Predicting paradigm shifts is important because often, the old paradigms and work and technology associated with them are unable to coexist simultaneously with the new ones. This can lead to a loss of time, opportunity, and money of those invested in the old viewpoints. If one could predict a shift in his field, then this new paradigm could be adopted quicker and losses will be subverted.
A method to identify pivotal papers of a paradigm shift within a field was published in 2015 by Lathabai et al. This method was termed Flow Vergence Gradient or FV Gradient. Thara Prabhakaran et al. attempted to extend this model from identification of paradigm shifts to prediction of paradigm shifts.
Other methods to detect pivotal scientific research papers exist such as the Anna Karennia Principle (AKP) and external literature-based verification, but these two suffer from time delay and bias as the former is based on the feedback of experts and the latter requires the paradigm in question to have been well developed. Another potential predictive measure is an estimation of ultimate citation impact or total citations ever possible, which was introduced in 2013 by Wang et al.
Prabhakaran, T., Lathabai, H.H., George, S. et al. Towards prediction of paradigm shifts from scientific literature. Scientometrics 117, 1611–1644 (2018). https://doi.org/10.1007/s11192-018-2931-3