This article discusses controversies about the Analysis of Competing Hypotheses (ACH) efficacy. The technique was developed by the Central Intelligence Agency (CIA) to mitigate cognitive biases and improve critical thinking, becoming one of the most popular analytical tools in the intelligence community. Despite its pervasiveness, ACH has some limitations and does not perform well in experimental tests. The findings suggest that the technique is not an effective way of mitigating analysts' biases and does not necessarily improve their reasoning. This might happen because of multiple factors (e.g. problems in implementing the technique, theoretical flaws, and vagueness of its instructions). However, this is not necessarily a reason to abolish the use of ACH in intelligence and investigative activities. With that in mind, the paper suggests some practical improvements that might lead to better results. These suggestions must be submitted to further experimental testing, which is related to the other aim of the article: to encourage the use of randomized experiments to test analytical techniques in the intelligence and investigative contexts.