Predicting with Logistic Regression

This visualization shows how different features predict the outcome of a football game.

---

---

---

---

---

---

---

---

---

---

---

---

---

___________________________________________________

On the left (top) side, there is a probability map which reflects whether the home or guest team will win the game. The more light-brown a part of the map is, the higher the likelihood that home team will win. The more teal the likelier the guest team will win.

Using the slider bar, you can select which football feature is displayed on the map. For example, if you selected the feature - total yards - than as you move towards the right of the map, total yards for the home team will increase. Similarly, as you move up the map, total yards for the guest team increases.

The plot on the right (bottom) side shows the probability of a home team victory for each feature. As you move the cursor over the probability maps these values will move up and down reflecting changes in the probability of the home team winning. Below the plot you can see the difference in estimate between the home and guest team for each feature. For example, under the total yards icon, a positive 200 means the home team earned 200 more yards than guest team. A negative value indicates that the guest team earned more yards than the home team.