我们说这场球多，历史上到底打出来多少次
模型给每场比赛一个进球分位——在它自己的联赛里，这场预计算多还是算少。这张表把已完赛比赛按联赛和分位摊开，只回答一件事：这么说的时候，球真的往那边走了几成。
- 样本
- 45,851 场
- 联赛
- 46 个
- 口径
- 外样本 逐月滚动重训
- 每格下限
- 50 场，不够就不给数
球更少球更多全库基准 50.1%
| 联赛 | 场次 | 场均 | 整体 ↓ | P0–10 | P10–25 | P25–50 | P50–75 | P75–90 | P90+ |
|---|---|---|---|---|---|---|---|---|---|
| 全库 | 45,851 | 2.67 | 50.1% | 43.8% | 46.0% | 47.9% | 51.5% | 53.6% | 59.0% |
 荷乙 | 351 | 3.29 | 65.8% | (37) | 68.0% | 67.0% | 63.6% | 61.5% | (36) |
 冰岛超 | 229 | 3.69 | 64.6% | (22) | (35) | 66.7% | 63.2% | (35) | (23) |
 德甲 | 1,391 | 3.19 | 61.6% | 52.5% | 51.9% | 59.8% | 63.5% | 69.7% | 72.9% |
 挪超 | 1,112 | 3.10 | 60.8% | 48.6% | 53.0% | 62.2% | 62.6% | 64.1% | 71.4% |
 澳超 | 765 | 3.11 | 60.4% | 61.8% | 55.6% | 58.1% | 63.9% | 60.9% | 62.3% |
 中超 | 398 | 3.17 | 60.3% | (39) | 45.0% | 57.6% | 68.0% | 66.7% | (40) |
 荷甲 | 1,417 | 3.10 | 60.1% | 50.3% | 57.3% | 55.6% | 59.0% | 67.6% | 76.8% |
 美职联 | 2,329 | 3.00 | 59.0% | 56.0% | 55.1% | 56.2% | 55.7% | 66.9% | 70.8% |
 德乙 | 1,401 | 3.01 | 58.9% | 55.0% | 56.7% | 58.6% | 58.6% | 60.0% | 66.0% |
 芬超 | 268 | 2.97 | 58.2% | (26) | (40) | 55.2% | 58.2% | (41) | (27) |
 英超 | 1,747 | 2.94 | 56.8% | 47.7% | 48.5% | 54.5% | 60.9% | 63.0% | 64.6% |
 沙特联 | 1,211 | 2.89 | 55.1% | 47.9% | 50.3% | 51.2% | 58.1% | 59.7% | 64.8% |
 土超 | 1,454 | 2.84 | 54.1% | 44.1% | 54.1% | 51.5% | 53.8% | 52.8% | 73.3% |
 法甲 | 1,519 | 2.82 | 54.0% | 47.0% | 52.2% | 47.4% | 56.8% | 56.1% | 69.7% |
 苏超 | 211 | 2.75 | 53.5% | (21) | (31) | 49.1% | 50.9% | (31) | (22) |
 瑞典超 | 1,119 | 2.82 | 53.5% | 39.6% | 53.0% | 48.6% | 53.9% | 59.5% | 70.5% |
 比甲 | 1,338 | 2.81 | 53.4% | 47.4% | 45.8% | 50.3% | 54.0% | 59.7% | 67.2% |
 墨超 | 236 | 2.88 | 53.0% | (23) | (35) | 49.1% | 49.1% | (36) | (24) |
 波兰甲 | 306 | 2.73 | 52.9% | (33) | (43) | 55.3% | 57.1% | (46) | (31) |
 智甲 | 1,135 | 2.70 | 52.0% | 46.9% | 49.4% | 51.8% | 51.8% | 54.7% | 57.9% |
 希腊超 | 208 | 2.63 | 51.4% | (19) | (32) | 40.4% | 53.8% | (32) | (21) |
 俄超 | 1,017 | 2.72 | 51.2% | 33.7% | 47.7% | 49.2% | 60.6% | 51.0% | 55.9% |
 葡超 | 1,431 | 2.64 | 50.7% | 42.0% | 39.7% | 43.9% | 55.0% | 60.3% | 67.4% |
 爱超 | 285 | 2.60 | 48.8% | (28) | (43) | 50.7% | 49.3% | (43) | (29) |
 英冠 | 2,533 | 2.55 | 48.1% | 42.7% | 45.3% | 45.8% | 49.8% | 50.8% | 55.5% |
 韩K2联 | 374 | 2.58 | 47.9% | (37) | 26.6% | 50.6% | 45.7% | 51.8% | (38) |
 意甲 | 1,739 | 2.56 | 47.8% | 42.8% | 39.5% | 48.7% | 47.1% | 53.3% | 56.3% |
 西甲 | 1,756 | 2.61 | 47.8% | 36.6% | 35.4% | 42.4% | 52.8% | 54.5% | 68.2% |
 日职乙 | 762 | 2.52 | 47.6% | 41.3% | 43.0% | 51.8% | 46.1% | 47.4% | 54.5% |
 日职甲 | 1,622 | 2.56 | 47.2% | 40.1% | 42.4% | 43.7% | 52.5% | 48.6% | 55.2% |
 巴甲 | 1,769 | 2.48 | 46.1% | 38.6% | 44.4% | 46.8% | 48.6% | 46.6% | 47.5% |
 韩K联 | 349 | 2.51 | 46.1% | (35) | 57.7% | 33.3% | 42.5% | 66.0% | (35) |
 意乙 | 1,699 | 2.47 | 46.0% | 50.9% | 45.9% | 44.7% | 44.0% | 47.1% | 48.2% |
 法乙 | 1,564 | 2.48 | 45.6% | 35.9% | 47.0% | 42.5% | 45.5% | 48.1% | 58.0% |
 巴拉圭甲 | 206 | 2.54 | 45.1% | (19) | (23) | 47.5% | 48.1% | (31) | (22) |
| 乌拉圭甲 | 320 | 2.34 | 43.8% | (26) | 52.7% | 42.9% | 46.3% | (49) | (33) |
 西乙 | 2,113 | 2.37 | 43.2% | 34.1% | 40.4% | 38.6% | 45.5% | 46.4% | 57.6% |
 厄瓜甲 | 417 | 2.38 | 42.0% | (41) | 39.7% | 49.0% | 42.3% | 33.3% | (42) |
 巴乙 | 1,778 | 2.15 | 37.3% | 28.2% | 38.2% | 38.3% | 38.2% | 37.8% | 39.3% |
 埃超 | 275 | 2.07 | 34.5% | (27) | (41) | 44.9% | 42.0% | (41) | (28) |
 阿甲 | 1,994 | 2.06 | 33.9% | 30.1% | 30.1% | 31.5% | 36.5% | 35.8% | 40.0% |
 南非超 | 235 | 2.11 | 33.6% | (23) | (35) | 32.2% | 23.7% | (31) | (28) |
 阿乙 | 1,003 | 1.88 | 30.7% | 30.0% | 26.7% | 25.1% | 36.3% | 37.3% | 27.7% |
「小 2.5 赢」和「小 2.0 不输」是同一件事——都是总进球 ≤ 2，所以只列一次。走盘（正好 2 球）在小 2.0 上退本金，算不输但不算赢；「小 2.0 赢」那一列才是总进球 ≤ 1。
括号里的数字是场次——不足 50 场的格子不给兑现率。一个 14 场算出来的 43% 和一个 5,461 场算出来的 56%，在表格里长得一模一样，而前者的误差是后者的二十倍。
怎么读
横着读一行，看的是「在这个联赛里，我们把比赛分档之后，球有没有真的按那个顺序走」。德甲从 P0–10 的 52.5% 一路走到 P90+ 的 72.9%，那一行是最干净的例子。
竖着读一列，看的是「同样是我们说球多，在哪个联赛更兑现」。同样落在 P90+，荷甲打出 76.8%，阿乙只有 27.7%——联赛的差距比分档的差距大得多。
有几行是平的，比如巴乙和阿乙——分档几乎不动。那说明在那些联赛，我们区分不出哪场球多，这也照实写在这里。
这张表不是什么
不含赔率。它只说历史上打出来多少次，不说这个价值不值得买。庄家也知道哪场球多，那部分早就定在价里了。
不是保证。每一格都是历史频率，不是下一场的答案。场次越少的格子，抖得越厉害。
不是拿自己的答案打分。分位用的是逐月滚动重训的模型——每场比赛的预测，都只用它开赛之前的数据算出来，它自己没有参与训练。
数据截至 2026-09-08。大 2.0 走盘（正好 2 球）计入「不输」。
AI 预测基于历史数据与模型推断，仅作为参考，不构成任何形式的投注建议。足球比赛存在不确定性，历史表现不代表未来结果，请自行控制风险。