{
 "slug": "consensus-watch-2026-08-24",
 "series": "weekly",
 "series_label": "WEEKLY · CONSENSUS",
 "issue": 4,
 "title": "Consensus Watch #4",
 "language": "en",
 "window_start": "2026-08-24T00:00:00+00:00",
 "window_end": "2026-08-31T00:00:00+00:00",
 "window_text": "Aug 24-30, 2026 UTC",
 "cutoff": "2026-08-31T16:00:00+00:00",
 "generated_at": "2026-09-02T12:15:22+00:00",
 "methodology": "v1.1 (2026-08-10)",
 "methodology_hash": "e66c7e8c864a2233",
 "source": "weekly_metrics_2026-08-24.json",
 "research_question": "if a model's forecast matches the consensus of the other models, is the forecast more reliable?",
 "method": "leave-one-out strict majority, >=3 directional peers",
 "base_field_hit": {
  "n": 4576,
  "hit_rate": 0.4294
 },
 "market_state": {
  "panel": "market_state_6p2",
  "week": {
   "start": "2026-08-24",
   "end": "2026-08-31"
  },
  "prev_week": {
   "start": "2026-08-17",
   "end": "2026-08-24"
  },
  "current": {
   "btc_net_pct": -0.07,
   "realized_vol_ann_pct": 30.85,
   "volume_top5_usd_bn": 19.7224,
   "volume_top5_wow_pct": -24.1,
   "avg_pairwise_corr": 0.85
  },
  "previous_as_published_issue3": {
   "btc_net_pct": 23.58,
   "realized_vol_ann_pct": 71.1,
   "volume_top5_usd_bn": 25.97,
   "avg_pairwise_corr": 0.81
  },
  "snapshot_line": "BTC net -0.07% (prior +23.58%) - ann vol 30.9% (was 71.1%) - TOP5 volume $19.72B, -24% w/w - pairwise corr 0.85 (was 0.81)",
  "exchange": "binance",
  "audit_filter_applied": true,
  "source": "single-exchange (binance) daily candle recompute, market_state_2026-08-24.json; prior-week values as published in issue #3",
  "audit_note": "Market-state row is computed on a single-exchange (binance) daily candle series, as in issue #3 after the audit that found the raw table mixes two exchanges; issue #3 values are as published.",
  "prev_week_control": {
   "published_issue3": {
    "btc_net_pct": 23.58,
    "realized_vol_ann_pct": 71.1,
    "volume_top5_usd_bn": 25.97,
    "avg_pairwise_corr": 0.81
   },
   "got": {
    "btc_net_pct": 23.58,
    "realized_vol_ann_pct": 65.02,
    "volume_top5_usd_bn": 25.9715,
    "avg_pairwise_corr": 0.81
   },
   "match": {
    "btc_net_pct": true,
    "realized_vol_ann_pct": false,
    "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": false,
   "note": "previous week of this run == published issue-#3 week; the four Snapshot pins must reproduce"
  },
  "regime_days": {
   "2026-08-24": {
    "move_pct": 1.61,
    "regime": "trend"
   },
   "2026-08-25": {
    "move_pct": -0.58,
    "regime": "flat"
   },
   "2026-08-26": {
    "move_pct": 0.63,
    "regime": "flat"
   },
   "2026-08-27": {
    "move_pct": 1.53,
    "regime": "trend"
   },
   "2026-08-28": {
    "move_pct": -2.91,
    "regime": "trend"
   },
   "2026-08-29": {
    "move_pct": 0.48,
    "regime": "flat"
   },
   "2026-08-30": {
    "move_pct": -0.7,
    "regime": "flat"
   }
  },
  "alive_slice_context": {
   "abs_move_1d_pct": {
    "A": 4.283,
    "B": 2.352
   },
   "btc_realized_vol_ann_pct_hourly": {
    "A": 58.72,
    "B": 37.9
   }
  }
 },
 "tables": {
  "pooled": {
   "agree": {
    "n": 3848,
    "hit_rate": 0.4119,
    "wilson_95": [
     0.3964,
     0.4275
    ]
   },
   "disagree": {
    "n": 36,
    "hit_rate": 0.4722,
    "wilson_95": [
     0.3199,
     0.6299
    ]
   },
   "lift_pp": -6.0,
   "thin_or_tied": 692,
   "by_margin": {
    "2": {
     "n": 21,
     "hit_rate": 0.619
    },
    "3": {
     "n": 456,
     "hit_rate": 0.4934
    },
    "4": {
     "n": 620,
     "hit_rate": 0.4758
    },
    "5": {
     "n": 924,
     "hit_rate": 0.3593
    },
    "6": {
     "n": 1827,
     "hit_rate": 0.3941
    }
   },
   "total_scoreable": 3884,
   "caveat": "Read with context. The disagree bucket is n=36 this week against 125 in issue #3 and 6 in issue #2; the two Wilson intervals overlap. This is a single week of dependent observations: not evidence about herding in general, and not a claim that the sign will hold next week."
  },
  "by_cell": [
   {
    "agree": {
     "n": 2374,
     "hit_rate": 0.4044,
     "wilson_95": [
      0.3848,
      0.4243
     ]
    },
    "disagree": {
     "n": 26,
     "hit_rate": 0.4231,
     "wilson_95": [
      0.2554,
      0.6105
     ]
    },
    "lift_pp": -1.9,
    "thin_or_tied": 462,
    "by_margin": {
     "2": {
      "n": 12,
      "hit_rate": 0.5833
     },
     "3": {
      "n": 308,
      "hit_rate": 0.474
     },
     "4": {
      "n": 350,
      "hit_rate": 0.4914
     },
     "5": {
      "n": 570,
      "hit_rate": 0.3596
     },
     "6": {
      "n": 1134,
      "hit_rate": 0.3792
     }
    },
    "n_subjects": 2862,
    "cell": "fh_1h_tf_1h",
    "fh": "1h",
    "tf": "1h"
   },
   {
    "agree": {
     "n": 601,
     "hit_rate": 0.4143,
     "wilson_95": [
      0.3756,
      0.4541
     ]
    },
    "disagree": {
     "n": 5,
     "hit_rate": 0.4,
     "wilson_95": [
      0.1176,
      0.7693
     ]
    },
    "lift_pp": 1.4,
    "thin_or_tied": 85,
    "by_margin": {
     "2": {
      "n": 3,
      "hit_rate": 1.0
     },
     "3": {
      "n": 68,
      "hit_rate": 0.5735
     },
     "4": {
      "n": 100,
      "hit_rate": 0.49
     },
     "5": {
      "n": 150,
      "hit_rate": 0.36
     },
     "6": {
      "n": 280,
      "hit_rate": 0.3714
     }
    },
    "n_subjects": 691,
    "cell": "fh_4h_tf_1h",
    "fh": "4h",
    "tf": "1h"
   },
   {
    "agree": {
     "n": 599,
     "hit_rate": 0.4157,
     "wilson_95": [
      0.3769,
      0.4556
     ]
    },
    "disagree": {
     "n": 4,
     "hit_rate": 0.75,
     "wilson_95": [
      0.3006,
      0.9544
     ]
    },
    "lift_pp": -33.4,
    "thin_or_tied": 109,
    "by_margin": {
     "2": {
      "n": 6,
      "hit_rate": 0.5
     },
     "3": {
      "n": 60,
      "hit_rate": 0.5333
     },
     "4": {
      "n": 120,
      "hit_rate": 0.45
     },
     "5": {
      "n": 126,
      "hit_rate": 0.3413
     },
     "6": {
      "n": 287,
      "hit_rate": 0.4077
     }
    },
    "n_subjects": 712,
    "cell": "fh_4h_tf_4h",
    "fh": "4h",
    "tf": "4h"
   },
   {
    "agree": {
     "n": 115,
     "hit_rate": 0.5304,
     "wilson_95": [
      0.4397,
      0.6192
     ]
    },
    "disagree": {
     "n": 1,
     "hit_rate": 1.0,
     "wilson_95": [
      0.2065,
      1.0
     ]
    },
    "lift_pp": -47.0,
    "thin_or_tied": 27,
    "by_margin": {
     "3": {
      "n": 12,
      "hit_rate": 0.3333
     },
     "4": {
      "n": 40,
      "hit_rate": 0.5
     },
     "5": {
      "n": 42,
      "hit_rate": 0.5
     },
     "6": {
      "n": 21,
      "hit_rate": 0.7619
     }
    },
    "n_subjects": 143,
    "cell": "fh_1d_tf_1d",
    "fh": "1d",
    "tf": "1d"
   },
   {
    "agree": {
     "n": 159,
     "hit_rate": 0.4151,
     "wilson_95": [
      0.3414,
      0.4928
     ]
    },
    "disagree": {
     "n": 0,
     "hit_rate": null,
     "wilson_95": [
      null,
      null
     ]
    },
    "lift_pp": null,
    "thin_or_tied": 9,
    "by_margin": {
     "3": {
      "n": 8,
      "hit_rate": 0.5
     },
     "4": {
      "n": 10,
      "hit_rate": 0.0
     },
     "5": {
      "n": 36,
      "hit_rate": 0.25
     },
     "6": {
      "n": 105,
      "hit_rate": 0.5048
     }
    },
    "n_subjects": 168,
    "cell": "fh_1d_tf_4h",
    "fh": "1d",
    "tf": "4h"
   }
  ],
  "by_margin": [
   {
    "margin": 2,
    "n": 21,
    "hit_rate": 0.619
   },
   {
    "margin": 3,
    "n": 456,
    "hit_rate": 0.4934
   },
   {
    "margin": 4,
    "n": 620,
    "hit_rate": 0.4758
   },
   {
    "margin": 5,
    "n": 924,
    "hit_rate": 0.3593
   },
   {
    "margin": 6,
    "n": 1827,
    "hit_rate": 0.3941
   }
  ],
  "by_margin_week_over_week": [
   {
    "margin": 3,
    "issue3_hit_rate_pct": 56.4,
    "issue4_hit_rate": 0.4934,
    "issue4_n": 456
   },
   {
    "margin": 4,
    "issue3_hit_rate_pct": 48.9,
    "issue4_hit_rate": 0.4758,
    "issue4_n": 620
   },
   {
    "margin": 5,
    "issue3_hit_rate_pct": 54.4,
    "issue4_hit_rate": 0.3593,
    "issue4_n": 924
   },
   {
    "margin": 6,
    "issue3_hit_rate_pct": 56.5,
    "issue4_hit_rate": 0.3941,
    "issue4_n": 1827
   }
  ],
  "per_model": [
   {
    "agree": {
     "n": 572,
     "hit_rate": 0.4143,
     "wilson_95": [
      0.3747,
      0.4551
     ]
    },
    "disagree": {
     "n": 1,
     "hit_rate": 1.0,
     "wilson_95": [
      0.2065,
      1.0
     ]
    },
    "lift_pp": -58.6,
    "thin_or_tied": 78,
    "by_margin": {
     "2": {
      "n": 3,
      "hit_rate": 0.6667
     },
     "3": {
      "n": 70,
      "hit_rate": 0.4286
     },
     "4": {
      "n": 96,
      "hit_rate": 0.4688
     },
     "5": {
      "n": 142,
      "hit_rate": 0.3944
     },
     "6": {
      "n": 261,
      "hit_rate": 0.3985
     }
    },
    "model": "claude-fable-5"
   },
   {
    "agree": {
     "n": 531,
     "hit_rate": 0.4218,
     "wilson_95": [
      0.3805,
      0.4643
     ]
    },
    "disagree": {
     "n": 0,
     "hit_rate": null,
     "wilson_95": [
      null,
      null
     ]
    },
    "lift_pp": null,
    "thin_or_tied": 30,
    "by_margin": {
     "2": {
      "n": 1,
      "hit_rate": 1.0
     },
     "3": {
      "n": 49,
      "hit_rate": 0.551
     },
     "4": {
      "n": 83,
      "hit_rate": 0.5663
     },
     "5": {
      "n": 137,
      "hit_rate": 0.3504
     },
     "6": {
      "n": 261,
      "hit_rate": 0.387
     }
    },
    "model": "claude-opus-5"
   },
   {
    "agree": {
     "n": 572,
     "hit_rate": 0.4248,
     "wilson_95": [
      0.3849,
      0.4657
     ]
    },
    "disagree": {
     "n": 6,
     "hit_rate": 0.1667,
     "wilson_95": [
      0.0301,
      0.5635
     ]
    },
    "lift_pp": 25.8,
    "thin_or_tied": 151,
    "by_margin": {
     "2": {
      "n": 3,
      "hit_rate": 0.3333
     },
     "3": {
      "n": 80,
      "hit_rate": 0.525
     },
     "4": {
      "n": 93,
      "hit_rate": 0.4624
     },
     "5": {
      "n": 135,
      "hit_rate": 0.3852
     },
     "6": {
      "n": 261,
      "hit_rate": 0.4023
     }
    },
    "model": "deepseek-v4-pro"
   },
   {
    "agree": {
     "n": 544,
     "hit_rate": 0.4007,
     "wilson_95": [
      0.3604,
      0.4425
     ]
    },
    "disagree": {
     "n": 23,
     "hit_rate": 0.5217,
     "wilson_95": [
      0.3296,
      0.7076
     ]
    },
    "lift_pp": -12.1,
    "thin_or_tied": 188,
    "by_margin": {
     "2": {
      "n": 4,
      "hit_rate": 0.5
     },
     "3": {
      "n": 67,
      "hit_rate": 0.5075
     },
     "4": {
      "n": 89,
      "hit_rate": 0.4157
     },
     "5": {
      "n": 123,
      "hit_rate": 0.374
     },
     "6": {
      "n": 261,
      "hit_rate": 0.3793
     }
    },
    "model": "gemini-3.1-pro"
   },
   {
    "agree": {
     "n": 610,
     "hit_rate": 0.4082,
     "wilson_95": [
      0.3699,
      0.4477
     ]
    },
    "disagree": {
     "n": 1,
     "hit_rate": 1.0,
     "wilson_95": [
      0.2065,
      1.0
     ]
    },
    "lift_pp": -59.2,
    "thin_or_tied": 118,
    "by_margin": {
     "2": {
      "n": 4,
      "hit_rate": 0.75
     },
     "3": {
      "n": 86,
      "hit_rate": 0.4767
     },
     "4": {
      "n": 111,
      "hit_rate": 0.4775
     },
     "5": {
      "n": 148,
      "hit_rate": 0.3446
     },
     "6": {
      "n": 261,
      "hit_rate": 0.387
     }
    },
    "model": "gpt-5.6-sol"
   },
   {
    "agree": {
     "n": 451,
     "hit_rate": 0.4013,
     "wilson_95": [
      0.3571,
      0.4472
     ]
    },
    "disagree": {
     "n": 4,
     "hit_rate": 0.5,
     "wilson_95": [
      0.15,
      0.85
     ]
    },
    "lift_pp": -9.9,
    "thin_or_tied": 24,
    "by_margin": {
     "2": {
      "n": 3,
      "hit_rate": 0.6667
     },
     "3": {
      "n": 34,
      "hit_rate": 0.5588
     },
     "4": {
      "n": 54,
      "hit_rate": 0.5
     },
     "5": {
      "n": 99,
      "hit_rate": 0.303
     },
     "6": {
      "n": 261,
      "hit_rate": 0.3946
     }
    },
    "model": "grok-4.6"
   },
   {
    "agree": {
     "n": 568,
     "hit_rate": 0.4102,
     "wilson_95": [
      0.3705,
      0.4511
     ]
    },
    "disagree": {
     "n": 1,
     "hit_rate": 0.0,
     "wilson_95": [
      0.0,
      0.7935
     ]
    },
    "lift_pp": 41.0,
    "thin_or_tied": 103,
    "by_margin": {
     "2": {
      "n": 3,
      "hit_rate": 0.6667
     },
     "3": {
      "n": 70,
      "hit_rate": 0.4571
     },
     "4": {
      "n": 94,
      "hit_rate": 0.4574
     },
     "5": {
      "n": 140,
      "hit_rate": 0.35
     },
     "6": {
      "n": 261,
      "hit_rate": 0.41
     }
    },
    "model": "qwen-3.8-max"
   }
  ],
  "herd_worst_cards": [
   {
    "slot": "2026-08-26T12:01:00+00:00",
    "symbol": "XRP",
    "fh": "1h",
    "tf": "1h",
    "side": "short",
    "n_models": 7,
    "mean_conf": 68.9,
    "hit_rate": 0.4286
   },
   {
    "slot": "2026-08-27T15:01:00+00:00",
    "symbol": "SOL",
    "fh": "1h",
    "tf": "1h",
    "side": "long",
    "n_models": 7,
    "mean_conf": 68.9,
    "hit_rate": 0.0
   },
   {
    "slot": "2026-08-27T15:01:00+00:00",
    "symbol": "XRP",
    "fh": "1h",
    "tf": "1h",
    "side": "long",
    "n_models": 7,
    "mean_conf": 68.7,
    "hit_rate": 0.0
   },
   {
    "slot": "2026-08-24T16:01:00+00:00",
    "symbol": "SOL",
    "fh": "4h",
    "tf": "4h",
    "side": "long",
    "n_models": 6,
    "mean_conf": 68.3,
    "hit_rate": 0.0
   },
   {
    "slot": "2026-08-25T04:01:00+00:00",
    "symbol": "SOL",
    "fh": "4h",
    "tf": "1h",
    "side": "long",
    "n_models": 7,
    "mean_conf": 68,
    "hit_rate": 0.0
   }
  ],
  "counters": {
   "forecasts_total": 10290,
   "ok_in_gate": 9301,
   "out_of_gate": 978,
   "out_of_gate_by_fh": {
    "1w": 488,
    "1m": 490
   },
   "invalid": 11,
   "mature": 9301,
   "mature_directional": 4576,
   "mature_sideways": 4725,
   "pending_next_issue": 0,
   "pending_by_fh": {},
   "late_closes": 0,
   "uptime": [
    {
     "fh": "1h",
     "tf": "1h",
     "slots_seen": 168,
     "slots_expected": 168
    },
    {
     "fh": "4h",
     "tf": "4h",
     "slots_seen": 42,
     "slots_expected": 42
    },
    {
     "fh": "4h",
     "tf": "1h",
     "slots_seen": 42,
     "slots_expected": 42
    },
    {
     "fh": "1d",
     "tf": "1d",
     "slots_seen": 7,
     "slots_expected": 7
    },
    {
     "fh": "1d",
     "tf": "4h",
     "slots_seen": 7,
     "slots_expected": 7
    }
   ],
   "regime_days": {
    "2026-08-24": {
     "move_pct": 1.61,
     "regime": "trend"
    },
    "2026-08-25": {
     "move_pct": -0.58,
     "regime": "flat"
    },
    "2026-08-26": {
     "move_pct": 0.63,
     "regime": "flat"
    },
    "2026-08-27": {
     "move_pct": 1.53,
     "regime": "trend"
    },
    "2026-08-28": {
     "move_pct": -2.91,
     "regime": "trend"
    },
    "2026-08-29": {
     "move_pct": 0.48,
     "regime": "flat"
    },
    "2026-08-30": {
     "move_pct": -0.7,
     "regime": "flat"
    }
   },
   "grok_era": {
    "ids_in_window": [
     "grok-4.5",
     "grok-4.6"
    ],
    "n_grok45": 110,
    "n_grok46": 1360,
    "n_grok45_rows": 110,
    "n_grok46_rows": 1360,
    "first_grok46_slot": "2026-08-24T10:01:00+00:00",
    "last_grok45_slot": "2026-08-24T09:01:00+00:00",
    "first_grok45_slot": "2026-08-24T00:01:00+00:00",
    "flip_at": "2026-08-24T09:22:00+00:00",
    "flip_inside_window": true,
    "prev_window": {
     "n_grok45_rows": 1370,
     "n_grok46_rows": 0
    },
    "merged_model_id": "grok-4.6",
    "merged_label": "grok-4.6 (incl. 4.5-era, Aug 24 00:00-09:22 UTC)",
    "lineage_ids": [
     "grok-4.5",
     "grok-4.6"
    ],
    "lineage_source": "benchmarks.config.lineage_ids('grok-4.6') + MODEL_SUCCESSION[grok-4.6]=('grok-4.5',)",
    "lineage_warnings": [],
    "rows_remapped_a_plus_b": 1480,
    "note": "grok-4.6 went live 2026-08-24T09:22:00+00:00 -- INSIDE this window; all per-model aggregates use the lineage splice grok-4.5 -> grok-4.6 (one row, labelled 'grok-4.6 (incl. 4.5-era, Aug 24 00:00-09:22 UTC)'). Week A (prev window) is pure grok-4.5 and its numbers reproduce issue #3.",
    "cells_with_two_grok_rows": 0
   },
   "series_density": {
    "1d": {
     "days_with_slots": 7,
     "expected_days": 7,
     "full_daily_coverage": true,
     "in_fh_gate": true,
     "n_slots": 7,
     "n_rows": 490,
     "days": [
      "2026-08-24",
      "2026-08-25",
      "2026-08-26",
      "2026-08-27",
      "2026-08-28",
      "2026-08-29",
      "2026-08-30"
     ],
     "missing_days": []
    },
    "1h": {
     "days_with_slots": 7,
     "expected_days": 7,
     "full_daily_coverage": true,
     "in_fh_gate": true,
     "n_slots": 168,
     "n_rows": 5880,
     "days": [
      "2026-08-24",
      "2026-08-25",
      "2026-08-26",
      "2026-08-27",
      "2026-08-28",
      "2026-08-29",
      "2026-08-30"
     ],
     "missing_days": []
    },
    "1m": {
     "days_with_slots": 7,
     "expected_days": 7,
     "full_daily_coverage": true,
     "in_fh_gate": false,
     "n_slots": 7,
     "n_rows": 490,
     "days": [
      "2026-08-24",
      "2026-08-25",
      "2026-08-26",
      "2026-08-27",
      "2026-08-28",
      "2026-08-29",
      "2026-08-30"
     ],
     "missing_days": []
    },
    "1w": {
     "days_with_slots": 7,
     "expected_days": 7,
     "full_daily_coverage": true,
     "in_fh_gate": false,
     "n_slots": 7,
     "n_rows": 490,
     "days": [
      "2026-08-24",
      "2026-08-25",
      "2026-08-26",
      "2026-08-27",
      "2026-08-28",
      "2026-08-29",
      "2026-08-30"
     ],
     "missing_days": []
    },
    "4h": {
     "days_with_slots": 7,
     "expected_days": 7,
     "full_daily_coverage": true,
     "in_fh_gate": true,
     "n_slots": 42,
     "n_rows": 2940,
     "days": [
      "2026-08-24",
      "2026-08-25",
      "2026-08-26",
      "2026-08-27",
      "2026-08-28",
      "2026-08-29",
      "2026-08-30"
     ],
     "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-4.5",
     "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": [
     "qwen-3.7-max",
     "claude-opus-4.8"
    ],
    "tracked_total": 10,
    "canon_expected": {
     "current": 7,
     "archived_legacy": 4,
     "tracked_total": 11
    },
    "matches_canon_11_7_4": false,
    "roster_since": "2026-07-11T00:00:00+00:00",
    "note": "current = merged-lineage model ids emitting inside the window (grok-4.5 rows are spliced into grok-4.6); archived legacy = raw ids seen since 2026-07-11 that no longer emit in the window -- grok-4.5 is one of them by raw id even though its rows are spliced. Models retired before 2026-07-11 are not visible to this query."
   }
  }
 },
 "herd_note": "408 cells had 6+ directional models on one side; unanimous_cells_ge6 per weekly metrics",
 "smart_note": "Smart-vs-Outlook deferred: as-of weights pipeline not yet recording; will appear once weights are logged per-slot (methodology 3.2.3 fallback)",
 "key_finding": {
  "label": "KEY FINDING · OBSERVATION (one weekly window)",
  "text": "Herding ran at 99.1% of leave-one-out scoreable calls and the pooled lift came in at -6.0pp: agree-hit 41.2% against 47.2% on disagreement, on a disagree bucket of n=36.",
  "evidence_level": "observation"
 },
 "key_findings": [
  "**OBSERVATION -- 99.1% of the field's leave-one-out calls ran with the herd.** Of 3,884 LOO-scoreable directional calls, 3,848 (99.1%) sided with peers' majority (issue #3: 97.1%); agree-hit was 41.2% [39.6%, 42.8%] vs. 47.2% [32.0%, 63.0%] for the 36 disagreers, a pooled lift of -6.0pp. Descriptive, one week, not a claim of a durable edge.",
  "**The margin curve took a fourth shape.** 3 peers 49.3% (was 56.4%), 4 peers 47.6% (was 48.9%), 5 peers 35.9% (was 54.4%), full unanimity 39.4% (was 56.5%). 2 of the 4 tiers sit above the 42.9% field base; four issues in, the curve has still not repeated a shape.",
  "**Daily-horizon unanimity stayed above the field base.** 1d/1d margin-6 hit 76.2% (n=21) and 1d/4h margin-6 hit 50.5% (n=105); issue #3 printed 81.0% and 61.9% on the same two cells.",
  "**Herd failures did not disappear.** 408 cells had 6+ directional models on one side (issue #3: 437); the 5 worst sat at 68.0-68.9 mean confidence and resolved 0.0%-42.9%."
 ],
 "market_check": {
  "label": "MARKET CHECK · OBSERVATION (two adjacent calendar weeks)",
  "text": "OBSERVATION — mean |1d move| 4.28% -> 2.35% (-45% rel), BTC realized vol 58.7% -> 37.9% (ann., hourly); field directional accuracy 54.0% -> 42.9% (-11.1pp), 0 of 7 models improved, sim win-rate up for 0 of 7, field sim PnL +$889.37 -> -$473.59 (adjacent calendar weeks Aug 17-23 vs Aug 24-30; descriptive, one pair of weeks, not a claim).",
  "robustness": "Robustness: raw price-sign accuracy 53.5% -> 44.2% — the same direction as the trade-based hit rule.",
  "weeks": {
   "A": {
    "days": "2026-08-17..2026-08-23",
    "slots": [
     "2026-08-17T00:00:00+00:00",
     "2026-08-24T00:00:00+00:00"
    ],
    "cutoff": "2026-08-24T16:00:00+00:00"
   },
   "B": {
    "days": "2026-08-24..2026-08-30",
    "slots": [
     "2026-08-24T00:00:00+00:00",
     "2026-08-31T00:00:00+00:00"
    ],
    "cutoff": "2026-08-31T16:00:00+00:00"
   }
  },
  "price_source": "crypto_spot/1h@:00",
  "accuracy_field": {
   "A": {
    "n": 5130,
    "hit_rate": 0.5405,
    "wilson_95": [
     0.5269,
     0.5541
    ]
   },
   "B": {
    "n": 4576,
    "hit_rate": 0.4294,
    "wilson_95": [
     0.4151,
     0.4438
    ]
   },
   "delta_pp": -11.1
  },
  "accuracy_per_model": [
   {
    "model": "deepseek-v4-pro",
    "model_label": "deepseek-v4-pro",
    "A": {
     "n": 777,
     "hit_rate": 0.5161,
     "wilson_95": [
      0.481,
      0.5511
     ]
    },
    "B": {
     "n": 729,
     "hit_rate": 0.4458,
     "wilson_95": [
      0.4101,
      0.4821
     ]
    },
    "delta_pp": -7.0
   },
   {
    "model": "qwen-3.8-max",
    "model_label": "qwen-3.8-max",
    "A": {
     "n": 706,
     "hit_rate": 0.5227,
     "wilson_95": [
      0.4858,
      0.5593
     ]
    },
    "B": {
     "n": 672,
     "hit_rate": 0.4256,
     "wilson_95": [
      0.3887,
      0.4633
     ]
    },
    "delta_pp": -9.7
   },
   {
    "model": "gemini-3.1-pro",
    "model_label": "gemini-3.1-pro",
    "A": {
     "n": 798,
     "hit_rate": 0.5414,
     "wilson_95": [
      0.5067,
      0.5756
     ]
    },
    "B": {
     "n": 755,
     "hit_rate": 0.4424,
     "wilson_95": [
      0.4073,
      0.478
     ]
    },
    "delta_pp": -9.9
   },
   {
    "model": "gpt-5.6-sol",
    "model_label": "gpt-5.6-sol",
    "A": {
     "n": 688,
     "hit_rate": 0.5363,
     "wilson_95": [
      0.499,
      0.5733
     ]
    },
    "B": {
     "n": 729,
     "hit_rate": 0.4266,
     "wilson_95": [
      0.3912,
      0.4628
     ]
    },
    "delta_pp": -11.0
   },
   {
    "model": "grok-4.6",
    "model_label": "grok-4.6 (incl. 4.5-era, Aug 24 00:00-09:22 UTC)",
    "A": {
     "n": 710,
     "hit_rate": 0.5141,
     "wilson_95": [
      0.4773,
      0.5507
     ]
    },
    "B": {
     "n": 479,
     "hit_rate": 0.3987,
     "wilson_95": [
      0.3559,
      0.4432
     ]
    },
    "delta_pp": -11.5
   },
   {
    "model": "claude-fable-5",
    "model_label": "claude-fable-5",
    "A": {
     "n": 779,
     "hit_rate": 0.5648,
     "wilson_95": [
      0.5298,
      0.5992
     ]
    },
    "B": {
     "n": 651,
     "hit_rate": 0.4301,
     "wilson_95": [
      0.3926,
      0.4684
     ]
    },
    "delta_pp": -13.5
   },
   {
    "model": "claude-opus-5",
    "model_label": "claude-opus-5",
    "A": {
     "n": 672,
     "hit_rate": 0.5908,
     "wilson_95": [
      0.5532,
      0.6273
     ]
    },
    "B": {
     "n": 561,
     "hit_rate": 0.4242,
     "wilson_95": [
      0.384,
      0.4655
     ]
    },
    "delta_pp": -16.7
   }
  ],
  "raw_sign_field": {
   "A": {
    "n_scored": 5118,
    "hit_rate": 0.5348
   },
   "B": {
    "n_scored": 4567,
    "hit_rate": 0.4419
   }
  },
  "abs_move_1d": {
   "A": 4.283,
   "B": 2.352,
   "delta_pct_rel": -45.1
  },
  "btc_realized_vol_ann_pct_hourly": {
   "A": 58.72,
   "B": 37.9
  },
  "trading_field": {
   "A": {
    "n_trades": 5130,
    "wr": 0.4626,
    "pnl_net_usd": 889.37,
    "pnl_gross_usd": 1402.37
   },
   "B": {
    "n_trades": 4576,
    "wr": 0.3702,
    "pnl_net_usd": -473.59,
    "pnl_gross_usd": -15.99
   }
  },
  "verdict": {
   "models_with_hit_improved": "0 of 7",
   "models_with_rawsign_improved": "0 of 7",
   "models_with_wr_improved": "0 of 7",
   "field_hit_delta_pp": -11.1,
   "field_pnl_net": {
    "A": 889.37,
    "B": -473.59
   },
   "abs_move_delta": {
    "1h": {
     "A": 0.541,
     "B": 0.401,
     "delta_pct_rel": -25.9
    },
    "4h": {
     "A": 1.314,
     "B": 0.81,
     "delta_pct_rel": -38.4
    },
    "1d": {
     "A": 4.283,
     "B": 2.352,
     "delta_pct_rel": -45.1
    }
   },
   "btc_vol_ann_pct": {
    "A": 58.72,
    "B": 37.9
   },
   "note": "descriptive, two adjacent calendar weeks; 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: grok-4.5 rows are aggregated into grok-4.6 (flip 2026-08-24T09:22:00+00:00, inside week B). Week A (2026-08-17..2026-08-23) is pure grok-4.5 and reproduces the published issue #3; week B mixes the 4.5-era (Aug 24 00:00-09:22) with 4.6."
 },
 "why_it_matters": "MarketMania publishes forecasts from 7 frontier model lines into the same slots every week; a natural question for anyone reading the feed is whether a model's call is worth more when it agrees with what everyone else is saying, or whether that agreement is just everyone making the same mistake together. Consensus Watch answers that question empirically, one week at a time, using a leave-one-out design built specifically to avoid the obvious trap of comparing a model to an index that includes its own vote.",
 "how_to_read_this": "For every model M and every forecast it made, we rebuild that slot's consensus using only the OTHER active models in the same symbol / forecast-horizon (FH) / timeframe (TF) cell -- M's own call never counts toward its own consensus. M is scored agree if its side (long or short) matches the strict majority of those peers, and disagree if it sits alone against that majority; a cell needs at least 3 directional peers to count at all, and tied peer splits are excluded rather than forced either way. LOO peers are genuinely external to the model being scored, so the comparison below is not comparing a model to itself. The grok rows are one lineage-spliced line this issue (grok-4.6 (incl. 4.5-era, Aug 24 00:00-09:22)), so grok never counts as two peers in a cell.",
 "practical_implications": [
  "Four issues, four margin-curve readings. No consensus tier has shown a stable meaning, so a consensus filter still should not carry weight in a config.",
  "The pooled lift is -6.0pp this week against +17.1pp in issue #3 and -6.9pp in issue #2. Three readings, three magnitudes: treat the number as this week's weather, not an edge.",
  "Read the consensus number together with the market check on the same page. The field base moved -11.1pp week over week, and every consensus tier moves with it."
 ],
 "limitations": [
  "Single week (Aug 24-30, 2026 UTC). Every finding is descriptive for this window only.",
  "Observations are not independent: the same models watch overlapping symbol / FH / TF cells hour after hour. Wilson intervals here are descriptive, not inferential.",
  "Correlational, not causal: all models see the same market data, so 'agreement' and 'hit' can rise together simply because a slot was easy to read. LOO removes self-match bias, not this confound.",
  "The pooled lift rests on 36 disagreeing calls; per-model counts run 0-23. 692 calls (15.1% of the 4,576-call mature directional universe) were thin or tied and are excluded from every consensus table.",
  "This window ran 3 trend days of 7 (issue #3: 5 of 7); every comparison with issue #3 is a comparison across regimes as well as across weeks. Series density and lineage notes (wave 4): (1) rolling-1w forecasts emit daily since Aug 22, 2026 and rolling-1M daily since Aug 23, 2026, so inside this window (Aug 24-30) daily coverage is FULL for the first time in the series: the 1w series has slots on 7 of 7 days and the 1M series on 7 of 7 days; both sit outside this report's FH gate (1h/4h/1d) and appear only in the exclusions counter (1w 488, 1M 490). (2) The grok line flipped 4.5 -> 4.6 at Aug 24, 2026 09:22 UTC, INSIDE this window: every grok row in this issue is the lineage splice \"grok-4.6 (incl. 4.5-era, Aug 24 00:00-09:22)\" (110 rows from the 4.5 era, 1,360 from 4.6); issue #3 was pure grok-4.5, so the grok week-over-week row joins a spliced line to a single-model line. (3) qwen-3.8-max succeeded qwen-3.7-max on Aug 5, 2026 (fully before this window).",
  "Market-state row is computed on a single-exchange (binance) daily candle series, as in issue #3 after the audit that found the raw table mixes two exchanges; issue #3 values are as published.",
  "Research-to-date counter: directional forecasts resolved inside the FH gate (1h/4h/1d) since Jul 11, counted at the issue cutoff (Aug 31 16:00 UTC) -- the definition pinned in issue #3, which printed 33,809 at the Aug 24 16:00 cutoff. The pack reproduces that pin at the previous cutoff (control OK), so this issue's 38,385 is an additive step under one definition; counter deltas against issue #2 and earlier remain definitional."
 ],
 "notes": {
  "series_density_and_lineage_wave4": "Series density and lineage notes (wave 4): (1) rolling-1w forecasts emit daily since Aug 22, 2026 and rolling-1M daily since Aug 23, 2026, so inside this window (Aug 24-30) daily coverage is FULL for the first time in the series: the 1w series has slots on 7 of 7 days and the 1M series on 7 of 7 days; both sit outside this report's FH gate (1h/4h/1d) and appear only in the exclusions counter (1w 488, 1M 490). (2) The grok line flipped 4.5 -> 4.6 at Aug 24, 2026 09:22 UTC, INSIDE this window: every grok row in this issue is the lineage splice \"grok-4.6 (incl. 4.5-era, Aug 24 00:00-09:22)\" (110 rows from the 4.5 era, 1,360 from 4.6); issue #3 was pure grok-4.5, so the grok week-over-week row joins a spliced line to a single-model line. (3) qwen-3.8-max succeeded qwen-3.7-max on Aug 5, 2026 (fully before this window).",
  "grok_era": {
   "ids_in_window": [
    "grok-4.5",
    "grok-4.6"
   ],
   "n_grok45": 110,
   "n_grok46": 1360,
   "n_grok45_rows": 110,
   "n_grok46_rows": 1360,
   "first_grok46_slot": "2026-08-24T10:01:00+00:00",
   "last_grok45_slot": "2026-08-24T09:01:00+00:00",
   "first_grok45_slot": "2026-08-24T00:01:00+00:00",
   "flip_at": "2026-08-24T09:22:00+00:00",
   "flip_inside_window": true,
   "prev_window": {
    "n_grok45_rows": 1370,
    "n_grok46_rows": 0
   },
   "merged_model_id": "grok-4.6",
   "merged_label": "grok-4.6 (incl. 4.5-era, Aug 24 00:00-09:22 UTC)",
   "lineage_ids": [
    "grok-4.5",
    "grok-4.6"
   ],
   "lineage_source": "benchmarks.config.lineage_ids('grok-4.6') + MODEL_SUCCESSION[grok-4.6]=('grok-4.5',)",
   "lineage_warnings": [],
   "rows_remapped_a_plus_b": 1480,
   "note": "grok-4.6 went live 2026-08-24T09:22:00+00:00 -- INSIDE this window; all per-model aggregates use the lineage splice grok-4.5 -> grok-4.6 (one row, labelled 'grok-4.6 (incl. 4.5-era, Aug 24 00:00-09:22 UTC)'). Week A (prev window) is pure grok-4.5 and its numbers reproduce issue #3.",
   "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-4.5",
    "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": [
    "qwen-3.7-max",
    "claude-opus-4.8"
   ],
   "tracked_total": 10,
   "canon_expected": {
    "current": 7,
    "archived_legacy": 4,
    "tracked_total": 11
   },
   "matches_canon_11_7_4": false,
   "roster_since": "2026-07-11T00:00:00+00:00",
   "note": "current = merged-lineage model ids emitting inside the window (grok-4.5 rows are spliced into grok-4.6); archived legacy = raw ids seen since 2026-07-11 that no longer emit in the window -- grok-4.5 is one of them by raw id even though its rows are spliced. Models retired before 2026-07-11 are not visible to this query."
  },
  "series_density": {
   "1d": {
    "days_with_slots": 7,
    "expected_days": 7,
    "full_daily_coverage": true,
    "in_fh_gate": true,
    "n_slots": 7,
    "n_rows": 490,
    "days": [
     "2026-08-24",
     "2026-08-25",
     "2026-08-26",
     "2026-08-27",
     "2026-08-28",
     "2026-08-29",
     "2026-08-30"
    ],
    "missing_days": []
   },
   "1h": {
    "days_with_slots": 7,
    "expected_days": 7,
    "full_daily_coverage": true,
    "in_fh_gate": true,
    "n_slots": 168,
    "n_rows": 5880,
    "days": [
     "2026-08-24",
     "2026-08-25",
     "2026-08-26",
     "2026-08-27",
     "2026-08-28",
     "2026-08-29",
     "2026-08-30"
    ],
    "missing_days": []
   },
   "1m": {
    "days_with_slots": 7,
    "expected_days": 7,
    "full_daily_coverage": true,
    "in_fh_gate": false,
    "n_slots": 7,
    "n_rows": 490,
    "days": [
     "2026-08-24",
     "2026-08-25",
     "2026-08-26",
     "2026-08-27",
     "2026-08-28",
     "2026-08-29",
     "2026-08-30"
    ],
    "missing_days": []
   },
   "1w": {
    "days_with_slots": 7,
    "expected_days": 7,
    "full_daily_coverage": true,
    "in_fh_gate": false,
    "n_slots": 7,
    "n_rows": 490,
    "days": [
     "2026-08-24",
     "2026-08-25",
     "2026-08-26",
     "2026-08-27",
     "2026-08-28",
     "2026-08-29",
     "2026-08-30"
    ],
    "missing_days": []
   },
   "4h": {
    "days_with_slots": 7,
    "expected_days": 7,
    "full_daily_coverage": true,
    "in_fh_gate": true,
    "n_slots": 42,
    "n_rows": 2940,
    "days": [
     "2026-08-24",
     "2026-08-25",
     "2026-08-26",
     "2026-08-27",
     "2026-08-28",
     "2026-08-29",
     "2026-08-30"
    ],
    "missing_days": []
   }
  },
  "week_over_week_note": "Week-over-week columns compare back-to-back windows: Aug 17-23 (issue #3, and week A of this issue's alive slice) vs Aug 24-30 (this issue). \"Was\" values are the numbers published in issue #3; 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 (Aug 31 16:00 UTC) -- the definition pinned in issue #3, which printed 33,809 at the Aug 24 16:00 cutoff. The pack reproduces that pin at the previous cutoff (control OK), so this issue's 38,385 is an additive step under one definition; counter deltas against issue #2 and earlier remain definitional.",
  "market_state_audit": "Market-state row is computed on a single-exchange (binance) daily candle series, as in issue #3 after the audit that found the raw table mixes two exchanges; issue #3 values are as published.",
  "engine": "Engine 1.1 has powered the sandbox since Aug 18, i.e. before this 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"
 },
 "bibtex_key": "mm_consensus_2026w35",
 "citation": {
  "bibtex_key": "mm_consensus_2026w35",
  "title": "Consensus Watch #4: does agreeing with the crowd make an LLM's market call safer?",
  "author": "MarketMania Research",
  "year": 2026,
  "month": "September",
  "day": 2,
  "url": "https://marketmania.ai/research/reports/consensus-watch-2026-08-24.pdf",
  "note": "Methodology v1.1; window Aug 24-30, 2026 UTC; source weekly_metrics_2026-08-24.json"
 },
 "living_series_note": "**Issue #4.** Consensus Watch is a living weekly comparison; each issue appends one more week of leave-one-out agreement-vs-hit data. Issue #3 asked whether the agree-side lift survives a fourth week: this week it reads -6.0pp (agree 41.2% vs disagree 47.2% on n=36), and whether the 4-peer dip repeats: the 4-peer tier came in above the field base. The grok line is spliced this issue (grok-4.6 (incl. 4.5-era, Aug 24 00:00-09:22)) because the 4.5 -> 4.6 flip happened inside the window. Engine 1.1 has powered the sandbox since Aug 18, i.e. before this window; no cross-engine PnL comparisons are claimed.",
 "testing_next": [
  "Next issue: does the pooled agree-vs-disagree sign settle, or does it keep changing with the regime?",
  "Does any margin tier repeat its position against the field base two issues running?",
  "Monthly series: the agreement curve across regimes at monthly n."
 ],
 "related_research": [
  {
   "title": "Weekly Calibration #4",
   "url": "https://marketmania.ai/research/reports/weekly-calibration-2026-08-24.pdf"
  },
  {
   "title": "Weekly Model Watch #4",
   "url": "https://marketmania.ai/research/reports/model-watch-2026-08-24.pdf"
  },
  {
   "title": "Consensus Watch #3",
   "url": "https://marketmania.ai/research/reports/consensus-watch-2026-08-17.pdf"
  }
 ],
 "related_research_note": "The three weekly reports publish together as one issue each week; each links straight to the others' PDF and to its own previous issue. Direct links are the posting rule from wave 2 on.",
 "research_to_date": {
  "this_report": {
   "loo_scored_observations": 3884
  },
  "platform": {
   "as_of_cutoff": "2026-08-31",
   "resolved_forecasts": 38385,
   "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": 10,
   "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": 15
  },
  "source": "weekly_metrics_2026-08-24.json",
  "control_reproduces_wave3_pin": true,
  "pin_note": "Research-to-date counter: directional forecasts resolved inside the FH gate (1h/4h/1d) since Jul 11, counted at the issue cutoff (Aug 31 16:00 UTC) -- the definition pinned in issue #3, which printed 33,809 at the Aug 24 16:00 cutoff. The pack reproduces that pin at the previous cutoff (control OK), so this issue's 38,385 is an additive step under one definition; counter deltas against issue #2 and earlier remain definitional."
 },
 "site_entry": {
  "slug": "consensus-watch-2026-08-24",
  "series": "weekly",
  "seriesLabel": "Weekly · Consensus",
  "date": "September 2, 2026",
  "title": "Consensus Watch #4",
  "summary": "9,301 mature forecasts asked whether agreeing with the crowd makes an LLM market call safer. Herding ran at 99.1% of leave-one-out scoreable calls (issue #3: 97.1%), and the pooled lift came in at -6.0pp: agree-hit 41.2% against 47.2% on disagreement, on a disagree bucket of n=36 and a 42.9% field base. The margin curve read 49.3% / 47.6% / 35.9% / 39.4% from 3 peers to full unanimity. Field accuracy moved 54.0% -> 42.9% (-11.1pp) week over week, with 0 of 7 models improving.",
  "stats": [
   "3,884 peer-judged calls",
   "99.1% ran with the herd",
   "lift -6.0pp vs disagree"
  ],
  "manifest_key_finding": "Herding ran at 99.1% of leave-one-out scoreable calls and the pooled lift came in at -6.0pp: agree-hit 41.2% against 47.2% on disagreement (disagree bucket n=36), against a 42.9% field base.",
  "files": {
   "pdf": "/research/reports/consensus-watch-2026-08-24.pdf",
   "md": "/research/reports/consensus-watch-2026-08-24.md",
   "json": "/research/reports/consensus-watch-2026-08-24.json"
  }
 },
 "social": {
  "telegram": {
   "photo_caption_html": "🧭 <b>Consensus Watch #4</b> — weekly series (window Aug 24–30)\n\nResearch question: if a model's forecast matches the consensus of the other models, is it more reliable?\n\n3,884 leave-one-out scored calls, 7 model lines:\n• Herding ran at <b>99.1%</b> (issue #3: 97.1%) — agree-hit 41.2% vs <b>47.2%</b> when a model left the crowd (field base 42.9%, disagree n=36)\n• Margin curve: 3-peer 49.3%, 4-peer 47.6%, 5-peer 35.9%, unanimity 39.4%\n• Daily horizon: 1d/1d unanimity 76.2%, 1d/4h unanimity 50.5%\n• Market check: field accuracy 54.0% -> 42.9% (-11.1pp), 0 of 7 models improved; avg |1d| move 4.28% -> 2.35%\n\nNote: grok flipped 4.5 -> 4.6 inside this window; grok rows are one spliced line (4.5 era Aug 24 00:00-09:22 + 4.6)\n\n📄 <a href=\"https://marketmania.ai/research/reports/consensus-watch-2026-08-24.pdf\">Full report (PDF)</a> · <a href=\"https://marketmania.ai/research\">all research</a>\n\n#cryptotrading #LLM #AI #crypto #trading #consensus #research",
   "document_caption_html": "📄 Consensus Watch #4 — full report (PDF).\nWeb copy: <a href=\"https://marketmania.ai/research/reports/consensus-watch-2026-08-24.pdf\">consensus-watch-2026-08-24.pdf</a> · machine-readable: <a href=\"https://marketmania.ai/research/reports/consensus-watch-2026-08-24.json\">JSON</a> · <a href=\"https://marketmania.ai/research\">all research</a>\n\n#cryptotrading #LLM #AI #crypto #trading #consensus #research",
   "caption_len": 955,
   "document_caption_len": 402,
   "limit": 1024
  },
  "x": {
   "main": "Consensus Watch #4 is out.\n\n3,884 peer-judged LLM market calls: 99.1% ran with the herd, and agreeing scored 41.2% against 47.2% for the 36 calls that left it (-6.0pp).\n\nFull report: https://marketmania.ai/research/reports/consensus-watch-2026-08-24.pdf\n\n#AI #Crypto #Research #Consensus",
   "main_tco_len": 240,
   "reply": "Margin curve, w/w (hit):\n3 peers: 56.4 -> 49.3\n4 peers: 48.9 -> 47.6\n5 peers: 54.4 -> 35.9\nunanimity: 56.5 -> 39.4\nField base: 54.0 -> 42.9. LOO 3+ peers, methodology v1.1.\nData: https://marketmania.ai/research/reports/consensus-watch-2026-08-24.json",
   "reply_tco_len": 202,
   "limit_tco": 280,
   "replies_max": 1
  }
 }
}