Upload folder using huggingface_hub

README.md

@@ -7,7 +7,6 @@ tags:
   - flesh-and-blood
   - game-balance
   - educational
-  - sagemaker
 library_name: sklearn
 pipeline_tag: tabular-regression
 datasets:
@@ -18,25 +17,26 @@ metrics:
   - mae
 ---
 
-# Teklovossen Card Balance Predictor
+# FAB Prediction Model - Teklovossen Card Balance
 
-> **Educational Project — NOT affiliated with Legend Story Studios or the official Flesh and Blood game.**
+> **Educational Project - NOT affiliated with Legend Story Studios or the official Flesh and Blood game.**
 
-A scikit-learn RandomForest regression model that predicts **win rates** for custom
-Teklovossen (Mechanologist) card decks in the [Flesh and Blood](https://fabtcg.com/) TCG.
+A scikit-learn model that predicts **win rates** for custom Teklovossen (Mechanologist)
+card decks in the [Flesh and Blood](https://fabtcg.com/) trading card game.
 
-Trained on local using simulated gameplay data.
+This project explores how machine learning can be applied to game balance analysis
+using simulated gameplay data from a custom game engine.
 
 ## Model
 
 | Detail | Value |
 |--------|-------|
-| Type | `Pipeline(StandardScaler → RandomForestRegressor)` |
-| Task | Tabular regression — predict avg win rate |
+| Type | `Pipeline(StandardScaler + RandomForestRegressor)` |
+| Task | Tabular regression - predict average win rate |
 | Framework | scikit-learn 1.8.0 |
-| Training | local (local machine) |
+| Training data | Simulated gameplay (custom engine) |
 
-## Features (19 inputs)
+## Features (inputs)
 
 | # | Feature | Description |
 |---|---------|-------------|
@@ -44,7 +44,7 @@ Trained on local using simulated gameplay data.
 | 1 | Nano | Nano-themed card count |
 | 2 | Quantum | Quantum-themed card count |
 | 3 | Base | Base card count |
-| 4-8 | 0_cost … 4_plus_cost | Cost curve distribution |
+| 4-8 | 0_cost to 4_plus_cost | Cost curve distribution |
 | 9-11 | equipment, item, action | Card type counts |
 | 12 | avg_turns | Average game length |
 | 13 | avg_evos | Average Evo equipment used |
@@ -56,16 +56,16 @@ Trained on local using simulated gameplay data.
 
 ## Output
 
-`avg_win_rate` — float [0, 1].
+`avg_win_rate` - float in [0, 1] representing predicted deck win rate.
 
 ## Metrics
 
 | Metric | Value |
 |--------|-------|
-| Train R² | 0.7666 |
-| CV R² (mean ± std) | 0.1381 ± 0.2349 |
-| Train MSE | 0.0775 |
-| Train MAE | 0.163 |
+| Train R2 | 0.917 |
+| CV R2 (mean +/- std) | -1.3332 +/- 2.1072 |
+| Train MSE | 0.0049 |
+| Train MAE | 0.042 |
 | Samples | 9 |
 | Features | 17 |
 
@@ -80,15 +80,22 @@ model = joblib.load(path)
 
 # Example: a Nano-focused deck
 features = np.array([[9,0,0,2, 5,2,1,0, 9,9,12, 0.6, 9.4,13.6,7.3, 0.74, 6.7, 0,0]])
-print(model.predict(features))  # → predicted win rate
+print(model.predict(features))  # predicted win rate
 ```
 
+## About This Project
+
+This is a fan-made educational project exploring ML for TCG game balance.
+The model is trained on simulated gameplay data from a custom Flesh and Blood
+game engine featuring Teklovossen, a Mechanologist hero with custom Evo cards
+across four tech themes: AI, Nano, Quantum, and Biomancy.
+
 ## Limitations
 
 - Trained on **simulated** data, not real tournament results
-- Small dataset (9 samples) — rough estimates only
+- Small dataset (9 samples) - treat predictions as rough estimates
 - Simplified card mechanics in the simulation engine
-- Fan-made cards only
+- Fan-made cards only - does not cover the official card pool
 
 ## License
 

config.json

@@ -22,13 +22,13 @@
     "avg_quantum_charges"
   ],
   "sklearn_version": "1.8.0",
-  "trained_at": "2026-03-31T11:14:11.174996",
+  "trained_at": "2026-03-31T11:29:18.400928",
   "metrics": {
-    "train_r2": 0.7666,
-    "train_mse": 0.0775,
-    "train_mae": 0.163,
-    "cv_r2_mean": 0.1381,
-    "cv_r2_std": 0.2349,
+    "train_r2": 0.917,
+    "train_mse": 0.0049,
+    "train_mae": 0.042,
+    "cv_r2_mean": -1.3332,
+    "cv_r2_std": 2.1072,
     "n_samples": 9,
     "n_features": 17,
     "training_mode": "local"

model.joblib

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:56cacbb15ec89759267022bc2e609577274ed94b41c2bef967e09a54f08a0251
-size 13306
+oid sha256:1d9b4c7b57ee1ca549288182b3481698c5044ed55d5a7163bd83f409ae01e6d4
+size 16618