arXiv cs.CY· David Almog, Romain Gauriot, Lionel Page, Daniel Martin·· 11 小时前AI 评分31
算法复核如何改变人类判断：来自职业网球鹰眼系统的证据
Human Responses to Algorithmic Correction: Evidence from Centre Court
AI 导读
基于职业网球 Hawk-Eye 记录的研究发现，在线内 20 mm 范围内，误判“in”的比例上升 8.6 个百分点，误判“out”的比例下降 4.1 个百分点，且同一场比赛中判罚会向避免此前被推翻的错误偏移。研究未发现在线内 100 mm 范围内初始准确率下降的证据，并提出一个将人类反应与复核收益、补救成本一并纳入的投资框架，指出固定人类决策会误估复核的组织回报。
正文 · 原文
Abstract:We study how algorithmic review changes error composition using Hawk-Eye records from professional tennis. Within 20 mm of the line, mistaken "in" call rates increase by 8.6 percentage points, while mistaken "out" call rates decrease by 4.1 points. Within matches, calls shift toward avoiding previously overturned errors. We find no evidence that initial accuracy deteriorates within 100 mm of the line, despite concerns that workers may rely on correction as a safety net. We derive an investment framework incorporating human responses alongside correction benefits and remedy costs, showing that holding human decisions fixed can misstate the organizational returns to review.
| Subjects: | Machine Learning (cs.LG); Computers and Society (cs.CY); General Economics (econ.GN) |
| Cite as: | arXiv:2401.16754 [cs.LG] |
| (or arXiv:2401.16754v5 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2401.16754 arXiv-issued DOI via DataCite |
Submission history
From: Daniel Martin [view email]
[v1]
Tue, 30 Jan 2024 05:22:45 UTC (3,390 KB)
[v2]
Mon, 19 Feb 2024 04:38:06 UTC (3,292 KB)
[v3]
Wed, 12 Feb 2025 23:28:07 UTC (3,614 KB)
[v4]
Sun, 18 Oct 2026 04:45:51 UTC (1,510 KB)
[v5]
Wed, 4 Nov 2026 21:24:08 UTC (1,520 KB)
来源：arXiv cs.CY · arxiv.org