | Predictive Evidence |
|---|---|---|---|
| Cycle Structure | High | freshness/coverage | primary/diagnostic evidence |
| Market Capitulation | High | freshness/coverage | primary/diagnostic evidence |
| Publication Quality | High | freshness/coverage | low/context-only |
| Social Capitulation | High | freshness/coverage | low/context-only |
Confidence is split conceptually: social/news freshness can be high while promotion confidence remains low/context-only.
?
Compares today's component readings with the historical archive so you can see whether a signal is unusually extreme or ordinary.
Historical Percentiles
Historical percentiles will appear after calibration data is available.
?
Shows the highest total score realistically possible today, plus blockers that prevent the model from reaching a stronger bottom signal.
Score Ceiling Today
Maximum Plausible Today75.70
- market drawdown is 49%; severe threshold starts near 60% and deep bottom threshold near 75%
- cycle age is 8.0 months since top; mature-cycle bottom threshold starts around 12 months
?
Lists safety checks that keep the system quiet even if some individual signals look interesting. This is designed to avoid noisy false alerts.
Alert Suppression
- No suppressions recorded.