Third reference bot 2026
This bot (written in python_monte_carlo) implements a strong-hole-card strategy for begging, similarly to reference-2-2026, and runs a Monte Carlo simulation to decide which card to toss. Running 100 games against the default skeleton bot got the following score distribution: 
As we can see, this bot destroys the skeleton. But interestingly enough, it performs slightly worse than tossing the lowest card (which is what reference-2-2026 does)! We leave it as an exercise for the reader to brainstorm about why this might be the case (hehe).