L2 — Bowler vs Batter-Type × Phase

Mirror of L2 bowler-type-phase from the bowling side: how a bowler's figures differ vs right-handers and left-handers across powerplay / middle / death phases.

Period: IPL 2026. Sample-size floor: ≥30 balls bowled. Suppressed cells render the URI with the parent L1 link instead of metrics.

Conditioning dimensions

batter_type — RHB or LHB, classified from `batsman.battingstyle`. "Unknown" surfaces as a separate bucket.
phase — same definitions as the batter-side methodology.

Sample-size floor + rationale

≥30 balls bowled per cell.

Doubles the §3.1 bowling floor since the 2 × 3 grid is fewer cells than UC1; we can afford a tighter floor.

Suppression rule

Same as L2 bowler-type-phase. Suppressed cells preserve URI; APIs return `suppressed: true`.

Update cadence

Same — post-match incremental + daily full rebuild.

Known limitations

Single-season corpus thinness applies (see notes on the player-side methodology).
Per-bowler match-state is folded into the period aggregate; bowler-in-state is a separate methodology (UC7).

Change log

2026-05-13 · Phase 1 launch. Floor: ≥30 balls bowled.

Consumed by

UC2: /players/{slug}/bowling/vs/{batter-type}/phase/{phase}/