Introducing a new and improved pwOBA+ (Batters)

This will be the final predictive stat that I introduce. So far, I’ve introduced four new and improved pitching metrics (pwOBAcon+, pBB%+, pK%+, and pwOBA+) and two new and improved batting metrics (pwOBAcon+ and pBB%+).

In addition to pwOBAcon+ and pBB%+, regressed hit-by-pitch percent and regressed strikeout percent are encompassed by pwOBA+.

pwOBA = (pBB%*wBB)+(HBP% regressed*wHBP)+((1-HBP% regressed-pBB%-K% regressed)*pwOBAcon)

pwOBA+ = pwOBA/league average non-bunt wOBA*100

Here are the lowest single-season predictive weighted on-base average plus marks since 2015

1. Jeff Mathis 2019 (80.4)
2. Chris Owings 2019 (82.5)
3. Daniel Castro 2016 (83.4)
4. Keon Broxton 2019 (83.7)
5. Adam Engel 2017 (83.7)
6. Martín Maldonado 2021 (83.9)
7. Arismendy Alcántara 2017 (84.0)
8. Billy Hamilton 2019 (84.1)
9. Juan Graterol 2017 (84.1)
10. Luis Sardiñas 2015 (84.3)

Highest

1. Mike Trout 2018 (129.7)
2. Mike Trout 2019 (129.6)
3. Mookie Betts 2018 (126.5)
4. Mike Trout 2015 (125.8)
5. Mike Trout 2017 (125.5)
6. Aaron Judge 2018 (124.3)
7. Mike Trout 2016 (124.2)
8. David Ortiz 2016 (124.1)
9. Bryce Harper 2015 (122.7)
10. José Bautista 2015 (121.1)

In looking at how predictive 2018 pwOBA was of 2019 non-bunt wOBA compared to 2018 non-bunt wOBA and 2018 non-bunt xwOBA, one can see that pwOBA is far more predictive of future non-bunt wOBA than non-bunt xwOBA and non-bunt wOBA are. It is also the most stable of the three metrics.

Instead of using non-bunt xwOBA to identity batters that one predicts to hit better or worse moving forward, one should consult pwOBA, as it is more stable and predictive.

Featured image- Creator: Sean M. Haffey | Credit: Getty Images