Update README.md
Browse files
README.md
CHANGED
|
@@ -109,3 +109,19 @@ x = {
|
|
| 109 |
pred = model.predict(x, verbose=0)["dwp"][0] # ΔWP for [go, FG, punt]
|
| 110 |
actions = ["Go for it", "Field goal", "Punt"]
|
| 111 |
print(dict(zip(actions, pred)))
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 109 |
pred = model.predict(x, verbose=0)["dwp"][0] # ΔWP for [go, FG, punt]
|
| 110 |
actions = ["Go for it", "Field goal", "Punt"]
|
| 111 |
print(dict(zip(actions, pred)))
|
| 112 |
+
|
| 113 |
+
## Intended use & limitations
|
| 114 |
+
|
| 115 |
+
Purpose: Educational decision assistant for 4th‑down scenarios.
|
| 116 |
+
|
| 117 |
+
Not for wagering or real‑time coaching; relies on historical
|
| 118 |
+
probabilities and inherits assumptions from nfl4th.
|
| 119 |
+
|
| 120 |
+
Weather and wp_punt imputation introduce simplifications; apply domain
|
| 121 |
+
judgment when using recommendations.
|
| 122 |
+
|
| 123 |
+
## Citation
|
| 124 |
+
|
| 125 |
+
If you use this model, please cite the original project:
|
| 126 |
+
|
| 127 |
+
Villagran, M. (2025). NFL 4th-Down WinPct Recommender. MIT License.
|