Win Probability Added (WPA) for O’s hitters

[Frobby]

As most posters know, WPA for hitters is calculated by looking at a hitter’s at bats and how the outcome affected the team’s probability of winning the game.   A home run when losing 8-0 in the 9th counts for almost nothing, while a humble walk in a tie game in the 9th counts for a lot.   Here’s the WPA for all Orioles hitters with at least 100 PA:

Rutschman 1.52

Mancini 0.46

Urias 0.44

Hays 0.36

Mullins 0.11

McKenna 0.08

Ssntander 0.02

Mountcastle -0.07

Odor -0.12

Chirinos -0.39

Nevin -1.00

Mateo -1.56

I’d say both Santander and Mateo are lower than I would have guessed.   And, it’s interesting to see that Odor has had a negligible negative affect.

Edited August 30, 2022 by Frobby

[LA2]

10 minutes ago, Frobby said:

As most posters know, WPA for hitters is calculated by looking at a hitter’s at bats and how the outcome affected the team’s probability of winning the game.   A home run when losing 8-0 in the 9th counts for almost nothing, while a humble walk in a tie game in the 9th counts for a lot.   Here’s the WPA for all Orioles hitters with at least 100 PA:

Rutschman 1.52

Mancini 0.46

Urias 0.44

Hays 0.36

Mullins 0.11

McKenna 0.08

Ssntander 0.02

Mountcastle -0.07

Odor -0.12

Chirinos -0.39

Nevin -1.00

Mateo -1.56

I’d say both Santander and Mateo are lower than I would have guessed.   And, it’s interesting to see that Odor has had a negligible negative affect.

Mateo being the worst is shocking. But it must be because of his poor first half and because WPA is purely based on at-bats, whereas so much of Mateo's impact is his defensive play and speed on the base paths. 

[Can_of_corn]

Nevin scoring that badly with such little playing time is notable.

[Aristotelian]

Interesting, I would have guessed Santander higher, and that number from Adley is insane. 

[Jim'sKid26]

This may be the stat that explains why most folks in the know are not picking the O's to get a WC spot. 

Pretty stark contrast between Adley and everybody else.

[Pickles]

To clarify, this is calculated in relation to their actual performance?  Meaning Rutschman has been worth 1.5 extra wins beyond his base production due to "clutchness", correct?

[Frobby]

17 minutes ago, Aristotelian said:

Interesting, I would have guessed Santander higher, and that number from Adley is insane. 

Well, it’s not really insane.  It’s 48th in MLB.  Aaron Judge leads at +5.31, more than triple what Adley has.   Manny, by the way, is third at 4.37.

[Hank Scorpio]

Sure could use a boost offensively, huh? 

[Frobby]

22 minutes ago, Pickles said:

To clarify, this is calculated in relation to their actual performance?  Meaning Rutschman has been worth 1.5 extra wins beyond his base production due to "clutchness", correct?

It’s not related to Adley’s base performance.   It’s just the cumulative impact on the team’s chances of winning as the result of each at bat, given the game situation.   For example, Adley’s most impactful at bat was on July 4, one out in the bottom of the ninth, runner on 1st, losing 6-5.  At that moment, based on all like situations in baseball history, the Orioles had a 10% (.10) chance of winning that game.  Adley hit a game-tying double.  As a result, following that play the O’s now had a 61% (.61) chance of winning the game.  So, Adley’s WPA for that at bat was .510 (.61 - .10).   

[LA2]

Continuing to read along the same line of stats in Baseball-Reference.com, I spotted very good baserunning stats for Mateo, Mullins, McKenna, Adley, and Urias. Mateo and Mullins are tied for the league lead in stolen bases (28) with a decent success rate (78%). The others haven't stolen much at all, but are good at taking an extra base. And in almost all cases, they are performing better than in 2021. Probably an underrated dimension of the Orioles' success this season.

[Pickles]

21 minutes ago, Frobby said:

It’s not related to Adley’s base performance.   It’s just the cumulative impact on the team’s chances of winning as the result of each at bat, given the game situation.   For example, Adley’s most impactful at bat was on July 4, one out in the bottom of the ninth, runner on 1st, losing 6-5.  At that moment, based on all like situations in baseball history, the Orioles had a 10% (.10) chance of winning that game.  Adley hit a game-tying double.  As a result, following that play the O’s now had a 61% (.61) chance of winning the game.  So, Adley’s WPA for that at bat was .510 (.61 - .10).   

Got it.  Thanks for the clarification.

[Filmstudy]

2 hours ago, Frobby said:

It’s not related to Adley’s base performance.   It’s just the cumulative impact on the team’s chances of winning as the result of each at bat, given the game situation.   For example, Adley’s most impactful at bat was on July 4, one out in the bottom of the ninth, runner on 1st, losing 6-5.  At that moment, based on all like situations in baseball history, the Orioles had a 10% (.10) chance of winning that game.  Adley hit a game-tying double.  As a result, following that play the O’s now had a 61% (.61) chance of winning the game.  So, Adley’s WPA for that at bat was .510 (.61 - .10).   

I think the answer to the original question is that the baselines for WP are based on league average performance from that point forward in the game.  That said, I don't know if they use a 3-year rolling average or just 2022, or whatever to model win probabilities by game/base/out situation.

Edited August 30, 2022 by Filmstudy