Tactician expert system (instant, hard) Neural policy-gradient players (instant, trained by final VP) Boosted policy-gradient players (instant, trained by final VP) PPO actor-critic players (instant, clipped updates from final VP) AlphaZero policy/value player (slow, full-game PUCT at every decision) Factorization-machine players (instant, learns feature interactions) Deck-specialist evolved players (instant, one historical model per deck) Evolved parameter players (instant, trained by final victory points) Brute-force checkpoint search (slow, 3 card plays ahead) XGBoost ranker with Tactician fallback (instant, experimental) Random forest with Tactician fallback (instant, experimental) Instinct bots (instant, easier) Ollama LLM seats (slow, battery-heavy)