What Is WPA (Win Probability Added)?
WPA (Win Probability Added) quantifies how much each plate appearance or play shifts a team's probability of winning. For example, in a tied game with two outs and bases loaded in the bottom of the ninth, the home team's win probability is roughly 35 percent. A walk-off hit pushes it to 100 percent, yielding a WPA of +0.65 for the batter. A strikeout sends the game to extras and produces negative WPA. The concept originated in the 1970s when Eldon and Harlan Mills created win probability tables for MLB. It gained widespread use in the 2000s with the sabermetrics movement. In NPB, data sites began calculating and publishing WPA in the late 2010s, establishing it as a new axis for player evaluation.
Top WPA Performers in NPB
Players with high WPA in NPB tend to overlap with those considered clutch hitters. In the 2023 season, Orix's Yuma Tongu posted a .307 batting average with 17 home runs and a WPA of +4.2, finishing as the Pacific League batting champion while ranking among the league leaders in clutch performance. In the Central League, Hanshin's Yusuke Oyama recorded a WPA of +3.8, contributing significantly to the Tigers' first league title in 18 years under the ARE slogan. Among pitchers, Yoshinobu Yamamoto posted a WPA of +5.1 in 2023, an exceptionally high figure for a pitcher. Historically, Sadaharu Oh's 1964 season with 55 home runs and Hiromitsu Ochiai's 1985 Triple Crown season are estimated to have exceeded +7.0 WPA.
How WPA Is Calculated and Its Limitations
Calculating WPA requires a win probability table derived from four factors: inning, out count, base-runner situation, and score differential. In NPB, these tables are built from roughly 10 years of game data, covering approximately 858 games and 120,000 plate appearances per season. The win probability shift for each plate appearance outcome (single, double, strikeout, double play, etc.) is computed, and the season total becomes a player's WPA. However, WPA has several limitations. First, it is highly context-dependent, favoring players who appear frequently in close games. Second, defensive and baserunning contributions are not captured in batting WPA. Third, opponent strength is not factored in. For these reasons, WPA is best used alongside comprehensive metrics such as WAR (Wins Above Replacement).
Related books are also helpful
Practical Applications and Future Outlook
Since the early 2020s, more NPB teams have begun using WPA for tactical decisions. Pinch-hitting and bullpen management timing are evaluated by expected WPA to maximize win probability. In 2024, SoftBank reportedly introduced a proprietary player evaluation system incorporating WPA, optimizing deployment of closer Livan Moinelo (WPA +2.8). Real-time win probability displays for fans are also spreading; during the 2023 Japan Series, broadcast screens featured live win probability graphs. Going forward, integration with tracking data is expected to enable pitch-level WPA calculations and more refined defensive WPA metrics. As a metric that visualizes the value of each moment, WPA will continue to deepen data utilization across NPB.
The Relationship Between WPA and LI (Leverage Index)
Understanding WPA deeply requires the concept of LI (Leverage Index), which quantifies the criticality of each game situation. An average situation scores 1.0, while tight late-inning scenarios with runners on base can exceed 2.0. WPA is structurally higher for players who deliver in high-LI spots. A single in the top of the first with no runners yields roughly +0.03 WPA, but a game-tying hit in the bottom of the eighth with a runner on second and a one-run deficit can produce +0.25 or more. In NPB, Yakult's Munetaka Murakami batted .340 in situations with LI above 1.5 during the 2022 season, accumulating substantial WPA. Combining LI and WPA allows evaluation not merely of overall performance but of a player's true worth in decisive moments.
The Negative Side of WPA - The 'Anti-Clutch' Perspective
WPA swings sharply in both directions; a player with negative season WPA has effectively lowered the team's win probability. Relievers who fail in save situations and batters who make outs in high-leverage spots accumulate negative WPA rapidly. A closer who surrenders a walk-off hit in the ninth can lose -0.40 or more in a single play. In NPB, closers with frequent blown saves sometimes finish with negative WPA, and starters who suffer many tough-luck losses in low-run-support games can also fall below zero. Negative WPA does not necessarily indicate poor ability, however. Pitchers who frequently work in close games and top-of-the-order hitters face high-LI situations more often, where one failure produces outsized negative swings. Because WPA incorporates such randomness, examining multi-year trends rather than a single season is essential for fair evaluation.
Reading Iconic Japan Series Moments Through WPA
WPA moves most dramatically in the postseason, particularly the Japan Series, where each game's outcome directly determines the championship. In Game 5 of the 2014 Japan Series, SoftBank's Seiichi Uchikawa delivered a go-ahead RBI hit in the eighth inning for an estimated WPA of +0.38. In Game 7 of the 1992 Japan Series, Yakult's Toru Sugiura hit a walk-off home run in the tenth inning, with that single swing estimated at +0.55 WPA. Among pitchers, Nippon-Ham's Yu Darvish threw a complete-game shutout in Game 5 of the 2006 Japan Series, accumulating over +0.50 in pitching WPA across the game. Listing the top WPA plays in Japan Series history aligns precisely with the most memorable moments in NPB fans' collective memory. WPA provides numerical confirmation of just how dramatic those instants truly were.