研究团队开发了一款互动割草游戏，吸引超过3万人参与实验，以探索人类在路径规划问题上的认知能力差异。[1]参与者中，52%的人完成的路线在最优路径的5步以内，16%达到了完美方案，中位效率达到91%。[1]研究涉及的难度跨度很大——最简单的关卡为49个方格，存在14,589种不同解法，而最难的关卡接近200个方格，共有8,912人完成了全部6个阶段的挑战。[1]
性能差异并非源于投入的思考时间长短。[1]研究表明，思考时间长度仅能解释不到5%的性能差异，这意味着"想得久"并非取得更好成绩的决定因素。[1]真正区分优秀割草者和表现欠佳者的是决策策略：最优割草者在面临关键决策点时会停顿思考，随后在开放区域快速移动以完成任务；相比之下，表现较差的参与者往往匆匆开始，只有被困时才做出反应。[1]一位受访者Sarah描述了她的分解策略："我一块块处理……尽量把完成的部分留在身后。"[1]
Researchers conducted a large-scale study using an interactive lawn-mowing game to understand how humans approach path-planning problems. More than 30,000 participants engaged in the experiment, with 52 percent completing routes within five steps of the optimal path and 16 percent achieving perfect solutions [1]. The median efficiency across participants reached 91 percent, demonstrating that human performance remained robust even as problem complexity increased [1].
The study revealed that superior lawn-mowing performance depends less on the duration of deliberation than on strategic decision-making. Thinking time explained fewer than 5 percent of performance differences, with an R² value of 4.9 percent [1]. Instead, optimal performers distinguished themselves by pausing to think at critical decision points while moving quickly through open areas, whereas lower-performing participants rushed into decisions and only responded when encountering obstacles [1]. One participant, Sarah, described her approach as processing the lawn "in chunks" and trying to "keep the completed section behind me," illustrating how decomposition strategies improved outcomes [1]. Across the experiment's six levels, 8,912 participants completed all stages, including a final level with nearly 200 squares on a 14-by-14 grid [1].