This study drew on data mining from Powerlifting Championships recognized by the International Powerlifting Federation and compiled by the Openpowerlifting project. Analysis was made of the records of the best total score for each athlete by year from 2012 to 2021, evaluating the squat, bench press and deadlift disciplines. The annual progress (AP) of the athletes was then qualitatively codified using the difference between the results of consecutive years for each discipline and total. A positive difference was codified as 1 and a negative or equal result as 0, so obtaining 6 categories. AP was thus calculated for 32,539 men’s and 15,700 women’s samples from Classic powerlifting, and 14,181 men’s and 6,518 women’s samples from Equipped powerlifting. In order to evaluate these APs, decision tree analysis was generated using the CHAID growing method (Chi2 p<0.05), with AP as the dependent variable and Event, Age and Gender as factors. The most robust category was “Improvement of all disciplines and total” (IA), representing 46.1% of the global results. Age was the most relevant classification factor on the tree, with the youngest groups presenting the highest percentages of IA. Event was the next most significant factor, with Classic powerlifters showing higher percentages of IA than Equipped ones. Likewise, IA results within most weight classes were higher for Classic events than Equipped ones. The third important factor was Gender, with women attaining higher IAs than men, except for the 14–18-year-old group. Notably, the “No Improvements” category reflected a considerable percentage (11.7% of global results). The study suggests a sequence of annual outcomes in order of likelihood as follows: 1-Improvement of all disciplines and total, 2-Improvement of two disciplines and total, 3-No improvements, 4-Improvement of one discipline but not total, 5-Improvement of one discipline and total, and 6-Improvement of two disciplines but not total.