savant-extras

Baseball Savant leaderboard data — complements pybaseball.

pybaseball is great but many Baseball Savant leaderboards are missing or limited. savant-extras fills that gap with 17 leaderboards covering batting, pitching, catching, baserunning, fielding, and park factors — all as simple one-line function calls returning DataFrames.

Installation

pip install savant-extras

Quick Start

from savant_extras import (
    bat_tracking, pitch_tempo, arm_strength,
    pitch_movement, swing_take, catcher_throwing,
)

# Bat tracking with custom date range (2024+)
df = bat_tracking("2024-04-01", "2024-04-30")

# Pitcher pitch tempo
df = pitch_tempo(2024)

# Outfielder arm strength
df = arm_strength(2024, position="Outfielder")

# Slider movement
df = pitch_movement(2024, pitch_type="SL")

# Batter plate discipline
df = swing_take(2024)

# Catcher pop time & CS rate
df = catcher_throwing(2024)

All Functions

Every leaderboard function returns a pd.DataFrame. Most have a _range() variant for multi-season queries (adds a year column).

Batting

Pitching

Catching

Baserunning & Fielding

Park Factors

Columns returned: season, team, pf_5yr, pf_3yr, pf_1yr, pf_hr, pf_1b, pf_2b, pf_3b, pf_so, pf_bb, pf_fip. All factors: 100 = neutral, >100 = hitter-friendly, <100 = pitcher-friendly.

from savant_extras import park_factors, park_factors_range

# Single season
df = park_factors(2024)
print(df[df["team"] == "COL"][["team", "pf_5yr", "pf_hr"]])
#    team  pf_5yr  pf_hr
# 5   COL     116    131

# Multi-season (e.g. for model training)
df = park_factors_range(2020, 2025)
print(df.shape)  # 6 seasons × 30 teams = 180

Common Parameters

Multi-Season Queries

Most functions have a _range(start_year, end_year) variant:

from savant_extras import pitch_tempo_range, arm_strength_range

# Compare pitch tempo pre/post pitch clock
df = pitch_tempo_range(2022, 2024)

# 5 years of arm strength
df = arm_strength_range(2020, 2024)

Demo App

MLB Bat Tracking Dashboard — built with savant-extras (source)

Why savant-extras?

Known Issues

• swing_take(): Baseball Savant の Swing & Take リーダーボードの CSV エンドポイントに障害中（ヘッダーのみ、データ行なし）。現在は空の DataFrame が返ります。上流 API が復旧次第、コード変更なしで動作します。代替として batted_ball() や year_to_year() を使用してください。

Cloud Environment Notes

park_factors() and park_factors_range() fetch data from FanGraphs. In cloud environments (Kaggle, Google Colab, GitHub Actions), FanGraphs may return 403 errors. In that case, park_factors_range() returns an empty DataFrame with a warning instead of raising an exception.

Recommended workaround: pre-download locally and upload as a dataset file.

# Run locally and save
from savant_extras import park_factors_range
df = park_factors_range(2024, 2025)
df.to_csv("park_factors.csv", index=False)
# Upload park_factors.csv to your Kaggle dataset, then read it in the notebook

In Kaggle notebooks, pybaseball is not pre-installed. Install both explicitly:

!pip install savant-extras pybaseball

License

MIT