Race telemetry analysis: an Academy round read lap by lap

On Friday 2 October 2026, the JOBARD Racing Academy Series held its Round R3 at Kyalami, in the Mazda MX-5 ND Cup. We worked through the telemetry of one driver from the field, session by session and lap by lap, from the opening of practice to the chequered flag of the second race.

The aim is not to recount the evening, but to show what the data reveals that a finishing order never does. The driver remains anonymous: only the numbers matter, and the way to read them.

One evening, five lap times

A Sprint round follows a fixed structure: one hour of free practice, a ten-minute qualifying session that sets the grid for Race 1, twenty minutes of racing, then a second qualifying session and a second race.

The driver's best lap in each of these five sessions:

• Free practice: 2:05.214
• Qualifying 1: 2:07.089
• Race 1: 2:04.692
• Qualifying 2: 2:03.996
• Race 2: 2:03.699

The progression is almost perfect: just over one and a half seconds gained between practice and the final race, each session faster than the one before. Except one — precisely the session that sets the grid for Race 1.

Qualifying 1: ten minutes, one lap that counts

Kyalami measures 4.522 km and the MX-5 laps it in around 2:05. In ten minutes, after the out-lap, the session allows three flying laps, rarely more.

The telemetry from this qualifying session comes down to three lines. An out-lap in 2:39.087. A first flying lap invalidated by the game, completed in 2:56.274. Then a single valid flying lap in 2:07.089 — 2.4 seconds above what the same driver would set in the race half an hour later.

Result: third on the grid for Race 1.

The lesson lies in session management, not in outright pace. In a qualifying session this short, the first flying lap is an insurance lap, driven slightly below the limit to have a clean time secured; the attack comes afterwards. That plan is prepared before the evening itself, like the rest of the 48-hour pre-race preparation.

Race 1: eight laps within half a second

Race 1 lasts ten laps. The first, the standing-start lap, is completed in 2:19.626. Eight laps then follow, ranging between 2:04.692 and 2:05.223: a spread of 0.531 s from fastest to slowest, with a standard deviation of 0.19 s. That is a remarkably consistent pace.

A single lap breaks the sequence. The fifth, in 2:11.982, sits seven seconds above the pace. The telemetry does not say why; the driver does: a position battle. He finishes second, one place gained on his starting position, 0.513 s behind the winner.

The fastest lap of the race, 2:04.692, is his; the winner's fastest lap is 2:05.232. A race does not reward the quickest single lap but the sum of all laps, and the starting position weighs on that sum.

A note on reading the data: the average of these ten laps is 2:07.099, their median 2:05.048. The average blends the start, the battle, and the race pace, and no longer describes anything meaningful. To read a race, look at the median and the laps that deviate from it.

Race 2: the track had changed, not only the driver

In Race 2, the driver completes his best lap in 2:03.699, nearly one second quicker than in Race 1 (0.993 s precisely). He wins the race with a 6.017 s margin.

The temptation is to conclude that the driver has improved. The data calls for caution. The car was heavier: 30 litres of fuel at the start, compared with 17 in Race 1 — roughly ten kilograms more. The track temperature recorded by the telemetry is identical at the start of both races: 30 °C. And above all, every driver lapped faster in Race 2, by anywhere from three tenths to over two and a half seconds.

When the entire field improves at the same moment, the primary explanation is the track, not the driver. Comparing a Race 1 lap time to a Race 2 lap time is almost like comparing two different circuits. A fair comparison requires equal conditions: within the same session, or against a reference lap taken under the same conditions — which is precisely what reading a telemetry trace with its delta allows.

Three invalidated laps, none of them faster

In Race 2, the game invalidated three of the driver's ten laps, completed in 2:04.794, 2:05.385, and 2:05.136. In Race 1, only one, in 2:05.106.

The significant figure lies elsewhere: none of these laps is faster than the best valid lap from the same race. An invalidated lap is rarely a shortcut that pays off; most often it is a car placed slightly too wide, with no benefit whatsoever to the lap time.

In the Academy Series, track limits are enforced by Assetto Corsa EVO's automatic system, and the regulations provide for an official warning at the third infringement, followed by a drive-through at the fourth. An invalidated lap is the first signal, well before any penalty. If your invalidated laps consistently fall at the same point, the cause is rarely the track: it is often an apex taken too early, which throws the car beyond the white line on corner exit.

What the data shows, what the driver knows

Telemetry indicates where to look, rarely why. The fifth lap of Race 1 illustrates this: the data isolates seven seconds lost, but only the driver knows it was a position battle — a racing decision. In Race 2, a lap in 2:08.529, four seconds above pace, is associated with no incident in the official race report. The figure is there; the explanation is not yet: that is the question to raise at the debrief.

The incident count points in the same direction: thirteen in Race 1, including three contacts for the driver in question, four in Race 2, none for him. Race 2 was both quicker and cleaner.

Reading your own round in four questions

Everything above comes down to four questions:

• In which session, and under what conditions, was this lap time set?
• Which laps fall outside my pace window, and what happened in them?
• How many valid laps did my qualifying session give me?
• How many laps were invalidated, and at which point on track?

The next round, R4 at Laguna Seca on Friday 16 October, is an endurance format: a single sixty-minute race with a mandatory pit stop. Fuel and the pit window become full race variables in their own right.

All of this analysis comes from the telemetry recorded during the round by JOBARD Race Engineer: every timed or invalidated lap, pit stop passages, best lap progression, and fuel on board. The Discovery plan is free, with no session limit, and Assetto Corsa EVO is among the supported simulators.

Analyse your rounds with JOBARD Race Engineer →

Sources: JOBARD Race Engineer telemetry of the driver in question, sessions of 2 October 2026 at Kyalami (Mazda MX-5 ND Cup) · official race report for Round R3 of the JOBARD Racing Academy Series, published 3 October 2026 · Pre-season 2026 regulations, edition 1.7 of 3 October 2026.

In brief

What is a race telemetry analysis?

A lap-by-lap reading of session data: lap times, valid and invalidated laps, pit stop passages, fuel, then speed, braking, and acceleration. It shows where time is gained and lost, beyond the finishing order.

Why is the fastest lap not enough to win a race?

Because a race adds up every lap. A poor starting position or a lap lost in a battle matters more than half a second of pace per lap. In this round, the fastest lap of Race 1 did not belong to the winner.

Can you compare lap times from two races in the same evening?

With caution: fuel load and track condition change from one session to the next. Compare within the same session, or against a reference lap taken under the same conditions.

Does an invalidated lap save time?

Rarely. In this round, no invalidated lap bettered the best valid lap from the same race. It signals a line too close to the limit and, repeated, leads to a penalty.

How should you manage a ten-minute qualifying session?

First secure a clean lap, driven slightly below the limit, then attack with that time already banked. On a two-minute lap, the session offers little more than three flying laps.