Launch-Day Diffusion: Tracking Hacker News Impact on GitHub Stars for AI Tools

Preprint: https://doi.org/10.48550/arXiv.2511.04453
© 2025 Obada Kraishan. Text/figures: CC BY 4.0. Code: MIT.

Author: Obada Kraishan, Texas Tech University

🎯 Key Findings

Our analysis of 138 repository launches (2024-2025) reveals:

• Immediate Impact: Repositories gain an average of 121 stars within 24 hours, 189 stars within 48 hours, and 289 stars within a week of HN exposure
• Timing Matters: The difference between optimal and suboptimal posting hours is ~200 stars
• Best Window: 12-17 UTC consistently outperforms other time slots
• Show HN Paradox: The "Show HN" tag shows no statistical advantage after controlling for other factors

🚀 Quick Start

Prerequisites

python >= 3.8
pip install -r requirements.txt

Installation

git clone https://github.com/obadaKraishan/icwsm-hn-github.git
cd icwsm-hn-github
pip install -r requirements.txt

Environment Setup

cp .env.example .env
# Add your GitHub token (optional but recommended for higher rate limits)
echo "GH_TOKEN=your_github_token_here" >> .env

Run Complete Pipeline

# Run all scripts in sequence (takes ~5 minutes)
make demo

# Or run individual components
python src/01_collect_hn_posts.py --start 2024-01-01 --end 2025-01-01
python src/02_extract_github_repos.py
# ... etc

📊 Pipeline Overview

The system consists of 10 modular scripts that can be run independently or as a complete pipeline:

Script	Purpose	Output
01_collect_hn_posts.py	Fetch HN posts linking to GitHub	hn_posts.csv
02_extract_github_repos.py	Parse GitHub owner/repo from URLs	github_repos_from_hn.csv
03_github_repo_metadata.py	Fetch repository metadata	github_repos_metadata.csv
04_github_stars_timeseries.py	Get time-stamped star events	stars_timeseries.csv
05_build_event_windows.py	Align timeseries to HN post time	event_windows.csv
06_feature_engineering.py	Create modeling features/labels	features_labels.csv
07_event_study_plots.py	Generate event study curves	Event study figures
08_model_star_growth.py	Train predictive models	Model predictions & metrics
09_ablation_checks.py	Run robustness checks	Ablation estimates
10_make_report_txt.py	Generate summary report	REPORT.txt

📁 Project Structure

Launch-Day-Diffusion/
├── src/                    # Pipeline scripts
│   ├── figures/           # Generated plots
│   ├── raw/               # Raw API responses (JSONL)
│   ├── processed/         # Clean CSV files
│   └── summaries/         # Text summaries
├── requirements.txt       # Python dependencies
├── config.yaml           # Optional configuration
└── Makefile              # Automation commands

🔬 Methodology

Event Study Design

• Window: ±7 days around HN post time (t=0)
• Alignment: Hourly star counts aggregated to daily totals
• Labels: Δ24h, Δ48h, Δ168h star gains

Models

• Elastic Net: Interpretable linear relationships with L1/L2 regularization
• Gradient Boosting: Captures non-linear patterns and interactions
• Validation: 80/20 train-test split with 5-fold cross-validation

Statistical Tests

• OLS regression with heteroscedasticity-robust standard errors (HC1)
• Controls for baseline repository characteristics
• Multiple model specifications for robustness

📈 Sample Results

Event Study Curves

Model Performance

Model	Horizon	MAE	RMSE	R²
Gradient Boosting	48h	30.5	60.1	0.77*
Gradient Boosting	7d	92.5	182.0	0.48

*Includes day-0 momentum features

🔧 Configuration Options

Edit config.yaml or use command-line arguments:

# config.yaml
start_date: "2024-01-01"
end_date: "2025-01-01"
min_score: 10
query: "llm,gpt,rag,transformers,langchain"

📚 Cite

If you use this work, please cite the preprint:

APA Kraishan, O. (2025). Launch-Day Diffusion: Tracking Hacker News Impact on GitHub Stars for AI Tools. arXiv preprint arXiv:2511.04453. https://arxiv.org/abs/2511.04453

BibTeX @misc{kraishan2025launchday, title = {Launch-Day Diffusion: Tracking Hacker News Impact on GitHub Stars for AI Tools}, author = {Kraishan, Obada}, year = {2025}, eprint = {2511.04453}, archivePrefix = {arXiv}, primaryClass = {cs.SI}, url = {https://arxiv.org/abs/2511.04453}, note = {arXiv preprint} }

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request. Areas for extension:

• Additional social platforms (Reddit, Twitter/X)
• Different software ecosystems (npm, PyPI)
• Enhanced feature engineering
• Real-time monitoring capabilities

📄 License

• Code: MIT (see LICENSE).
• Text, figures, and the preprint PDF: Creative Commons CC BY 4.0.

🙏 Acknowledgments

• Hacker News Algolia API for search functionality
• GitHub REST API for repository data
• Texas Tech University College of Media and Communication

⚠️ Ethical Considerations

• All data collected via public APIs in compliance with terms of service
• No private or personal information collected
• Rate limiting implemented to respect API quotas
• Results represent associations, not causal effects

📧 Contact

Obada Kraishan - [email protected]
ORCID: 0009-0007-7180-8620
Website: https://okraishan.com/

---