Squiggles, Traces and the Truth: Decoding Your First Data Download
Photo: racing telemetry data screen throttle brake trace track day laptop in pit garage, via i.pinimg.com
The session ends. You come in, peeled out of the harness, and somebody hands you a laptop. On it is a graph with more colours than a BTCC livery and no obvious point of entry. Speed, throttle, brake, steering, gear, GPS position, lateral g, longitudinal g, and a channel called 'distance' that appears to be in metres but might be in feet.
That's the moment most track-day drivers close the laptop and go and look at the tyres instead. Which is a shame, because the data is genuinely useful — it's just badly introduced. Nobody sits you down and tells you what you're supposed to do with it.
So here's the version that would have helped.
Start with one corner, not one lap
The instinct is to overlay two full laps, see the mess, and give up. Don't. Zoom in.
Pick the corner you're least happy with. Not the one you're worst at — the one you're least sure about. The one where you come out and can't decide whether you braked too early or got on the power too late. That uncertainty is exactly what the data answers.
Set the window to about five seconds either side of the corner. Now you're looking at a manageable picture: a speed trace, a throttle trace, a brake trace. Three lines. That's the whole exercise.
The only three channels that matter at the start
Speed. The headline number. Where's your minimum? Is it at the apex, or forty metres before it? If your speed trace bottoms out early and then climbs through the corner, you were slow at turn-in and paid for it.
Throttle. This is the honest one. Throttle traces show hesitation, and hesitation is where lap time goes to die. A clean throttle trace steps up and stays up. A nervous one wiggles, lifts, re-applies, wiggles again — and every one of those wiggles is a car that wasn't settled.
Brake pressure. Where the trace goes from flat to vertical tells you your actual braking point in metres, not in vibes. Where it comes back up tells you how long you were trailing. If your pressure trace looks like a cliff, you're stamping and releasing. If it looks like a ramp, you're actually trailing — which is usually faster.
Everything else can wait.
What to ignore, deliberately
GPS-derived racing lines on a phone. At 1Hz or even 10Hz, a phone's GPS position is a suggestion, not a fact. The pretty overlay showing you a metre off the ideal line is often just satellite wobble. If you want line analysis, you need proper GPS — and even then it's a distraction in year one.
Lateral g figures. Entertaining, not actionable. You cannot drive to a g number.
Other people's reference laps. A YouTube overlay from a works driver in a car with three times your downforce is not a benchmark, it's an ambition. Compare yourself to yourself. That's the only fair comparison you have.
Predictive lap time. Useful in the car. Corrosive on the laptop. You'll spend forty minutes chasing a theoretical number instead of learning a corner.
The habits that make a log useful
Name your corners. Most software lets you build a lap with named segments. Do it once, at one circuit, and save it. Next time out, the analysis takes ninety seconds instead of twenty minutes.
Change one thing per session. If you change your line, your braking point and your tyre pressures at once, the data tells you nothing except that something happened. One variable. Write it in the notes field — yes, the notes field, the one nobody uses.
Save a reference lap early. Your best lap from session two is your benchmark for the day. Compare everything to that. Not to the session six flyer you got in traffic with a tow.
Write the debrief down. Two sentences after each session, in the car, on your phone. What felt good, what felt wrong. Data plus feel is coaching. Data alone is just decoration.
Kit, quickly
You don't need much. A phone running RaceChrono or Harry's LapTimer with a decent external GPS puck will do a surprising amount for well under £200, and it'll log the three channels above. Step up to an AiM Solo 2 or a VBOX and you get proper sampling rates, CAN data from the car and brake pressure if you've fitted a sensor — that's the point at which the traces get genuinely sharp.
What you absolutely do not need is more channels. Collecting twenty is easy. Reading three is a skill.
The point of all this
The data logger doesn't lie. That's the uncomfortable bit. It will tell you, without any sympathy whatsoever, that the corner you've been blaming on tyres is a corner you've been entering four mph slow for three sessions.
But it also doesn't judge. It just sits there, holding the truth in a chart, waiting for you to zoom in on five seconds of it and ask a single, answerable question.
Do that once and you'll never hand the laptop back unopened again.