The current publication by Chen and colleagues 1 presents a model aimed at predicting overall survival and disease-free survival in patients with gastric cancer. They argue that the current staging system, which uses routinely collected data, is limited in its ability to predict outcomes and, importantly, to predict potential benefits of chemotherapy. They focus on building a tumor-associated collagen signature (TACS) using data from the tumor microenvironment. Collagen is a major component of that tumor microenvironment, which can influence both cancer progression and response to chemotherapy.The authors used a sophisticated statistical modeling approach that allows consideration of large numbers of potential predictor variables, with limited overfitting of the models. (When models are overfit to a specific data set, they can be excellent for prediction in the current sample but not nearly as good at prediction in other populations.