Welcome back to “Stats All, Folks,” our season-long look at sabermetrics and analytics in baseball.
That’s enough pleasantries. It’s time for stats! Today’s edition: Win Expectancy and Win Probability Added.
What is Win Expectancy?
Win Expectancy (WE) projects a team’s chances of winning a game at any given point of a game. WE analyzes the inning, the score, the number of outs and the number of baserunners and compares that to how previous teams have fared in that situation.
Let’s say a club is trailing by one run in the seventh inning with one out and no one on base. That team’s WE is calculated by examining all previous games in which a team trailed by one run in the seventh with one out and no one on base, and finding the percentage of those teams that ended up winning the game (for example, 30%).
It’s important to note that WE does not take into account how good each team is, what pitchers are starting, or any other specifics about the teams. It also doesn’t factor in a home-field advantage. You and I both know that if Clayton Kershaw and the Los Angeles Dodgers were hosting the hapless Philadelphia Phillies, the Dodgers would be odds-on favorites to win. But WE treats them as having an equal chance of winning at the start of the game.
WE usually uses a sample of games from a certain time period, such as the past 10 years, rather than every game in baseball history. After all, today’s brand of baseball is a lot different than it was in the 1940s and 1950s, so it wouldn’t be very relevant to compare the two.
At the beginning of a game, each club has a 50% WE, but that WE changes with every play. Whenever that team’s pitchers record an out or its offense gets a runner on base or scores a run, its WE goes up. When its pitchers allow a run or a baserunner, or its offense hits into an out, the team’s WE goes down.
As you might expect, certain plays can lead to dramatic changes in Win Expectancy, while other plays barely move the needle. For instance, hitting a home run in the eighth inning to break a tie could increase a team’s WE by 25-30%. But hitting a home run in the eighth inning of a game you’re losing 10-1 might only change your WE by 1% or so, since you’re still very likely to lose.
What is Win Probability Added?
Win Probability Added (WPA) measures how much each individual hitter or pitcher changes his team’s Win Expectancy on a particular play. A player earns positive points for increasing his team’s WE and negative points for decreasing it.
Let’s say a hitter who blasts a go-ahead home run to lead off the eighth changes his team’s Win Expectancy from 58.6% to 79.6% (or, put in decimal terms, from .586 to .796). That hitter would earn a +.210 WPA. And the pitcher who gave up the home run would be charged with a -.210 WPA.
For a real-life example of WE and WPA in action, let’s look at one of the most memorable games in recent Orioles history: Game 2 of the 2014 AL Division Series, a.k.a. the Delmon Double game.
As the bottom of the eighth inning started, the Orioles trailed the Detroit Tigers, 6-3. The Orioles’ Win Expectancy stood at a meager 8.1%, according to FanGraphs’ box score. That means that in a sample of previous MLB games in which a team trailed by three runs with no outs and no one on base in the eighth, the trailing club won just 8.1% of the time. And after the Tigers’ Joba Chamberlain retired Alejandro De Aza to lead off of the eighth, that WE dropped to 5.5%.
Over the course of the inning, though, that WE steadily grew. An Adam Jones hit by pitch gave the Orioles a baserunner, slightly improving their WE from 5.5% to 8.5%. Jones was credited with a +.030 WPA. Nelson Cruz then singled, further increasing the WE to 14.0% (and giving Cruz a +.054 WPA for the play).
The barrage continued. Steve Pearce’s RBI single (+.106 WPA) bumped up the WE to 24.6% and chased Chamberlain from the game. Joakim Soria came in and poured gasoline on the fire. He walked J.J. Hardy (+.121 WPA) to load the bases. At that point, the Orioles had a 36.7% chance of winning — despite their rally, they were more likely to lose than win, based on historical precedent.
Delmon Young changed all that. His dramatic double into the left-field corner — plating all three runners — gave the Orioles a 7-6 lead as their WE skyrocketed to 88.3%. On one swing, Young had increased the Orioles’ chances of winning by more than 50%. He was credited with a whopping +.516 WPA on the play. Soria retired the final two batters, but the damage was done. The Orioles had an 84.1% WE, and they went on to win the game.
Just as Orioles hitters racked up positive WPAs in that eighth inning, the Tigers’ relievers compiled negative (and unsightly) WPAs. Chamberlain came in with his team having a 91.9% chance of winning, but by the time he left the game, it was down to 75.4%. So, his WPA for the game was -.165. And Soria was even worse. He dropped his team’s WE from 75.4% to 15.9%, saddling him with a -.595 WPA. That’s a pretty awful performance, to say the least.
Zach Britton’s historic WPA
It’s probably not a surprise to learn that the highest total WPA in the major leagues last season belonged to the best player in baseball, AL MVP Mike Trout. His cumulative 6.64 WPA shows how much he single-handedly improved the Angels’ chances of winning throughout the season.
But the second-highest WPA in the majors in 2016 might surprise you. It belonged to the Orioles’ Zach Britton, at 6.39.
That doesn’t mean Britton was the second-best player in baseball (although he was certainly among the best). What it means is that Britton, when he appeared in games, improved his team’s chances of winning more than anyone in baseball besides Trout.
Britton’s WPA was the highest single-season mark of any pitcher in Orioles history. It was also the second-best of any MLB pitcher in the last 16 years, behind only Zack Greinke’s 6.79 in 2015.
So how did Britton post such an historically high WPA? Well, a big factor was that he was a closer who appeared in a lot of tight games, when every swing of the bat could dramatically shift the Win Expectancy. When a pitcher is protecting a one-run lead in the ninth inning, each out he records improves his team’s WE anywhere from 5-15%, while each runner he allows to reach base can be equally damaging.
Britton was exceptionally good at getting batters out and keeping them off base, so he racked up a lot more positive-WPA plays than negative ones. Britton retired 201 batters and permitted only 56 baserunners. And, as Orioles fans are well aware, Britton was a perfect 47-for-47 in save opportunities. Not once did he blow a late-inning lead, which would’ve dealt serious damage to his WPA.
In the 2016 AL Cy Young voting, proponents of Britton pointed to his league-leading WPA as one of the biggest points in his favor. After all, even though he’d only pitched 67 innings, he’d had a huge influence on his team’s chances of winning — arguably more than anyone else in the majors. Alas, that argument wasn’t enough; Britton finished fourth in the voting.
Way back at the turn of this century, a venture capital guy named Tom Hicks,…
While the signings of players with major league experience to minor league contracts might not…
Question: Why wouldn't the O's grab both Max Scherzer and Justin Verlander since they could…
Tomoyuki Sugano, the 35-year-old Japanese right-hander who signed with the Orioles on Monday, appeared on…
Question: With the signing of Tomoyuki Sugano, it seems that if you prefer a smaller…
It’s been a busy offseason for the Orioles, and it’s not done. While the Orioles…
View Comments
And to think today's youth have turned to soccer claiming that baseball is boring,
Different people enjoy baseball in different ways. You can enjoy it in your way without being derisive of what other people find interesting.
You are absolutely correct Stacey. I apologize to everyone especially Paul.
Thanks, Boog. I know this stuff isn't for everyone. But with sabermetrics/analytics being such a key part of baseball these days, I think there's an audience that is interested in learning more about some of these statistics and trends. For me, at least, it helps shine light on what makes some teams successful and others not so much.
Before we analyze all of the advanced metrics, the Orioles should focus on the basics first; find a guy who wins more than he loses, and find someone who can control their ERA. Once we address that, then we can move on to more advanced categories.
Have a nice weekend all!
If you're saying the Orioles need some good pitchers, I think everyone agrees with that. But no good team operates by ignoring advanced statistics. They're an important part of building and evaluating your players and your team.
i don't need stats or metrics bs to know this Beckham kid is a great ballplayer also don't need stats to know if you don't swing the bat Davis you don't hit
I wouldn't be jumping on the "great" bandwagon yet with Beckham. Great start, but consistency is always the separator.