A measurement of piano playing

Beat sway

How much single beats are stretched and squeezed, after the shape of the phrase has been taken out.

What it sounds like

The same passage — Nocturne Op. 9/2, bars 2–3 — from the recordings that sit lowest and highest on this measurement. Every clip is a note list replayed on one sampled grand at one gain, so the only difference you are hearing is the playing.

Least

Jan Lisiecki, 2021 — bars 2–3, beat sway 8 ±%14 s
Van Cliburn, 1971 — bars 2–3, beat sway 8 ±%15 s

Most

Vladimir Feltsman, 2012 — bars 2–3, beat sway 17 ±%13 s
Aldo Ciccolini, 2005 — bars 2–3, beat sway 17 ±%16 s

How it is calculated

The same calculation as arch depth, applied to what is left once the one-bar trend is removed.

How much pianists differ on it

Every piece in the library, with the range across its recordings. The units mean the same thing in every row, except for basic tempo, which is in each piece’s own beat.

PieceRecordingsLowestMedianHighest
Bach Prelude in C274614
Gymnopédie 1233711
Pathétique II475711
Chopin Étude 10/1315815
Rachmaninoff Prelude296911
Brahms Op. 118/22881114
Mozart K. 5453771115
Nocturne Op. 9/23281117
Chopin Ballade 412111115
Schubert Op. 142/3991215
Fantaisie-Impromptu2071217
Nocturne Op. 48/127101320
Träumerei3191420
Ravel Pavane33101417
Schubert Op. 90/343101421
Chopin Étude 10/1219111418
Clair de lune27111620
Ravel Ondine61141924

How reliably it can be measured

A Disklavier capture — the pianist’s own key sensors, no microphone — was rendered to audio, transcribed as if it were a recording, and re-measured. The error the transcription step introduces, as a multiple of how far pianists actually differ on this dial, is 0.16. solid This dial survives transcription intact: where one recording sits on it can be argued from.