Sim racing can transfer useful skills to a real racing car. Max Verstappen and Chris Lulham provide recent examples, while motorsport simulator research shows that drivers adapt their control strategies to the cues available in a virtual environment. What public evidence cannot yet tell us is whether the fastest technique in iRacing, Le Mans Ultimate, Assetto Corsa Competizione or Assetto Corsa EVO matches the technique used in the corresponding real car.
There is good support for the transfer of circuit knowledge, setup understanding and aspects of driving technique. There is also strong reason to expect differences. A static home simulator removes much of the physical information available inside a racing car, particularly sustained acceleration and G-forces.
The unanswered question is more specific:
When a driver becomes fast in a simulator, what control strategy has that simulator taught them to use?
Driving behaviour can be measured, not just described
A 2020 study from Graz University of Technology examined behavioural validity in motorsport Driver-in-the-Loop simulation by comparing professional racing drivers on a real circuit and in a simulator.
The researchers used metrics derived from track and simulator data to quantify the distance between the driver’s behaviour in both environments. Absolute behavioural validity ranged from 40.47% to 59.29% across the drivers and simulator configurations tested. Activating the simulator’s motion system shifted behaviour by between 1.29% and 3.31% towards the characteristics measured on track.
Those percentages are measures produced by that specific methodology. They are not a percentage score for how “realistic” the simulator was. The useful finding is that driver behaviour changed with simulator configuration and that change could be quantified.
A 2026 study shows how drivers adapt to simulator cues
Further research published in Automotive Innovation in February 2026 examined motion cueing using amateur participants and two professional racing drivers in a four-degree-of-freedom simulator. The researchers assessed lap time and errors alongside objective driving characteristics such as steering wheel reversal rate and full-throttle ratio.
The professional drivers showed consistent preferences for the same motion cueing configuration, while preferences among amateur participants were distributed more widely. The researchers concluded that professional drivers with an adequate real-world reference are particularly valuable when tuning motorsport simulator motion systems.
The paper also discusses previous research showing that drivers can develop an internal model of scaled or filtered simulator cues and adapt their control actions accordingly.
For sim racing, that has a clear implication. After hundreds or thousands of hours, a driver learns what a particular reduction in steering torque means, how much visual movement precedes a loss of grip and what tyre sounds indicate that the car is approaching the limit.
That adaptation can be extremely effective without being identical to what happens inside a real racing car.
Max Verstappen says G-forces remain the biggest difference
Max Verstappen has extensive experience in both professional motorsport and sim racing, and his description of the difference is useful because he separates physical sensation from other areas of preparation.
In January 2025, Verstappen described G-forces as the biggest difference between home sim racing and driving a real car. Acceleration, braking and cornering all produce forces that he does not experience on his static simulator.
He also identified setup work and race strategy as areas with much greater similarity. His often-quoted estimate that sim racing can be “90, 95%” similar was made in that broader context, including setup, strategy and decision-making. It should not be interpreted as a claim that a consumer simulator reproduces vehicle physics with 90 to 95% accuracy.
Thousands of virtual Nürburgring laps transferred into useful real preparation
The Nürburgring Nordschleife provides a particularly clear example of transfer. Before driving the circuit in a real GT3 car, Verstappen had already completed thousands of simulator laps.
After testing a Ferrari 296 GT3 at the Nordschleife in May 2025, Verstappen explained that the virtual preparation had helped considerably. He already knew the circuit layout and could concentrate on the differences that appeared in reality, including grip levels, track conditions, asphalt, barriers and kerbs.
Before his 2026 Nürburgring 24 Hours programme, Verstappen again described simulator work as a central part of his preparation, after completing thousands of virtual laps and several simulated 24-hour races.
Once in the real car, his focus shifted to real grip levels, kerbs, G-forces and compressions. That is a concrete example of useful sim-to-real transfer. Circuit knowledge arrived already developed. The physical reference still had to be learned in the car.
Chris Lulham says the technique transfers better than the physical experience
Chris Lulham offers a different perspective because a large part of his development came through sim racing before moving into professional GT competition.
Ahead of his 2025 GT World Challenge programme, Lulham described the physical experience of driving the real car as very different while saying that the technique itself was very similar.
He specifically identified G-forces as the main difference and said that techniques learned during four years with Team Redline could be applied at the real circuit. His first GT World Challenge Endurance race exposed another set of differences.
In his post-race assessment, Lulham identified traffic management, the physical intensity of racing, tyre pickup and dust as important parts of his learning process. Those variables could significantly change the balance and feeling of the real car.
Core driving techniques can transfer well while real-world conditions still require substantial recalibration.
Lulham’s progression shows that sim-to-real transfer can reach professional GT racing
By 2026, Lulham was competing in the Pro Cup of GT World Challenge Europe for Mercedes-AMG Team Verstappen Racing.
At Brands Hatch in May, Lulham and Dani Juncadella finished second in Race 2, a result recorded in the official GT World Challenge Europe results. This does not validate a particular simulator physics model. It does establish that a driver whose development included substantial sim racing experience has successfully transferred into high-level professional GT competition.
Verstappen looks beyond raw sim racing lap time
Verstappen has also explained what he considered important when evaluating Lulham for the move into real racing.
He acknowledged that many sim racers can produce a fast lap, then pointed to other abilities: understanding setup changes, adapting to different conditions and switching quickly between different cars.
Those were areas in which Verstappen considered Lulham particularly strong.
This suggests a broader way of thinking about sim-to-real transfer. Raw pace is one part of the skill set. Adaptability, technical feedback and understanding why a setup change affects the car may also transfer.
What can we verify about iRacing?
Any investigation into iRacing needs to account for the fact that its vehicle models continue to change. For 2026 Season 3, iRacing substantially updated the traction-control system used by its GT3 class.
According to the official 2026 Season 3 release notes, the TC parameters were tuned using feedback from a real-world professional GT driver. ABS parameters were also updated. The same build enabled series-specific sporting regulations for GT3, including rulesets that can disable TC, ABS or both.
Later in the season, Patch 3 continued making corrections to GT3 traction-control reporting and vehicle setups. These development notes tell us what iRacing changed. They do not provide a public real-car telemetry comparison showing that iRacing requires more trail braking, a faster brake release or a particular amount of combined braking and steering.
That leaves a testable question:
Does the fastest iRacing lap require the same brake release as the real car?
A proper comparison would record physical pedal force, brake pressure at steering onset, release rate, brake and steering overlap and the point at which braking reaches zero.
Le Mans Ultimate has changed its tyres using real drivers and data
Studio 397 states on the official Le Mans Ultimate physics overview that the simulator uses an advanced brush tyre model designed to represent slip angles and effects such as flat spots.
The physics engine runs at 400 Hz, with that output translated directly into force feedback. The simulator also uses Real Road 2.0 for track evolution. In February 2025, Studio 397 announced a revised LMGT3 tyre model developed using real-world drivers and data.
The update changed load sensitivity, the slip curve and transient thermal grip loss. Studio 397 said feedback from its real-world driver testers had focused on creating a more stable and predictable operating window.
The development continued into 2026. Le Mans Ultimate 1.4 was released on 28 July 2026. Its official release notes include adjustments to Hypercar cold-tyre grip intended to better match the 2026 real-world compounds, along with further tyre, vehicle and Balance of Performance changes. None of these official notes establish that LMU makes a driver correct the steering more frequently than in reality. That can be investigated directly.
Steering wheel reversal rate, already used in professional simulator research, would provide a useful starting point. The same driver’s steering activity could be compared between LMU and the real car through similar corners.
ACC gives us a controlled GT3-focused platform
Assetto Corsa Competizione was developed around GT World Challenge racing.
On its official ACC product page, Kunos states that the physics model simulates tyre grip, aerodynamics, engine parameters, suspension and the electronic systems that determine vehicle balance. That makes ABS and traction control natural areas for a sim-to-real comparison.
Kunos described that release in its official 2025 Season Update announcement as a content update adding new GT World Challenge liveries and updated drivers rather than a major physics rewrite.
The interesting experiment is straightforward:
Does a professional GT driver use ACC’s ABS and traction control in the same way they use those systems in the real car?
The useful measurements would include brake force, time at peak pressure, ABS intervention, throttle position at first TC intervention, partial-throttle duration and the timing of full throttle.
Assetto Corsa EVO requires version-specific conclusions
Assetto Corsa EVO remains in Early Access, so its physics cannot be treated as a fixed target.
Version 0.4 introduced a new tyre damping method, revised tyre inertia and slip ratio, expanded thermal and wear behaviour and a grip model with progressive combined-slip limits, according to the official 0.4 update notes.
Update 0.6, released on 15 April 2026, added the Ferrari 296 GT3 and further revised suspension modelling. Kunos specifically described improvements to the way cars communicate load transfer, attitude and grip. The update also expanded telemetry support.
Those changes are documented in the official Assetto Corsa EVO 0.6 announcement. The current major build is Early Access 0.8, released on 8 July 2026. Any telemetry test performed now should therefore identify its build explicitly.
David Perel provides a real-driver reference inside AC EVO
Kunos has also published an official Assetto Corsa EVO lap guide with professional racing driver David Perel.
At Donington Park, Perel discusses braking points, positioning the car, sacrificing one part of a sequence to improve the next and rotating the car early to find grip through Melbourne Hairpin and Goddards. The guide is useful because it shows a professional driver applying recognisable real-world driving concepts inside the simulator.
It still leaves the quantitative question open. Telemetry would be required to determine whether the amount of braking, steering and resulting yaw matches the physical car.
What a SIM TO REALITY test should actually measure
The available evidence points towards a test based on driver behaviour rather than simulator reputation.
The same professional driver should use a real car and the closest practical virtual equivalents. Hardware, seating position, field of view and control calibration need to remain as consistent as possible. The brake pedal deserves particular attention.
A 100% brake value in simulator telemetry is a calibrated digital input. It cannot be assumed to represent the same physical pedal force used in the racing car.For that reason, the experiment should measure the force applied by the driver’s foot alongside the simulator’s brake channel.
The 2026 Driver-in-the-Loop research also provides a reason to give the driver sufficient adaptation time. Drivers can learn simulator-specific cues and adjust their control strategy, so the first few laps would mix technique with unfamiliarity.
Brake release may be more revealing than braking distance
For iRacing, one of the most useful comparisons could be the shape of the braking trace during corner entry.
Measurements should include:
- Physical brake-pedal force
- Peak brake input
- Brake pressure when steering begins
- Brake-release rate
- Brake and steering overlap
- Brake input at or near the apex
This would allow the test to answer whether a fast iRacing lap actually requires a different brake-release strategy.
Until that comparison exists, claims that iRacing “teaches more trail braking” remain unverified.
LMU can be tested through steering corrections
Le Mans Ultimate’s official documentation gives us information about the tyre model, physics frequency and development process. It does not tell us whether the resulting fastest driving technique requires more steering activity.
A useful comparison would record:
- Steering angle
- Steering rate
- Number of direction reversals
- Countersteer events
- Yaw rate
- Sideslip angle where reliable telemetry is available
The steering wheel reversal rate used in the 2026 academic study gives this part of the experiment an established research reference.
ACC can be tested through ABS and traction control usage
ACC’s GT3 focus makes electronic driver aids a logical target.
The test could compare how much time the professional driver spends in ABS, how quickly brake pressure is released after intervention and how throttle application interacts with traction control.
The result would answer a useful question without assuming the outcome:
Does the simulator reward the same relationship with ABS and TC as the real GT3 car?
AC EVO can be tested through rotation and combined inputs
Kunos’ documented changes to combined slip, suspension and load-transfer communication make vehicle rotation an obvious measurement area for AC EVO.
The comparison could examine yaw rate relative to brake release, throttle and steering angle.
If the professional driver generates a similar vehicle response with similar inputs in the simulator and real car, that would be meaningful evidence.
A different input pattern would identify an area for further investigation without requiring a subjective verdict about which one “feels” better.
Two driving modes could expose simulator-specific technique
A useful experiment would ask the driver to complete two separate stints.
- REAL STYLE: drive using the technique they would normally use in the physical racing car.
- SIM MAX: adapt fully to the simulator and use whatever technique produces the fastest repeatable lap.
The difference between those traces would show whether optimising for the software requires a change in behaviour.
If the inputs remain similar, the behavioural correlation is strong in that area. If the driver finds time by changing brake release, steering activity, slip or use of electronics, the telemetry shows exactly where the virtual optimum has moved.
Where does the driver release the brake? How much brake remains as steering begins? How often is the wheel corrected? When does full throttle arrive? How much does the driver rely on ABS and TC? Which of those inputs change when the driver stops following the real-car reference and starts optimising for the simulator?
Those measurements can show what each simulator changes about the way the driver controls the car.
See you on the track!
This website uses affiliate links which may earn a commission at no additional cost to you.
