diff --git a/README.md b/README.md index 85ac05d47bf859adc571f2ff6fb977a4b9a4a76c..97fbcd9181c37baaaf3d4342aa9285629c5ef8fc 100644 --- a/README.md +++ b/README.md @@ -1,15 +1,11 @@ # Predicting Ice Hockey Goals Using Random Forest and XGBoost ## Intro & Background - -This project explores the prediction of goals in professional ice hockey games using machine learning techniques. -The main goal is to determine which features most effectively predict goal-scoring events, with a particular focus on shot attempts. -Predicting goals is valuable for teams, analysts, and broadcasters to better understand game dynamics. -While traditional stats like shot counts have been used, machine learning offers deeper, data-driven insights. - -Ice hockey is a fast-paced sport with frequent possession changes and rapid movement, making goal prediction challenging. -The dataset contains in-game event data including player positions, shot locations, man-power situations, and contextual game details. -Understanding basic ice hockey rules and terms like "manpower situation" or "xG" (expected goals) will help interpret the models. +This project explores predicting goals in professional ice hockey using machine learning, focusing on identifying features—especially shot +attempts—that most effectively forecast goal-scoring events. Accurate goal prediction is valuable for teams, analysts, and broadcasters aiming +to better understand game dynamics. While traditional stats like shot counts provide some insight, machine learning enables deeper, data-driven analysis. +The fast-paced nature of ice hockey, with constant possession changes and rapid movement, makes prediction complex. The dataset includes in-game events such as player positions, +shot locations, manpower situations, and other contextual game details. Familiarity with hockey concepts like "manpower situation" and "xG" (expected goals) is helpful for interpreting model outcomes. ## Algorithms