DA/PA and TA/PA: A New Method of Analyzing Outcomes

1. The "Soul" of the Metric: Why We Built This

Modern sabermetrics has largely converged on two approaches: crude composite ratios like OPS+ (which simply adds On-Base Percentage and Slugging Percentage across mismatched denominators) or long-term run expectancy models like wOBA and RE24. Linear-weight models operate by assigning fractional coefficients derived from historical macro-environments (0.69 for a walk, 0.88 for a single, 2.01 for a home run). While powerful for macroeconomic team projection, these regression weights strip away the physical reality of what happened on the diamond.

Traditional stats evaluate players in an artificial, sterile vacuum. Neither OPS+ nor wOBA cares that bases were loaded when a batter hit a home run, nor do they care whether a runner was advanced from second to third under pressure. They treat a solo home run down by 10 runs in July identical to a grand slam in a one-run game, and they treat a passive walk identical whether the bases are empty or loaded.

What is "Action"?

To solve this, we propose the concept of Action—the physical currency of baseball. Rather than translating outcomes into decimal run probabilities, an Action Unit represents a tangible event achieved on the field: every base gained on offense (1, 2, 3, 4) or every out recorded on defense (+1.0 out, +2.0 double play) is treated as a whole, invariant unit.

Monitoring baseball through Action fundamentally changes how we view both sides of the ball:

On Offense: We stop asking "What was the historical run expectancy of that hit?" and instead ask "How many physical bases and advancement units did this hitter force on the defense?"
On Pitching: We stop viewing pitchers as passive damage-avoiders trying to minimize negative ERA/FIP penalties, and instead model them as active Out Generators whose primary goal is producing positive Action Units (outs, strikeouts, and double plays) that destroy offensive momentum.

The Analytical Progression: From Action to Rate to Plus (`+`) Index

To evaluate players fairly across different workloads, lineup positions, and ballparks, our methodology follows a strict three-step progression:

Gross Counting Action (DA and TA): First, we measure the raw total Action Units a player accumulates: Direct Action (DA) from plate outcomes alone, and Total Action (TA) when incorporating baserunning pressure and situational advancement.
Per-Opportunity Efficiency (DA/PA and TA/PA): Because counting totals favor hitters with 650+ plate appearances, we divide gross Action by Plate Appearances (PA). This establishes a true rate of per-opportunity efficiency, revealing which players maximize every single trip to the plate.
Park-Normalized Plus Index (DA/PA+ and TA/PA+): Finally, to compare players across different stadiums and run environments, we scale per-PA efficiency against the league average and adjust for stadium Park Factor (PF). On this index, 100.0 represents exact league average; every point above or below represents a one-percent deviation in park-adjusted Action efficiency (e.g., a 125.0 index indicates a player is 25% more efficient per plate appearance than the league average).

2. The Core Philosophy: The Dual-Lens Pairing Principle

A cornerstone of the DA/PA and TA/PA framework is the realization that no single metric can simultaneously measure pure player skill in a vacuum and real-world contextual impact on the field.

To solve this, we introduce the Paired Evaluation Ecosystem:

Direct Action (`DA/PA`)	Total Action (`TA/PA`)
Pure Player Skill & Execution	Contextual & Situational Impact
"Who is out there doing a great job playing baseball at the plate?"	"Who produced the most total action given the opportunities they were placed in?"
Immune to Lineup/Situational Luck	Captures Baserunning & Advancement Chaos

2.1 Direct Action per Plate Appearance (`DA/PA`)

What it measures: A hitter’s pure, unadulterated ability to generate bases and productive outcomes at the plate.
Why it matters: DA/PA answers the practitioner's foundational question: "Who is simply out there doing a great job playing baseball?" It rewards clean hits (1B, 2B, 3B, HR), plate discipline (BB, HBP), and unselfish team execution (SF, SH).
Vacuum Purity: Because DA/PA is independent of baserunners on base prior to the plate appearance, it evaluates hitters fairly across any lineup spot.

2.2 Total Action per Plate Appearance (`TA/PA`)

What it measures: A hitter's comprehensive real-world footprint on the game by bridging plate production with dynamic baserunning pressure, stolen base equity, and situational runner advancement.
Why it matters: TA/PA is explicitly beholden to the opportunities a player is placed in. It acknowledges that baseball is not played in a vacuum. When a player is placed in a high-leverage opportunity, TA/PA measures whether they compounded that opportunity into actual runs and baserunning chaos.

2.3 Why `DA/PA` and `TA/PA` MUST Be Paired

Evaluating a player with only one half of the system creates analytical blind spots:

Using TA/PA Alone: Discounts opportunity bias. A player batting fourth in a powerhouse lineup will naturally encounter more high-action-point-gaining situations than an elite hitter stranded on a rebuilding team. Looking at TA/PA in isolation risks confusing opportunity volume with individual skill.
Using DA/PA Alone: Discounts dynamic situational value. It misses the baserunning catalyst who consistently steals second, takes extra bases on balls in the dirt, and scores from second on singles (actions that directly win games).
The Paired Synthesis (TA - DA Surplus): By pairing both stats, evaluators instantly identify player archetypes:
- High DA / High TA: The complete franchise cornerstone.
- High DA / Moderate TA: The station-to-station pure hitter (elite bat, minimal base-path pressure).
- Moderate DA / High TA: The situational catalyst whose baserunning and situational awareness generate massive field pressure beyond their raw plate skills.

3. The Pitching Counterpart: Active Out-Generation (`PDA/PA` & `PTA/PA`)

Traditional pitching models (ERA, FIP, SIERA) treat pitching as passive damage avoidance where pitchers start at zero and accumulate negative penalties. Our framework models the pitcher as an active Out Generator.

Just as hitters require paired evaluation (DA vs. TA), pitchers are evaluated through a dual-lens framework:

Pitcher Direct Action (PDA/PA): Measures foundational contact management and out generation. It credits +1.0 per Out Recorded and +2.0 per Double Play Induced, minus physical base penalties allowed (1B, 2B, 3B, HR, BB, HBP). Unlike FIP (which sets outs to 0.00 value), PDA/PA properly rewards ground-ball sinkerballers who consistently erase baserunners via double plays.
Pitcher Total Action (PTA/PA): Measures a pitcher's absolute run-prevention ceiling by adding an explicit Strikeout Dominance Bonus (+0.5 per K, totaling +1.5 Action Units per strikeout) and incorporating baserunning control (stolen bases allowed vs. caught stealing induced). Strikeouts eliminate 100% of contact and advancement risk; pairing PDA/PA with PTA/PA separates contact managers from dominant bullpen closers.

4. Intellectual Honesty: Boundaries & What This Stat Is NOT

To maintain intellectual rigor, a foundational statistical methodology must clearly define what it is not designed to do:

Analytical Task	Recommended Metric	Why Our Action Framework (`DA/TA`) Takes a Backseat
Macro Run-Expectancy Simulation (10,000 Games)	wOBA / RE24	Empirical linear weights are mathematically fitted to historical team runs scored across seasons. Integer Action Units prioritize physical transparency over fractional macro-fitting.
Pure Contact Physics & Launch-Angle True Talent	Statcast `xwOBA` / Barrel%	`DA/PA` evaluates what actually occurred on the field (results + discipline), whereas `xwOBA` isolates underlying exit velocity and launch angle independently of defense.
Context-Neutral Lineup Comparison (Un-Paired)	`DA/PA+` (Alone)	`TA/PA+` should never be used alone to compare players across vastly different team contexts without pairing it alongside `DA/PA+`.

5. Formal Mathematical Specification

5.1 Offensive Formulas

DA = (1 * 1B) + (2 * 2B) + (3 * 3B) + (4 * HR) + (1 * BB) + (1 * HBP) + (0.5 * SF) + (0.5 * SH)

TA = DA + (1.0 * SB) - (1.0 * CS) + Advancement Bonus

Advancement Bonus = 0.5 * MAX(0, R - HR - SB)

5.2 Pitching Formulas

PDA = (Outs Recorded * 1.0) + (Double Plays Induced * 2.0) - Base Penalties

PTA = PDA + (0.5 * SO) - (0.5 * SB Allowed) + (0.5 * CS Induced)

Where Base Penalties are defined as: Base Penalties = (1 * 1B) + (2 * 2B) + (3 * 3B) + (4 * HR) + (1 * BB) + (1 * HBP)

5.3 Universal Plus (`+`) Index Normalization (Park Adjusted)

All metrics are converted into a 100-indexed plus score normalized to league average and adjusted for stadium Park Factor (PF):

Metric+ = (Player Metric per PA / (League Avg Metric per PA * Park Factor)) * 100

100.0 = Exactly league average after park adjustment.
120.0 = +20% more action-efficient per plate appearance than league average.

6. Minimum Qualification Standards

To prevent small-sample distortions (such as pinch-running specialists inflating TA/PA+), official leaderboards enforce canonical qualification thresholds:

Hitters: >= 3.1 PA per Team Game Played (~502 PA full season).
Starting Pitchers: >= 1.0 IP per Team Game Played (~162 IP full season).
Relief Pitchers: >= 100 Batters Faced (BF).