Monday, November 9, 2026

How is team wOBA differential related to winning percentage?

I had bases.chat, the baseball AI program created by SABR member Ted Ingram, calculate wOBA and wOBA allowed for all teams over the years 2021-2025 combined. Information about wOBA is at the end of this post.

This graph shows the relationship.

 Image

The regression equation or the trend line equation has an intercept of .4999. Almost exactly .500. That would mean that a team with a zero differential would be a .500 team. The intercept that I usually get if I do this with OPS differential is also about .500.

The r-squared of .94 means that 94% of the variation in winning percentage is explained by wOBA differential.

Tom Tango created wOBA (Weighted On-Base Average). Here is one version of the formula
 
wOBA = 
 
(0.690×uBB + 0.722×HBP + 0.888×1B + 1.271×2B + 1.616×3B + 2.101×HR) 
 
divided by 
 
(AB + BB – IBB + SF + HBP) 
 
uBB means unintentional walks. The coefficients on the events in the numerator vary slightly from year to year. 
 
For further information: 

 
Guts! This has the annual coefficients for wOBA from FanGraphs
 
Was wOBA Actually Invented Nearly 100 Years Ago? This is another article from FanGraphs. Discusses the work of Baseball Magazine editor F.C. Lane. He came up with values for batting events that were close to what we see with linear weights.

No comments:

Post a Comment