want play-by-play data from Retrosheet as tables (CSV, SQL, pandas) and need Chadwick's
well-known columns, such as cwevent's 97 standard fields and 67 extended ones;
work somewhere you cannot or would rather not compile C: a managed notebook, a locked-down
work laptop, Windows, a CI job, a container with no build tools;
are reproducing or extending work that used the Chadwick tools and want the same
definitions for events, runner advances, and counts;
want to call the parser from Python without shelling out to a separate program;
want one pip install and nothing else: there are no dependencies.
process many seasons repeatedly and speed matters more than convenience: per core the C tools
are about 10 to 40 times faster, and they can also be run on several cores
(see compared with Chadwick);
want ready-made tables with no tool at all: Retrosheet publishes its own CSV downloads,
which use Retrosheet's columns rather than Chadwick's;
want pitch-by-pitch tracking data (velocity, spin, location): Retrosheet event files
record the result of each pitch (ball, strike, foul...), not measurements. That data comes from
other sources.
need live or current-season feeds: chadwickpy reads files you give it.