Predictive analytics for the 2019 College Baseball World Series

Who should win the 2019 College Baseball World Series? Analytics and markets can help answer this question.

Ken Massey tracks a number of baseball rankings. However, only some of these numbers are predictive going forward.

My research in college football has shown that using margin of victory is essential for predictions. A method like RPI on the Massey site is immediately not considered.

A method like Elo can be predictive, but only if margin of victory is used in the calculations, not just wins and losses.

Two methods I know use margin of victory are Dokter Entropy and Massey Ratings. Here are the ranks of the 8 World Series teams by these two numbers (Dokter, then Massey).

  • Vanderbilt: 1st, 1st
  • Mississippi State: 3rd, 2nd
  • Arkansas: 4th, 4th
  • Texas Tech: 6th, 8th
  • Louisville: 8th, 5th
  • Michigan: 30th, 17th
  • Auburn: 35th, 11th
  • Florida State: 37th, 10th

The top 5 teams (Vanderbilt, Mississippi State, Arkansas, Texas Tech and Louisville) are the primary contenders, as they rank in the top 10 by both metrics. Michigan, Auburn and Florida State are on the outside looking in.

The markets agree with this assessment. These are the odds at Caesars.

These odds translate into these implied win probabilities.

  • Vanderbilt, 21.0%
  • Arkansas, 21.0%
  • Mississippi State, 18.7%
  • Texas Tech, 9.3%
  • Louisville, 9.3%
  • Florida State, 7.6%
  • Auburn, 7.6%
  • Michigan, 5.3%

These odds have probably moved since June 10th.

Speak Your Mind


This site uses Akismet to reduce spam. Learn how your comment data is processed.