Stochastic simulation of March Madness

• Layne Bradshaw
• "Monte Carlo" stochastic simulation of March Madness 2016 tournament
• How often does every team win? How often will the n-seeded win?

Approach

• gather stats (win/loss, historical, seeding, ....)
• simulate matches and decide who wins through random numbers
• simulate whole tournament repeatedly

Objectives

1. gather stats
2. simulate individual games
3. run through all 64 games
4. run production (many times)
5. compare to real results

Q&A

1. How accurate can you be? What would count as success? Predicting the winning team on average.