It has been a pretty long time since I made a post and a lot of stuff has happened, but a new statistical though has brought me back. So here is my latest thought on how to predict playoff games correctly.
This goes back to the idea of Pitching Performance Value, but this time evaluated with WAR. PPV would evaluate a pitcher's performance after they have pitched. This new stat will make a prediction of the game winner before the game starts.
 |
| Former AL East rivals will go head to head tomorrow |
So if an average pitcher pitches against an average lineup and an above average pitcher pitches against an above average lineup the box score is likely to appear the same. For example if Wei Yin Chen and Sonny Gray were to go head to head the box score may look the same. Wei Yin Chen (Orioles) is an average pitcher against an average lineup and Sonny Gray (Athletics) is an above average pitcher against an above average lineup.
To figure out who might win this game we will look at a ratio of the pitcher's WAR and the offense's WAR as a whole. So take the WAR of the starter and divide by the summation of the WAR of the opposing lineup. Do this for both pitchers and lineups and the greater value is the pitcher who will win. So lets do this for tomorrows game between the Athletics and Royals
The Royals lineup has a WAR of 18.3 and the Athletics lineup has a WAR of 21.3. Jon Lester (Athletics) has a WAR of 4.8 and James Sheilds has a WAR of 3.4.
Royals Ratio = 3.4/ 21.3 = 0.15
Athletics Ratio = 4.8/18.3 = 0.26
The Athletics will win the game tomorrow. It is hard to say what the final score will be, but according to the overall talent of the players and pitchers Oakland has an absolute advantage. Baseball is unpredictable but i'm just trying to quantify it.
Using the WAR ratio I used for the Oakland vs Kansas City game the Giants appear to have a slight advantage.
