 This paper compares two methods for predicting crop yield responses to climate change, ordinary least squares linear regression, OLS, and boosted regression trees, BRTs. The authors found that BRTs provided a higher level of accuracy than OLS when predicting crop yield responses to climate change. Additionally, BRTs were able to detect breakpoints in the relationship between climate and crop yield, which OLS was unable to do. The authors tested their model predictions against real-world data from India and found that BRTs predicted a lower negative impact of climate change on crop yields than OLS. They concluded that caution should be taken when interpreting the results of single-model analyses, as these models may be overly sensitive to the region's climate and agricultural practices. This article was authored by Baal Shia Singh Sidhu, Zia Merabi, Navin Ramankatti, and others.