{
  "slug": "model-watch-2026-08-03",
  "series": "weekly",
  "series_label": "WEEKLY · MODEL WATCH",
  "issue": 1,
  "title": "Weekly Model Watch #1",
  "language": "en",
  "window_start": "2026-08-03T00:00:00+00:00",
  "window_end": "2026-08-10T00:00:00+00:00",
  "window_text": "Aug 3-9, 2026 UTC",
  "cutoff": "2026-08-10T16:00:00+00:00",
  "generated_at": "2026-08-11T14:24:23.984020+00:00",
  "methodology": "v1.1 (2026-08-10)",
  "methodology_hash": "e66c7e8c864a2233",
  "source": "weekly_metrics_2026-08-03.json",
  "hit_rule": "direction: exit tp1/tp2 -> hit, sl -> miss, expiry -> sign of gross pnl",
  "bibtex_key": "mm_model_watch_2026w32",
  "research_question": "who led the week, 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": 4042,
    "hit_rate": 0.4243
  },
  "key_finding": {
    "label": "KEY FINDING · OBSERVATION (one weekly window)",
    "text": "The week's only trend reversal went uncalled: zero of 8 models caught BTC's +2.1% Monday turn.",
    "evidence_level": "observation"
  },
  "tables": {
    "leaderboard": {
      "rows": [
        {
          "model": "qwen-3.8-max",
          "legacy": false,
          "n": 447,
          "coverage": 0.5392,
          "hit_rate": 0.4407,
          "wilson_95": {
            "lo": 0.3954,
            "hi": 0.4871
          }
        },
        {
          "model": "gemini-3.1-pro",
          "legacy": false,
          "n": 629,
          "coverage": 0.4831,
          "hit_rate": 0.4372,
          "wilson_95": {
            "lo": 0.3989,
            "hi": 0.4762
          }
        },
        {
          "model": "gpt-5.6-sol",
          "legacy": false,
          "n": 612,
          "coverage": 0.469,
          "hit_rate": 0.4297,
          "wilson_95": {
            "lo": 0.3911,
            "hi": 0.4693
          }
        },
        {
          "model": "qwen-3.7-max",
          "legacy": true,
          "n": 263,
          "coverage": 0.5596,
          "hit_rate": 0.4221,
          "wilson_95": {
            "lo": 0.3639,
            "hi": 0.4824
          }
        },
        {
          "model": "grok-4.5",
          "legacy": false,
          "n": 607,
          "coverage": 0.4669,
          "hit_rate": 0.4217,
          "wilson_95": {
            "lo": 0.3831,
            "hi": 0.4614
          }
        },
        {
          "model": "deepseek-v4-pro",
          "legacy": false,
          "n": 377,
          "coverage": 0.2889,
          "hit_rate": 0.4164,
          "wilson_95": {
            "lo": 0.3678,
            "hi": 0.4668
          }
        },
        {
          "model": "claude-fable-5",
          "legacy": false,
          "n": 642,
          "coverage": 0.492,
          "hit_rate": 0.4143,
          "wilson_95": {
            "lo": 0.3768,
            "hi": 0.4528
          }
        },
        {
          "model": "claude-opus-5",
          "legacy": false,
          "n": 465,
          "coverage": 0.3563,
          "hit_rate": 0.4086,
          "wilson_95": {
            "lo": 0.3648,
            "hi": 0.4539
          }
        }
      ],
      "title_awarded": false,
      "title_rule": ">=5pp gap AND non-overlapping 95% CI, N>=10",
      "movement_vs_prior_week": "first issue -- starts next week (issue #2)",
      "top4_spread_pp": 1.86,
      "note": "hit-rate rank and Brier rank diverge for several models -- see block8.calibration_bridge"
    },
    "ticker": {
      "rows": [
        {
          "symbol": "SOL",
          "n": 852,
          "hit_rate": 0.4824,
          "wilson_95": {
            "lo": 0.449,
            "hi": 0.516
          }
        },
        {
          "symbol": "XRP",
          "n": 800,
          "hit_rate": 0.465,
          "wilson_95": {
            "lo": 0.4307,
            "hi": 0.4996
          }
        },
        {
          "symbol": "BNB",
          "n": 782,
          "hit_rate": 0.4361,
          "wilson_95": {
            "lo": 0.4017,
            "hi": 0.471
          }
        },
        {
          "symbol": "BTC",
          "n": 884,
          "hit_rate": 0.3835,
          "wilson_95": {
            "lo": 0.352,
            "hi": 0.416
          }
        },
        {
          "symbol": "ETH",
          "n": 724,
          "hit_rate": 0.3481,
          "wilson_95": {
            "lo": 0.3143,
            "hi": 0.3835
          }
        }
      ],
      "best": "SOL",
      "hardest": "ETH",
      "best_vs_hardest_ci_overlap": false,
      "adjacent_pairs_all_overlap": true
    },
    "pairs_top": [
      {
        "model": "qwen-3.8-max",
        "symbol": "SOL",
        "n": 104,
        "hit_rate": 0.5288
      },
      {
        "model": "qwen-3.8-max",
        "symbol": "XRP",
        "n": 96,
        "hit_rate": 0.5208
      },
      {
        "model": "deepseek-v4-pro",
        "symbol": "SOL",
        "n": 80,
        "hit_rate": 0.5
      },
      {
        "model": "gpt-5.6-sol",
        "symbol": "XRP",
        "n": 119,
        "hit_rate": 0.4958
      },
      {
        "model": "claude-opus-5",
        "symbol": "SOL",
        "n": 110,
        "hit_rate": 0.4909
      },
      {
        "model": "gemini-3.1-pro",
        "symbol": "XRP",
        "n": 141,
        "hit_rate": 0.4894
      }
    ],
    "pairs_worst": [
      {
        "model": "gpt-5.6-sol",
        "symbol": "ETH",
        "n": 113,
        "hit_rate": 0.3451
      },
      {
        "model": "gemini-3.1-pro",
        "symbol": "ETH",
        "n": 108,
        "hit_rate": 0.3426
      },
      {
        "model": "deepseek-v4-pro",
        "symbol": "BTC",
        "n": 81,
        "hit_rate": 0.3333
      },
      {
        "model": "claude-opus-5",
        "symbol": "ETH",
        "n": 73,
        "hit_rate": 0.3288
      },
      {
        "model": "qwen-3.8-max",
        "symbol": "ETH",
        "n": 88,
        "hit_rate": 0.2841
      }
    ],
    "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": "2026-08-03T08:00:00+00:00",
          "new_side": "long",
          "move_8h_pct": 2.1,
          "n_callers": 0,
          "first_caller": null,
          "caller_models": []
        }
      ],
      "scope_v1": "BTC/ETH only; expansion planned after 2-3 issues"
    },
    "cross_tf": {
      "fh_4h_tf_4h_vs_1h": {
        "agree": {
          "n": 441,
          "hit_rate": 0.3968,
          "wilson_95": [
            0.3522,
            0.4432
          ],
          "by_regime": {
            "trend": {
              "n": 0,
              "hit_rate": null
            },
            "flat": {
              "n": 441,
              "hit_rate": 0.3968
            }
          }
        },
        "disagree": {
          "n": 7,
          "hit_rate": 0.7143,
          "wilson_95": [
            0.3589,
            0.9178
          ],
          "by_regime": {
            "trend": {
              "n": 0,
              "hit_rate": null
            },
            "flat": {
              "n": 7,
              "hit_rate": 0.7143
            }
          }
        },
        "mixed_pairs": 208,
        "base": {
          "n": 665,
          "hit_rate": 0.4165
        },
        "lift_pp": -2.0
      },
      "fh_1d_tf_1d_vs_4h": {
        "agree": {
          "n": 53,
          "hit_rate": 0.3396,
          "wilson_95": [
            0.2269,
            0.4741
          ],
          "by_regime": {
            "trend": {
              "n": 0,
              "hit_rate": null
            },
            "flat": {
              "n": 53,
              "hit_rate": 0.3396
            }
          }
        },
        "disagree": {
          "n": 0,
          "hit_rate": null,
          "wilson_95": [
            null,
            null
          ],
          "by_regime": {
            "trend": {
              "n": 0,
              "hit_rate": null
            },
            "flat": {
              "n": 0,
              "hit_rate": null
            }
          }
        },
        "mixed_pairs": 25,
        "base": {
          "n": 78,
          "hit_rate": 0.3846
        },
        "lift_pp": -4.5
      }
    },
    "cross_fh": {
      "fh_4h_confirmed_by_1h": {
        "agree": {
          "n": 410,
          "hit_rate": 0.3976,
          "wilson_95": [
            0.3514,
            0.4457
          ],
          "by_regime": {
            "trend": {
              "n": 0,
              "hit_rate": null
            },
            "flat": {
              "n": 410,
              "hit_rate": 0.3976
            }
          }
        },
        "disagree": {
          "n": 17,
          "hit_rate": 0.7059,
          "wilson_95": [
            0.4687,
            0.8672
          ],
          "by_regime": {
            "trend": {
              "n": 0,
              "hit_rate": null
            },
            "flat": {
              "n": 17,
              "hit_rate": 0.7059
            }
          }
        },
        "mixed_pairs": 238,
        "base": {
          "n": 665,
          "hit_rate": 0.4165
        },
        "lift_pp": -1.9
      },
      "fh_1d_confirmed_by_4h": {
        "agree": {
          "n": 44,
          "hit_rate": 0.3636,
          "wilson_95": [
            0.2378,
            0.5113
          ],
          "by_regime": {
            "trend": {
              "n": 0,
              "hit_rate": null
            },
            "flat": {
              "n": 44,
              "hit_rate": 0.3636
            }
          }
        },
        "disagree": {
          "n": 1,
          "hit_rate": 0.0,
          "wilson_95": [
            0.0,
            0.7935
          ],
          "by_regime": {
            "trend": {
              "n": 0,
              "hit_rate": null
            },
            "flat": {
              "n": 1,
              "hit_rate": 0.0
            }
          }
        },
        "mixed_pairs": 33,
        "base": {
          "n": 78,
          "hit_rate": 0.3846
        },
        "lift_pp": -2.1
      }
    },
    "regime_days": {
      "2026-08-03": {
        "move_pct": 0.09,
        "regime": "flat"
      },
      "2026-08-04": {
        "move_pct": 0.93,
        "regime": "flat"
      },
      "2026-08-05": {
        "move_pct": 0.98,
        "regime": "flat"
      },
      "2026-08-06": {
        "move_pct": -0.57,
        "regime": "flat"
      },
      "2026-08-07": {
        "move_pct": 0.88,
        "regime": "flat"
      },
      "2026-08-08": {
        "move_pct": 0.16,
        "regime": "flat"
      },
      "2026-08-09": {
        "move_pct": 0.09,
        "regime": "flat"
      }
    },
    "side_mix": [
      {
        "model": "claude-fable-5",
        "legacy": false,
        "n": 1305,
        "long": 0.3785,
        "short": 0.1134,
        "sideways": 0.508,
        "long_short_ratio": 3.3377
      },
      {
        "model": "claude-opus-5",
        "legacy": false,
        "n": 1305,
        "long": 0.2812,
        "short": 0.0751,
        "sideways": 0.6437,
        "long_short_ratio": 3.7443
      },
      {
        "model": "deepseek-v4-pro",
        "legacy": false,
        "n": 1305,
        "long": 0.1686,
        "short": 0.1203,
        "sideways": 0.7111,
        "long_short_ratio": 1.4015
      },
      {
        "model": "gemini-3.1-pro",
        "legacy": false,
        "n": 1302,
        "long": 0.3172,
        "short": 0.1659,
        "sideways": 0.5169,
        "long_short_ratio": 1.912
      },
      {
        "model": "gpt-5.6-sol",
        "legacy": false,
        "n": 1305,
        "long": 0.305,
        "short": 0.164,
        "sideways": 0.531,
        "long_short_ratio": 1.8598
      },
      {
        "model": "grok-4.5",
        "legacy": false,
        "n": 1300,
        "long": 0.2685,
        "short": 0.1985,
        "sideways": 0.5331,
        "long_short_ratio": 1.3526
      },
      {
        "model": "qwen-3.7-max",
        "legacy": true,
        "n": 470,
        "long": 0.2957,
        "short": 0.2638,
        "sideways": 0.4404,
        "long_short_ratio": 1.1209
      },
      {
        "model": "qwen-3.8-max",
        "legacy": false,
        "n": 829,
        "long": 0.3932,
        "short": 0.146,
        "sideways": 0.4608,
        "long_short_ratio": 2.6932
      }
    ],
    "long_short_ratio_field": {
      "value": 2.178,
      "method": "mean of each model's own long-share/short-share ratio (8 models, unweighted)"
    },
    "consensus_skew": {
      "cells_ge5_same_side": 471,
      "n_mature": 2901,
      "hit_rate": 0.4171
    },
    "fail_of_week": {
      "model": "qwen-3.7-max",
      "symbol": "BTC",
      "side": "long",
      "confidence": 85,
      "slot": "2026-08-04T19:01:00+00:00",
      "fh": "1h",
      "tf": "1h",
      "exit_reason": "expiry",
      "pnl_net_usd": -0.28,
      "cross_reference": "7-of-7 unanimous long, same slot -- see Consensus Watch #1"
    },
    "counters": {
      "forecasts_total": 9194,
      "ok_in_gate": 9121,
      "invalid": 3,
      "out_of_gate_1w": 70,
      "mature": 9121,
      "mature_directional": 4042,
      "mature_sideways": 5079,
      "pending_next_issue": 0,
      "late_closes": 0
    },
    "uptime": [
      {
        "fh": "1h",
        "tf": "1h",
        "slots_seen": 165,
        "slots_expected": 168
      },
      {
        "fh": "4h",
        "tf": "4h",
        "slots_seen": 41,
        "slots_expected": 42
      },
      {
        "fh": "4h",
        "tf": "1h",
        "slots_seen": 41,
        "slots_expected": 42
      },
      {
        "fh": "1d",
        "tf": "1d",
        "slots_seen": 7,
        "slots_expected": 7
      },
      {
        "fh": "1d",
        "tf": "4h",
        "slots_seen": 7,
        "slots_expected": 7
      }
    ],
    "ok_rate_by_model": {
      "qwen-3.8-max": 0.9988,
      "claude-fable-5": 1.0,
      "grok-4.5": 1.0,
      "claude-opus-5": 1.0,
      "deepseek-v4-pro": 1.0,
      "gemini-3.1-pro": 0.9985,
      "gpt-5.6-sol": 1.0,
      "qwen-3.7-max": 1.0
    },
    "calibration_bridge": {
      "best_calibrated_model": "qwen-3.8-max",
      "brier": 0.2713,
      "note": "full calibration analysis in Weekly Calibration #1"
    }
  },
  "key_findings": [
    "OBSERVATION -- No weekly title awarded. The top four models sit within ~1.9pp of hit-rate (qwen-3.8-max 44.1%, gemini-3.1-pro 43.7%, gpt-5.6-sol 43.0%, qwen-3.7-max 42.2%) with overlapping 95% CIs -- the title rule (>=5pp gap AND non-overlapping 95% CI) was not met. qwen-3.8-max leads descriptively.",
    "Ticker of the week: SOL; hardest: ETH. SOL was the field's most readable coin (48.2%, n=852) and ETH the hardest (34.8%, n=724) -- the widest gap in the table, and the one split where 95% CIs land clean of each other (44.9-51.6% vs. 31.4-38.4%); every adjacent pair in the ranking overlaps its neighbor. Pair of the week: qwen-3.8-max x SOL 52.9% (n=104) -- the same model also owns the worst pair (x ETH, 28.4%, n=88).",
    "Self-agreement added nothing this week. When a model's two timeframes (or neighboring horizons) agreed on direction, hit-rates ran 1.9-4.5pp BELOW the weekly base; the rare intra-model disagreements did far better (71.4% / 70.6%) but on just 7 and 17 calls -- descriptive only, not a strategy.",
    "A one-regime week. All 7 days were flat (max |BTC daily move| 0.98%, methodology flat/trend threshold 1.5%); the field skewed long over short (average of each model's own long:short ratio, ~2.2:1) in a sideways-heavy tape (55.7% of mature calls); its single most confident miss was a conf-85 BTC long (qwen-3.7-max, Aug 4)."
  ],
  "practical_implications": [
    "Do not read this week's leaderboard as a skill ranking. The top four models overlap in both point estimate (within ~1.9pp) and 95% CI, and the title rule was not met -- qwen-3.8-max's lead is descriptive, for this window only.",
    "Treat SOL as this week's most-readable ticker and ETH as the hardest with real caution: it is one week, and the field also missed its only reversal outright (0 of 8 models), which points to a general weak spot at turns rather than an ETH-specific one.",
    "Do not use within-model cross-TF/cross-FH agreement as a confidence booster yet. In every tested cell, agreement underperformed the base rate this week (by 1.9-4.5pp); the small number of disagreements did better but on too few calls (7, 17) to act on."
  ],
  "limitations": [
    "Single week (Aug 3-9, 2026 UTC). Every finding in this issue is descriptive for this window only; no claim is made about next week or about any model's underlying skill.",
    "This week sat inside a single market regime: all 7 days were classified flat (max |BTC daily move| 0.98%). The trend-vs-flat regime control called for by methodology (Block 5) could not be exercised this issue -- there were no trend days to compare against.",
    "Observations inside one window are not independent: the same models watch overlapping symbol/FH/TF cells hour after hour, so agree/disagree calls made close in time are correlated, not i.i.d. draws.",
    "95% Wilson CIs shown throughout are descriptive, not inferential, for the reason above -- read them as a range, not a formal coverage guarantee.",
    "The reversal detector (Block 4) is v1 and BTC/ETH-only. Its underlying price series is reconstructed from trade entry prices (median per symbol-slot), not an independent tick feed, per this week's metrics notes. One event this week is not enough to characterize detector performance either way.",
    "Cross-TF and cross-FH disagreement cells are tiny this week (n=0, 1, 7, 17) -- too small to rank models on; cells below N=10 are marked insufficient and are never used to rank models.",
    "This is issue #1 of Weekly Model Watch -- there is no prior issue to compare against, so no week-over-week deltas are reported (Block 1's 'movement vs prior week' begins with issue #2)."
  ],
  "testing_next": [
    "Next issue: does the reversal blind spot repeat, and does cross-TF agreement stay flat-to-negative?",
    "Monthly test: cross-TF/FH lifts with regime control across mixed weeks (trend vs. flat), block bootstrap."
  ],
  "related_research": [
    {
      "title": "Stability Index Run 2",
      "url": "marketmania.ai/research/reports/si-run-2.pdf"
    },
    {
      "title": "Weekly Calibration #1",
      "url": "marketmania.ai/research/reports/weekly-calibration-2026-08-03.pdf"
    },
    {
      "title": "Consensus Watch #1",
      "url": "marketmania.ai/research/reports/consensus-watch-2026-08-03.pdf"
    }
  ],
  "related_research_note": "These three weekly reports -- Consensus Watch, Weekly Calibration and Weekly Model Watch -- publish together as one wave each week; links are stable.",
  "research_to_date": {
    "this_report": {
      "scored_observations": 4042
    },
    "platform": {
      "as_of_cutoff": "2026-08-10",
      "resolved_forecasts": 14151,
      "resolved_forecasts_since": "2026-07-11",
      "models_tracked": 9,
      "models_tracked_detail": "7 current frontier + 2 archived legacy generations",
      "assets": 5,
      "forecast_horizons": 5,
      "cadence": "hourly",
      "published_research_reports": 4
    },
    "source": "platform_counters_2026-08-10"
  },
  "citation": {
    "bibtex_key": "mm_model_watch_2026w32",
    "title": "Weekly Model Watch #1: weekly leaderboard, ticker/pair reads and cross-confirmation, Aug 3-9 2026",
    "author": "MarketMania Research",
    "year": 2026,
    "month": "August",
    "day": 10,
    "url": "https://marketmania.ai/research"
  }
}