The Lap Analysis Workflow, the Data Channel Reference, and the Setup Correlation guide tell you what to do. This page covers what goes wrong - the errors that appear in every debrief, at every circuit, across every level of club racing. Some are beginner mistakes. Several are made by experienced drivers who should know better. All of them have appeared in real post-session analysis at SupaTune Motorsport in Porsche Club GB.
Why the reference lap is everything
At SupaTune Motorsport, the reference lap used for driver analysis is frequently a lap set by Colin Tester - founder, team manager, Grade S ARDS instructor, and the driver widely known in Porsche Club GB circles as the King of Brands Hatch. Colin won the Porsche Club GB Championship outright in 2022 and has accumulated more than 100 first-place finishes since gaining his race licence in 2007. His coaching experience spans everyone from developing club racers through to Jake Hill, Jodie Sloss, Theo Paphitis and Paul Hollywood - and he has sat alongside Valentino Rossi at Brands Hatch GP, guiding him around the circuit.
In practice, this means Colin will often take a driver's own car out during a test day to set a benchmark lap - same car, same tyre set, same conditions. The driver then analyses their data against his. This is a categorically more useful reference than the driver's own personal best, because it separates what the car and circuit allow from what the driver is currently extracting. A PB comparison measures improvement against your own ceiling. A Colin lap comparison measures the gap to what is actually possible.
The reference lap concept is not unique to SupaTune - but having access to a Grade S instructor with a championship-winning record to set that reference in your own car, at your own circuit, on your own tyres, is. The most common mistake in club racing data analysis is not a channel interpretation error. It is using the wrong reference lap to begin with.
Reference and Comparison Errors
The mistakes that corrupt the analysis before a single channel trace is opened. If the reference is wrong, everything that follows is wrong.
The mistake
A personal best lap is the fastest lap you have driven. It is not the fastest lap the car and circuit allow. Comparing against your own PB tells you how much you have improved relative to yourself - it tells you nothing about how much time remains available.
A driver who is 2.5 s off the circuit benchmark and improves by 0.8 s looks like they are making progress. Against a reference lap set by an experienced driver in the same car, the 1.7 s gap that remains is suddenly visible - and the data shows exactly where it is being lost.
The correct approach
When a coached reference lap is available - set by an instructor in your own car, on the same tyres, in the same session - use it. Always. The sector delta against that reference is the most honest performance measurement available at club level.
If no coached reference exists, use the fastest clean lap from the most experienced driver in the same car and class that weekend. A reference from a car in a different class, on different tyres, or with a significantly different setup is not valid - the variables are not controlled. See the Lap Analysis Workflow for reference lap selection criteria.
The mistake
A lap recorded in session 1 on cold tyres and a damp track is not a valid reference for a lap recorded in session 3 on rubber-scrubbed dry tarmac. The numbers will look different - but the track changed, not the driver. This error is most damaging when it leads to a setup change based on a cross-condition comparison. The setup change was not needed - the conditions explained the difference.
Track temperature affects lap times by more than most club racing setup changes. A circuit that has rubbered in through the morning will be 0.5-1.5 s faster in the afternoon with no changes to the car at all.
The correct approach
Always record air temperature, track temperature, and session number alongside your data. If you do not have track temperature, use the best comparable lap from the same class and same day as your reference point - not a historical circuit record, which may come from a different tyre specification, weather, era, or surface condition.
For cross-session comparisons that need to be valid, use a car running in the same configuration as a floating reference throughout the day. If both cars improve at the same rate between sessions, the track improved. If your car improves more than the reference at a specific corner, the setup change contributed something real at that corner specifically.
The mistake
A lap that is significantly faster than the others in the session is almost never representative. Traffic cleared ahead at exactly the right moment, the tyre was at peak temperature for exactly that lap, the driver had a clean exit from the previous corner onto the back straight. The data from that lap looks brilliant. It is also unrepeatable.
Using an anomalous fast lap as the reference for all subsequent comparisons sets an artificial ceiling. Everything else looks slower by comparison not because the driver regressed but because the reference is wrong.
The correct approach
Three clean laps minimum before drawing any conclusion. The reference lap should be a lap that represents consistent, repeatable performance - not the single best lap of the session. In AiM Race Studio, check the lap time distribution across the session before selecting a reference. A lap that sits more than 0.3-0.4 s faster than the next three laps is likely anomalous.
The theoretical fastest lap (TFL) tool in AiM Race Studio is useful here - if the single fast lap is close to the TFL, it may be representative of the car's true potential. If the TFL is significantly faster still, the lap was not as good as it looks.
The mistake
An AiM EVO4S and a Motorsport VBOX HD2 will give different absolute G values at the same event. The MEMS sensors are calibrated differently, mounted in different positions, and processed through different firmware. A driver showing 1.4 g on an EVO4S and another showing 1.6 g on a VBOX at the same corner in the same session are not necessarily cornering at different levels - the instruments may simply read differently.
This error is most damaging when it leads to a coaching conclusion - "driver B is generating more lateral grip" - that is actually an instrument difference.
The correct approach
Lap time, sector time, and broad speed shape compare reasonably well across different GPS-based systems - useful for macro analysis of where time is being lost. Precise transient comparisons, particularly braking onset and release timing, are more trustworthy when both cars use the same logger family, similar mounting practice, and the same filtering approach. A 10 Hz system and a higher-rate system will not give identical braking onset detail even if the GPS physics are the same. Use G channels for shape comparison across hardware, not absolute values or precise timing.
Within the same hardware platform — two EVO4S units or two VBOX units of the same variant in identical mounting positions — G channel comparison is valid. Comparing a 10 Hz VBOX against a 25 Hz VBOX GNSS variant introduces an additional variable in transient resolution even though both are VBOX systems. See the Data Channel Reference for full detail on cross-hardware comparison.
Test Day and Setup Process Errors
Mistakes made before the analysis starts - in how the test day is structured, how setup changes are evaluated, and how consumable degradation is tracked.
The mistake
The most expensive mistake in club racing. If the lateral G trace shows the driver is consistently at 80% of available grip mid-corner - across multiple laps, in multiple sessions - changing the front anti-roll bar will not fix it. The car has more grip available than the driver is using. Setup changes in this situation compound technique issues by changing the reference feel the driver is already working from.
The honest benchmark is the reference lap from the same car, same corner, same day - not an absolute G ceiling. Corner radius, banking, fuel load, tyre age, pressure, ambient temperature, and logger mounting all move the peak Ay number around. A driver who is materially below the lateral G shown by the benchmark lap at the same corner in the same conditions has grip available. That is a technique problem. No setup change addresses it.
The correct approach
Establish whether the driver is at the grip limit before any setup discussion begins. Peak lateral G at medium-speed corners is the primary check. If the driver is more than 15-20% below the expected ceiling for the tyre and conditions, setup work is premature - coaching work from the Lap Analysis Workflow produces more lap time than any setup change at this stage.
The setup discussion becomes valid when the driver is consistently within 10-15% of the grip ceiling and the channel traces show they are at the limit in a specific phase of specific corners. Below this threshold, the data is telling you the driver needs coaching, not the car needs adjusting.
The mistake
A setup change made between qualifying and the race - or between sessions at a race weekend rather than a dedicated test day - is made without the minimum three clean laps needed to validate it. The driver goes into the race on a setup that has not been confirmed in the data. If the race is then slower than expected, it is impossible to tell whether the track conditions changed, the tyres were in a different state, or the setup change made things worse.
This is particularly common with tyre pressure adjustments made on a feeling from qualifying. The feeling may be real. It may also be track temperature, traffic, or the driver's own warm-up routine.
The correct approach
At a race weekend, setup changes between sessions should be conservative and based on channel data from the previous session - not on driver feel alone. If the data from qualifying clearly shows a specific setup issue (understeer plateau at a specific corner, exit instability at a specific corner), one targeted change is justified. Two or more changes are not - there is no time to validate them.
The discipline: any change made without a validation run is a hypothesis, not a confirmed improvement. Enter the race knowing the change is untested and drive accordingly for the first lap until the car's behaviour is understood. The full test day methodology is in Part 3: Setup Correlation.
The mistake
A setup change that fixes the worst sector will often make two other sectors slightly worse. Stiffening the front ARB to gain rotation at slow corners increases understeer at fast corners. Adding rear toe-in for stability on the back straight costs rotation at the hairpin. If the analysis only looks at the target corner, the net lap time effect of the change appears positive - until the sector delta for the whole lap is calculated and the gains are smaller than expected, or absent entirely.
This is also a race strategy question. Which corners are the overtaking opportunities on this circuit? A setup that costs time at a corner where no-one ever passes may be worth keeping if it gains time at the corner before the main straight.
The correct approach
Always calculate the full-lap sector delta before and after a setup change - not just the delta at the target corner. A change that gains 0.15 s at corner 3 but costs 0.08 s at corners 6 and 9 is a net gain of only 0.07 s, not 0.15 s. The decision whether to keep the change depends on the net figure and on circuit-specific context.
Understanding the compromise is as important as finding the improvement. Document which setup works at which circuit and why - a front ARB position that is correct at Silverstone National may not be correct at Brands Hatch Indy. The data from each event belongs in a circuit-specific setup file, not a single universal sheet. See Part 3: Setup Correlation FAQ Q5 for the full approach.
The mistake
A test day that evaluates three setup changes across four runs may actually be evaluating three setup changes plus progressive tyre degradation - and attributing the degradation effect to the setup changes. Lap 3 of run 1 on fresh tyres and lap 3 of run 4 on tyres with 20 additional laps are not the same comparison, even if the session conditions are identical.
Brake and disc condition has a similar effect. As pad and disc temperature, friction characteristics, and fluid condition change through a day, the driver may feel a different initial bite, pedal travel, or stability under braking. That can move braking points even when technique has not changed - and it shows in the speed trace as a shift in brake application point that has no corresponding change in entry speed.
The correct approach
Record tyre lap count at the start of every run. Record tyre temperatures at all four corners within 90 seconds of the car stopping after every run - inner, middle, and outer edge. A tyre that is going off shows a characteristic pattern in lateral G: peak Ay drops progressively lap-to-lap even with identical driving technique. If you see this pattern, note it as a tyre degradation effect before attributing it to the setup change being evaluated.
For brake wear, note the pedal feel and braking point consistency in the driver debrief after each run. A significant change in braking point - visible in the speed trace - without a corresponding change in entry speed is a brake condition indicator, not a technique regression.
The mistake
In championships with limited tyre allocation — Boxster Cup drivers are limited to a maximum of two sets of Pirelli Trofeo R (4 front, 4 rear) for the entire season (Reg 44.3) — the performance delta between a new set and a well-used set is significant and is rarely accounted for in the data analysis. A driver who sets their best lap time on new tyres in qualifying and then races on a set with 30 laps on it may see 0.8–1.2 s per lap difference that has nothing to do with driving or setup.
More importantly, the car may need a different setup to extract the best from a worn tyre. A worn tyre runs hotter and has less lateral stiffness - the optimal camber, pressure, and ARB balance for a worn tyre may differ from the new tyre optimum.
The correct approach
Track tyre age on every set and include it in the session log alongside lap times and setup notes. When comparing lap times across sessions, note the tyre age at the start of each run. In AiM Race Studio, look at peak lateral G on the known worn set versus the new set at the same corner - the worn set will show a lower, earlier Ay peak before the tyre overheats and loses grip.
With a worn tyre, avoid default assumptions about setup direction. Look first at hot pressures, shoulder temperature spread across the tread, steering response from the driver, and the shape of the lateral G trace. Some worn sets want more pressure support as the carcass softens; some want less as heat build increases. Some want more front platform; some want more compliance. The data and the temperature spread tell you which way to go - a recipe does not. Confirm any adjustment with lateral G at the front-limited corners before and after, with tyre age and hot pressures recorded for both runs.
The mistake
When looking at tyre data or trying to understand why lap 1 is slower than lap 3, the outlap is almost always omitted from the analysis sequence. This gives an incomplete picture of how the tyre was being managed and makes it impossible to compare tyre strategies between drivers or between runs.
A driver who pushes hard on the outlap scrubs heat into the tyre faster and benefits earlier in the run - but may have compromised tyre life. A driver who runs a conservative outlap may look slow on laps 1 and 2 but be quicker on laps 5 and 6. Without the outlap in the dataset, neither strategy is visible.
The correct approach
Always include the outlap in the analysis sequence when evaluating tyre behaviour. In AiM Race Studio, start the lap sequence from the pit lane exit - not from the first timed lap. The lateral G and tyre temperature progression from the outlap through the first three timed laps tells you everything about how the tyre is being brought in and whether the strategy is consistent lap-to-lap.
The outlap is also where brake temperature management begins. A driver who brakes hard on the outlap to build brake temperature faster may be doing so deliberately - the longitudinal G trace on the outlap will show this clearly and it should be noted in the session debrief as intentional, not as a technique error.
Data Interpretation Errors
Mistakes made in the debrief - reading the channels incorrectly, drawing wrong conclusions from valid data, or missing real events because of how the data is displayed.
The mistake
A spike on the lateral G trace at the apex of a corner could mean the driver corrected a slide, the car hit a kerb, the IMU moved in its mount, or the data file has a corruption artefact. Opening the channel traces without watching the video first means drawing conclusions from a spike without knowing what actually happened. The conclusion may be completely wrong.
This error is compounded when the channel analysis produces a coaching point - "you corrected a slide at corner 7" - that the video would have immediately shown was actually a kerb strike, not a handling event.
The correct approach
Watch the video from the fastest lap before opening a single channel trace. Note the moments where something happened - a correction, a kerb, a piece of traffic, a moment of hesitation. These are your anchors for the channel analysis. When the Ay trace shows a spike, you already know whether it was a correction or a kerb because you saw it.
The VBOX HD2 video overlay with channel data burnt in - speed, G, lap time - exists precisely for this reason. It is the ground truth against which every channel is validated. Data does not tell you what happened. It tells you what the sensors recorded. The video tells you what actually happened.
The mistake
The most common misread in club racing data analysis. Distance marker alone tells you almost nothing useful about braking technique. A reference lap that shows a braking point further from the corner does not mean the driver simply "braked earlier" - they may be arriving faster and needing more distance to scrub the speed, braking more efficiently and reaching the same apex with less distance used, or accepting a higher minimum speed through the corner altogether. Three completely different situations, identical on a distance marker comparison.
Looking at the distance marker alone produces wrong coaching conclusions. "You need to brake later" or "you need to brake earlier" both miss the point if entry speed and braking efficiency have not been checked first.
The correct approach
Always check three things together before drawing any conclusion about a braking point comparison: entry speed at the point of brake application, brake onset and release shape in the longitudinal G trace, and minimum corner speed at the apex. The distance marker is the last thing to look at, not the first.
Trace the speed delta back to where it originates. It is frequently not at the corner where the sector time shows the loss - a driver who is slow out of corner 3 will arrive at corner 4 at a lower speed and appear to brake from a different position, even if their braking technique at corner 4 is identical to the reference. The full methodology is in step 3 of the Lap Analysis Workflow and FAQ Q2 and Q4.
The mistake
AiM Race Studio and VBOX Tools both apply display smoothing to channel traces by default. At the standard smoothing level this is appropriate - it removes high-frequency IMU noise and makes the traces readable. But increasing the smoothing level to make the traces look cleaner also removes real events. A brief mid-corner correction that appears as a small spike in the raw Ay trace disappears entirely under heavy smoothing. The debrief then misses a real handling event that happened on every lap.
The opposite error also occurs - turning off smoothing entirely and then trying to interpret raw 200 Hz IMU data as if every oscillation is a real event. Most of the high-frequency content at 200 Hz is road surface and mounting vibration, not vehicle dynamics.
The correct approach
Keep sample rate and display smoothing conceptually separate. Log at the native channel rate you need for the analysis - see the Data Channel Reference for recommended rates per channel. For debrief work, use light to moderate display smoothing rather than raw data or heavy filtering.
If an event appears in the raw trace and survives light smoothing, and lines up with something visible in the video, it is likely real. If it vanishes with modest smoothing, it is probably noise from the IMU mount or road surface. When a specific event needs closer inspection, zoom in on the time axis rather than removing smoothing - you get more temporal resolution on the event without reintroducing the noise floor of the raw data.
The mistake
GPS speed and wheel speed are not the same channel. GPS speed is ground speed - it does not know about wheelspin. Wheel speed measures hub rotation and converts to road speed using a tyre rolling circumference constant entered at setup. If that constant is wrong, every wheel speed reading is systematically wrong by the same proportion across the entire session - and every derived calculation (braking point distance, sector time via wheel speed) inherits that error.
The two channels are often displayed together in AiM Race Studio and can look nearly identical at first glance. The difference only becomes apparent under acceleration or when a circumference error has been made.
The correct approach
At constant speed on a straight, GPS speed and wheel speed should agree within 1-2 km/h. A consistent offset everywhere - including at constant speed - is a wheel circumference constant error. A divergence that only occurs under acceleration is real wheelspin. Do not confuse the two.
The full GPS vs wheel speed diagnostic, including how to calculate and correct the circumference constant, is covered in the Configuration Pitfalls section of the Data Channel Reference. This is one of the most common pre-session configuration errors and one of the least visible - the data looks valid until it is cross-checked.
The mistake
A session analysis that produces a coaching point - "you lose 0.3 s at corner 7 every lap" - is valid only if the session it was drawn from was representative. A session run in cooling conditions at the end of a day, with green track, after a safety car period, or as the first session of a wet-to-dry transition is not representative. The coaching conclusion from that session may be correct but the magnitude of the deficit is wrong - and adjusting setup or technique based on the wrong magnitude produces the wrong response.
The correct approach
Before using a session's data as the basis for coaching or setup conclusions, check whether the session was representative. The fastest lap in the session against the circuit benchmark is the quick check - if the session was 2 s off the record in conditions that should be representative, something unusual was happening. Check lap time progression through the session - a session where lap times improved continuously was a green track session, not a steady-state performance measurement.
The most reliable coaching and setup data comes from the middle laps of a settled session - after the track has rubbered in, after the tyres have come to operating temperature, and before degradation begins. Use those laps. Everything else is context, not data.
Reference Lap Questions
The questions that come up when a driver first uses a coached reference lap for analysis.
This is the right question and it matters. A reference lap from a driver with a very different style - earlier braking, later throttle, different line - produces a speed trace that looks different even if the lap times are similar. Comparing technique directly against a different style risks coaching toward the reference driver's approach rather than toward the fastest approach for this driver in this car.
The solution is to focus on outcomes rather than technique. Compare apex speeds, sector times, and minimum corner speeds - not braking points and throttle application points. A driver who arrives at the same apex speed via a different route is not doing anything wrong. The apex speed is the outcome. The technique that produces it is secondary.
Where a reference lap from an instructor like Colin is particularly useful is in exposing outcomes that differ. If his apex speed at corner 7 is consistently 8 km/h higher than yours across multiple laps, that is a real gap regardless of how either of you got there. The technique discussion starts from that gap, not from a comparison of braking points.
Three checks. First, the lap time - does it sit within 0.3-0.4 s of the fastest lap of that session? A reference lap that is significantly slower than the session best was not a maximum effort lap and the channel data will reflect that.
Second, the speed trace - is it smooth and continuous with no obvious flat spots, abrupt speed changes, or anomalous sections? A flat spot in the speed trace mid-corner usually means traffic was encountered or the driver braked to avoid something. That lap is not clean for the affected corner.
Third, the lateral G trace - are the Ay arcs smooth and consistent with the corner geometry? An Ay spike or early drop mid-corner usually means a correction event. One correction in an otherwise clean lap is acceptable for reference purposes at corners that were not affected. A lap with multiple correction events is not a valid reference.
In AiM Race Studio, overlay three laps from the reference driver and compare the speed trace. Where the three laps agree closely, the reference is repeatable and clean. Where they diverge, something was different on one of the laps and that section should not be used for analysis.
Not for direct lap time comparison - the circuits are different and the absolute times mean nothing. But a reference lap from a different circuit is useful for two specific purposes.
First, technique signature comparison. The shape of the longitudinal G trace under braking, the onset rate of lateral G at turn-in, the smoothness of the Ax trace under acceleration - these are driver characteristics that appear consistently at every circuit. A driver who uses the brakes as an on/off switch at Brands Hatch will show the same signature at Silverstone. Identifying the technique signature from one circuit confirms what to look for at every other circuit.
Second, the theoretical fastest lap structure. If a driver consistently has a weak sector 2 at every circuit they attend - not just one - that is a systemic pattern rather than a circuit-specific issue. Cross-circuit analysis exposes systemic patterns that circuit-specific analysis misses.
Start with the sector that shows the largest delta and work inward from there. A reference driver who is faster everywhere is almost always faster by different amounts in different sectors - there is always one sector where the gap is largest. That sector contains the highest-value coaching opportunity.
Within that sector, look at minimum corner speed at each corner. The corner with the largest minimum speed deficit is the primary coaching target. One corner. One conclusion. Do not try to explain the whole lap at once.
The discipline that prevents analysis paralysis: you are allowed one coaching point per session. The reference driver may be 2.5 s faster. Your job at the end of the next session is to be 2.0 s slower, having closed the gap at one specific corner by working on one specific thing. Trying to close 2.5 s in one session by addressing everything at once closes nothing. The structured approach from Part 1 applies regardless of how large the gap is.
Yes - one session is enough to identify the primary gap and the primary coaching target. The reference lap comparison does not require multiple sessions of data to be useful. What it requires is a clean reference lap and a minimum of three clean laps from the driver being analysed in the same conditions.
With one session, the conclusions you can draw are: where is the biggest sector delta, what is the minimum corner speed gap at the worst corner, and is the primary issue at entry, apex, or exit. Those three conclusions are enough to set one coaching target for the next session. That is exactly what the analysis should produce.
What you cannot do reliably from one session is distinguish a systemic pattern from a circuit-specific issue, evaluate whether a setup change worked, or draw conclusions about tyre management strategy. Those require multiple sessions. For an initial driver assessment from a single session, the reference lap comparison gives you everything you need to start.