The cumulative redundancy bias (CRB) refers to people's difficulty in ignoring the redundancy in cumulatively presented information. When people consider which of two competing agents is better, they are influenced by the sequence of events that led to their accumulative total performance. If one agent was ahead most of the time, people consider this agent better – even if the agents are tied eventually. However, we show that an opposite performance-slope bias (PSB) emerges when participants focus on the performance trajectories of agents. When a trailing agent is substantially catching up to the leading agent, people judge the former as the better agent. In four experiments where we manipulated the magnitude of performance slope difference between two agents, we obtained both effects. In Experiments 1 and 2, a large slope difference between agents overshadowed other factors that may have influenced the reversal from CRB to PSB. In Experiments 3 and 4 we showed conclusively with the same design, but for different real-life contexts that the CRB emerges when the slope difference between agents is small while it reverses as the PSB when the slope difference is large. The four experiments demonstrate individuals' ability for flexible cue utilization.

累积冗余偏差（CRB）是指人们难以忽略累积呈现的信息中的冗余。当人们考虑两个竞争代理人中哪一个更好时，他们会受到导致其累积总绩效的事件顺序的影响。如果一个代理人大部分时间领先，人们就会认为这个代理人更好——即使代理人最终打成平手。然而，我们表明，当参与者关注代理人的绩效轨迹时，会出现相反的绩效斜率偏差 (PSB)。当落后的代理人基本上赶上领先的代理人时，人们会判断前者是更好的代理人。在我们操纵两个代理之间的性能斜率差异大小的四个实验中，我们获得了两种效果。在实验 1 和 2 中，代理之间的较大斜率差异掩盖了可能影响从 CRB 到 PSB 逆转的其他因素。在实验 3 和 4 中，我们最终展示了相同的设计，但对于不同的现实生活环境，当代理之间的斜率差异较小时，CRB 出现，而​​当斜率差异较大时，它反转为 PSB。这四个实验证明了个人灵活利用线索的能力。