Baseball Lineup Optimizer

A lineup of 9 baseball players can be ordered in 362880 different ways. But which way is best?

Enter player ratings for contact ability, power, plate discipline, and speed (0-10 with 10 being the best). The best lineup is the one that produces the most runs consistently.

Theoretical optimization uses a formula based on player ratings to weight the probabilities of making an out, advancing runners, getting on base, and driving in runs. Players at the top of the order receive more plate appearances, and certain sequences of batters generate more runs.

All 362880 possible lineups are evaluated to find the highest run expectation. Each inning sequence is evaluated producing a decimal number for run probability, with the first inning receiving a higher weighting.

Simulation mode uses a random number generator to simulate every at bat. This adds the possibility of double plays, sacrifice bunts, stolen bases, intentional walks, etc. The results are more realistic, but higher variance when the sample size is small. But this mode may take an extremely long time to run due to the vast number of calculations.

For instance, if based on player ratings, a player is expected to hit a homer 5% of the time. In theoretical mode, the player will hit exactly 35 homers every 700 at bats. While in simulation mode, the player may hit anywhere from about 10 to 50 or more. With a larger sample size, their true talent level will emerge as the numbers even out.