Baseball Mogul 18 Preview: Historical Sims

Previous versions of Baseball Mogul had a problem with extreme outliers. Babe Ruth is the best example. Baseball Mogul’s rating system assumes that player performance almost always falls within about 4 standard deviations of the league average: this should cover 99.994% of all players.
​
Unfortunately, Babe Ruth hit home runs at a rate that was 4-6 standard deviations above the rest of the league. In a simulation that produces a realistic distribution of outcomes, the chance of a result that is 6 deviations above average is about 1 in a billion (1,013,594,692 to be exact).

---

Babe Ruth might able to hit 60 homers in 1987, when the league average was quite high, but he’ll never hit 60 homers in 1921 – even though this is what he did in real life. This means that if you simmed 1927 in Baseball Mogul 2017, Babe Ruth only hit 20-30 home runs.

Baseball Mogul 2018 fixes this. It’s complicated, but I re-wrote the rating calculation code to compress the distribution of player ratings in seasons when statistical output (e.g. home runs) is below the historical average. Then I adjusted the sim to correctly handle these new values.

Actual 1927 AL Home Run Leaders

1927 Simmed in Baseball Mogul 2017

1927 Simmed in Baseball Mogul 2018