We can use regression analysis to inform our predictions. Regression analysis is a data modeling technique that helps us understand how different variables change with one another. A regression model must incorporate at least one dependent (or outcome) and independent (or predictor) variable, although several of both can be included. We can use regression models to predict outcomes based on the data that is available to us. As the Shu strategist, you will be predicting your army's performance in battle across different courses of action, based on what you know about past conflicts.
Note
Note that the regression models we create will predict the Shu army's future performance rating based on several conditions of battle. Recall that the past battle ratings were recorded by Zhuge Liang, who was an expert in assessing the army's combat performance. Therefore, Zhuge Liang's ratings can be considered a valid measure for predicting future performance, since they are equivalent to the actual performance of the Shu forces.
While it is technically possible to directly predict the result of battle (victory or defeat) from our dataset, this would require the use of advanced regression techniques that are beyond the scope of this book. Thus, we will focus on predicting outcomes based on performance ratings via the most common regression methods.