{
"slug": "model-watch-monthly-2026-09",
"series": "monthly",
"series_label": "MONTHLY · MODEL WATCH",
"issue": 2,
"title": "Monthly Model Watch #2",
"language": "en",
"window_start": "2026-09-01T00:00:00+00:00",
"window_end": "2026-10-01T00:00:00+00:00",
"window_text": "Sep 1-30, 2026 UTC",
"cutoff": "2026-10-01T16:00:00+00:00",
"generated_at": "2026-10-02T09:18:46+00:00",
"methodology": "v1.1 (2026-08-10)",
"methodology_hash": "e66c7e8c864a2233",
"source": "monthly_metrics_2026-09.json",
"hit_rule": "direction: exit tp1/tp2 -> hit, sl -> miss, expiry -> sign of gross pnl",
"bibtex_key": "mm_modelwatch_monthly_2026m09",
"research_question": "who led the month, which tickers were readable, who saw the turn first, and does a model agreeing with itself across timeframes/horizons make its call stronger?",
"base_field_hit": {
"n": 20682,
"hit_rate": 0.455
},
"market_state": {
"panel": "market_state_6p2",
"month": {
"start": "2026-09-01",
"end": "2026-10-01",
"days": 30
},
"prev_month_partial": {
"start": "2026-08-01",
"end": "2026-09-01",
"days": 31,
"partial": false
},
"current": {
"btc_net_pct": 6.42,
"realized_vol_ann_pct": 41.17,
"volume_top5_usd_bn": 80.0142,
"volume_top5_mom_pct": 21.3,
"avg_pairwise_corr": 0.78
},
"previous_partial_month": {
"btc_net_pct": 24.95,
"realized_vol_ann_pct": 41.53,
"volume_top5_usd_bn": 65.9696,
"avg_pairwise_corr": 0.79
},
"snapshot_line": "BTC net +6.42% (prior +24.95%) - ann vol 41.2% (was 41.5%) - TOP5 volume $80.01B, +21% m/m - pairwise corr 0.78 (was 0.79)",
"exchange": "binance",
"audit_filter_applied": true,
"source": "single-exchange (binance) daily candle recompute, market_state_2026-09.json; the prior-month block is the pack's own prev_month (Aug 1-31)",
"audit_note": "Market-state row is computed on a single-exchange (binance) daily candle series, as in the weekly series after the audit that found the raw table mixes two exchanges. The prior-month block is the full August: it covers 31 days (Aug 1-31), and the TOP5 volume figure compares a 30-day sum with a 31-day sum -- a level difference, not a like-for-like change.",
"weeks": {
"W1": {
"btc_net_pct": 3.42,
"realized_vol_ann_pct": 43.35,
"regime": "calm"
},
"W2": {
"btc_net_pct": -4.36,
"realized_vol_ann_pct": 19.68,
"regime": "calm"
},
"W3": {
"btc_net_pct": 5.64,
"realized_vol_ann_pct": 51.01,
"regime": "volatile"
},
"W4": {
"btc_net_pct": 4.06,
"realized_vol_ann_pct": 52.25,
"regime": "volatile"
}
},
"volatility_rule": "volatile if realized_vol_ann_pct >= median(W1..W4) else calm",
"median_realized_vol_ann_pct": 47.18,
"delta_month_vs_prev_partial": {
"btc_net_pct": -18.53,
"realized_vol_ann_pct": -0.36,
"avg_daily_range_pct": 0.06,
"volume_top5_usd_pct": 21.3,
"avg_pairwise_corr": -0.01,
"note": "September (30 days) vs August (31 days, the full previous month); volume is a sum, so volume_top5_usd_pct compares a 30-day sum with a 31-day sum (not normalised per day)"
},
"controls": {
"w4_matches_published": {
"pins_source": "market_state_2026-09-21.json -> 'current' (the published weekly #8 week, Sep 21-27)",
"block": "W4",
"published": {
"btc_net_pct": 4.06,
"realized_vol_ann_pct": 52.25,
"volume_top5_usd_bn": 22.4265,
"avg_pairwise_corr": 0.71
},
"got": {
"btc_net_pct": 4.06,
"realized_vol_ann_pct": 52.25,
"volume_top5_usd_bn": 22.4265,
"avg_pairwise_corr": 0.71
},
"match": {
"btc_net_pct": true,
"realized_vol_ann_pct": true,
"volume_top5_usd_bn": true,
"avg_pairwise_corr": true
},
"tolerance": {
"btc_net_pct": 0.05,
"realized_vol_ann_pct": 0.5,
"volume_top5_usd_bn": 0.05,
"avg_pairwise_corr": 0.02
},
"all_match": true
},
"w3_matches_published_previous": {
"pins_source": "market_state_2026-09-21.json -> 'previous' (the weekly #8 run's prev week = weekly #7, Sep 14-20)",
"block": "W3",
"published": {
"btc_net_pct": 5.64,
"realized_vol_ann_pct": 51.01,
"volume_top5_usd_bn": 18.7386,
"avg_pairwise_corr": 0.93
},
"got": {
"btc_net_pct": 5.64,
"realized_vol_ann_pct": 51.01,
"volume_top5_usd_bn": 18.7386,
"avg_pairwise_corr": 0.93
},
"match": {
"btc_net_pct": true,
"realized_vol_ann_pct": true,
"volume_top5_usd_bn": true,
"avg_pairwise_corr": true
},
"tolerance": {
"btc_net_pct": 0.05,
"realized_vol_ann_pct": 0.5,
"volume_top5_usd_bn": 0.05,
"avg_pairwise_corr": 0.02
},
"all_match": true
},
"note": "W4 and W3 of this run are exactly the two weeks the weekly #8 market panel computed, with the same audit filter, so both must reproduce. Non-fatal: a MISMATCH means the exchange filter or the symbol set moved -- send the log, do not publish the row."
},
"regime_days": {
"2026-09-01": {
"move_pct": -1.5,
"regime": "trend"
},
"2026-09-02": {
"move_pct": -0.06,
"regime": "flat"
},
"2026-09-03": {
"move_pct": 4.98,
"regime": "trend"
},
"2026-09-04": {
"move_pct": -1.97,
"regime": "trend"
},
"2026-09-05": {
"move_pct": 0.24,
"regime": "flat"
},
"2026-09-06": {
"move_pct": 0.58,
"regime": "flat"
},
"2026-09-07": {
"move_pct": -1.52,
"regime": "trend"
},
"2026-09-08": {
"move_pct": -0.72,
"regime": "flat"
},
"2026-09-09": {
"move_pct": -0.25,
"regime": "flat"
},
"2026-09-10": {
"move_pct": -2.14,
"regime": "trend"
},
"2026-09-11": {
"move_pct": 0.85,
"regime": "flat"
},
"2026-09-12": {
"move_pct": 0.0,
"regime": "flat"
},
"2026-09-13": {
"move_pct": -0.57,
"regime": "flat"
},
"2026-09-14": {
"move_pct": 1.79,
"regime": "trend"
},
"2026-09-15": {
"move_pct": -3.35,
"regime": "trend"
},
"2026-09-16": {
"move_pct": 0.86,
"regime": "flat"
},
"2026-09-17": {
"move_pct": 0.21,
"regime": "flat"
},
"2026-09-18": {
"move_pct": 5.82,
"regime": "trend"
},
"2026-09-19": {
"move_pct": 0.46,
"regime": "flat"
},
"2026-09-20": {
"move_pct": 0.13,
"regime": "flat"
},
"2026-09-21": {
"move_pct": 6.39,
"regime": "trend"
},
"2026-09-22": {
"move_pct": -0.4,
"regime": "flat"
},
"2026-09-23": {
"move_pct": -2.04,
"regime": "trend"
},
"2026-09-24": {
"move_pct": 0.02,
"regime": "flat"
},
"2026-09-25": {
"move_pct": -0.46,
"regime": "flat"
},
"2026-09-26": {
"move_pct": 0.36,
"regime": "flat"
},
"2026-09-27": {
"move_pct": 0.2,
"regime": "flat"
},
"2026-09-28": {
"move_pct": -1.19,
"regime": "flat"
},
"2026-09-29": {
"move_pct": 0.22,
"regime": "flat"
},
"2026-09-30": {
"move_pct": -0.22,
"regime": "flat"
}
},
"alive_slice_context": {
"abs_move_1d_pct": {
"A": 1.979,
"B": 1.905
},
"btc_realized_vol_ann_pct_hourly": {
"A": 36.92,
"B": 34.88
}
}
},
"key_finding": {
"label": "KEY FINDING · OBSERVATION (one monthly window)",
"text": "Field accuracy went 46.3% -> 45.5% (-0.8pp) from August to September with 1 of 7 models improving, a single model line led the four weeks, and no monthly title was awarded.",
"evidence_level": "observation"
},
"tables": {
"leaderboard": {
"rows": [
{
"model": "gemini-3.1-pro",
"n": 3284,
"hit_rate": 0.4851,
"wilson_95": [
0.468,
0.5022
],
"coverage": 0.5762,
"model_label": "gemini-3.1-pro",
"insufficient": false
},
{
"model": "claude-opus-5",
"n": 2532,
"hit_rate": 0.4538,
"wilson_95": [
0.4345,
0.4732
],
"coverage": 0.4445,
"model_label": "claude-opus-5",
"insufficient": false
},
{
"model": "claude-fable-5",
"n": 2945,
"hit_rate": 0.4537,
"wilson_95": [
0.4357,
0.4717
],
"coverage": 0.5167,
"model_label": "claude-fable-5",
"insufficient": false
},
{
"model": "deepseek-v4-pro",
"n": 3352,
"hit_rate": 0.4526,
"wilson_95": [
0.4358,
0.4695
],
"coverage": 0.5907,
"model_label": "deepseek-v4-pro",
"insufficient": false
},
{
"model": "qwen-3.8-max",
"n": 3165,
"hit_rate": 0.4471,
"wilson_95": [
0.4298,
0.4645
],
"coverage": 0.5597,
"model_label": "qwen-3.8-max",
"insufficient": false
},
{
"model": "gpt-5.6-sol",
"n": 3133,
"hit_rate": 0.4446,
"wilson_95": [
0.4273,
0.4621
],
"coverage": 0.5499,
"model_label": "gpt-5.6-sol",
"insufficient": false
},
{
"model": "grok-4.6",
"n": 2271,
"hit_rate": 0.4439,
"wilson_95": [
0.4235,
0.4644
],
"coverage": 0.3995,
"model_label": "grok-4.6",
"insufficient": false
}
],
"title_awarded": false,
"title_rule": ">=5pp gap AND non-overlapping 95% CI, N>=10",
"top2_gap_pp": 3.1,
"movement_vs_prior_window": [
{
"model": "gemini-3.1-pro",
"hit_rate_prior_window": 0.4761,
"n_prior_window": 3033,
"hit_rate_monthly2": 0.4851,
"delta_pp": 0.9,
"rank_monthly2": 1
},
{
"model": "claude-opus-5",
"hit_rate_prior_window": 0.4779,
"n_prior_window": 2281,
"hit_rate_monthly2": 0.4538,
"delta_pp": -2.4,
"rank_monthly2": 2
},
{
"model": "claude-fable-5",
"hit_rate_prior_window": 0.4673,
"n_prior_window": 2812,
"hit_rate_monthly2": 0.4537,
"delta_pp": -1.4,
"rank_monthly2": 3
},
{
"model": "deepseek-v4-pro",
"hit_rate_prior_window": 0.4554,
"n_prior_window": 2578,
"hit_rate_monthly2": 0.4526,
"delta_pp": -0.3,
"rank_monthly2": 4
},
{
"model": "qwen-3.8-max",
"hit_rate_prior_window": 0.4565,
"n_prior_window": 3001,
"hit_rate_monthly2": 0.4471,
"delta_pp": -0.9,
"rank_monthly2": 5
},
{
"model": "gpt-5.6-sol",
"hit_rate_prior_window": 0.4586,
"n_prior_window": 2970,
"hit_rate_monthly2": 0.4446,
"delta_pp": -1.4,
"rank_monthly2": 6
},
{
"model": "grok-4.6",
"hit_rate_prior_window": 0.4527,
"n_prior_window": 2708,
"hit_rate_monthly2": 0.4439,
"delta_pp": -0.9,
"rank_monthly2": 7
}
],
"prior_window_note": "prior = month.prior_window.model_watch.leaderboard, the full August computed by the pack at the prior cutoff",
"note": "No monthly title is awarded this issue. Title rule: **>=5pp gap AND non-overlapping 95% CI, N>=10** per side. gemini-3.1-pro's lead over claude-opus-5 is 3.1pp and their CIs overlap, so neither condition is met."
},
"ticker": {
"rows": [
{
"symbol": "XRP",
"n": 4308,
"hit_rate": 0.4705,
"wilson_95": [
0.4556,
0.4854
],
"prior_hit_rate": 0.4673,
"prior_n": 3835
},
{
"symbol": "BTC",
"n": 3987,
"hit_rate": 0.4597,
"wilson_95": [
0.4443,
0.4752
],
"prior_hit_rate": 0.4648,
"prior_n": 4068
},
{
"symbol": "ETH",
"n": 3973,
"hit_rate": 0.4538,
"wilson_95": [
0.4384,
0.4693
],
"prior_hit_rate": 0.4287,
"prior_n": 3695
},
{
"symbol": "BNB",
"n": 3994,
"hit_rate": 0.4512,
"wilson_95": [
0.4358,
0.4666
],
"prior_hit_rate": 0.4807,
"prior_n": 3857
},
{
"symbol": "SOL",
"n": 4420,
"hit_rate": 0.4403,
"wilson_95": [
0.4257,
0.455
],
"prior_hit_rate": 0.4733,
"prior_n": 3928
}
],
"best": "XRP",
"hardest": "SOL",
"all_above_51pct": false,
"prior_note": "prior = month.prior_window.model_watch.by_symbol"
},
"pairs_top": [
{
"model": "gemini-3.1-pro",
"symbol": "SOL",
"n": 702,
"hit_rate": 0.4943
},
{
"model": "gemini-3.1-pro",
"symbol": "BTC",
"n": 640,
"hit_rate": 0.4906
},
{
"model": "gemini-3.1-pro",
"symbol": "XRP",
"n": 673,
"hit_rate": 0.4889
},
{
"model": "grok-4.6",
"symbol": "XRP",
"n": 482,
"hit_rate": 0.4855
},
{
"model": "gemini-3.1-pro",
"symbol": "ETH",
"n": 622,
"hit_rate": 0.4759
},
{
"model": "gemini-3.1-pro",
"symbol": "BNB",
"n": 647,
"hit_rate": 0.4745
}
],
"pairs_worst": [
{
"model": "deepseek-v4-pro",
"symbol": "SOL",
"n": 691,
"hit_rate": 0.4327
},
{
"model": "grok-4.6",
"symbol": "BTC",
"n": 407,
"hit_rate": 0.4324
},
{
"model": "gpt-5.6-sol",
"symbol": "SOL",
"n": 657,
"hit_rate": 0.4247
},
{
"model": "qwen-3.8-max",
"symbol": "BNB",
"n": 594,
"hit_rate": 0.4209
},
{
"model": "grok-4.6",
"symbol": "SOL",
"n": 503,
"hit_rate": 0.4095
}
],
"reversal_events": {
"definition": "4h grid, trend=sign of prior 24h, counter-move >=2.0% (BTC/ETH) sustained 8h; caller: directional hit, FH 4h/1d, slot in [T-12h, T+FH]",
"events": [
{
"symbol": "BTC",
"t": "2026-09-23T08:00:00+00:00",
"new_side": "short",
"move_8h_pct": -2.9,
"n_callers": 1,
"first_caller": {
"model": "gemini-3.1-pro",
"slot": "2026-09-24T00:01:00+00:00",
"fh": "1d",
"tf": "4h",
"confidence": 65
},
"caller_models": [
"gemini-3.1-pro"
]
},
{
"symbol": "ETH",
"t": "2026-09-03T08:00:00+00:00",
"new_side": "long",
"move_8h_pct": 4.61,
"n_callers": 0,
"first_caller": null,
"caller_models": []
},
{
"symbol": "ETH",
"t": "2026-09-04T08:00:00+00:00",
"new_side": "short",
"move_8h_pct": -2.13,
"n_callers": 0,
"first_caller": null,
"caller_models": []
},
{
"symbol": "ETH",
"t": "2026-09-11T08:00:00+00:00",
"new_side": "long",
"move_8h_pct": 4.19,
"n_callers": 11,
"first_caller": {
"model": "gemini-3.1-pro",
"slot": "2026-09-12T00:01:00+00:00",
"fh": "1d",
"tf": "1d",
"confidence": 70
},
"caller_models": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
]
},
{
"symbol": "ETH",
"t": "2026-09-11T16:00:00+00:00",
"new_side": "short",
"move_8h_pct": -2.11,
"n_callers": 0,
"first_caller": null,
"caller_models": []
},
{
"symbol": "ETH",
"t": "2026-09-20T12:00:00+00:00",
"new_side": "long",
"move_8h_pct": 2.22,
"n_callers": 17,
"first_caller": {
"model": "gemini-3.1-pro",
"slot": "2026-09-20T00:01:00+00:00",
"fh": "1d",
"tf": "1d",
"confidence": 75
},
"caller_models": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
]
},
{
"symbol": "ETH",
"t": "2026-09-23T08:00:00+00:00",
"new_side": "short",
"move_8h_pct": -3.73,
"n_callers": 6,
"first_caller": {
"model": "gpt-5.6-sol",
"slot": "2026-09-23T08:01:00+00:00",
"fh": "4h",
"tf": "1h",
"confidence": 67
},
"caller_models": [
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
]
}
],
"scope_v1": "BTC/ETH only",
"status": "ok",
"first_call_timing": {
"rows": [
{
"symbol": "BTC",
"t": "2026-09-23T08:00:00+00:00",
"first_call_slot": "2026-09-24T00:01:00+00:00",
"lag_minutes": 961,
"relation": "after_turn"
},
{
"symbol": "ETH",
"t": "2026-09-11T08:00:00+00:00",
"first_call_slot": "2026-09-12T00:01:00+00:00",
"lag_minutes": 961,
"relation": "after_turn"
},
{
"symbol": "ETH",
"t": "2026-09-20T12:00:00+00:00",
"first_call_slot": "2026-09-20T00:01:00+00:00",
"lag_minutes": -719,
"relation": "before_turn"
},
{
"symbol": "ETH",
"t": "2026-09-23T08:00:00+00:00",
"first_call_slot": "2026-09-23T08:01:00+00:00",
"lag_minutes": 1,
"relation": "after_turn"
}
],
"note": "slot times of the first counted call against the detector's turn time; the caller window is [T-12h, T+FH], so a counted call can follow the turn"
}
},
"cross_tf": {
"fh_4h_tf_4h_vs_1h": {
"agree": {
"n": 2305,
"hit_rate": 0.4256,
"wilson_95": [
0.4056,
0.4459
],
"by_regime": {
"trend": {
"n": 1026,
"hit_rate": 0.5517
},
"flat": {
"n": 1279,
"hit_rate": 0.3245
}
}
},
"disagree": {
"n": 68,
"hit_rate": 0.5735,
"wilson_95": [
0.4552,
0.684
],
"by_regime": {
"trend": {
"n": 15,
"hit_rate": 0.6667
},
"flat": {
"n": 53,
"hit_rate": 0.5472
}
}
},
"mixed_pairs": 761,
"base": {
"n": 3139,
"hit_rate": 0.4393
},
"lift_pp": -1.4
},
"fh_1d_tf_1d_vs_4h": {
"agree": {
"n": 423,
"hit_rate": 0.3475,
"wilson_95": [
0.3037,
0.3941
],
"by_regime": {
"trend": {
"n": 165,
"hit_rate": 0.3394
},
"flat": {
"n": 258,
"hit_rate": 0.3527
}
}
},
"disagree": {
"n": 21,
"hit_rate": 0.5714,
"wilson_95": [
0.3655,
0.7553
],
"by_regime": {
"trend": {
"n": 6,
"hit_rate": 0.5
},
"flat": {
"n": 15,
"hit_rate": 0.6
}
}
},
"mixed_pairs": 221,
"base": {
"n": 665,
"hit_rate": 0.3624
},
"lift_pp": -1.5
}
},
"cross_fh": {
"fh_4h_confirmed_by_1h": {
"agree": {
"n": 2059,
"hit_rate": 0.4415,
"wilson_95": [
0.4202,
0.463
],
"by_regime": {
"trend": {
"n": 948,
"hit_rate": 0.5601
},
"flat": {
"n": 1111,
"hit_rate": 0.3402
}
}
},
"disagree": {
"n": 92,
"hit_rate": 0.5109,
"wilson_95": [
0.4104,
0.6105
],
"by_regime": {
"trend": {
"n": 28,
"hit_rate": 0.5357
},
"flat": {
"n": 64,
"hit_rate": 0.5
}
}
},
"mixed_pairs": 982,
"base": {
"n": 3139,
"hit_rate": 0.4393
},
"lift_pp": 0.2
},
"fh_1d_confirmed_by_4h": {
"agree": {
"n": 341,
"hit_rate": 0.3636,
"wilson_95": [
0.3144,
0.416
],
"by_regime": {
"trend": {
"n": 134,
"hit_rate": 0.3881
},
"flat": {
"n": 207,
"hit_rate": 0.3478
}
}
},
"disagree": {
"n": 28,
"hit_rate": 0.4286,
"wilson_95": [
0.2651,
0.6093
],
"by_regime": {
"trend": {
"n": 13,
"hit_rate": 0.3077
},
"flat": {
"n": 15,
"hit_rate": 0.5333
}
}
},
"mixed_pairs": 296,
"base": {
"n": 665,
"hit_rate": 0.3624
},
"lift_pp": 0.1
}
},
"regime_days": {
"2026-09-01": {
"move_pct": -1.5,
"regime": "trend"
},
"2026-09-02": {
"move_pct": -0.06,
"regime": "flat"
},
"2026-09-03": {
"move_pct": 4.98,
"regime": "trend"
},
"2026-09-04": {
"move_pct": -1.97,
"regime": "trend"
},
"2026-09-05": {
"move_pct": 0.24,
"regime": "flat"
},
"2026-09-06": {
"move_pct": 0.58,
"regime": "flat"
},
"2026-09-07": {
"move_pct": -1.52,
"regime": "trend"
},
"2026-09-08": {
"move_pct": -0.72,
"regime": "flat"
},
"2026-09-09": {
"move_pct": -0.25,
"regime": "flat"
},
"2026-09-10": {
"move_pct": -2.14,
"regime": "trend"
},
"2026-09-11": {
"move_pct": 0.85,
"regime": "flat"
},
"2026-09-12": {
"move_pct": 0.0,
"regime": "flat"
},
"2026-09-13": {
"move_pct": -0.57,
"regime": "flat"
},
"2026-09-14": {
"move_pct": 1.79,
"regime": "trend"
},
"2026-09-15": {
"move_pct": -3.35,
"regime": "trend"
},
"2026-09-16": {
"move_pct": 0.86,
"regime": "flat"
},
"2026-09-17": {
"move_pct": 0.21,
"regime": "flat"
},
"2026-09-18": {
"move_pct": 5.82,
"regime": "trend"
},
"2026-09-19": {
"move_pct": 0.46,
"regime": "flat"
},
"2026-09-20": {
"move_pct": 0.13,
"regime": "flat"
},
"2026-09-21": {
"move_pct": 6.39,
"regime": "trend"
},
"2026-09-22": {
"move_pct": -0.4,
"regime": "flat"
},
"2026-09-23": {
"move_pct": -2.04,
"regime": "trend"
},
"2026-09-24": {
"move_pct": 0.02,
"regime": "flat"
},
"2026-09-25": {
"move_pct": -0.46,
"regime": "flat"
},
"2026-09-26": {
"move_pct": 0.36,
"regime": "flat"
},
"2026-09-27": {
"move_pct": 0.2,
"regime": "flat"
},
"2026-09-28": {
"move_pct": -1.19,
"regime": "flat"
},
"2026-09-29": {
"move_pct": 0.22,
"regime": "flat"
},
"2026-09-30": {
"move_pct": -0.22,
"regime": "flat"
}
},
"side_mix": [
{
"model": "claude-fable-5",
"n": 5700,
"long": 0.34,
"short": 0.1767,
"sideways": 0.4833,
"ls_ratio": 1.92
},
{
"model": "claude-opus-5",
"n": 5696,
"long": 0.299,
"short": 0.1455,
"sideways": 0.5555,
"ls_ratio": 2.05
},
{
"model": "deepseek-v4-pro",
"n": 5675,
"long": 0.3299,
"short": 0.2608,
"sideways": 0.4093,
"ls_ratio": 1.26
},
{
"model": "gemini-3.1-pro",
"n": 5699,
"long": 0.3239,
"short": 0.2523,
"sideways": 0.4238,
"ls_ratio": 1.28
},
{
"model": "gpt-5.6-sol",
"n": 5697,
"long": 0.2777,
"short": 0.2722,
"sideways": 0.4501,
"ls_ratio": 1.02
},
{
"model": "grok-4.6",
"n": 5684,
"long": 0.193,
"short": 0.2065,
"sideways": 0.6005,
"ls_ratio": 0.93
},
{
"model": "qwen-3.8-max",
"n": 5655,
"long": 0.3326,
"short": 0.2271,
"sideways": 0.4403,
"ls_ratio": 1.46
}
],
"long_short_ratio_field": {
"value": 1.42,
"method": "mean of the per-model long-share / short-share ratios"
},
"consensus_skew": {
"cells_ge5_same_side": 2473,
"n_mature": 15689,
"hit_rate": 0.45
},
"fail_of_month": {
"model": "gpt-5.6-sol",
"symbol": "BNB",
"side": "short",
"confidence": 84,
"slot": "2026-09-16T09:01:00+00:00",
"fh": "1h",
"tf": "1h",
"exit_reason": "expiry",
"pnl_net_usd": -0.67
},
"counters": {
"forecasts_total": 44100,
"ok_in_gate": 39806,
"out_of_gate": 4185,
"out_of_gate_by_fh": {
"1w": 2096,
"1m": 2089
},
"invalid": 109,
"mature": 39806,
"mature_directional": 20682,
"mature_sideways": 19124,
"pending_next_issue": 0,
"pending_by_fh": {},
"late_closes": 0,
"uptime": [
{
"fh": "1h",
"tf": "1h",
"slots_seen": 720,
"slots_expected": 720
},
{
"fh": "4h",
"tf": "4h",
"slots_seen": 180,
"slots_expected": 180
},
{
"fh": "4h",
"tf": "1h",
"slots_seen": 180,
"slots_expected": 180
},
{
"fh": "1d",
"tf": "1d",
"slots_seen": 30,
"slots_expected": 30
},
{
"fh": "1d",
"tf": "4h",
"slots_seen": 30,
"slots_expected": 30
}
],
"regime_days": {
"2026-09-01": {
"move_pct": -1.5,
"regime": "trend"
},
"2026-09-02": {
"move_pct": -0.06,
"regime": "flat"
},
"2026-09-03": {
"move_pct": 4.98,
"regime": "trend"
},
"2026-09-04": {
"move_pct": -1.97,
"regime": "trend"
},
"2026-09-05": {
"move_pct": 0.24,
"regime": "flat"
},
"2026-09-06": {
"move_pct": 0.58,
"regime": "flat"
},
"2026-09-07": {
"move_pct": -1.52,
"regime": "trend"
},
"2026-09-08": {
"move_pct": -0.72,
"regime": "flat"
},
"2026-09-09": {
"move_pct": -0.25,
"regime": "flat"
},
"2026-09-10": {
"move_pct": -2.14,
"regime": "trend"
},
"2026-09-11": {
"move_pct": 0.85,
"regime": "flat"
},
"2026-09-12": {
"move_pct": 0.0,
"regime": "flat"
},
"2026-09-13": {
"move_pct": -0.57,
"regime": "flat"
},
"2026-09-14": {
"move_pct": 1.79,
"regime": "trend"
},
"2026-09-15": {
"move_pct": -3.35,
"regime": "trend"
},
"2026-09-16": {
"move_pct": 0.86,
"regime": "flat"
},
"2026-09-17": {
"move_pct": 0.21,
"regime": "flat"
},
"2026-09-18": {
"move_pct": 5.82,
"regime": "trend"
},
"2026-09-19": {
"move_pct": 0.46,
"regime": "flat"
},
"2026-09-20": {
"move_pct": 0.13,
"regime": "flat"
},
"2026-09-21": {
"move_pct": 6.39,
"regime": "trend"
},
"2026-09-22": {
"move_pct": -0.4,
"regime": "flat"
},
"2026-09-23": {
"move_pct": -2.04,
"regime": "trend"
},
"2026-09-24": {
"move_pct": 0.02,
"regime": "flat"
},
"2026-09-25": {
"move_pct": -0.46,
"regime": "flat"
},
"2026-09-26": {
"move_pct": 0.36,
"regime": "flat"
},
"2026-09-27": {
"move_pct": 0.2,
"regime": "flat"
},
"2026-09-28": {
"move_pct": -1.19,
"regime": "flat"
},
"2026-09-29": {
"move_pct": 0.22,
"regime": "flat"
},
"2026-09-30": {
"move_pct": -0.22,
"regime": "flat"
}
},
"grok_era": {
"ids_in_window": [
"grok-4.6"
],
"n_grok45": 0,
"n_grok46": 6300,
"n_grok45_rows": 0,
"n_grok46_rows": 6300,
"first_grok46_slot": "2026-09-01T00:01:00+00:00",
"last_grok45_slot": null,
"first_grok45_slot": null,
"flip_at": "2026-08-24T09:22:00+00:00",
"flip_inside_window": false,
"prev_window": {
"n_grok45_rows": 4520,
"n_grok46_rows": 1570
},
"merged_model_id": "grok-4.6",
"merged_label": "grok-4.6",
"lineage_ids": [
"grok-4.5",
"grok-4.6"
],
"lineage_source": "benchmarks.config.lineage_ids(head) for each of the 7 heads + MODEL_SUCCESSION[claude-opus-5=('claude-opus-4.8',), grok-4.6=('grok-4.5',), qwen-3.8-max=('qwen-3.7-max',)]",
"lineage_warnings": [],
"rows_remapped_a_plus_b": 4520,
"note": "window MONTH [2026-09-01T00:00:00+00:00, 2026-10-01T00:00:00+00:00): grok-4.5 rows 0, grok-4.6 rows 6300; flip 2026-08-24T09:22:00+00:00 is NOT inside this window. All per-model aggregates use the lineage splice grok-4.5 -> grok-4.6 (one row, labelled 'grok-4.6'). See lineage_era for the same facts across all 7 heads.",
"cells_with_two_grok_rows": 0
},
"lineage_era": {
"window_key": "MONTH",
"heads": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"lineage_map": {
"claude-opus-4.8": "claude-opus-5",
"grok-4.5": "grok-4.6",
"qwen-3.7-max": "qwen-3.8-max"
},
"legacy_by_head": {
"claude-fable-5": [],
"claude-opus-5": [
"claude-opus-4.8"
],
"deepseek-v4-pro": [],
"gemini-3.1-pro": [],
"gpt-5.6-sol": [],
"grok-4.6": [
"grok-4.5"
],
"qwen-3.8-max": [
"qwen-3.7-max"
]
},
"lineage_source": "benchmarks.config.lineage_ids(head) for each of the 7 heads + MODEL_SUCCESSION[claude-opus-5=('claude-opus-4.8',), grok-4.6=('grok-4.5',), qwen-3.8-max=('qwen-3.7-max',)]",
"lineage_resolved_heads": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"lineage_warnings": [],
"per_head": {
"claude-fable-5": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "claude-fable-5"
},
"claude-opus-5": {
"legacy_ids": [
"claude-opus-4.8"
],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"n_rows_remapped_prev_window": {
"claude-opus-4.8": 0
},
"by_legacy_id": {
"claude-opus-4.8": {
"n_rows_remapped": {
"MONTH": 0,
"prev_window": 0
},
"first_slot": null,
"last_slot": null,
"first_slot_prev_window": null,
"last_slot_prev_window": null
}
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "claude-opus-5"
},
"deepseek-v4-pro": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "deepseek-v4-pro"
},
"gemini-3.1-pro": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "gemini-3.1-pro"
},
"gpt-5.6-sol": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "gpt-5.6-sol"
},
"grok-4.6": {
"legacy_ids": [
"grok-4.5"
],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"n_rows_remapped_prev_window": {
"grok-4.5": 4520
},
"by_legacy_id": {
"grok-4.5": {
"n_rows_remapped": {
"MONTH": 0,
"prev_window": 4520
},
"first_slot": null,
"last_slot": null,
"first_slot_prev_window": "2026-08-01T00:01:00+00:00",
"last_slot_prev_window": "2026-08-24T09:01:00+00:00"
}
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "grok-4.6"
},
"qwen-3.8-max": {
"legacy_ids": [
"qwen-3.7-max"
],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"n_rows_remapped_prev_window": {
"qwen-3.7-max": 870
},
"by_legacy_id": {
"qwen-3.7-max": {
"n_rows_remapped": {
"MONTH": 0,
"prev_window": 870
},
"first_slot": null,
"last_slot": null,
"first_slot_prev_window": "2026-08-01T00:01:00+00:00",
"last_slot_prev_window": "2026-08-05T09:01:00+00:00"
}
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "qwen-3.8-max"
}
},
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"rows_remapped_window": 0,
"rows_remapped_prev_window": 5390,
"note": "Only the 7 current models are tracked; earlier versions are folded into their successors' lineage. Every per-model aggregate in this pack sees the merged head id; the raw id survives in model_raw. Labels and era spans are generated from the pulled rows, not hardcoded.",
"cells_with_two_rows_per_head": {
"claude-fable-5": 0,
"claude-opus-5": 0,
"deepseek-v4-pro": 0,
"gemini-3.1-pro": 0,
"gpt-5.6-sol": 0,
"grok-4.6": 0,
"qwen-3.8-max": 0
}
},
"series_density": {
"1d": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": true,
"n_slots": 30,
"n_rows": 2100,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"1h": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": true,
"n_slots": 720,
"n_rows": 25200,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"1m": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": false,
"n_slots": 30,
"n_rows": 2100,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"1w": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": false,
"n_slots": 30,
"n_rows": 2100,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"4h": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": true,
"n_slots": 180,
"n_rows": 12600,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
}
},
"model_roster": {
"current": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"current_count": 7,
"current_raw_ids_in_window": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"archived_legacy": [
"grok-4.5",
"qwen-3.7-max",
"claude-opus-4.8"
],
"archived_legacy_count": 3,
"archived_legacy_detail": [
{
"model": "grok-4.5",
"n_rows_since": 11128,
"first_slot": "2026-07-15T18:01:00+00:00",
"last_slot": "2026-08-24T09:01:00+00:00"
},
{
"model": "qwen-3.7-max",
"n_rows_since": 7479,
"first_slot": "2026-07-15T18:01:00+00:00",
"last_slot": "2026-08-05T09:01:00+00:00"
},
{
"model": "claude-opus-4.8",
"n_rows_since": 6315,
"first_slot": "2026-07-15T18:01:00+00:00",
"last_slot": "2026-07-30T08:01:00+00:00"
}
],
"archived_excluding_lineage_merged": [],
"tracked_total": 10,
"canon_expected": {
"current": 7,
"archived_legacy": 4,
"tracked_total": 11
},
"matches_canon_11_7_4": false,
"roster_rule_w213": "7 models tracked (current line-up; earlier versions folded into their successors' lineage)",
"canon_expected_w213": {
"current": 7
},
"heads_w213": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"matches_canon_7_current": true,
"heads_missing_in_window": [],
"heads_unexpected_in_window": [],
"roster_since": "2026-07-11T00:00:00+00:00",
"note": "current = merged-lineage model ids emitting inside the window (legacy ids are spliced into their successor); archived legacy = raw ids seen since 2026-07-11 that no longer emit in the window. Models retired before 2026-07-11 are not visible to this query. The canon checks are DESCRIPTIVE: if the data says otherwise, the data wins and is printed."
}
},
"uptime": [
{
"fh": "1h",
"tf": "1h",
"slots_seen": 720,
"slots_expected": 720
},
{
"fh": "4h",
"tf": "4h",
"slots_seen": 180,
"slots_expected": 180
},
{
"fh": "4h",
"tf": "1h",
"slots_seen": 180,
"slots_expected": 180
},
{
"fh": "1d",
"tf": "1d",
"slots_seen": 30,
"slots_expected": 30
},
{
"fh": "1d",
"tf": "4h",
"slots_seen": 30,
"slots_expected": 30
}
],
"ok_rate_by_model": {
"qwen-3.8-max": 0.9906,
"deepseek-v4-pro": 0.996,
"claude-fable-5": 0.9998,
"claude-opus-5": 0.9994,
"grok-4.6": 0.9975,
"gemini-3.1-pro": 0.9998,
"gpt-5.6-sol": 0.9995
},
"calibration_bridge": {
"best_calibrated_model": "qwen-3.8-max",
"brier": 0.2701,
"hit_rate_leader": "gemini-3.1-pro"
},
"asset_league": {
"by_symbol": [
{
"n": 4308,
"hit_rate": 0.4705,
"wilson_95": [
0.4556,
0.4854
],
"insufficient": false,
"symbol": "XRP",
"rank": 1
},
{
"n": 3987,
"hit_rate": 0.4597,
"wilson_95": [
0.4443,
0.4752
],
"insufficient": false,
"symbol": "BTC",
"rank": 2
},
{
"n": 3973,
"hit_rate": 0.4538,
"wilson_95": [
0.4384,
0.4693
],
"insufficient": false,
"symbol": "ETH",
"rank": 3
},
{
"n": 3994,
"hit_rate": 0.4512,
"wilson_95": [
0.4358,
0.4666
],
"insufficient": false,
"symbol": "BNB",
"rank": 4
},
{
"n": 4420,
"hit_rate": 0.4403,
"wilson_95": [
0.4257,
0.455
],
"insufficient": false,
"symbol": "SOL",
"rank": 5
}
],
"by_symbol_fh": {
"BTC": {
"fh_1h": {
"n": 2538,
"hit_rate": 0.4831,
"wilson_95": [
0.4637,
0.5025
],
"insufficient": false
},
"fh_4h": {
"n": 1243,
"hit_rate": 0.4401,
"wilson_95": [
0.4127,
0.4678
],
"insufficient": false
},
"fh_1d": {
"n": 206,
"hit_rate": 0.2913,
"wilson_95": [
0.2335,
0.3567
],
"insufficient": false
}
},
"ETH": {
"fh_1h": {
"n": 2525,
"hit_rate": 0.4689,
"wilson_95": [
0.4495,
0.4884
],
"insufficient": false
},
"fh_4h": {
"n": 1223,
"hit_rate": 0.4399,
"wilson_95": [
0.4123,
0.4679
],
"insufficient": false
},
"fh_1d": {
"n": 225,
"hit_rate": 0.36,
"wilson_95": [
0.3001,
0.4246
],
"insufficient": false
}
},
"SOL": {
"fh_1h": {
"n": 2720,
"hit_rate": 0.4544,
"wilson_95": [
0.4358,
0.4732
],
"insufficient": false
},
"fh_4h": {
"n": 1403,
"hit_rate": 0.4355,
"wilson_95": [
0.4098,
0.4616
],
"insufficient": false
},
"fh_1d": {
"n": 297,
"hit_rate": 0.3333,
"wilson_95": [
0.2821,
0.3888
],
"insufficient": false
}
},
"BNB": {
"fh_1h": {
"n": 2499,
"hit_rate": 0.487,
"wilson_95": [
0.4674,
0.5066
],
"insufficient": false
},
"fh_4h": {
"n": 1249,
"hit_rate": 0.3947,
"wilson_95": [
0.368,
0.4221
],
"insufficient": false
},
"fh_1d": {
"n": 246,
"hit_rate": 0.374,
"wilson_95": [
0.3159,
0.436
],
"insufficient": false
}
},
"XRP": {
"fh_1h": {
"n": 2726,
"hit_rate": 0.4762,
"wilson_95": [
0.4575,
0.4949
],
"insufficient": false
},
"fh_4h": {
"n": 1343,
"hit_rate": 0.4676,
"wilson_95": [
0.4411,
0.4943
],
"insufficient": false
},
"fh_1d": {
"n": 239,
"hit_rate": 0.4226,
"wilson_95": [
0.3617,
0.486
],
"insufficient": false
}
}
},
"by_week_symbol_published": {
"W1": [
{
"symbol": "BNB",
"n": 916,
"hit_rate": 0.4727,
"wilson_95": [
0.4406,
0.5051
],
"rank": 1
},
{
"symbol": "XRP",
"n": 909,
"hit_rate": 0.4422,
"wilson_95": [
0.4103,
0.4747
],
"rank": 2
},
{
"symbol": "ETH",
"n": 982,
"hit_rate": 0.4308,
"wilson_95": [
0.4001,
0.4619
],
"rank": 3
},
{
"symbol": "BTC",
"n": 780,
"hit_rate": 0.4064,
"wilson_95": [
0.3725,
0.4413
],
"rank": 4
},
{
"symbol": "SOL",
"n": 933,
"hit_rate": 0.3869,
"wilson_95": [
0.3562,
0.4186
],
"rank": 5
}
],
"W2": [
{
"symbol": "BNB",
"n": 963,
"hit_rate": 0.4694,
"wilson_95": [
0.438,
0.5009
],
"rank": 1
},
{
"symbol": "XRP",
"n": 937,
"hit_rate": 0.4216,
"wilson_95": [
0.3903,
0.4534
],
"rank": 2
},
{
"symbol": "BTC",
"n": 891,
"hit_rate": 0.4141,
"wilson_95": [
0.3822,
0.4468
],
"rank": 3
},
{
"symbol": "SOL",
"n": 875,
"hit_rate": 0.36,
"wilson_95": [
0.3289,
0.3924
],
"rank": 4
},
{
"symbol": "ETH",
"n": 802,
"hit_rate": 0.2955,
"wilson_95": [
0.265,
0.328
],
"rank": 5
}
],
"W3": [
{
"symbol": "ETH",
"n": 1116,
"hit_rate": 0.5717,
"wilson_95": [
0.5425,
0.6004
],
"rank": 1
},
{
"symbol": "XRP",
"n": 1108,
"hit_rate": 0.5036,
"wilson_95": [
0.4742,
0.533
],
"rank": 2
},
{
"symbol": "BTC",
"n": 1163,
"hit_rate": 0.503,
"wilson_95": [
0.4743,
0.5317
],
"rank": 3
},
{
"symbol": "SOL",
"n": 1156,
"hit_rate": 0.4948,
"wilson_95": [
0.4661,
0.5236
],
"rank": 4
},
{
"symbol": "BNB",
"n": 1055,
"hit_rate": 0.4626,
"wilson_95": [
0.4327,
0.4927
],
"rank": 5
}
],
"W4": [
{
"symbol": "XRP",
"n": 1165,
"hit_rate": 0.4996,
"wilson_95": [
0.4709,
0.5282
],
"rank": 1
},
{
"symbol": "BTC",
"n": 919,
"hit_rate": 0.4886,
"wilson_95": [
0.4564,
0.5209
],
"rank": 2
},
{
"symbol": "SOL",
"n": 1190,
"hit_rate": 0.4874,
"wilson_95": [
0.4591,
0.5158
],
"rank": 3
},
{
"symbol": "ETH",
"n": 891,
"hit_rate": 0.4579,
"wilson_95": [
0.4254,
0.4907
],
"rank": 4
},
{
"symbol": "BNB",
"n": 793,
"hit_rate": 0.3821,
"wilson_95": [
0.3489,
0.4164
],
"rank": 5
}
]
},
"by_week_symbol": {
"W1": [
{
"n": 811,
"hit_rate": 0.5055,
"wilson_95": [
0.4712,
0.5399
],
"insufficient": false,
"symbol": "BNB",
"rank": 1
},
{
"n": 771,
"hit_rate": 0.4527,
"wilson_95": [
0.4178,
0.4879
],
"insufficient": false,
"symbol": "XRP",
"rank": 2
},
{
"n": 821,
"hit_rate": 0.4348,
"wilson_95": [
0.4013,
0.469
],
"insufficient": false,
"symbol": "ETH",
"rank": 3
},
{
"n": 622,
"hit_rate": 0.4196,
"wilson_95": [
0.3814,
0.4588
],
"insufficient": false,
"symbol": "BTC",
"rank": 4
},
{
"n": 807,
"hit_rate": 0.4027,
"wilson_95": [
0.3694,
0.4369
],
"insufficient": false,
"symbol": "SOL",
"rank": 5
}
],
"W2": [
{
"n": 963,
"hit_rate": 0.4694,
"wilson_95": [
0.438,
0.5009
],
"insufficient": false,
"symbol": "BNB",
"rank": 1
},
{
"n": 937,
"hit_rate": 0.4216,
"wilson_95": [
0.3903,
0.4534
],
"insufficient": false,
"symbol": "XRP",
"rank": 2
},
{
"n": 891,
"hit_rate": 0.4141,
"wilson_95": [
0.3822,
0.4468
],
"insufficient": false,
"symbol": "BTC",
"rank": 3
},
{
"n": 875,
"hit_rate": 0.36,
"wilson_95": [
0.3289,
0.3924
],
"insufficient": false,
"symbol": "SOL",
"rank": 4
},
{
"n": 802,
"hit_rate": 0.2955,
"wilson_95": [
0.265,
0.328
],
"insufficient": false,
"symbol": "ETH",
"rank": 5
}
],
"W3": [
{
"n": 1116,
"hit_rate": 0.5717,
"wilson_95": [
0.5425,
0.6004
],
"insufficient": false,
"symbol": "ETH",
"rank": 1
},
{
"n": 1108,
"hit_rate": 0.5036,
"wilson_95": [
0.4742,
0.533
],
"insufficient": false,
"symbol": "XRP",
"rank": 2
},
{
"n": 1163,
"hit_rate": 0.503,
"wilson_95": [
0.4743,
0.5317
],
"insufficient": false,
"symbol": "BTC",
"rank": 3
},
{
"n": 1156,
"hit_rate": 0.4948,
"wilson_95": [
0.4661,
0.5236
],
"insufficient": false,
"symbol": "SOL",
"rank": 4
},
{
"n": 1055,
"hit_rate": 0.4626,
"wilson_95": [
0.4327,
0.4927
],
"insufficient": false,
"symbol": "BNB",
"rank": 5
}
],
"W4": [
{
"n": 1165,
"hit_rate": 0.4996,
"wilson_95": [
0.4709,
0.5282
],
"insufficient": false,
"symbol": "XRP",
"rank": 1
},
{
"n": 919,
"hit_rate": 0.4886,
"wilson_95": [
0.4564,
0.5209
],
"insufficient": false,
"symbol": "BTC",
"rank": 2
},
{
"n": 1190,
"hit_rate": 0.4874,
"wilson_95": [
0.4591,
0.5158
],
"insufficient": false,
"symbol": "SOL",
"rank": 3
},
{
"n": 891,
"hit_rate": 0.4579,
"wilson_95": [
0.4254,
0.4907
],
"insufficient": false,
"symbol": "ETH",
"rank": 4
},
{
"n": 793,
"hit_rate": 0.3821,
"wilson_95": [
0.3489,
0.4164
],
"insufficient": false,
"symbol": "BNB",
"rank": 5
}
]
},
"by_model_symbol": {
"claude-fable-5": {
"BTC": {
"n": 578,
"hit_rate": 0.4602
},
"ETH": {
"n": 552,
"hit_rate": 0.4583
},
"SOL": {
"n": 658,
"hit_rate": 0.4377
},
"BNB": {
"n": 569,
"hit_rate": 0.4482
},
"XRP": {
"n": 588,
"hit_rate": 0.466
}
},
"claude-opus-5": {
"BTC": {
"n": 479,
"hit_rate": 0.4509
},
"ETH": {
"n": 498,
"hit_rate": 0.4558
},
"SOL": {
"n": 533,
"hit_rate": 0.439
},
"BNB": {
"n": 479,
"hit_rate": 0.453
},
"XRP": {
"n": 543,
"hit_rate": 0.4696
}
},
"deepseek-v4-pro": {
"BTC": {
"n": 664,
"hit_rate": 0.4578
},
"ETH": {
"n": 660,
"hit_rate": 0.4591
},
"SOL": {
"n": 691,
"hit_rate": 0.4327
},
"BNB": {
"n": 652,
"hit_rate": 0.4617
},
"XRP": {
"n": 685,
"hit_rate": 0.4526
}
},
"gemini-3.1-pro": {
"BTC": {
"n": 640,
"hit_rate": 0.4906
},
"ETH": {
"n": 622,
"hit_rate": 0.4759
},
"SOL": {
"n": 702,
"hit_rate": 0.4943
},
"BNB": {
"n": 647,
"hit_rate": 0.4745
},
"XRP": {
"n": 673,
"hit_rate": 0.4889
}
},
"gpt-5.6-sol": {
"BTC": {
"n": 609,
"hit_rate": 0.4401
},
"ETH": {
"n": 607,
"hit_rate": 0.4366
},
"SOL": {
"n": 657,
"hit_rate": 0.4247
},
"BNB": {
"n": 602,
"hit_rate": 0.4468
},
"XRP": {
"n": 658,
"hit_rate": 0.4742
}
},
"grok-4.6": {
"BTC": {
"n": 407,
"hit_rate": 0.4324
},
"ETH": {
"n": 428,
"hit_rate": 0.4416
},
"SOL": {
"n": 503,
"hit_rate": 0.4095
},
"BNB": {
"n": 451,
"hit_rate": 0.4501
},
"XRP": {
"n": 482,
"hit_rate": 0.4855
}
},
"qwen-3.8-max": {
"BTC": {
"n": 610,
"hit_rate": 0.4738
},
"ETH": {
"n": 606,
"hit_rate": 0.4455
},
"SOL": {
"n": 676,
"hit_rate": 0.4334
},
"BNB": {
"n": 594,
"hit_rate": 0.4209
},
"XRP": {
"n": 679,
"hit_rate": 0.461
}
}
},
"sol_hypothesis": {
"symbol": "SOL",
"month_rank": 5,
"month_n": 4420,
"month_hit_rate": 0.4403,
"weeks_ranked_first": [],
"weeks_rank": {
"W1": 5,
"W2": 4,
"W3": 4,
"W4": 3
},
"weeks_rank_published": {
"W1": 5,
"W2": 4,
"W3": 4,
"W4": 3
},
"best_fh": "fh_1h",
"best_fh_hit_rate": 0.4544,
"note": "SOL: month rank 5 of 5 symbols (n=4420, hit_rate=0.4403, wilson_95=[0.4257, 0.455]); rank 1 in 0 of 4 weeks (none); highest hit_rate by horizon fh_1h (n=2720, hit_rate=0.4544). Ranks from the month pull split by slot week; published ranks from the weekly packs' ticker_of_week."
},
"rank_rule": "rank 1 = highest hit_rate; ties broken by n desc; empty cells rank last",
"source": "month_pull_split_by_slot_week (cutoff 2026-10-01T16:00:00+00:00)",
"by_week_symbol_published_source": "weeks.W*.model_watch.ticker_of_week (frozen weekly cutoffs)"
},
"rank_stability": {
"by_week": {
"W1": [
{
"model": "gemini-3.1-pro",
"rank": 1,
"n": 761,
"hit_rate": 0.4652,
"wilson_95": [
0.43,
0.5007
],
"insufficient": false
},
{
"model": "deepseek-v4-pro",
"rank": 2,
"n": 724,
"hit_rate": 0.442,
"wilson_95": [
0.4062,
0.4784
],
"insufficient": false
},
{
"model": "claude-fable-5",
"rank": 3,
"n": 628,
"hit_rate": 0.4283,
"wilson_95": [
0.3902,
0.4674
],
"insufficient": false
},
{
"model": "qwen-3.8-max",
"rank": 4,
"n": 697,
"hit_rate": 0.4247,
"wilson_95": [
0.3885,
0.4617
],
"insufficient": false
},
{
"model": "claude-opus-5",
"rank": 5,
"n": 545,
"hit_rate": 0.4239,
"wilson_95": [
0.383,
0.4657
],
"insufficient": false
},
{
"model": "gpt-5.6-sol",
"rank": 6,
"n": 689,
"hit_rate": 0.4165,
"wilson_95": [
0.3803,
0.4537
],
"insufficient": false
},
{
"model": "grok-4.6",
"rank": 7,
"n": 476,
"hit_rate": 0.3761,
"wilson_95": [
0.3337,
0.4204
],
"insufficient": false
}
],
"W2": [
{
"model": "gemini-3.1-pro",
"rank": 1,
"n": 673,
"hit_rate": 0.4175,
"wilson_95": [
0.3808,
0.4552
],
"insufficient": false
},
{
"model": "deepseek-v4-pro",
"rank": 2,
"n": 727,
"hit_rate": 0.414,
"wilson_95": [
0.3788,
0.4502
],
"insufficient": false
},
{
"model": "gpt-5.6-sol",
"rank": 3,
"n": 709,
"hit_rate": 0.4006,
"wilson_95": [
0.3651,
0.4371
],
"insufficient": false
},
{
"model": "qwen-3.8-max",
"rank": 4,
"n": 707,
"hit_rate": 0.4003,
"wilson_95": [
0.3648,
0.4368
],
"insufficient": false
},
{
"model": "grok-4.6",
"rank": 5,
"n": 518,
"hit_rate": 0.3958,
"wilson_95": [
0.3546,
0.4385
],
"insufficient": false
},
{
"model": "claude-fable-5",
"rank": 6,
"n": 624,
"hit_rate": 0.3654,
"wilson_95": [
0.3285,
0.4039
],
"insufficient": false
},
{
"model": "claude-opus-5",
"rank": 7,
"n": 510,
"hit_rate": 0.3647,
"wilson_95": [
0.3241,
0.4073
],
"insufficient": false
}
],
"W3": [
{
"model": "gemini-3.1-pro",
"rank": 1,
"n": 883,
"hit_rate": 0.5289,
"wilson_95": [
0.4959,
0.5616
],
"insufficient": false
},
{
"model": "claude-fable-5",
"rank": 2,
"n": 799,
"hit_rate": 0.5232,
"wilson_95": [
0.4885,
0.5576
],
"insufficient": false
},
{
"model": "claude-opus-5",
"rank": 3,
"n": 711,
"hit_rate": 0.5162,
"wilson_95": [
0.4795,
0.5527
],
"insufficient": false
},
{
"model": "qwen-3.8-max",
"rank": 4,
"n": 819,
"hit_rate": 0.5079,
"wilson_95": [
0.4737,
0.5421
],
"insufficient": false
},
{
"model": "gpt-5.6-sol",
"rank": 5,
"n": 821,
"hit_rate": 0.497,
"wilson_95": [
0.4628,
0.5311
],
"insufficient": false
},
{
"model": "grok-4.6",
"rank": 6,
"n": 629,
"hit_rate": 0.496,
"wilson_95": [
0.4571,
0.535
],
"insufficient": false
},
{
"model": "deepseek-v4-pro",
"rank": 7,
"n": 936,
"hit_rate": 0.484,
"wilson_95": [
0.4521,
0.516
],
"insufficient": false
}
],
"W4": [
{
"model": "gemini-3.1-pro",
"rank": 1,
"n": 794,
"hit_rate": 0.5139,
"wilson_95": [
0.4791,
0.5485
],
"insufficient": false
},
{
"model": "claude-opus-5",
"rank": 2,
"n": 620,
"hit_rate": 0.4758,
"wilson_95": [
0.4368,
0.5151
],
"insufficient": false
},
{
"model": "grok-4.6",
"rank": 3,
"n": 537,
"hit_rate": 0.4693,
"wilson_95": [
0.4274,
0.5116
],
"insufficient": false
},
{
"model": "gpt-5.6-sol",
"rank": 4,
"n": 744,
"hit_rate": 0.4677,
"wilson_95": [
0.4321,
0.5037
],
"insufficient": false
},
{
"model": "claude-fable-5",
"rank": 5,
"n": 736,
"hit_rate": 0.4674,
"wilson_95": [
0.4316,
0.5035
],
"insufficient": false
},
{
"model": "qwen-3.8-max",
"rank": 6,
"n": 750,
"hit_rate": 0.4453,
"wilson_95": [
0.4101,
0.4811
],
"insufficient": false
},
{
"model": "deepseek-v4-pro",
"rank": 7,
"n": 777,
"hit_rate": 0.4389,
"wilson_95": [
0.4044,
0.474
],
"insufficient": false
}
]
},
"month_rank": [
{
"model": "gemini-3.1-pro",
"rank": 1,
"n": 3284,
"hit_rate": 0.4851,
"wilson_95": [
0.468,
0.5022
],
"insufficient": false
},
{
"model": "claude-opus-5",
"rank": 2,
"n": 2532,
"hit_rate": 0.4538,
"wilson_95": [
0.4345,
0.4732
],
"insufficient": false
},
{
"model": "claude-fable-5",
"rank": 3,
"n": 2945,
"hit_rate": 0.4537,
"wilson_95": [
0.4357,
0.4717
],
"insufficient": false
},
{
"model": "deepseek-v4-pro",
"rank": 4,
"n": 3352,
"hit_rate": 0.4526,
"wilson_95": [
0.4358,
0.4695
],
"insufficient": false
},
{
"model": "qwen-3.8-max",
"rank": 5,
"n": 3165,
"hit_rate": 0.4471,
"wilson_95": [
0.4298,
0.4645
],
"insufficient": false
},
{
"model": "gpt-5.6-sol",
"rank": 6,
"n": 3133,
"hit_rate": 0.4446,
"wilson_95": [
0.4273,
0.4621
],
"insufficient": false
},
{
"model": "grok-4.6",
"rank": 7,
"n": 2271,
"hit_rate": 0.4439,
"wilson_95": [
0.4235,
0.4644
],
"insufficient": false
}
],
"per_model": {
"claude-fable-5": {
"ranks": [
3,
6,
2,
5
],
"n": [
628,
624,
799,
736
],
"hit_rates": [
0.4283,
0.3654,
0.5232,
0.4674
],
"mean_rank": 4.0,
"best": 2,
"worst": 6,
"range": 4,
"month_rank": 3,
"month_hit_rate": 0.4537
},
"claude-opus-5": {
"ranks": [
5,
7,
3,
2
],
"n": [
545,
510,
711,
620
],
"hit_rates": [
0.4239,
0.3647,
0.5162,
0.4758
],
"mean_rank": 4.25,
"best": 2,
"worst": 7,
"range": 5,
"month_rank": 2,
"month_hit_rate": 0.4538
},
"deepseek-v4-pro": {
"ranks": [
2,
2,
7,
7
],
"n": [
724,
727,
936,
777
],
"hit_rates": [
0.442,
0.414,
0.484,
0.4389
],
"mean_rank": 4.5,
"best": 2,
"worst": 7,
"range": 5,
"month_rank": 4,
"month_hit_rate": 0.4526
},
"gemini-3.1-pro": {
"ranks": [
1,
1,
1,
1
],
"n": [
761,
673,
883,
794
],
"hit_rates": [
0.4652,
0.4175,
0.5289,
0.5139
],
"mean_rank": 1.0,
"best": 1,
"worst": 1,
"range": 0,
"month_rank": 1,
"month_hit_rate": 0.4851
},
"gpt-5.6-sol": {
"ranks": [
6,
3,
5,
4
],
"n": [
689,
709,
821,
744
],
"hit_rates": [
0.4165,
0.4006,
0.497,
0.4677
],
"mean_rank": 4.5,
"best": 3,
"worst": 6,
"range": 3,
"month_rank": 6,
"month_hit_rate": 0.4446
},
"grok-4.6": {
"ranks": [
7,
5,
6,
3
],
"n": [
476,
518,
629,
537
],
"hit_rates": [
0.3761,
0.3958,
0.496,
0.4693
],
"mean_rank": 5.25,
"best": 3,
"worst": 7,
"range": 4,
"month_rank": 7,
"month_hit_rate": 0.4439
},
"qwen-3.8-max": {
"ranks": [
4,
4,
4,
6
],
"n": [
697,
707,
819,
750
],
"hit_rates": [
0.4247,
0.4003,
0.5079,
0.4453
],
"mean_rank": 4.5,
"best": 4,
"worst": 6,
"range": 2,
"month_rank": 5,
"month_hit_rate": 0.4471
}
},
"spearman_consecutive": [
{
"pair": "W1-W2",
"rho": 0.5357
},
{
"pair": "W2-W3",
"rho": -0.1071
},
{
"pair": "W3-W4",
"rho": 0.5714
}
],
"spearman_vs_month": {
"W1": 0.75,
"W2": 0.0714,
"W3": 0.75,
"W4": 0.3929
},
"top_model_by_week": {
"W1": "gemini-3.1-pro",
"W2": "gemini-3.1-pro",
"W3": "gemini-3.1-pro",
"W4": "gemini-3.1-pro"
},
"n_distinct_leaders": 1,
"rank_rule": "rank = position in the pack's leaderboard, which w201 sorts by hit_rate desc",
"spearman_note": "rho computed here (no scipy on the box); ties get average ranks",
"source": "weekly packs model_watch.leaderboard.rows (frozen cutoffs) + MONTH pack",
"widths_mm_used": [
40,
17,
17,
17,
17,
20,
24,
22
]
},
"regime_weeks": {
"per_week": {
"W1": {
"label_window": "Aug 31-Sep 6",
"days_in_week": 7,
"days_with_price": 6,
"mean_abs_1d_move_pct": 1.555,
"max_abs_1d_move_pct": 4.98,
"trend_days": 3,
"flat_days": 3,
"daily_moves_pct": {
"2026-09-01": -1.5,
"2026-09-02": -0.06,
"2026-09-03": 4.98,
"2026-09-04": -1.97,
"2026-09-05": 0.24,
"2026-09-06": 0.58
},
"label": "volatile"
},
"W2": {
"label_window": "Sep 7-13",
"days_in_week": 7,
"days_with_price": 7,
"mean_abs_1d_move_pct": 0.864,
"max_abs_1d_move_pct": 2.14,
"trend_days": 2,
"flat_days": 5,
"daily_moves_pct": {
"2026-09-07": -1.52,
"2026-09-08": -0.72,
"2026-09-09": -0.25,
"2026-09-10": -2.14,
"2026-09-11": 0.85,
"2026-09-12": 0.0,
"2026-09-13": -0.57
},
"label": "calm"
},
"W3": {
"label_window": "Sep 14-20",
"days_in_week": 7,
"days_with_price": 7,
"mean_abs_1d_move_pct": 1.803,
"max_abs_1d_move_pct": 5.82,
"trend_days": 3,
"flat_days": 4,
"daily_moves_pct": {
"2026-09-14": 1.79,
"2026-09-15": -3.35,
"2026-09-16": 0.86,
"2026-09-17": 0.21,
"2026-09-18": 5.82,
"2026-09-19": 0.46,
"2026-09-20": 0.13
},
"label": "volatile"
},
"W4": {
"label_window": "Sep 21-27",
"days_in_week": 7,
"days_with_price": 7,
"mean_abs_1d_move_pct": 1.41,
"max_abs_1d_move_pct": 6.39,
"trend_days": 2,
"flat_days": 5,
"daily_moves_pct": {
"2026-09-21": 6.39,
"2026-09-22": -0.4,
"2026-09-23": -2.04,
"2026-09-24": 0.02,
"2026-09-25": -0.46,
"2026-09-26": 0.36,
"2026-09-27": 0.2
},
"label": "calm"
}
},
"week_labels": {
"W1": "volatile",
"W2": "calm",
"W3": "volatile",
"W4": "calm"
},
"per_model": {
"claude-fable-5": {
"calm": {
"n": 1360,
"hit_rate": 0.4206,
"wilson_95": [
0.3946,
0.447
],
"insufficient": false
},
"volatile": {
"n": 1325,
"hit_rate": 0.4906,
"wilson_95": [
0.4637,
0.5175
],
"insufficient": false
},
"diff_pp": 7.0
},
"claude-opus-5": {
"calm": {
"n": 1130,
"hit_rate": 0.4257,
"wilson_95": [
0.3971,
0.4547
],
"insufficient": false
},
"volatile": {
"n": 1173,
"hit_rate": 0.4834,
"wilson_95": [
0.4549,
0.512
],
"insufficient": false
},
"diff_pp": 5.8
},
"deepseek-v4-pro": {
"calm": {
"n": 1504,
"hit_rate": 0.4269,
"wilson_95": [
0.4021,
0.452
],
"insufficient": false
},
"volatile": {
"n": 1551,
"hit_rate": 0.4771,
"wilson_95": [
0.4523,
0.502
],
"insufficient": false
},
"diff_pp": 5.0
},
"gemini-3.1-pro": {
"calm": {
"n": 1467,
"hit_rate": 0.4697,
"wilson_95": [
0.4442,
0.4953
],
"insufficient": false
},
"volatile": {
"n": 1530,
"hit_rate": 0.5085,
"wilson_95": [
0.4835,
0.5335
],
"insufficient": false
},
"diff_pp": 3.9
},
"gpt-5.6-sol": {
"calm": {
"n": 1453,
"hit_rate": 0.435,
"wilson_95": [
0.4097,
0.4606
],
"insufficient": false
},
"volatile": {
"n": 1397,
"hit_rate": 0.4717,
"wilson_95": [
0.4457,
0.4979
],
"insufficient": false
},
"diff_pp": 3.7
},
"grok-4.6": {
"calm": {
"n": 1055,
"hit_rate": 0.4332,
"wilson_95": [
0.4036,
0.4633
],
"insufficient": false
},
"volatile": {
"n": 1032,
"hit_rate": 0.4535,
"wilson_95": [
0.4233,
0.484
],
"insufficient": false
},
"diff_pp": 2.0
},
"qwen-3.8-max": {
"calm": {
"n": 1457,
"hit_rate": 0.4235,
"wilson_95": [
0.3983,
0.449
],
"insufficient": false
},
"volatile": {
"n": 1422,
"hit_rate": 0.4789,
"wilson_95": [
0.453,
0.5049
],
"insufficient": false
},
"diff_pp": 5.5
}
},
"field": {
"calm": {
"n": 9426,
"hit_rate": 0.4339,
"wilson_95": [
0.4239,
0.4439
],
"insufficient": false
},
"volatile": {
"n": 9430,
"hit_rate": 0.4818,
"wilson_95": [
0.4717,
0.4919
],
"insufficient": false
},
"diff_pp": 4.8
},
"rule": "label = 'volatile' if mean_abs_1d_move_pct >= median of the four weeks else 'calm'",
"median_mean_abs_1d_move_pct": 1.482,
"diff_definition": "diff_pp = volatile hit_rate - calm hit_rate, in percentage points",
"trend_day_rule": "w201 regime_days: |BTC 1d move| >= 1.5% -> trend, else flat",
"rows_outside_weeks": 1826
},
"prior_window_model_watch": {
"leaderboard": [
{
"model": "claude-opus-5",
"n": 2281,
"hit_rate": 0.4779,
"wilson_95": [
0.4574,
0.4984
],
"coverage": 0.3889,
"insufficient": false
},
{
"model": "gemini-3.1-pro",
"n": 3033,
"hit_rate": 0.4761,
"wilson_95": [
0.4584,
0.4939
],
"coverage": 0.5177,
"insufficient": false
},
{
"model": "claude-fable-5",
"n": 2812,
"hit_rate": 0.4673,
"wilson_95": [
0.4489,
0.4858
],
"coverage": 0.48,
"insufficient": false
},
{
"model": "gpt-5.6-sol",
"n": 2970,
"hit_rate": 0.4586,
"wilson_95": [
0.4407,
0.4765
],
"coverage": 0.5064,
"insufficient": false
},
{
"model": "qwen-3.8-max",
"n": 3001,
"hit_rate": 0.4565,
"wilson_95": [
0.4388,
0.4744
],
"coverage": 0.5166,
"insufficient": false
},
{
"model": "deepseek-v4-pro",
"n": 2578,
"hit_rate": 0.4554,
"wilson_95": [
0.4362,
0.4747
],
"coverage": 0.4397,
"insufficient": false
},
{
"model": "grok-4.6",
"n": 2708,
"hit_rate": 0.4527,
"wilson_95": [
0.4341,
0.4715
],
"coverage": 0.4621,
"insufficient": false
}
],
"by_symbol": [
{
"symbol": "BNB",
"n": 3857,
"hit_rate": 0.4807,
"wilson_95": [
0.4649,
0.4965
]
},
{
"symbol": "SOL",
"n": 3928,
"hit_rate": 0.4733,
"wilson_95": [
0.4577,
0.4889
]
},
{
"symbol": "XRP",
"n": 3835,
"hit_rate": 0.4673,
"wilson_95": [
0.4515,
0.4831
]
},
{
"symbol": "BTC",
"n": 4068,
"hit_rate": 0.4648,
"wilson_95": [
0.4496,
0.4802
]
},
{
"symbol": "ETH",
"n": 3695,
"hit_rate": 0.4287,
"wilson_95": [
0.4128,
0.4447
]
}
],
"by_model_symbol": {
"claude-fable-5": {
"BTC": {
"n": 598,
"hit_rate": 0.4599
},
"ETH": {
"n": 529,
"hit_rate": 0.4405
},
"SOL": {
"n": 604,
"hit_rate": 0.4851
},
"BNB": {
"n": 582,
"hit_rate": 0.4759
},
"XRP": {
"n": 499,
"hit_rate": 0.4729
}
},
"claude-opus-5": {
"BTC": {
"n": 484,
"hit_rate": 0.4669
},
"ETH": {
"n": 430,
"hit_rate": 0.4651
},
"SOL": {
"n": 482,
"hit_rate": 0.5041
},
"BNB": {
"n": 447,
"hit_rate": 0.4855
},
"XRP": {
"n": 438,
"hit_rate": 0.4658
}
},
"deepseek-v4-pro": {
"BTC": {
"n": 552,
"hit_rate": 0.4656
},
"ETH": {
"n": 522,
"hit_rate": 0.4119
},
"SOL": {
"n": 499,
"hit_rate": 0.4629
},
"BNB": {
"n": 500,
"hit_rate": 0.472
},
"XRP": {
"n": 505,
"hit_rate": 0.4653
}
},
"gemini-3.1-pro": {
"BTC": {
"n": 647,
"hit_rate": 0.4683
},
"ETH": {
"n": 552,
"hit_rate": 0.4511
},
"SOL": {
"n": 626,
"hit_rate": 0.4665
},
"BNB": {
"n": 568,
"hit_rate": 0.4965
},
"XRP": {
"n": 640,
"hit_rate": 0.4969
}
},
"gpt-5.6-sol": {
"BTC": {
"n": 631,
"hit_rate": 0.4643
},
"ETH": {
"n": 574,
"hit_rate": 0.4164
},
"SOL": {
"n": 572,
"hit_rate": 0.4755
},
"BNB": {
"n": 604,
"hit_rate": 0.4669
},
"XRP": {
"n": 589,
"hit_rate": 0.4686
}
},
"grok-4.6": {
"BTC": {
"n": 544,
"hit_rate": 0.4779
},
"ETH": {
"n": 512,
"hit_rate": 0.4199
},
"SOL": {
"n": 535,
"hit_rate": 0.4486
},
"BNB": {
"n": 569,
"hit_rate": 0.4728
},
"XRP": {
"n": 548,
"hit_rate": 0.4416
}
},
"qwen-3.8-max": {
"BTC": {
"n": 612,
"hit_rate": 0.4526
},
"ETH": {
"n": 576,
"hit_rate": 0.4045
},
"SOL": {
"n": 610,
"hit_rate": 0.4721
},
"BNB": {
"n": 587,
"hit_rate": 0.4957
},
"XRP": {
"n": 616,
"hit_rate": 0.4562
}
}
},
"side_mix": [
{
"model": "claude-fable-5",
"n": 5858,
"long": 0.3564,
"short": 0.1236,
"sideways": 0.52
},
{
"model": "claude-opus-5",
"n": 5865,
"long": 0.2921,
"short": 0.0968,
"sideways": 0.6111
},
{
"model": "deepseek-v4-pro",
"n": 5863,
"long": 0.2555,
"short": 0.1842,
"sideways": 0.5603
},
{
"model": "gemini-3.1-pro",
"n": 5859,
"long": 0.3173,
"short": 0.2004,
"sideways": 0.4823
},
{
"model": "gpt-5.6-sol",
"n": 5865,
"long": 0.2781,
"short": 0.2283,
"sideways": 0.4936
},
{
"model": "grok-4.6",
"n": 5860,
"long": 0.2396,
"short": 0.2225,
"sideways": 0.5379
},
{
"model": "qwen-3.8-max",
"n": 5809,
"long": 0.3217,
"short": 0.1949,
"sideways": 0.4834
}
],
"symbols": [
"BTC",
"ETH",
"SOL",
"BNB",
"XRP"
],
"models": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"side_mix_source": "ok in-gate rows of the window (w201 model_watch.side_mix definition)",
"symbols_compared": [
"XRP",
"BTC",
"ETH",
"BNB",
"SOL"
],
"symbols_prior_window_all": [
"BTC",
"ETH",
"SOL",
"BNB",
"XRP"
],
"symbols_retired_not_compared": [],
"coin_set_rule": "only the current five assets are compared; coins the prior window carried but the month does not are listed here by name only"
},
"verdict_block": {
"rule": "Verdict rule: each row states one reading of the month; a week agrees when it shows the same reading (same sign, or the same name). Confirmed = 3 or 4 of the 4 weeks agree with the month; Inconclusive = 2; Rejected = 0 or 1. Weekly cells come from the frozen weekly packs (7-line roster); the month is the pooled Sep 1-30 pull. W1 straddles the month edge (Aug 31).",
"header": [
"Metric",
"W1 Aug 31-Sep 6",
"W2 Sep 7-13",
"W3 Sep 14-20",
"W4 Sep 21-27",
"Month",
"Verdict"
],
"widths_mm": [
36,
26.5,
22,
22,
22,
18,
27.5
],
"weeks_source": "frozen weekly packs (weeks.W1..W4 of the monthly pack, 7-line roster)",
"month_source": "the pooled Sep 1-30 pull (month block)",
"rows": [
{
"metric": "Leader (hit-rate)",
"weeks": {
"W1": {
"value": "gemini-3.1",
"reading": "gemini-3.1-pro"
},
"W2": {
"value": "gemini-3.1",
"reading": "gemini-3.1-pro"
},
"W3": {
"value": "gemini-3.1",
"reading": "gemini-3.1-pro"
},
"W4": {
"value": "gemini-3.1",
"reading": "gemini-3.1-pro"
}
},
"month": {
"value": "gemini-3.1",
"reading": "gemini-3.1-pro"
},
"agree_count": 4,
"verdict": "Confirmed (4/4)"
},
{
"metric": "Most readable coin",
"weeks": {
"W1": {
"value": "BNB",
"reading": "BNB"
},
"W2": {
"value": "BNB",
"reading": "BNB"
},
"W3": {
"value": "ETH",
"reading": "ETH"
},
"W4": {
"value": "XRP",
"reading": "XRP"
}
},
"month": {
"value": "XRP",
"reading": "XRP"
},
"agree_count": 1,
"verdict": "Rejected (1/4)"
},
{
"metric": "Field hit below 50%",
"weeks": {
"W1": {
"value": "42.8%",
"reading": "below 50%"
},
"W2": {
"value": "39.6%",
"reading": "below 50%"
},
"W3": {
"value": "50.7%",
"reading": "at or above 50%"
},
"W4": {
"value": "46.8%",
"reading": "below 50%"
}
},
"month": {
"value": "45.5%",
"reading": "below 50%"
},
"agree_count": 3,
"verdict": "Confirmed (3/4)"
}
]
}
},
"key_findings": [
"**OBSERVATION -- No monthly title.** gemini-3.1-pro leads the month descriptively at 48.5% [46.8%, 50.2%]; the gap to #2 (claude-opus-5, 45.4%) is 3.1pp. 1 of 7 models improved against August, and a single model line held the weekly top spot across the four weeks.",
"**No ticker cleared 51%.** XRP is the month's most readable coin (47.0%, n=4,308) and SOL the hardest at 44.0%; 0 of the 5 assets cleared 51% over the month. Pair of the month: gemini-3.1-pro x SOL 49.4% (n=702); worst: grok-4.6 x SOL 40.9% (n=503).",
"**Mean weekly rank ran 1.00 to 5.25 across the four weeks.** claude-opus-5 swung across 5 rank positions and gemini-3.1-pro across 0. Consecutive-week Spearman came in at W1-W2 0.54, W2-W3 -0.11, W3-W4 0.57.",
"**The field read volatile weeks better.** Split by the pack's own volatility rule the field hit 43.4% (n=9,426) in the two calm weeks against 48.2% (n=9,430) in the two volatile weeks, a difference of +4.8pp. Cross-TF/FH lifts ran -1.5pp to +0.2pp. On trend days, 4h TF-agreement hit 55.2% (n=1,026) against 32.5% (n=1,279) on flat days."
],
"market_check": {
"label": "MARKET CHECK · OBSERVATION (September against August)",
"text": "OBSERVATION — mean |1d move| 1.98% -> 1.91% (-4% rel), BTC realized vol 36.9% -> 34.9% (ann., hourly); field directional accuracy 46.3% -> 45.5% (-0.8pp), 1 of 7 models improved, sim win-rate up for 6 of 7, field sim PnL -$1054.79 -> -$2554.15 (Aug 1-31 vs Sep 1-30; descriptive, one pair of windows, not a claim).",
"robustness": "Robustness: raw price-sign accuracy 46.0% -> 48.4% — the opposite direction to the trade-based hit rule.",
"weeks": {
"A": {
"days": "2026-08-01..2026-08-31",
"slots": [
"2026-08-01T00:00:00+00:00",
"2026-09-01T00:00:00+00:00"
],
"cutoff": "2026-09-01T16:00:00+00:00",
"partial": false
},
"B": {
"days": "2026-09-01..2026-09-30",
"slots": [
"2026-09-01T00:00:00+00:00",
"2026-10-01T00:00:00+00:00"
],
"cutoff": "2026-10-01T16:00:00+00:00"
}
},
"price_source": "crypto_spot/1h@:00",
"accuracy_field": {
"A": {
"n": 19383,
"hit_rate": 0.4633,
"wilson_95": [
0.4563,
0.4703
]
},
"B": {
"n": 20682,
"hit_rate": 0.455,
"wilson_95": [
0.4483,
0.4618
]
},
"delta_pp": -0.8
},
"accuracy_per_model": [
{
"model": "gemini-3.1-pro",
"model_label": "gemini-3.1-pro",
"A": {
"n": 3033,
"hit_rate": 0.4761,
"wilson_95": [
0.4584,
0.4939
]
},
"B": {
"n": 3284,
"hit_rate": 0.4851,
"wilson_95": [
0.468,
0.5022
]
},
"delta_pp": 0.9
},
{
"model": "deepseek-v4-pro",
"model_label": "deepseek-v4-pro",
"A": {
"n": 2578,
"hit_rate": 0.4554,
"wilson_95": [
0.4362,
0.4747
]
},
"B": {
"n": 3352,
"hit_rate": 0.4526,
"wilson_95": [
0.4358,
0.4695
]
},
"delta_pp": -0.3
},
{
"model": "grok-4.6",
"model_label": "grok-4.6",
"A": {
"n": 2708,
"hit_rate": 0.4527,
"wilson_95": [
0.4341,
0.4715
]
},
"B": {
"n": 2271,
"hit_rate": 0.4439,
"wilson_95": [
0.4235,
0.4644
]
},
"delta_pp": -0.9
},
{
"model": "qwen-3.8-max",
"model_label": "qwen-3.8-max",
"A": {
"n": 3001,
"hit_rate": 0.4565,
"wilson_95": [
0.4388,
0.4744
]
},
"B": {
"n": 3165,
"hit_rate": 0.4471,
"wilson_95": [
0.4298,
0.4645
]
},
"delta_pp": -0.9
},
{
"model": "claude-fable-5",
"model_label": "claude-fable-5",
"A": {
"n": 2812,
"hit_rate": 0.4673,
"wilson_95": [
0.4489,
0.4858
]
},
"B": {
"n": 2945,
"hit_rate": 0.4537,
"wilson_95": [
0.4357,
0.4717
]
},
"delta_pp": -1.4
},
{
"model": "gpt-5.6-sol",
"model_label": "gpt-5.6-sol",
"A": {
"n": 2970,
"hit_rate": 0.4586,
"wilson_95": [
0.4407,
0.4765
]
},
"B": {
"n": 3133,
"hit_rate": 0.4446,
"wilson_95": [
0.4273,
0.4621
]
},
"delta_pp": -1.4
},
{
"model": "claude-opus-5",
"model_label": "claude-opus-5",
"A": {
"n": 2281,
"hit_rate": 0.4779,
"wilson_95": [
0.4574,
0.4984
]
},
"B": {
"n": 2532,
"hit_rate": 0.4538,
"wilson_95": [
0.4345,
0.4732
]
},
"delta_pp": -2.4
}
],
"raw_sign_field": {
"A": {
"n_scored": 19304,
"hit_rate": 0.4599
},
"B": {
"n_scored": 20654,
"hit_rate": 0.4837
}
},
"abs_move_1d": {
"A": 1.979,
"B": 1.905,
"delta_pct_rel": -3.7
},
"btc_realized_vol_ann_pct_hourly": {
"A": 36.92,
"B": 34.88
},
"trading_field": {
"A": {
"n_trades": 19383,
"wr": 0.3576,
"pnl_net_usd": -1054.79,
"pnl_gross_usd": 883.51
},
"B": {
"n_trades": 20682,
"wr": 0.384,
"pnl_net_usd": -2554.15,
"pnl_gross_usd": -485.95
}
},
"verdict": {
"models_with_hit_improved": "1 of 7",
"models_with_rawsign_improved": "7 of 7",
"models_with_wr_improved": "6 of 7",
"field_hit_delta_pp": -0.8,
"field_pnl_net": {
"A": -1054.79,
"B": -2554.15
},
"abs_move_delta": {
"1h": {
"A": 0.325,
"B": 0.356,
"delta_pct_rel": 9.5
},
"4h": {
"A": 0.724,
"B": 0.737,
"delta_pct_rel": 1.8
},
"1d": {
"A": 1.979,
"B": 1.905,
"delta_pct_rel": -3.7
}
},
"btc_vol_ann_pct": {
"A": 36.92,
"B": 34.88
},
"note": "descriptive, two adjacent windows; hit rule = methodology v1.1 trade-based; raw_sign = price-sign robustness check (closes at :00 vs slots at :01)"
},
"lineage_note": "lineage splice ACTIVE for all 7 heads: [('claude-opus-4.8', 'claude-opus-5'), ('grok-4.5', 'grok-4.6'), ('qwen-3.7-max', 'qwen-3.8-max')]. Window MONTH remapped 0 row(s), the previous window 5390. Labels are generated from the pulled rows (see lineage_era).",
"prior_window_partial": false
},
"why_it_matters": "MarketMania scores 7 model lines against the same market, hour after hour. Monthly Model Watch asks the weekly questions at monthly n -- who led, which tickers were readable, who saw the turn first, does self-agreement help -- and then asks the three questions a single week cannot answer: which coin the field reads best over a month, how far a weekly rank moves inside one month, and how the field's accuracy differs between the calm and volatile weeks the pack labels. Each block gets its own table, its own N and its own caveats.",
"how_to_read_this": "Every block below is scored on the same pool: matured directional calls inside the 1h/4h/1d FH gate, hit by the methodology v1.1 direction rule, over the whole month at one cutoff (Thu Oct 1, 2026, 16:00 UTC). The monthly title is awarded only on a >=5pp gap AND non-overlapping 95% Wilson CIs with N>=10 per side -- a descriptive lead is not a title. Reversals use the 4h grid with a >=2.0% counter-move sustained 8h, BTC/ETH only in v1. Cross-TF/FH blocks ask whether a model agreeing with itself does any better, and are split trend vs flat. The by-week blocks come from the four frozen weekly packs at their own published cutoffs. No model line is lineage-spliced this month (the prior window carries the grok and qwen splices).",
"practical_implications": [
"On this month's leaderboard no title was awarded, the top two CIs overlap, and a single model line led the four Mon-Sun weeks from Aug 31 to Sep 27.",
"XRP ranked #1 for the month on n=4,308; the week-by-week league shows it ranked #1 in 1 of the four published weeks.",
"The field differs by +4.8pp between the two regimes on n=9,426 calm vs n=9,430 volatile; the rule that draws the line is the pack's own four-week median, recomputed every month."
],
"limitations": [
"The prior window is the full August: every \"prior\" column, the gray series of the page-1 chart and the market-check A column describe Aug 1-31. It is window A of the same pull, slots 2026-08-01 to 2026-09-01 (first slot actually seen 2026-08-01 00:01 UTC), matured at the prior cutoff Tue Sep 1, 2026, 16:00 UTC under the same engine, the same FH gate and the same hit rule as September -- 19,383 directional calls against 20,682 in the month. Month-over-month deltas are a step between two windows of different length (31 and 30 days).",
"Two cutoffs live in this issue. The month is frozen at Thu Oct 1, 2026, 16:00 UTC; the four weekly blocks are frozen at their own published cutoffs (W1 Aug 31-Sep 6, W2 Sep 7-13, W3 Sep 14-20, W4 Sep 21-27). A forecast still pending at its week's cutoff is immature in the weekly pack and mature in the month pull, so the month does not equal the sum of the weeks (20,682 vs 19,544 directional, plus 1,826 on the 3 remainder days); W1 also opens on Aug 31, a day the month pull does not hold, so its pack counts 4,520 directional calls against 3,832 for Sep 1-6 inside the month pull -- that gap is arithmetic, not a data problem.",
"Observations inside one window are not independent, and the 95% Wilson CIs shown throughout are descriptive, not inferential. Ranks are ordinal: a 0.3pp gap and a 6pp gap both read as one rank position, so the rank-stability table prints the hit rate next to every rank.",
"The reversal detector remains v1 and BTC/ETH-only, on price series reconstructed from trade entry prices; 7 events in this window still cannot characterise detector performance. Cross-TF and cross-FH disagreement cells are small (n=68, 21, 92, 28); cells under N=10 are insufficient and never used to rank models.",
"Model lines in this window: claude-fable-5, claude-opus-5, deepseek-v4-pro, gemini-3.1-pro, gpt-5.6-sol, grok-4.6, qwen-3.8-max. Series density and lineage notes (monthly #2): (1) rolling-1w forecasts emit daily since Aug 22, 2026 and rolling-1M daily since Aug 23, 2026, so across this 30-day window daily coverage is FULL as measured -- the 1w series has slots on 30 of 30 days and the 1M series on 30 of 30 days; both sit outside this report's FH gate (1h/4h/1d) and appear only in the exclusions counter (1w 2,096, 1M 2,089). (2) No model line is lineage-spliced inside this window. (3) Every per-model aggregate in this issue sees the merged head id; the raw id is kept in the pack, and 0 rows were remapped inside the month against 5,390 in the prior window.",
"Market-state row is computed on a single-exchange (binance) daily candle series, as in the weekly series after the audit that found the raw table mixes two exchanges. The prior-month block is the full August: it covers 31 days (Aug 1-31), and the TOP5 volume figure compares a 30-day sum with a 31-day sum -- a level difference, not a like-for-like change.",
"Research-to-date counter: directional forecasts resolved inside the FH gate (1h/4h/1d) since Jul 11, counted at the issue cutoff (Thu Oct 1, 2026, 16:00 UTC) -- the D2 definition pinned in the weekly series; this issue prints 59,755. The pack carries no prior-cutoff pin for a monthly window (the prior-cutoff pin is null by construction), so the control this issue is the weekly reproduction: the W4 block of this pack (Sep 21-27) reproduces the published weekly #8 on 7 of 7 checks, and the prior window of this pack (Aug 1-31, prior cutoff Tue Sep 1, 2026, 16:00 UTC) reproduces the published Monthly #1 on 7 of 7 checks. Counter deltas against pre-wave-3 issues remain definitional."
],
"notes": {
"series_density_and_lineage_monthly2": "Series density and lineage notes (monthly #2): (1) rolling-1w forecasts emit daily since Aug 22, 2026 and rolling-1M daily since Aug 23, 2026, so across this 30-day window daily coverage is FULL as measured -- the 1w series has slots on 30 of 30 days and the 1M series on 30 of 30 days; both sit outside this report's FH gate (1h/4h/1d) and appear only in the exclusions counter (1w 2,096, 1M 2,089). (2) No model line is lineage-spliced inside this window. (3) Every per-model aggregate in this issue sees the merged head id; the raw id is kept in the pack, and 0 rows were remapped inside the month against 5,390 in the prior window.",
"grok_era": {
"ids_in_window": [
"grok-4.6"
],
"n_grok45": 0,
"n_grok46": 6300,
"n_grok45_rows": 0,
"n_grok46_rows": 6300,
"first_grok46_slot": "2026-09-01T00:01:00+00:00",
"last_grok45_slot": null,
"first_grok45_slot": null,
"flip_at": "2026-08-24T09:22:00+00:00",
"flip_inside_window": false,
"prev_window": {
"n_grok45_rows": 4520,
"n_grok46_rows": 1570
},
"merged_model_id": "grok-4.6",
"merged_label": "grok-4.6",
"lineage_ids": [
"grok-4.5",
"grok-4.6"
],
"lineage_source": "benchmarks.config.lineage_ids(head) for each of the 7 heads + MODEL_SUCCESSION[claude-opus-5=('claude-opus-4.8',), grok-4.6=('grok-4.5',), qwen-3.8-max=('qwen-3.7-max',)]",
"lineage_warnings": [],
"rows_remapped_a_plus_b": 4520,
"note": "window MONTH [2026-09-01T00:00:00+00:00, 2026-10-01T00:00:00+00:00): grok-4.5 rows 0, grok-4.6 rows 6300; flip 2026-08-24T09:22:00+00:00 is NOT inside this window. All per-model aggregates use the lineage splice grok-4.5 -> grok-4.6 (one row, labelled 'grok-4.6'). See lineage_era for the same facts across all 7 heads.",
"cells_with_two_grok_rows": 0
},
"model_ids_line": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"model_roster": {
"current": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"current_count": 7,
"current_raw_ids_in_window": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"archived_legacy": [
"grok-4.5",
"qwen-3.7-max",
"claude-opus-4.8"
],
"archived_legacy_count": 3,
"archived_legacy_detail": [
{
"model": "grok-4.5",
"n_rows_since": 11128,
"first_slot": "2026-07-15T18:01:00+00:00",
"last_slot": "2026-08-24T09:01:00+00:00"
},
{
"model": "qwen-3.7-max",
"n_rows_since": 7479,
"first_slot": "2026-07-15T18:01:00+00:00",
"last_slot": "2026-08-05T09:01:00+00:00"
},
{
"model": "claude-opus-4.8",
"n_rows_since": 6315,
"first_slot": "2026-07-15T18:01:00+00:00",
"last_slot": "2026-07-30T08:01:00+00:00"
}
],
"archived_excluding_lineage_merged": [],
"tracked_total": 10,
"canon_expected": {
"current": 7,
"archived_legacy": 4,
"tracked_total": 11
},
"matches_canon_11_7_4": false,
"roster_rule_w213": "7 models tracked (current line-up; earlier versions folded into their successors' lineage)",
"canon_expected_w213": {
"current": 7
},
"heads_w213": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"matches_canon_7_current": true,
"heads_missing_in_window": [],
"heads_unexpected_in_window": [],
"roster_since": "2026-07-11T00:00:00+00:00",
"note": "current = merged-lineage model ids emitting inside the window (legacy ids are spliced into their successor); archived legacy = raw ids seen since 2026-07-11 that no longer emit in the window. Models retired before 2026-07-11 are not visible to this query. The canon checks are DESCRIPTIVE: if the data says otherwise, the data wins and is printed."
},
"series_density": {
"1d": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": true,
"n_slots": 30,
"n_rows": 2100,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"1h": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": true,
"n_slots": 720,
"n_rows": 25200,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"1m": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": false,
"n_slots": 30,
"n_rows": 2100,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"1w": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": false,
"n_slots": 30,
"n_rows": 2100,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
},
"4h": {
"days_with_slots": 30,
"expected_days": 30,
"full_daily_coverage": true,
"in_fh_gate": true,
"n_slots": 180,
"n_rows": 12600,
"days": [
"2026-09-01",
"2026-09-02",
"2026-09-03",
"2026-09-04",
"2026-09-05",
"2026-09-06",
"2026-09-07",
"2026-09-08",
"2026-09-09",
"2026-09-10",
"2026-09-11",
"2026-09-12",
"2026-09-13",
"2026-09-14",
"2026-09-15",
"2026-09-16",
"2026-09-17",
"2026-09-18",
"2026-09-19",
"2026-09-20",
"2026-09-21",
"2026-09-22",
"2026-09-23",
"2026-09-24",
"2026-09-25",
"2026-09-26",
"2026-09-27",
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"missing_days": []
}
},
"month_over_month_note": "Month-over-month columns compare back-to-back windows: Aug 1-31 vs Sep 1-30 (this issue). \"Was\" values are the pack's own prior-window block -- window A of the same pull, same engine, same FH gate and same hit rule, matured at the prior cutoff Tue Sep 1, 2026, 16:00 UTC. The prior month is the full August (31 days against 30 in September), and it holds 19,383 directional calls against 20,682 in September. Deltas are computed on unrounded rates.",
"research_to_date_pin": "Research-to-date counter: directional forecasts resolved inside the FH gate (1h/4h/1d) since Jul 11, counted at the issue cutoff (Thu Oct 1, 2026, 16:00 UTC) -- the D2 definition pinned in the weekly series; this issue prints 59,755. The pack carries no prior-cutoff pin for a monthly window (the prior-cutoff pin is null by construction), so the control this issue is the weekly reproduction: the W4 block of this pack (Sep 21-27) reproduces the published weekly #8 on 7 of 7 checks, and the prior window of this pack (Aug 1-31, prior cutoff Tue Sep 1, 2026, 16:00 UTC) reproduces the published Monthly #1 on 7 of 7 checks. Counter deltas against pre-wave-3 issues remain definitional.",
"market_state_audit": "Market-state row is computed on a single-exchange (binance) daily candle series, as in the weekly series after the audit that found the raw table mixes two exchanges. The prior-month block is the full August: it covers 31 days (Aug 1-31), and the TOP5 volume figure compares a 30-day sum with a 31-day sum -- a level difference, not a like-for-like change.",
"engine": "Engine 1.1 has powered the sandbox since Aug 18, i.e. before this window and inside the prior window; no cross-engine PnL comparisons are claimed.",
"report_template": "PDF built with the canonical mm_report.py template module; preview card with report_preview.render_preview",
"lineage_era": {
"window_key": "MONTH",
"heads": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"lineage_map": {
"claude-opus-4.8": "claude-opus-5",
"grok-4.5": "grok-4.6",
"qwen-3.7-max": "qwen-3.8-max"
},
"legacy_by_head": {
"claude-fable-5": [],
"claude-opus-5": [
"claude-opus-4.8"
],
"deepseek-v4-pro": [],
"gemini-3.1-pro": [],
"gpt-5.6-sol": [],
"grok-4.6": [
"grok-4.5"
],
"qwen-3.8-max": [
"qwen-3.7-max"
]
},
"lineage_source": "benchmarks.config.lineage_ids(head) for each of the 7 heads + MODEL_SUCCESSION[claude-opus-5=('claude-opus-4.8',), grok-4.6=('grok-4.5',), qwen-3.8-max=('qwen-3.7-max',)]",
"lineage_resolved_heads": [
"claude-fable-5",
"claude-opus-5",
"deepseek-v4-pro",
"gemini-3.1-pro",
"gpt-5.6-sol",
"grok-4.6",
"qwen-3.8-max"
],
"lineage_warnings": [],
"per_head": {
"claude-fable-5": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "claude-fable-5"
},
"claude-opus-5": {
"legacy_ids": [
"claude-opus-4.8"
],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"n_rows_remapped_prev_window": {
"claude-opus-4.8": 0
},
"by_legacy_id": {
"claude-opus-4.8": {
"n_rows_remapped": {
"MONTH": 0,
"prev_window": 0
},
"first_slot": null,
"last_slot": null,
"first_slot_prev_window": null,
"last_slot_prev_window": null
}
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "claude-opus-5"
},
"deepseek-v4-pro": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "deepseek-v4-pro"
},
"gemini-3.1-pro": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "gemini-3.1-pro"
},
"gpt-5.6-sol": {
"legacy_ids": [],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"n_rows_remapped_prev_window": {},
"by_legacy_id": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "gpt-5.6-sol"
},
"grok-4.6": {
"legacy_ids": [
"grok-4.5"
],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"n_rows_remapped_prev_window": {
"grok-4.5": 4520
},
"by_legacy_id": {
"grok-4.5": {
"n_rows_remapped": {
"MONTH": 0,
"prev_window": 4520
},
"first_slot": null,
"last_slot": null,
"first_slot_prev_window": "2026-08-01T00:01:00+00:00",
"last_slot_prev_window": "2026-08-24T09:01:00+00:00"
}
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "grok-4.6"
},
"qwen-3.8-max": {
"legacy_ids": [
"qwen-3.7-max"
],
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"n_rows_remapped_prev_window": {
"qwen-3.7-max": 870
},
"by_legacy_id": {
"qwen-3.7-max": {
"n_rows_remapped": {
"MONTH": 0,
"prev_window": 870
},
"first_slot": null,
"last_slot": null,
"first_slot_prev_window": "2026-08-01T00:01:00+00:00",
"last_slot_prev_window": "2026-08-05T09:01:00+00:00"
}
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00",
"last_successor_slot": "2026-09-30T23:01:00+00:00",
"n_rows_successor": 6300,
"spliced_in_window": false,
"label": "qwen-3.8-max"
}
},
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"rows_remapped_window": 0,
"rows_remapped_prev_window": 5390,
"note": "Only the 7 current models are tracked; earlier versions are folded into their successors' lineage. Every per-model aggregate in this pack sees the merged head id; the raw id survives in model_raw. Labels and era spans are generated from the pulled rows, not hardcoded.",
"cells_with_two_rows_per_head": {
"claude-fable-5": 0,
"claude-opus-5": 0,
"deepseek-v4-pro": 0,
"gemini-3.1-pro": 0,
"gpt-5.6-sol": 0,
"grok-4.6": 0,
"qwen-3.8-max": 0
}
},
"frozen_cutoffs": "Two cutoffs live in this issue. The month is frozen at Thu Oct 1, 2026, 16:00 UTC; the four weekly blocks are frozen at their own published cutoffs (W1 Aug 31-Sep 6, W2 Sep 7-13, W3 Sep 14-20, W4 Sep 21-27). A forecast still pending at its week's cutoff is immature in the weekly pack and mature in the month pull, so the month does not equal the sum of the weeks (20,682 vs 19,544 directional, plus 1,826 on the 3 remainder days); W1 also opens on Aug 31, a day the month pull does not hold, so its pack counts 4,520 directional calls against 3,832 for Sep 1-6 inside the month pull -- that gap is arithmetic, not a data problem.",
"partial_prior_window": "The prior window is the full August: every \"prior\" column, the gray series of the page-1 chart and the market-check A column describe Aug 1-31. It is window A of the same pull, slots 2026-08-01 to 2026-09-01 (first slot actually seen 2026-08-01 00:01 UTC), matured at the prior cutoff Tue Sep 1, 2026, 16:00 UTC under the same engine, the same FH gate and the same hit rule as September -- 19,383 directional calls against 20,682 in the month. Month-over-month deltas are a step between two windows of different length (31 and 30 days)."
},
"living_series_note": "**Issue #2.** Monthly Model Watch is a living monthly comparison; each issue appends one more calendar month of leaderboard, asset-league, rank-stability and regime data. This second issue reads September against the full August: gemini-3.1-pro at the top (48.5%; 47.6% in the prior window), 4 of 7 qualifying turns found a caller, 0 of 5 tickers above 51%, a single model line leading the four weeks and a calm-vs-volatile field difference of +4.8pp. No model line is lineage-spliced this month (the prior window carries the grok and qwen splices). Engine 1.1 has powered the sandbox since Aug 18, i.e. before this window and inside the prior window; no cross-engine PnL comparisons are claimed.",
"testing_next": [
"Monthly #3 (October): the same three series over Oct 1-31, cut on Nov 1 plus the maturity lag, so its month-over-month comparison runs against this September.",
"Monthly Config Watch #2: the same September window through the config search engine (w298), published as the fourth monthly report of this issue.",
"Quarterly Stability Index: Run 3, the next repeat-prompt stability run, is planned for the first half of October and is the first cross-quarter point on model stability."
],
"related_research": [
{
"title": "Monthly Consensus Watch #2",
"url": "https://marketmania.ai/research/reports/consensus-watch-monthly-2026-09.pdf"
},
{
"title": "Monthly Calibration #2",
"url": "https://marketmania.ai/research/reports/calibration-monthly-2026-09.pdf"
},
{
"title": "Weekly Model Watch #8",
"url": "https://marketmania.ai/research/reports/model-watch-2026-09-21.pdf"
}
],
"related_research_note": "The three monthly reports publish together as one issue each month; each links straight to the others' PDF and to the last weekly issue of the same series. Direct links are the posting rule from wave 2 on.",
"research_to_date": {
"this_report": {
"scored_observations": 20682
},
"platform": {
"as_of_cutoff": "2026-10-01",
"resolved_forecasts": 59755,
"resolved_forecasts_definition": "directional forecasts resolved inside the FH gate (1h/4h/1d) since Jul 11, at the issue cutoff",
"resolved_forecasts_since": "2026-07-11",
"models_tracked": 7,
"models_tracked_detail": "7 models tracked (current line-up; earlier versions folded into their successors' lineage)",
"assets": 5,
"forecast_horizons": 5,
"cadence": "hourly",
"published_research_reports": 41
},
"source": "monthly_metrics_2026-09.json",
"control_reproduces_wave3_pin": null,
"pin_note": "Research-to-date counter: directional forecasts resolved inside the FH gate (1h/4h/1d) since Jul 11, counted at the issue cutoff (Thu Oct 1, 2026, 16:00 UTC) -- the D2 definition pinned in the weekly series; this issue prints 59,755. The pack carries no prior-cutoff pin for a monthly window (the prior-cutoff pin is null by construction), so the control this issue is the weekly reproduction: the W4 block of this pack (Sep 21-27) reproduces the published weekly #8 on 7 of 7 checks, and the prior window of this pack (Aug 1-31, prior cutoff Tue Sep 1, 2026, 16:00 UTC) reproduces the published Monthly #1 on 7 of 7 checks. Counter deltas against pre-wave-3 issues remain definitional.",
"control_last_week_reproduces_weekly8": {
"pins_source": "weekly_metrics_2026-09-21.json (the published weekly #8 pack)",
"note": "W4 of this run is the weekly #8 window with the weekly #8 frozen cutoff, so every pin must reproduce exactly. A mismatch means the window or the cutoff moved -- non-fatal here, but the monthly numbers must not be published until it is explained.",
"checks": {
"counters.mature_directional": {
"pin": 4958,
"got": 4958,
"match": true,
"tolerance": 0
},
"calibration.field.hit_rate": {
"pin": 0.4683,
"got": 0.4683,
"match": true,
"tolerance": 5e-05
},
"calibration.field.gap_pp": {
"pin": 16.1,
"got": 16.1,
"match": true,
"tolerance": 0.05
},
"consensus.loo_pooled.lift_pp": {
"pin": -14.0,
"got": -14.0,
"match": true,
"tolerance": 0.05
},
"model_watch.leaderboard.rows[0].model": {
"pin": "gemini-3.1-pro",
"got": "gemini-3.1-pro",
"match": true,
"tolerance": 0
},
"model_watch.leaderboard.rows[0].hit_rate": {
"pin": 0.5139,
"got": 0.5139,
"match": true,
"tolerance": 5e-05
},
"research_to_date.canonical_value_at_cutoff": {
"pin": 57929,
"got": 57929,
"match": true,
"tolerance": 0
}
},
"all_match": true,
"n_checks": 7,
"n_match": 7,
"last_week_key": "W4"
}
},
"citation": {
"bibtex_key": "mm_modelwatch_monthly_2026m09",
"title": "Monthly Model Watch #2: monthly leaderboard, asset league, rank stability and regime split, Sep 1-30 2026",
"author": "MarketMania Research",
"year": 2026,
"month": "October",
"day": 2,
"url": "https://marketmania.ai/research/reports/model-watch-monthly-2026-09.pdf",
"note": "Methodology v1.1; window Sep 1-30, 2026 UTC; source monthly_metrics_2026-09.json"
},
"site_entry": {
"slug": "model-watch-monthly-2026-09",
"series": "monthly",
"seriesLabel": "Monthly · Model Watch",
"date": "October 2, 2026",
"title": "Monthly Model Watch #2",
"summary": "20,682 directional calls over one calendar month, three questions the weekly series cannot answer: which coin the field reads best over a month, how far a weekly rank moves inside one month, and how accuracy differs between the calm and the volatile weeks. No monthly title — gemini-3.1-pro led at 48.5% (+0.9pp vs August) against claude-opus-5 at 45.4%, a gap of 3.1pp, while a single model line held the weekly top spot across the four weeks. XRP was the most readable ticker at 47.0% and SOL the hardest at 44.0%, with 0 of 5 assets above 51 percent. Calm weeks hit 43.4% against 48.2% in the volatile ones (+4.8pp). Reversals: 4 of 7 qualifying turns found a caller. Field accuracy moved 46.3% -> 45.5% (-0.8pp) with 1 of 7 models improving.",
"stats": [
"20,682 directional calls",
"No monthly title",
"field -0.8pp m/m"
],
"manifest_key_finding": "Field accuracy moved 46.3% -> 45.5% (-0.8pp) against August with 1 of 7 models improving; gemini-3.1-pro led the month at 48.5%, a single model line led the four weeks, and no monthly title was awarded.",
"files": {
"pdf": "/research/reports/model-watch-monthly-2026-09.pdf",
"md": "/research/reports/model-watch-monthly-2026-09.md",
"json": "/research/reports/model-watch-monthly-2026-09.json"
}
},
"social": {
"telegram": {
"photo_caption_html": "📊 Monthly Model Watch #2 — monthly series (window Sep 1–30)\n\nWho led, what was readable, who saw the turn? 20,682 directional calls, 7 lines, one month:\n• No monthly title — gemini-3.1-pro leads at 48.5% (+0.9pp m/m) against claude-opus-5 at 45.4%, a 3.1pp gap\n• Rank stability: a single line topped the four weeks\n• Asset league: XRP most readable (47.0%); SOL hardest at 44.0% — 0 of 5 above 51%\n• Calm vs volatile weeks: field 43.4% vs 48.2% (+4.8pp)\n• Reversals: 4 of 7 qualifying turns found a caller\n• Market check: field 46.3% -> 45.5% (-0.8pp), 1 of 7 improved\n\nNote: prior window = the full August (Aug 1-31), same rules as September\n\n📄 Full report (PDF) · all research\n\n#cryptotrading #LLM #AI #crypto #trading #models #research #Monthly",
"document_caption_html": "📄 Monthly Model Watch #2 — full report (PDF).\nWeb copy: model-watch-monthly-2026-09.pdf · machine-readable: JSON · all research\n\n#cryptotrading #LLM #AI #crypto #trading #models #research #Monthly",
"caption_len": 921,
"document_caption_len": 415,
"limit": 1024
},
"x": {
"main": "Monthly Model Watch #2 is out.\n\n20,682 directional LLM calls over September: gemini-3.1-pro led at 48.5% and topped every one of the four weeks. XRP was the most readable coin.\n\nFull report: https://marketmania.ai/research/reports/model-watch-monthly-2026-09.pdf\n\n#AI #Crypto #Research #LLM",
"main_tco_len": 242,
"reply": "Hit-rate, August -> September:\ngemini-3.1: 47.6 -> 48.5\nopus-5: 47.8 -> 45.4\nfable-5: 46.7 -> 45.4\ndeepseek: 45.5 -> 45.3\nqwen-3.8: 45.6 -> 44.7\ngpt-5.6: 45.9 -> 44.5\ngrok-4.6: 45.3 -> 44.4\nField: 46.3 -> 45.5. Calm 43.4% vs volatile 48.2%.\nData: https://marketmania.ai/research/reports/model-watch-monthly-2026-09.json",
"reply_tco_len": 270,
"limit_tco": 280,
"replies_max": 1
}
},
"monthly_window": {
"start": "2026-09-01T00:00:00+00:00",
"end": "2026-10-01T00:00:00+00:00",
"text": "Sep 1-30, 2026 UTC",
"cutoff": "2026-10-01T16:00:00+00:00",
"cutoff_text": "Thu Oct 1, 2026, 16:00 UTC",
"days": 30,
"label": "Sep 1-30"
},
"prior_window": {
"start": "2026-08-01T00:00:00+00:00",
"end": "2026-09-01T00:00:00+00:00",
"text": "Aug 1-31, 2026",
"printed_as": "Aug 1-31",
"cutoff_text": "Tue Sep 1, 2026, 16:00 UTC",
"partial": false,
"note": "August is a full month: 2026-08-01 .. 2026-09-01, 31 days",
"applicable": true,
"not_applicable_reason": null,
"source": "month.prior_window -- window A of the same pull, matured at the prior cutoff with the same gate, maturity and hit rule",
"effective_start": "2026-08-01T00:01:00+00:00",
"counters": {
"forecasts_total": 42654,
"ok_in_gate": 40979,
"mature": 40979,
"mature_directional": 19383,
"mature_sideways": 21596
},
"definitions": {
"maturity": "horizon expired AND trade closed by the PRIOR cutoff (the same rule the current window uses at its own cutoff)",
"hit": "direction: exit tp1/tp2 -> hit, sl -> miss, expiry -> sign of gross pnl",
"fh_gate": [
"1h",
"4h",
"1d"
],
"min_n": 10,
"lineage": "rows spliced by the generalised 7-head lineage map before any aggregate",
"note": "identical definitions to the current window; only the window and the cutoff differ"
}
},
"weeks_index": {
"W1": {
"label": "Aug 31-Sep 6",
"issue": 5,
"published_slug_suffix": [
"2026-08-31T00:00:00Z",
"2026-09-07T00:00:00Z"
],
"bounds": [
"2026-08-31T00:00:00Z",
"2026-09-07T00:00:00Z"
],
"regime": "volatile"
},
"W2": {
"label": "Sep 7-13",
"issue": 6,
"published_slug_suffix": [
"2026-09-07T00:00:00Z",
"2026-09-14T00:00:00Z"
],
"bounds": [
"2026-09-07T00:00:00Z",
"2026-09-14T00:00:00Z"
],
"regime": "calm"
},
"W3": {
"label": "Sep 14-20",
"issue": 7,
"published_slug_suffix": [
"2026-09-14T00:00:00Z",
"2026-09-21T00:00:00Z"
],
"bounds": [
"2026-09-14T00:00:00Z",
"2026-09-21T00:00:00Z"
],
"regime": "volatile"
},
"W4": {
"label": "Sep 21-27",
"issue": 8,
"published_slug_suffix": [
"2026-09-21T00:00:00Z",
"2026-09-28T00:00:00Z"
],
"bounds": [
"2026-09-21T00:00:00Z",
"2026-09-28T00:00:00Z"
],
"regime": "calm"
}
},
"roster_rule": "7 models tracked (current line-up; earlier versions folded into their successors' lineage)",
"lineage_labels": {
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"spliced_in_window": [],
"cell_form": {},
"families": []
},
"controls": {
"last_week_reproduces_weekly8": {
"pins_source": "weekly_metrics_2026-09-21.json (the published weekly #8 pack)",
"note": "W4 of this run is the weekly #8 window with the weekly #8 frozen cutoff, so every pin must reproduce exactly. A mismatch means the window or the cutoff moved -- non-fatal here, but the monthly numbers must not be published until it is explained.",
"checks": {
"counters.mature_directional": {
"pin": 4958,
"got": 4958,
"match": true,
"tolerance": 0
},
"calibration.field.hit_rate": {
"pin": 0.4683,
"got": 0.4683,
"match": true,
"tolerance": 5e-05
},
"calibration.field.gap_pp": {
"pin": 16.1,
"got": 16.1,
"match": true,
"tolerance": 0.05
},
"consensus.loo_pooled.lift_pp": {
"pin": -14.0,
"got": -14.0,
"match": true,
"tolerance": 0.05
},
"model_watch.leaderboard.rows[0].model": {
"pin": "gemini-3.1-pro",
"got": "gemini-3.1-pro",
"match": true,
"tolerance": 0
},
"model_watch.leaderboard.rows[0].hit_rate": {
"pin": 0.5139,
"got": 0.5139,
"match": true,
"tolerance": 5e-05
},
"research_to_date.canonical_value_at_cutoff": {
"pin": 57929,
"got": 57929,
"match": true,
"tolerance": 0
}
},
"all_match": true,
"n_checks": 7,
"n_match": 7,
"last_week_key": "W4"
},
"month_vs_weeks": {
"month": 20682,
"sum_weeks": 19544,
"per_week": {
"W1": 4520,
"W2": 4468,
"W3": 5598,
"W4": 4958
},
"remainder_days": [
"2026-09-28",
"2026-09-29",
"2026-09-30"
],
"remainder_days_n": 3,
"remainder_mature_directional_month_pull": 1826,
"remainder_by_day": {
"2026-09-29": 668,
"2026-09-28": 772,
"2026-09-30": 386
},
"sum_weeks_plus_remainder": 21370,
"month_minus_sum_weeks_plus_remainder": -688,
"month_pull_by_week": {
"W1": 3832,
"W2": 4468,
"W3": 5598,
"W4": 4958
},
"note": "These WILL differ, and the difference is expected: each weekly pack is computed at its own frozen cutoff (week_end + 16h), while the MONTH pack is computed at 2026-10-01T16:00Z. A forecast still pending at its week's cutoff (trade not yet closed) is immature in the weekly pack and mature in the MONTH pack, so month >= sum_weeks + remainder. remainder_days_n counts the calendar days of the month outside the four Mon-Sun weeks (2026-09-28, 2026-09-29, 2026-09-30). month_pull_by_week is the same month pull split by slot week at the MONTH cutoff -- compare THAT with per_week to see the cutoff effect alone. W1 starts before the month (2026-08-31T00:00:00+00:00), so its pack also counts slots that are not in the MONTH pull and the inequality above need not hold for it."
},
"lineage_splice_counts": {
"MONTH": {
"claude-opus-4.8": 0,
"grok-4.5": 0,
"qwen-3.7-max": 0
},
"W1": {
"claude-opus-4.8": 0,
"grok-4.5": 0,
"qwen-3.7-max": 0
},
"W2": {
"claude-opus-4.8": 0,
"grok-4.5": 0,
"qwen-3.7-max": 0
},
"W3": {
"claude-opus-4.8": 0,
"grok-4.5": 0,
"qwen-3.7-max": 0
},
"W4": {
"claude-opus-4.8": 0,
"grok-4.5": 0,
"qwen-3.7-max": 0
}
},
"lineage_splice_detail": {
"MONTH": {
"total_window": 0,
"total_prev_window": 5390,
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"per_head": {
"claude-fable-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
},
"claude-opus-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
},
"deepseek-v4-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
},
"gemini-3.1-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
},
"gpt-5.6-sol": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
},
"grok-4.6": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
},
"qwen-3.8-max": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-01T00:01:00+00:00"
}
}
},
"W1": {
"total_window": 0,
"total_prev_window": 110,
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"per_head": {
"claude-fable-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
},
"claude-opus-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
},
"deepseek-v4-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
},
"gemini-3.1-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
},
"gpt-5.6-sol": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
},
"grok-4.6": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
},
"qwen-3.8-max": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-08-31T00:01:00+00:00"
}
}
},
"W2": {
"total_window": 0,
"total_prev_window": 0,
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"per_head": {
"claude-fable-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
},
"claude-opus-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
},
"deepseek-v4-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
},
"gemini-3.1-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
},
"gpt-5.6-sol": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
},
"grok-4.6": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
},
"qwen-3.8-max": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-07T00:01:00+00:00"
}
}
},
"W3": {
"total_window": 0,
"total_prev_window": 0,
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"per_head": {
"claude-fable-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
},
"claude-opus-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
},
"deepseek-v4-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
},
"gemini-3.1-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
},
"gpt-5.6-sol": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
},
"grok-4.6": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
},
"qwen-3.8-max": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-14T00:01:00+00:00"
}
}
},
"W4": {
"total_window": 0,
"total_prev_window": 0,
"labels": {
"claude-fable-5": "claude-fable-5",
"claude-opus-5": "claude-opus-5",
"deepseek-v4-pro": "deepseek-v4-pro",
"gemini-3.1-pro": "gemini-3.1-pro",
"gpt-5.6-sol": "gpt-5.6-sol",
"grok-4.6": "grok-4.6",
"qwen-3.8-max": "qwen-3.8-max"
},
"per_head": {
"claude-fable-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
},
"claude-opus-5": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"claude-opus-4.8": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
},
"deepseek-v4-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
},
"gemini-3.1-pro": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
},
"gpt-5.6-sol": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
},
"grok-4.6": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"grok-4.5": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
},
"qwen-3.8-max": {
"legacy_ids_seen_in_window": [],
"n_rows_remapped": {
"qwen-3.7-max": 0
},
"first_legacy_slot": null,
"last_legacy_slot": null,
"first_successor_slot": "2026-09-21T00:01:00+00:00"
}
}
}
},
"d2_at_cutoff": {
"window_key": "MONTH",
"cutoff": "2026-10-01T16:00:00+00:00",
"canonical_definition": "D2_directional_only",
"canonical_value_at_cutoff": 59755,
"since": "2026-07-11T00:00:00+00:00",
"d1_d3_at_cutoff": {
"D1_expired_ok_in_gate": 122919,
"D2_directional_only": 59755,
"D3_weekly_rule_mature": 122770
},
"marker": "RTD-D2-MONTHLY=59755",
"note": "this is the monthly RTD pin -- the number that goes into the Monthly plain boxes"
},
"prior_window_summary": {
"applicable": true,
"start": "2026-08-01T00:00:00+00:00",
"end": "2026-09-01T00:00:00+00:00",
"effective_start": "2026-08-01T00:01:00+00:00",
"cutoff": "2026-09-01T16:00:00+00:00",
"partial": false,
"note": "August is a full month: forecasts and candles from 2026-08-01",
"counters": {
"forecasts_total": 42654,
"ok_in_gate": 40979,
"mature": 40979,
"mature_directional": 19383,
"mature_sideways": 21596
},
"lineage_splice_counts": {
"claude-opus-4.8": 0,
"grok-4.5": 4520,
"qwen-3.7-max": 870
},
"field": {
"hit_rate": {
"month": 0.455,
"prior": 0.4633,
"delta_pp": -0.8
},
"gap_pp": {
"month": 17.0,
"prior": 16.3,
"delta_pp": 0.7
},
"brier": {
"month": 0.2796,
"prior": 0.2784
},
"mean_conf": {
"month": 62.6,
"prior": 62.6
},
"n": {
"month": 20682,
"prior": 19383
}
},
"consensus": {
"lift_pp": {
"month": -9.7,
"prior": 1.9,
"delta_pp": -11.6
},
"n_scoreable": {
"month": 17625,
"prior": 16089
},
"herd_share_pct": {
"month": 99.2,
"prior": 98.7
}
},
"leader": {
"month": "gemini-3.1-pro",
"month_hit_rate": 0.4851,
"prior": "claude-opus-5",
"prior_hit_rate": 0.4779
},
"per_model": {
"claude-fable-5": {
"hit_month": 0.4537,
"hit_prior": 0.4673,
"hit_delta_pp": -1.4,
"gap_month": 15.6,
"gap_prior": 13.9,
"gap_delta_pp": 1.7,
"n_month": 2945,
"n_prior": 2812
},
"claude-opus-5": {
"hit_month": 0.4538,
"hit_prior": 0.4779,
"hit_delta_pp": -2.4,
"gap_month": 15.9,
"gap_prior": 13.2,
"gap_delta_pp": 2.7,
"n_month": 2532,
"n_prior": 2281
},
"deepseek-v4-pro": {
"hit_month": 0.4526,
"hit_prior": 0.4554,
"hit_delta_pp": -0.3,
"gap_month": 15.3,
"gap_prior": 15.4,
"gap_delta_pp": -0.1,
"n_month": 3352,
"n_prior": 2578
},
"gemini-3.1-pro": {
"hit_month": 0.4851,
"hit_prior": 0.4761,
"hit_delta_pp": 0.9,
"gap_month": 18.0,
"gap_prior": 18.8,
"gap_delta_pp": -0.8,
"n_month": 3284,
"n_prior": 3033
},
"gpt-5.6-sol": {
"hit_month": 0.4446,
"hit_prior": 0.4586,
"hit_delta_pp": -1.4,
"gap_month": 23.5,
"gap_prior": 21.8,
"gap_delta_pp": 1.7,
"n_month": 3133,
"n_prior": 2970
},
"grok-4.6": {
"hit_month": 0.4439,
"hit_prior": 0.4527,
"hit_delta_pp": -0.9,
"gap_month": 15.8,
"gap_prior": 15.1,
"gap_delta_pp": 0.7,
"n_month": 2271,
"n_prior": 2708
},
"qwen-3.8-max": {
"hit_month": 0.4471,
"hit_prior": 0.4565,
"hit_delta_pp": -0.9,
"gap_month": 14.7,
"gap_prior": 14.7,
"gap_delta_pp": 0.0,
"n_month": 3165,
"n_prior": 3001
}
},
"per_symbol": {
"BNB": {
"hit_month": 0.4512,
"hit_prior": 0.4807,
"hit_delta_pp": -3.0,
"n_month": 3994,
"n_prior": 3857
},
"BTC": {
"hit_month": 0.4597,
"hit_prior": 0.4648,
"hit_delta_pp": -0.5,
"n_month": 3987,
"n_prior": 4068
},
"ETH": {
"hit_month": 0.4538,
"hit_prior": 0.4287,
"hit_delta_pp": 2.5,
"n_month": 3973,
"n_prior": 3695
},
"SOL": {
"hit_month": 0.4403,
"hit_prior": 0.4733,
"hit_delta_pp": -3.3,
"n_month": 4420,
"n_prior": 3928
},
"XRP": {
"hit_month": 0.4705,
"hit_prior": 0.4673,
"hit_delta_pp": 0.3,
"n_month": 4308,
"n_prior": 3835
}
},
"log_line": "prior_window: n_mature_directional=19383, effective_start=2026-08-01T00:01:00+00:00, field hit=0.4633 gap=16.3, loo lift=1.9",
"note_definitions": "the prior block is window A of the MONTH pull, matured at the prior cutoff with the same rules; delta = month - prior in percentage points"
},
"prior_reproduces_monthly1": {
"pins_source": "the published Monthly #1 JSONs calibration-monthly-2026-08.json, consensus-watch-monthly-2026-08.json, model-watch-monthly-2026-08.json (window 2026-08-01 .. 2026-09-01, cutoff 2026-09-01T16:00Z)",
"note": "month.prior_window of this run has the window, the cutoff and the definitions of the month Monthly #1 published; each check is printed and stored as a fact. A mismatch is not fatal and nothing is adjusted to match.",
"applicable": true,
"d2_cutoff_used": "2026-09-01T16:00:00+00:00",
"d2_slot_bound": "2026-09-01T00:00:00+00:00",
"checks": {
"prior_window.counters.mature_directional": {
"pin": 19383,
"got": 19383,
"match": true,
"tolerance": 0,
"published_path": "calibration-monthly-2026-08.json $.tables.counters.mature_directional"
},
"prior_window.calibration.field.hit_rate": {
"pin": 0.4633,
"got": 0.4633,
"match": true,
"tolerance": 5e-05,
"published_path": "calibration-monthly-2026-08.json $.tables.calibration_main[7].hit_rate"
},
"prior_window.calibration.field.gap_pp": {
"pin": 16.3,
"got": 16.3,
"match": true,
"tolerance": 0.05,
"published_path": "calibration-monthly-2026-08.json $.tables.calibration_main[7].gap_pp"
},
"prior_window.consensus.loo_pooled.lift_pp": {
"pin": 1.9,
"got": 1.9,
"match": true,
"tolerance": 0.05,
"published_path": "consensus-watch-monthly-2026-08.json $.tables.pooled.lift_pp"
},
"prior_window.model_watch.leaderboard[0].model": {
"pin": "claude-opus-5",
"got": "claude-opus-5",
"match": true,
"tolerance": 0,
"published_path": "model-watch-monthly-2026-08.json $.tables.leaderboard.rows[0].model"
},
"prior_window.model_watch.leaderboard[0].hit_rate": {
"pin": 0.4779,
"got": 0.4779,
"match": true,
"tolerance": 5e-05,
"published_path": "model-watch-monthly-2026-08.json $.tables.leaderboard.rows[0].hit_rate"
},
"research_to_date.at_prev_cutoff.D2_directional_only": {
"pin": 39073,
"got": 39073,
"match": true,
"tolerance": 0,
"published_path": "model-watch-monthly-2026-08.json $.research_to_date.platform.resolved_forecasts"
}
},
"all_match": true,
"n_checks": 7,
"n_match": 7
},
"note": "controls are DESCRIPTIVE and never fatal; they are printed in the log as well",
"builder_prior_vs_published_monthly1": {
"source": "month.prior_window of this pack vs the published Monthly #1 JSONs (calibration-monthly-2026-08.json, consensus-watch-monthly-2026-08.json, model-watch-monthly-2026-08.json)",
"applicable": true,
"checks": [
{
"check": "field hit rate",
"pack_prior_window": 0.4633,
"published_monthly1": 0.4633,
"tolerance": 5e-05,
"match": true,
"published_path": "calibration-monthly-2026-08.json tables.calibration_main[Field].hit_rate"
},
{
"check": "field gap pp",
"pack_prior_window": 16.3,
"published_monthly1": 16.3,
"tolerance": 0.05,
"match": true,
"published_path": "calibration-monthly-2026-08.json tables.calibration_main[Field].gap_pp"
},
{
"check": "field Brier",
"pack_prior_window": 0.2784,
"published_monthly1": 0.2784,
"tolerance": 5e-05,
"match": true,
"published_path": "calibration-monthly-2026-08.json tables.calibration_main[Field].brier"
},
{
"check": "LOO lift pp",
"pack_prior_window": 1.9,
"published_monthly1": 1.9,
"tolerance": 0.05,
"match": true,
"published_path": "consensus-watch-monthly-2026-08.json tables.pooled.lift_pp"
},
{
"check": "herd share pct",
"pack_prior_window": 98.7,
"published_monthly1": 98.7,
"tolerance": 0.05,
"match": true,
"published_path": "consensus-watch-monthly-2026-08.json tables.pooled agree.n / (agree.n + disagree.n)"
},
{
"check": "leader model",
"pack_prior_window": "claude-opus-5",
"published_monthly1": "claude-opus-5",
"tolerance": null,
"match": true,
"published_path": "model-watch-monthly-2026-08.json tables.leaderboard.rows[0].model"
},
{
"check": "leader hit rate",
"pack_prior_window": 0.4779,
"published_monthly1": 0.4779,
"tolerance": 5e-05,
"match": true,
"published_path": "model-watch-monthly-2026-08.json tables.leaderboard.rows[0].hit_rate"
},
{
"check": "mature directional",
"pack_prior_window": 19383,
"published_monthly1": 19383,
"tolerance": 0,
"match": true,
"published_path": "calibration-monthly-2026-08.json tables.counters.mature_directional"
}
],
"n_checks": 8,
"n_match": 8,
"all_match": true,
"note": "printed in the build log as facts; never a source of a printed number"
}
}
}