Positional Adjustment in CPBL
How much is a catcher's bat worth compared to a DH? Applying Tom Tango's Position Switch method to 36 years of CPBL data.

Not all positions are created equal. A shortstop who hits .250 is more valuable than a first baseman who hits .250—because the shortstop is doing his job at a much harder defensive position. Positional adjustment quantifies this difference in runs.
The standard approach uses MLB values developed by Tom Tango: catchers get +12.5 runs, while DHs get -17.5 runs per 162 games. But should CPBL use the same numbers? The defensive demands, player development paths, and league context differ significantly.
We decided to find out.
The Position Switch Method
Tom Tango's Position Switch Analysis is elegantly simple: track players who change positions between seasons and measure how their offensive production changes. If a player moves from catcher to first base and their wRC+ jumps, that tells us something about the defensive burden each position carries.
"The Position Switch method measures defensive burden indirectly through offense—when a player moves to an easier position, their energy shifts from defense to hitting."
Using 36 years of CPBL data (1990-2025), we identified 313 cases where a player changed their primary position between consecutive seasons. Each case provides a data point: Position A vs Position B, with the wRC+ difference measuring the relative defensive burden.
The Methodology Ensemble
A single method can be noisy. To get robust estimates, we combined three approaches:
313 consecutive-season switches with PA weighting and outlier winsorization. Weighted Least Squares to solve for positional values.
Causal inference approach using ML models to control for confounders (age, era, team quality) and isolate the position effect.
Position-level wRC+ averages across 3,467 player-seasons, with Mixed Effects Models (player as random effect).
We also attempted a game-level analysis with 28,098 position switches—but the results were nonsensical (r=0.18 correlation with MLB). Single-game wOBA is simply too noisy to detect positional effects. Season-level analysis was far more stable.
The Results
The ensemble produced a clear defensive spectrum for CPBL:
| Position | CPBL | MLB | Diff |
|---|---|---|---|
| CCatcher | +10.9 | +12.5 | -1.6 |
| SSShortstop | +4.9 | +7.5 | -2.6 |
| 2BSecond Base | +4.2 | +2.5 | +1.7 |
| 3BThird Base | +0.8 | +2.5 | -1.7 |
| CFCenter Field | -0.7 | +2.5 | -3.2 |
| LFLeft Field | -0.9 | -7.5 | +6.6 |
| RFRight Field | -3.5 | -7.5 | +4.0 |
| 1BFirst Base | -7.4 | -12.5 | +5.1 |
| DHDesignated Hitter | -8.4 | -17.5 | +9.1 |
Values in runs per 162 games
Key Findings
The CPBL shows a compressed spectrum. The gap between the hardest and easiest positions is 19.3 runs in CPBL versus 30 runs in MLB—about 35% narrower. This makes sense: with smaller rosters and less specialization, CPBL teams can't afford pure defensive specialists at premium positions.
Center field isn't premium in CPBL. While MLB gives CF a +2.5 adjustment (same as SS and 2B), CPBL data shows CF at -0.7—essentially neutral. This may reflect smaller outfields in Taiwanese stadiums and different outfield defensive demands.
The DH penalty is smaller. CPBL DHs get -8.4 versus MLB's -17.5. In a league where positional flexibility is valued and the talent pool is shallower, the opportunity cost of dedicating a lineup spot to a pure hitter is lower.
What This Means for WAR
These adjustments are now integrated into myCPBL's WAR calculations. A CPBL catcher who hits league-average gets +10.9 runs of positional credit per 162 games, while a DH gets -8.4 runs. The narrower range means positional adjustments have less impact on WAR rankings in CPBL compared to MLB.
Consider two players with identical wRC+ of 100 (league average): if one is a catcher and one is a DH, the catcher is worth 19.3 more runs over a full season—roughly 2 wins. That's the defensive spectrum at work.
Positional adjustment is one of the more abstract components of WAR. But it captures something real: the physical and mental burden of playing different positions, reflected in how much offensive energy a player can spare. Our CPBL-specific values ensure we're measuring this burden in the right context.
Position Switch Analysis + Double ML + Mixed Effects Models on 1990-2025 CPBL data. Sample: 3,467 player-seasons, 313 position switches. Game-level analysis excluded due to excessive noise.