Zandvoort FIA F4 Case Study: 55 Laps, 0.61s Faster Repeatable Pace
Zandvoort FIA F4 Case Study: 55 Laps, 0.61s Faster Repeatable Pace
A personal best is useful, but it can also lie to you.
One clean lap can come from a perfect tow, a better track state, one unusually good corner, or simply putting together a lap that is difficult to repeat. For driver development, the more important question is whether the driver's normal pace is actually moving.
This case study follows one anonymous FIA F4 driver across 55 valid archived laps at Zandvoort between August 20 and August 24, 2026. The sample includes offline testing, open practice, lone qualifying and race running. Valor recorded the laps, input traces and modeled turn-loss data throughout the program.
The headline result was a 1:34.810 best lap, but that is not the main result. The stronger signal is that representative pace improved by 0.611 seconds, modeled turn loss fell 17.5% across comparable early and late samples, and part of that improvement carried into race running.
The program at a glance
The archived sample contained:
55 valid laps total
6 offline-testing laps
6 practice laps
4 lone-qualifying laps
39 race laps
First recorded lap: 1:36.376
Best recorded lap: 1:34.810
Peak first-lap-to-best-lap improvement: 1.566 seconds
The driver began with normal pace centered in the mid-1:36 range. By the end of the program, normal laps were repeatedly appearing in the mid-1:35s, with a race best of 1:35.081 and the 1:34.810 practice lap showing the upper edge of the pace available.
Why we did not use PB versus PB as the main comparison
A fastest lap is important, but it is a poor standalone measure of development.
To reduce the effect of a single exceptional lap, we compared the median of the first ten normal laps with the median of the last ten normal laps in the sample.
First 10 median: 1:36.250 Last 10 median: 1:35.639 Representative pace gain: 0.611 seconds
That is the number that matters most in this analysis. The driver's typical execution moved by more than six tenths.
The improvement was not perfectly linear. There were slower laps, race interruptions and sessions where the driver gave time back. That is normal. Driver development rarely looks like a smooth descending line. What matters is where the center of the performance distribution moves over time.
Here, it moved clearly.
Turn loss moved with the stopwatch
Valor's modeled turn loss gives another view of the same development.
Across the first ten-lap sample, average modeled turn loss was 2.182 seconds. Across the final ten-lap sample, it was 1.799 seconds.
That is a reduction of 0.383 seconds, or 17.5%.
The best lap took that progression further. The first recorded lap carried approximately 2.400 seconds of modeled turn loss. On the 1:34.810 lap, modeled loss fell to 0.984 seconds.
That is roughly a 59% reduction in modeled corner loss from the first archived lap to the best archived lap.
This does not mean Valor "created" 59% more performance, and it does not mean every modeled tenth converts one-for-one into official lap time. Turn-loss modeling is a diagnostic comparison tool. The important observation is that the reduction happened alongside a clear improvement in both peak and repeatable lap pace.
The first test session established the baseline
The opening offline-testing block contained six valid laps.
Best: 1:36.119 Median: 1:36.250 Average modeled turn loss: 2.259s
The laps were already reasonably clustered. This was not a driver starting ten seconds off pace and simply learning where the circuit went. The opportunity was in refining execution across several corners.
The first recorded lap, a 1:36.376, carried meaningful loss in multiple areas at once:
Turn 8: 0.310s
Turn 12: 0.415s
Turn 7: 0.257s
Turn 13: 0.302s
Turn 1: 0.301s
No single corner explained the entire lap. That becomes important later, because the eventual gain also did not come from one heroic corner.
The practice breakthrough
Across six archived practice laps, the driver reached:
Best: 1:34.810 Median: 1:35.767 Average modeled turn loss: 1.739s
The 1:34.810 was the fastest lap of the entire analysis window, but the practice median is arguably more valuable. It shows the driver was not merely capable of one lap in the 1:34s; the surrounding pace had also moved well below the original 1:36.250 testing median.
On the best lap, total modeled turn loss dropped below one second to 0.984s.
What the 1:34.810 lap looked like
The strongest feature of the lap was not a radical change in one braking zone. Several previously expensive corners became small-loss corners at the same time.
Turn 7 fell from approximately 0.257s of loss on the first recorded lap to 0.009s on the PB. The PB entered the corner at about 129.1 mph, carried a minimum around 108.2 mph, and exited around 113.1 mph. Peak brake was only about 26.8%, which fits the character of a high-speed commitment corner where excessive deceleration can cost more than it saves.
Turn 8 improved from 0.310s of loss to only 0.022s. On the PB, entry speed was approximately 118.9 mph, minimum speed 100.8 mph, and exit speed 105.7 mph, with peak brake around 34.6%. That is a gain of approximately 0.288s in one corner compared with the first archived lap.
Turn 12 improved from 0.415s to 0.162s, a gain of roughly 0.253s. The best lap reached the corner at about 68.4 mph, dipped to approximately 58.7 mph, and exited around 84.2 mph. The useful change here was not simply a higher minimum speed; the overall corner sequence became less costly.
Turn 13 moved from 0.302s to 0.146s, recovering another 0.156s. The PB entered at approximately 103.5 mph, carried around 87.4 mph minimum speed and exited at about 95.9 mph.
Those four corners alone account for a substantial portion of the observed reduction in modeled loss.
The four largest corner improvements
Comparing the first archived lap with the 1:34.810 best lap:
Turn 8: 0.310s → 0.022s — 0.288s recovered
Turn 12: 0.415s → 0.162s — 0.253s recovered
Turn 7: 0.257s → 0.009s — 0.248s recovered
Turn 13: 0.302s → 0.146s — 0.156s recovered
The pattern matters. The biggest gains came from different corner types and different parts of the lap. This was not one braking marker being moved ten meters later. Time was removed from a combination of entry execution, minimum-speed management and exits.
The difficult corners did not disappear
A good case study should show what was still imperfect.
On the 1:34.810 lap, Turn 1 still showed about 0.184s of modeled loss. The car entered around 140.6 mph, reached approximately 61.9 mph minimum speed and exited around 83.2 mph, with peak brake close to 79.6%.
Turns 9, 10, 12 and 13 also remained measurable opportunities:
Turn 9: approximately 0.136s loss
Turn 10: approximately 0.174s loss
Turn 12: approximately 0.162s loss
Turn 13: approximately 0.146s loss
That is one reason this lap is useful for development. A 1:34.810 was a significant step forward, but the telemetry still showed a credible path to more pace instead of pretending the lap was perfect.
Qualifying pace
The lone-qualifying sample contained four valid laps.
Best: 1:35.835 Median: 1:35.960 Average modeled turn loss: 2.001s
The qualifying sample did not match the absolute practice PB, but it remained materially stronger than the opening test median. That distinction is useful: peak pace had improved, but reproducing the absolute best lap on demand was still a separate skill.
The most important test: did the improvement survive race conditions?
Across 39 archived race laps, the driver recorded:
Race best: 1:35.081 Race median: 1:35.917 Average modeled turn loss: 1.967s
The race median was approximately 0.333s faster than the original offline-testing median of 1:36.250 despite the additional variability inherent in race running.
The latest larger race block was stronger again:
Latest race-block median: 1:35.757 Average modeled turn loss: 1.858s
This is the part of the dataset that makes the progression more credible. The improvement was not trapped inside a low-fuel practice lap. A meaningful portion of it appeared in repeated race laps.
The race laps were not uniformly faster. Traffic, battles, mistakes and race context create noise. Some archived race laps were far slower than representative pace. That is exactly why median pace is more informative than simply averaging every lap in the session.
What changed in the driver's performance profile
By the end of the program, three things were true at the same time:
1. The ceiling moved. The fastest lap improved from the original 1:36-range running to 1:34.810.
2. The floor moved. The driver was producing far more laps in the mid-1:35 range instead of relying on an isolated PB.
3. The telemetry moved. Modeled corner loss fell alongside the stopwatch, and the largest reductions appeared across several different corners.
That combination is much stronger evidence of driver development than a PB alone.
What this does — and does not — prove about Valor
This is an observational case study, not a controlled experiment.
The driver had access to Valor's live telemetry ecosystem during the program, including comparison data and coaching information. But there was no control group, no identical before-and-after track state, and no way to isolate every variable that affects lap time.
We therefore do not claim that Valor caused every tenth of the improvement.
What the data does demonstrate is that Valor can capture a detailed development record across multiple sessions and make changes visible that the stopwatch alone hides.
Instead of only saying "the driver gained 1.566 seconds," we can separate:
peak improvement from repeatable improvement,
practice pace from race pace,
total lap time from corner-level loss,
one exceptional lap from a shifting pace distribution,
and the corners that changed most from the corners that remain open opportunities.
That is the purpose of the system.
Final scorecard
Analysis window: Aug. 20–24, 2026 Car: FIA F4 Circuit: Zandvoort GP Valid archived laps: 55 First recorded lap: 1:36.376 Best lap: 1:34.810 Peak improvement: 1.566s First-10 median: 1:36.250 Last-10 median: 1:35.639 Repeatable pace gain: 0.611s First-10 modeled turn loss: 2.182s Last-10 modeled turn loss: 1.799s Representative loss reduction: 17.5% First-lap modeled turn loss: 2.400s Best-lap modeled turn loss: 0.984s First-lap-to-best loss reduction: ~59% Race best: 1:35.081 Race median: 1:35.917 Latest larger race-block median: 1:35.757
The takeaway
The 1:34.810 gets attention, but it is not the strongest result in this dataset.
The stronger result is that over 55 valid laps, the driver's normal pace moved by about 0.61 seconds, modeled turn loss fell 17.5% across representative samples, several major corner losses were substantially reduced, and the improvement carried into race running.
That is what driver development looks like when you can see more than the final lap time.
Explore Valor Baselines Download Valor
Comments