Matchup Simulator

The Simulation Engine

This tool does not simply roll dice or randomly pick a winner. It runs a highly advanced predictive analytics engine powered by a Bivariate Poisson Distribution, a statistical model used globally to predict the probability of a given number of events happening in a fixed interval of time.

1. Skater Impact Scores

Players are evaluated based on their positional responsibilities rather than raw goals. Forwards are measured by their total offensive production (Goals + Assists per Game). Defensemen are measured by a proprietary "Shutdown" metric that weighs Blocked Shots and Plus/Minus per Game, actively suppressing the opponent's Expected Goals.

2. Bayesian Smoothing & Missing Eras

To prevent small sample sizes from breaking the math (e.g., a player with 1 goal in 1 career game looking like a superstar), the engine applies "Bayesian Smoothing," adding a baseline of league-average games to balance the data. Furthermore, for historical players from the 1998-2006 era where data availability gaps exist for specific statistical categories (blocks, +/-, and shots), the engine seamlessly injects a mathematically neutral era-adjustment so legends of the past are not unfairly penalized.

3. Special Teams & Discipline

Penalty Minutes (PIMs) are dynamically calculated. Teams that take more penalties actively grant their opponents more Power Play opportunities in the simulation, increasing the likelihood of PPGs (Power Play Goals) or SHGs (Short-Handed Goals) being assigned to individual players in the box score.

4. The Final Matrix

The engine combines the Home Team's Offensive Impact, the Away Team's Defensive Shutdown metrics, Goaltending Save Percentages, and Discipline modifiers to calculate an exact "Expected Goals" baseline for both teams. It then simulates all three periods and overtime, generating goal scorers and assists based on individual player weights.