What CPI is — and isn’t
The Court Pace Index (CPI) has obvious limitations. It doesn’t fully account for every variable that affects perceived speed, but it’s more dynamic than Court Pace Rating (CPR), which is measured by firing a no-spin ball at a uniform angle at the surface under controlled conditions.
CPR is fine as a standardised metric — but no professional hits a perfectly flat ball (spin rates are very high), and spin and angle of incidence affect bounce and perceived speed differently on different surfaces. Taking a mixture of CPI, adjusted ace rates, rally length, and relative air density — while still occasionally noisy — is probably the best overall proxy for court speed and conditions.
Formula
CPI = 100(1−μ) + 150(0.81−e)
where μ is the coefficient of friction and e the coefficient of restitution. CPI values are taken directly from official ATP/TennisTV CPI charts. Figures from 2012–2015 aren’t widely available. Where charts disagree (CPI drifts within a tournament as grit wears and weather varies), the tournament average is preferred, or multiple day readings are averaged.
Reading the numbers
- Ace rate, breaks of serve and return points won are main-draw figures only.
- CPI values are almost always centre/stadium court readings.
- Sea-level clay is basically always slow, so CPI changes there are small and usually weather-related.
- Elevation is taken from topographic maps at the precise location of the main stadium.
- Source screenshots are linked where available — hover any CPI value (or tap, on mobile).
Relative air density (RAD)
RAD measures how dense or “heavy” the air is, which affects ball speed and bounce. It’s calculated from air pressure, temperature and humidity across actual tournament dates and playing hours (11am–10pm), expressed as a percentage of standard sea-level density (1.225 kg/m³ at 15°C).
RAD explains conditions differences CPI misses: Toronto and Montreal use near-identical courts but Toronto consistently plays faster (altitude and temperature). Madrid’s high ace rate and short rallies aren’t captured by CPI alone — its thin air at 562m is. The 2023 US Open heatwave is visible in that year’s RAD. Indoor events (Paris, ATP Finals) are excluded — indoor temperatures differ meaningfully from outdoor readings.
Overall speed rating (OSR)
A composite of CPI, ace rate, rally length and RAD with surface-specific adjustments (clay −5, grass +5). Requires at least 3 of 4 metrics; each is scored 0–100 by continuous interpolation, then averaged. AFAIK it’s the only court-speed rating that explicitly accounts for weather and altitude. Its main limitation: it’s a relative measure across this set of tournaments, not an absolute scale. Recent years with all four metrics are most accurate.
Quirks worth knowing
- The ATP Finals sample is just 8 players in round-robin — more prone to skew. Its ace rate has been extremely high since moving to Turin (altitude), yet medium-length rallies remain common: the top 8 are usually elite returners and baseliners, so anything returnable gets extended.
- Montreal/Toronto alternates every other year.
- Some events (Bogotá WTA, Gstaad ATP) use high-altitude balls; no Masters or Slam does — not even Madrid. For comparison: Bogotá’s daytime RAD is 0.76–0.78 against Madrid’s 0.91 average.
Contact
Questions: @mattracquet or infotheracquet [at] gmail [dot] com.