正在加载盘面数据…
竞彩决策台 三源融合 · 赔率分析与串关
⚠ 本工具仅供研究参考 · 彩票有风险 · 投注需理性

🧪 模型实验室

回测 · 训练 · 监控
📈 回测中心 🤖 模型训练 📡 监控信号

当前视图:监控信号。长任务完成后将在此处显示结果。

📡 监控信号

observation-only · 只读不下注
ℹ️ 说明
本面板集中展示未经样本外证据升级的观测型信号(串关闭环、凯利离散度、Dixon-Coles 比分回测、周度 binomial 一致性、市场边际地图)。这些信号仅供研究观测,不改变任何仓位或下注决策,与详情页决策核心物理隔离。
🎫 串关闭环追踪
只读 parlay_records
1428
累计串关(目标 60)
72
覆盖天数(目标 30)
1279/149
已结算 / 待结算
7.1%
命中率(已结算)

样本量已达统计门槛,命中率/ROI 可作参考(仍为小仓位娱乐口径)。

market/玩法条数已结算胜/负命中率盈亏ROI
beidan/double110/1 0.0% -30.00 -100.0%
beidan/goals69610/61 0.0% -488.00 -100.0%
beidan/hcap24823115/216 6.5% +10347.76 +559.9%
beidan/htft100/0 — +0.00 —
beidan/score69631/62 1.6% -205.44 -40.8%
beidan/single110/1 0.0% -20.00 -100.0%
beidan/spf37327730/247 10.8% +1764.80 +75.3%
jingcai/double220/2 0.0% -60.00 -100.0%
jingcai/goals65650/61 0.0% -488.00 -93.8%
jingcai/hcap23322519/203 8.4% -662.48 -36.8%
jingcai/htft331/2 33.3% +57.00 +38.0%
jingcai/score64640/58 0.0% -464.00 -90.6%
jingcai/single222/0 100.0% +12.60 +31.5%
jingcai/spf29728423/258 8.1% -472.50 -19.0%

半全场为模型折算口径(非官方兑付),与胜平负官方盘口分列。

📐 周度一致性监控(binomial)
实盘 vs 回测预期

窗口 7 天 · market=jingcai · 样本 326 行 · 状态: 一致

cohort判定
all✅ [all] 一致:实盘 0.5566 (95%CI 0.4893~0.6219) 覆盖回测预期 0.5223,n=212
standard✅ [standard] 样本不足 n=0<30,暂不判定(实盘 NA vs 预期 0.7221)

跳过:无方向 0 / 放弃 0 / 赛果未齐 114

🗺️ 市场边际地图
市场边际地图-20260829.json

全量样本外分层命中率/ROI,用于观测「哪些切片系统性有边际」。全集期望值为负属正常(庄家抽水),高 ROI 冷门切片样本极小不可外推。

47.7%
胜平负 命中率(n=24731, ROI -6.0%)
51.6%
让球 命中率(n=909, ROI -10.7%)
胜平负 · 按赔率段
切片n命中率ROI
>=5.0(冷门)425.0%+81.9%
3.0-5.010335.0%+23.6%
<2.0(大热)824564.0%-4.5%
2.0-3.01637939.6%-7.1%
胜平负 · 按置信段
切片n命中率ROI
0.40-0.50248846.3%-3.8%
>=0.60474170.2%-4.2%
0.50-0.60325656.1%-4.6%
<0.401424638.6%-7.8%
胜平负 · 按方向
切片n命中率ROI
负815546.7%-5.4%
平252234.3%-5.8%
胜1405450.7%-6.4%
胜平负 · 按联赛层级
切片n命中率ROI
五大联赛343948.0%-4.0%
其他2129247.7%-6.3%
胜平负 · 赔率段×联赛层级
切片n命中率ROI
3.0-5.0|五大450.0%+82.8%
>=5.0(冷门)|其他425.0%+81.9%
3.0-5.0|其他9934.3%+20.4%
2.0-3.0|五大230740.4%-3.8%
<2.0(大热)|其他711764.1%-4.5%
<2.0(大热)|五大112863.4%-4.7%
2.0-3.0|其他1407239.5%-7.7%

实盘追踪:已落地 55 场(已结算 55),总命中率 54.5%。样本极小,仅供对照观测。

🎲 凯利离散度(KD)
kelly_dispersion_synthetic_20260827.md · 生产默认关

Stage2 Kelly 离散度 focused synthetic report

This report is synthetic only. It is not a football-data historical backtest and makes no performance claim.

Safety contract

  • Input is structured bookmaker 1X2 odds only.
  • Avg/Max-style aggregates and malformed odds are excluded.
  • Only matching handicap=0 and one timestamp are eligible; the synthetic timestamp stands in for T−24 and is not a source publication timestamp.
  • At least three valid independent bookmakers are required.
  • *_std is population standard deviation (denominator n); both KD thresholds use kelly_max_std across directions (no average-std fallback).
  • KD-LOW emits a dynamic consensus variant only; KD-HIGH emits a risk label with no outcome.
  • Both KD rules are observation-only and forced to D; even a high synthetic statistic is never reported as recommended enablement.

Synthetic observations

fixturevalid companieseligibleprob overall stdKelly max stdoverall rangevariants
------::---:---:---:---:---
low-consensus3True0.00.00.0KD-LOW-v1=win
high-dispersion3True0.1330650.9407930.455717KD-HIGH-v1=high-dispersion
aggregate-and-malformed3True0.00.00.0KD-LOW-v1=win

Interpretation

  • This fixture confirms the safety gates and variant plumbing only.
  • No historical hit rate, EV, evidence grade upgrade, or production recommendation is inferred.
  • config.yaml keeps rules.kelly_dispersion.enabled: false; the current main decision remains unchanged.
  • The production registry currently receives no company_odds from the main ensemble path; KD requires an explicit structured panel and therefore remains observation-only.
📊 Dixon-Coles τ 比分回测(DC)
DC比分回测-20260828.md · 生产默认关

Dixon-Coles τ 回测报告(M2)

日期:2026-08-28 | ρ=-0.08 (config dc.rho) | ξ 时间衰减未接入 (M3)

一、概览

  • CSV: 112 个; 有效: 32641; 剔除(<3家): 6726
  • 面板分布: 3×4364, 4×5201, 5×4771, 6×18305

二、比分 top-5

方法场次命中命中率Δpp
-------------------------------
独立泊松326411658850.8%—
DC326411669751.2%++0.33
  • 目标 +≥1pp → ❌ 未达标

三、大小球 2.5

方法命中命中率Δpp
-------------------------
独立泊松1794355.0%—
DC1841356.4%++1.44

四、裁决

  • 保持 dc.enabled=false (比分提升 +0.33pp)