Safety Events

​

Safety Cars

​

Schema

​

Sample Data

​

Dataset Statistics

• Total Records: 364 safety car periods
• First Usage: 1973 Canadian Grand Prix
• Common Causes: Accidents, debris, weather conditions

​

Red Flags

​

Schema

​

Resume Codes

• Y - Yes, race resumed
• N - No, race ended under red flag
• R - Race restarted
• S - Special circumstances

​

Sample Data

​

Dataset Statistics

• Total Records: 98 red flag incidents
• First Usage: 1950
• Common Causes: Weather, major accidents, track safety issues

​

Virtual Safety Car

​

Structure

{
    "Race Name": [lap_numbers],
    "2015 Belgian Grand Prix": [20, 21],
    "2015 British Grand Prix": [32, 33, 34],
    "2015 Hungarian Grand Prix": [41, 42]
}

• Keys: Race name (Grand Prix title)
• Values: Array of lap numbers when VSC was active

​

Dataset Statistics

• Format: JSON
• First VSC: Introduced in 2015
• Purpose: Track safety without physical safety car

​

Example Queries

​

Count safety car deployments per race

import pandas as pd

safety_cars = pd.read_csv('safety_cars.csv')
deployments = safety_cars['Race'].value_counts()
print(deployments.head(10))

​

Find longest safety car periods

longest_sc = safety_cars.nlargest(10, 'FullLaps')
print(longest_sc[['Race', 'Cause', 'FullLaps']])

​

Analyze red flag outcomes

red_flags = pd.read_csv('red_flags.csv')
resumed_count = red_flags['Resumed'].value_counts()
print("Red flag outcomes:")
print(resumed_count)

​

Find races with multiple safety cars

multiple_sc = safety_cars.groupby('Race').size()
races_with_multiple = multiple_sc[multiple_sc > 1].sort_values(ascending=False)
print(races_with_multiple.head(10))

​

Parse VSC data

import json

with open('virtual_safety_car_estimates.json', 'r') as f:
    vsc_data = json.load(f)

# Count VSC laps per race
for race, laps in vsc_data.items():
    print(f"{race}: {len(laps)} VSC laps")

​

Analyze safety interventions by year

# Extract year from race name
safety_cars['year'] = safety_cars['Race'].str.extract(r'(\d{4})')
sc_by_year = safety_cars.groupby('year').size()
print(sc_by_year)

​

Compare safety car vs red flag frequency

print(f"Safety car deployments: {len(safety_cars)}")
print(f"Red flag incidents: {len(red_flags)}")
print(f"Ratio: {len(safety_cars) / len(red_flags):.2f}:1")

​

Find common safety car causes

cause_counts = safety_cars['Cause'].value_counts()
print("Top causes for safety car:")
print(cause_counts.head(10))

​

Identify races with both SC and red flag

sc_races = set(safety_cars['Race'])
rf_races = set(red_flags['Race'])
both = sc_races & rf_races
print(f"Races with both safety car and red flag: {len(both)}")
for race in sorted(both):
    print(f"  - {race}")

​

Safety Intervention Types

​

Safety Car (SC)

• Physical car leads the field
• Drivers must slow down and bunch up
• No overtaking allowed
• Pit lane usually open

​

Virtual Safety Car (VSC)

• Introduced in 2015
• Drivers must maintain delta time (slow down by percentage)
• No physical safety car on track
• More flexible than full SC

​

Red Flag

• Race completely stopped
• Cars return to pit lane or grid
• Used for serious incidents or conditions
• Race may or may not resume

​

Notes

• Safety car data is more complete for modern races
• VSC estimates are based on lap time analysis (not official FIA data)
• Red flags can occur during any session (practice, qualifying, race)
• Multiple safety car periods can occur in a single race
• \N represents missing or unavailable data
• Race names should be matched with the races.csv table for full context
• Weather-related interventions are common at specific circuits (e.g., Belgium, Japan)