Skip to main content
The pit_stops.csv table contains detailed information about every pit stop made during Formula 1 races.

​
Schema

FieldTypeDescription
raceIdintegerForeign key to races.csv
driverIdintegerForeign key to drivers.csv
stopintegerPit stop number for this driver in this race (1, 2, 3, etc.)
lapintegerLap number when the pit stop occurred
timetimeTime of day when the pit stop occurred (HH:MM:SS)
durationstringDuration of the pit stop in seconds
millisecondsintegerDuration of the pit stop in milliseconds
This table uses a composite key of (raceId, driverId, stop) rather than a single unique identifier.

​
Sample Data

raceIddriverIdstoplaptimedurationmilliseconds
2581001114:01:3449.11149111
2587911714:20:4628.48228482
2585711814:22:3543.74543745
2587111814:23:0021.99221992

​
Relationships

References:
  • pit_stops.raceId → races.raceId
  • pit_stops.driverId → drivers.driverId

​
Dataset Statistics

  • Total Records: 22,145 pit stops
  • Date Range: Pit stop data available from 2011 onwards
  • Stops per Race: Varies (typically 1-3 per driver)

​
Example Queries

​
Get all pit stops for a race

import pandas as pd

pit_stops = pd.read_csv('pit_stops.csv')
race_stops = pit_stops[pit_stops['raceId'] == 841]
race_stops_sorted = race_stops.sort_values('lap')
print(race_stops_sorted[['driverId', 'stop', 'lap', 'duration']])

​
Find fastest pit stops

fastest_stops = pit_stops.nsmallest(10, 'milliseconds')
print(fastest_stops[['raceId', 'driverId', 'duration', 'milliseconds']])

​
Analyze pit stop strategies

# Count stops per driver in a race
race_stops = pit_stops[pit_stops['raceId'] == 1000]
strategy_counts = race_stops.groupby('driverId')['stop'].max()
print("Drivers by number of stops:")
print(strategy_counts.value_counts().sort_index())

​
Calculate average pit stop duration by race

avg_duration = pit_stops.groupby('raceId')['milliseconds'].mean()
print(avg_duration.head())

​
Find pit stop lap distribution

import matplotlib.pyplot as plt

race_stops = pit_stops[pit_stops['raceId'] == 1000]
plt.hist(race_stops['lap'], bins=20)
plt.xlabel('Lap Number')
plt.ylabel('Number of Pit Stops')
plt.title('Pit Stop Distribution Throughout Race')
plt.show()

​
Identify unusual pit stop durations

# Stops longer than 10 seconds (potential issues)
slow_stops = pit_stops[pit_stops['milliseconds'] > 10000]
print(slow_stops[['raceId', 'driverId', 'lap', 'duration']].head(20))

​
Compare pit stop performance by constructor

results = pd.read_csv('results.csv')

stops_with_constructor = pit_stops.merge(
    results[['raceId', 'driverId', 'constructorId']], 
    on=['raceId', 'driverId']
)

avg_by_constructor = stops_with_constructor.groupby('constructorId')['milliseconds'].mean()
print(avg_by_constructor.sort_values().head(10))

​
Analyze undercut/overcut strategies

# Find drivers who pitted earlier/later than competitors
race_stops = pit_stops[pit_stops['raceId'] == 1000]
first_stops = race_stops[race_stops['stop'] == 1]
first_stops_sorted = first_stops.sort_values('lap')
print(first_stops_sorted[['driverId', 'lap', 'time']])

​
Calculate pit stop time trend over seasons

races = pd.read_csv('races.csv')

stops_with_year = pit_stops.merge(races[['raceId', 'year']], on='raceId')
avg_by_year = stops_with_year.groupby('year')['milliseconds'].mean()

print("Average pit stop duration by year:")
print(avg_by_year)

​
Pit Stop Records

  • Fastest pit stop: Typically under 2 seconds (modern era)
  • Typical duration: 2-3 seconds for tire change (2010s-present)
  • Common strategies:
    • One-stop: Single pit stop during race
    • Two-stop: Two pit stops (most common)
    • Three-stop: Three or more stops (rare, usually weather-related)

​
Notes

  • Pit stop timing data is available from 2011 onwards
  • Duration includes time from pit entry to pit exit
  • Very long pit stops (30+ seconds) often indicate mechanical issues or penalties
  • Pit stop strategy significantly affects race outcomes
  • time field shows time of day (race-local time)
  • Red flag periods may affect pit stop strategies
  • Modern pit stops (2018+) average around 2-2.5 seconds
  • Regulations on number of mechanics allowed affect pit stop times
  • The stop field helps identify one-stop, two-stop, etc. strategies

Build docs developers (and LLMs) love

Get started for freeTalk to us