BitcoinBTC Predictor

Analysis — Direction

Recomputed from btc.db on every request · 2026-11-18 15:12:04 UTC

The phase filter applies to all sections. Only evaluated UP/DOWN rows are included. Strength = max(p_up, p_down). Accuracy uses stored correct (missing values are recomputed; ties predict UP). Brier = mean((p_up − outcome)²), where UP = 1 and DOWN = 0; lower is better (50/50 baseline: 0.25). Empty groups show —. Invalid probability rows excluded: 0; correct values recomputed: 0.

Phase comparison

PhaseModelNCorrectAccuracyBrier scoreAvg prediction strength
Phase 1GPT1697242.6%0.260154.7%
Phase 1Jev1697745.6%0.385179.7%

Direction accuracy by realized volatility

Retrospective / conditional on realized volatility: subsequent-hour RV is known only afterward. This is exploratory and not tradable prospectively. Phase 4 thresholds were calibrated after the direction experiments. Observations from different models for the same hour are not independent.

RV = 100 × sqrt(sum(log(C_i / C_(i−1))²)), exactly twelve consecutive 5m returns including the preceding close. QUIET: RV < 0.227425610%; NORMAL: 0.227425610% ≤ RV < 0.370664156%; ACTIVE: RV ≥ 0.370664156%.

GPT / OpenAI: ACTIVE 35.5% (N=62), non-ACTIVE 46.7% (N=107); difference -11.2 percentage points. Jev: ACTIVE 41.9% (N=62), non-ACTIVE 47.7% (N=107); difference -5.7 percentage points.

Missing RV excluded: 0; invalid probability rows excluded: 0. ALL includes only RV-assigned rows. Stored actual_return is in percent; missing/invalid returns are excluded only from magnitude metrics (Return N). No interpolation or network fetching occurs in this page.

PhaseModelEligible NRV assignedMissing RV
phase1openai1691690
phase1jev1691690
PhaseModelActual regimeNCorrectAccuracyBrierReturn NMean |return|Median |return|Mean RV
phase1GPT / OpenAIALL1697242.6%0.26011690.2520%0.1787%0.3532%
phase1GPT / OpenAIQUIET502040.0%0.2679500.1361%0.1094%0.1727%
phase1GPT / OpenAINORMAL573052.6%0.2457570.2088%0.1829%0.2947%
phase1GPT / OpenAIACTIVE622235.5%0.2669620.3853%0.2608%0.5527%
phase1GPT / OpenAINON-ACTIVE1075046.7%0.25611070.1748%0.1460%0.2377%
phase1JevALL1697745.6%0.38511690.2520%0.1787%0.3532%
phase1JevQUIET501938.0%0.4582500.1361%0.1094%0.1727%
phase1JevNORMAL573256.1%0.3273570.2088%0.1829%0.2947%
phase1JevACTIVE622641.9%0.3794620.3853%0.2608%0.5527%
phase1JevNON-ACTIVE1075147.7%0.38841070.1748%0.1460%0.2377%

Calibration analysis

Blue bar: actual accuracy. Black marker: ideal accuracy, equal to the mean prediction strength in that bucket. Buckets include their lower bound and exclude their upper bound, except 100% is included in the last bucket.

GPT

Strength bucketNAccuracyIdeal (mean strength)Actual / ideal · 0–100%
50%–55%8843.2%53.0%
55%–60%7943.0%56.5%
60%–70%20.0%61.0%
70%–80%0--—
80%–90%0--—
90%–100%0--—

Jev

Strength bucketNAccuracyIdeal (mean strength)Actual / ideal · 0–100%
50%–55%1346.2%52.5%
55%–60%1323.1%56.7%
60%–70%1947.4%63.3%
70%–80%2462.5%74.4%
80%–90%3345.5%85.0%
90%–100%6743.3%93.4%

Threshold analysis

Keep predictions whose strength is at least the threshold (not Jev's separate confidence field). Compare accuracy and Brier together with N; higher thresholds leave fewer observations. These are historical subsets, not evidence of future trading profit.

GPT

Minimum strengthNAccuracyBrier score
≥ 50%16942.6%0.2601
≥ 55%8142.0%0.2657
≥ 60%20.0%0.3722
≥ 70%0--
≥ 80%0--
≥ 90%0--

Jev

Minimum strengthNAccuracyBrier score
≥ 50%16945.6%0.3851
≥ 55%15645.5%0.3961
≥ 60%14347.6%0.4060
≥ 70%12447.6%0.4251
≥ 80%10044.0%0.4685
≥ 90%6743.3%0.5016

Jev confidence investigation

Probability margin = abs(p_up − p_down). Evaluated Jev rows with valid confidence only. Missing or invalid confidence excluded: 0.

Paired N169
Pearson r0.999718
Mean absolute difference0.491 percentage points

If confidence repeats the probability margin, r approaches 1 and the absolute difference approaches 0. The table groups pairs by margin (lower bound included; 100% included in the last group).

Probability marginNMean marginMean Jev confidence
0–10%135.1%4.8%
10–20%1514.3%14.3%
20–30%1224.8%24.8%
30–40%836.0%36.1%
40–50%745.1%45.3%
50–60%1452.4%52.4%
60–70%1464.6%64.7%
70–80%1974.0%74.1%
80–90%5285.0%85.1%
90–100%1592.8%92.6%