This page documents the raw JSON files served from the TracingInsights data repos. tif1 loads these files and maps their abbreviated keys to fastf1-compatible PascalCase DataFrame columns. See the Data Schema Reference for the columns of
session.laps, lap.telemetry, and the other public APIs.How This Data Is Generated
The current season’s repository uses two main extraction scripts. Older season repositories hold earlier versions, some under other names such astelR.py. The scripts share the same output format for laptimes.json, but produce different session-level files.
Data sources:
- FastF1: base lap timing, telemetry, weather, race control messages, circuit info, driver info.
- Ergast/Jolpica: official Race lap times and positions (used by
LapTimes.pyto overwrite FastF1 RaceLapTimeandPosition). - OpenF1: mini-sector segment data and session keys (used by
LapTimes.pyforms1/ms2/ms3). - MultiViewer API: circuit corner geometry fallback when FastF1 circuit info is unavailable.
Weather Data (weather.json)
Environmental conditions recorded approximately once per minute during the session.
Why Weather Matters in F1
Weather has a strong effect on lap times. Cold air reduces engine power, hot tracks reduce tire grip, and rain changes braking distances. Wind direction affects DRS (Drag Reduction System) effectiveness and how the car handles in turns.Race Control Messages (rcm.json)
Messages sent by Race Control (the officials) to teams about track status, penalties, and incidents.
What This Means
When a yellow flag is shown in marshal sector 2, the data can show{cat: "Flag", msg: "YELLOW IN TRACK SECTOR 2", flag: "YELLOW", scope: "Sector", sector: 2.0}. This tells the teams that drivers must slow down in that sector.
Session State (session_info.json)
The session state that fastf1 reads from the live-timing streams besides the laps. The keys and the column names are the fastf1 ones. A value that fastf1 cannot load is null. The data repositories have this file for every session from 2018.
tif1 reads this file for the six attributes in the table. Before 2018, tif1 derives the values from the race control messages, the laps, and the schedule.
Telemetry Data ({lap}_tel.json)
Samples from the car’s sensors, merged with the car position. In the 2024 Monaco Race data, samples are about 0.13 s apart (about 7.5 samples per second). Each lap is stored in its own {lap}_tel.json file, with the fields under a top-level tel key.
Basic Car Metrics
Driver and Car Ahead Information
tif1 replaces the
"None" sentinel in both fields with a missing value: NaN or None on pandas, and null on polars. In tif1 telemetry frames, DriverAhead holds driver-number strings and DistanceToDriverAhead is a float column.
Acceleration Vectors (G-Forces)
The car experiences forces in three directions simultaneously. The extraction script computes these values using gradient analysis of position, speed, and distance data. These are not raw IMU sensor values. They are mathematically derived and heavily processed.Position Data (3D Coordinates)
The coordinates use units of 1/10 m, so the path length through
x and y is about 10 times distance. Car data comes at roughly 240 ms intervals and position data at roughly 220 ms. Each file merges the two streams, and values are interpolated between the measurements.Other Telemetry Fields
The format is
Year-EventName-Session-Driver 3 letter code-Lap Number, for example 2025-United States Grand Prix-Race-VER-9.
For 2026 pre-season testing, the EventName part has no spaces or hyphens:
PreSeasonTesting1 for Pre-Season Testing 1, and PreSeasonTesting for Pre-Season Testing.Driver Data (drivers.json)
Static information about each driver that does not change during a session. The file holds one drivers list with one object per driver.
Lap Times Data (laptimes.json)
Timing and performance information for each completed lap, stored per driver. session_laptimes.json holds the same keys for all drivers in one file, in alphabetical driver order. tif1 prefers that file when it loads laps, and falls back to the per-driver files.
Core Timing Information
Sector Times (Track Divided into 3 Sections)
The track is divided into 3 sectors. Sector times show where a driver gained or lost time.Speed Trap Data (Measured at Fixed Points)
Modern F1 tracks have speed guns that measure top speed at specific locations:Tire Information
Mini-Sector Segment Times (OpenF1)
Encoded digit strings show the colour classification of each mini-sector in a sector. These fields are available only when OpenF1 data can be matched to the session. Otherwise the value is"None".
Each character is one mini-sector. The number of mini-sectors in a sector depends on the circuit. Each character uses the following colour map:
Qualifying Segment
Position and Status
Pit Stop Information
A pit lap is either an in-lap (driving into the pits) or an out-lap (driving out of the pits).Data Quality and Accuracy
Driver/Team Reference
Per-Lap Weather
When weather data can be matched to individual laps, the following fields are appended tolaptimes.json. These use the same abbreviations as weather.json.
Track Status Reference
Circuit/Corner Data (corners.json)
Information about the physical corners of the racetrack layout.
Some newer files, for example 2026, add more keys such as
CornerLetter, MarshalSectorNumber, TrackX, and TrackY. tif1 reads only the keys in this table.
Ergast Data (TracingInsights-Archive/Stats)
The Stats repository stores the Ergast tables as raw Jolpica responses, one JSON file per table and round, every page merged. tif1.ergast reads it first and asks the Jolpica API only for what it cannot answer.
slug is the race name of events.json in lower case with spaces replaced by hyphens, so tif1 resolves it by round through that file. A lap whose timings were split across two Jolpica pages appears twice in laptimes.json; tif1 merges the timings. The season-level driverPoints.json and teamPoints.json files are stale and are not read.
Where to Find More Information
Official FastF1 Documentation: https://docs.fastf1.dev/- Detailed API documentation
- Examples and tutorials
- Data accuracy information
- Troubleshooting guides