Deadlock Winner Prediction

Overview

The goal of this project is to predict the winning team in the new MOBA game Deadlock by Valve, based solely on the teams picks.

Installation

1. Clone the repository:

git clone https://github.com/1adore1/deadlock-winner-prediction.git
cd deadlock-winner-prediction

2. Install required libraries:

pip install -r requirements.txt

Usage

1. Preprocess the dataset by running (takes about 1 minute for 10k matches):

python preprocess.py

2. Train the model with:

python train_model.py

3. Make predictions on your input picks:

python predict.py

Models Used

• Random Forest
• Gradient Boosting
• Logistic Regression

Data

The dataset includes match history from Deadlock, that was parsed from tracklock.gg.

Dataset contains:

• Team picks: Heroes chosen by each team.
• Game results: Outcome of the match (win/loss).

---