[1] Talent Identification and the Bypassing of Traditional Feeder Series
Esports competitions operate as the primary recruitment pipeline for modern factory racing programs, bypassing the capital-intensive karting and junior formula categories. Sim-native drivers demonstrate immediate real-world pace because modern simulation accurately replicates physical vehicle dynamics.
Nissan established the precedent in 2008 with the GT Academy, transitioning drivers from Gran Turismo directly into the physical Nissan Micra Cup and eventually elevating drivers like Thanaroj Thanasitnitikate and Abhinay Bikkani to international podiums.2 Porsche subsequently formalized this pipeline by designating the Porsche Coanda Esports Racing Team as its third official factory motorsport squad, operating alongside its Penske sportscar (LMDh) and TAG Heuer Formula E teams.36
The Porsche program operates from the 324-square-meter Porsche Esports Performance Center (PEPC) located at the Trilux Light Campus in Cologne-Ossendorf.38 The facility houses six professional racing simulators alongside workstations for performance engineers, isolating drivers from light reflections and background noise to optimize cognitive processing.38 Drivers including Joshua Rogers, Mitchell deJong, Mack Bakkum, Tommy Østgaard, and Martin Krönke operate as contracted Porsche Esports works drivers, mirroring their 2020 virtual 24 Hours of Le Mans victory where real and virtual professionals shared a single vehicle entry.37
Individual career trajectories confirm the absolute transferability of sim racing skills to physical racecraft. Rajah Caruth transitioned from the iRacing Ignite Series to the NASCAR Craftsman Truck Series.51 Discovered by NASCAR’s Drive for Diversity program at age 16, Caruth was placed in a physical car at the Bojangles Summer Shootout, immediately winning two heat races.51 By 2024, Caruth secured a historic victory at Las Vegas Motor Speedway in the Truck Series, establishing a direct link between virtual training and national touring series victories.53
Similarly, James Baldwin transitioned from winning the 2019 World’s Fastest Gamer championship directly into a full season of the British GT Championship with Jenson Team Rocket.21 In his physical debut season, Baldwin secured a race win, four pole positions, and three podium finishes.21 He currently serves as a Simulator Driver for the Mercedes-AMG PETRONAS F1 Team while simultaneously competing in the physical 24 Hours of Spa and securing class wins in the GT World Challenge Europe at Jeddah.21
[2] Hardware Architecture: Motion Fidelity and Vestibular Cues [source]
The physical architecture of a simulator dictates the volume of transferable skill. Commercial and entry-level simulators rely on linked actuators (Stewart platforms), which introduce mechanical lag and artificial motion by failing to rotate around the vehicle's true center of mass.28
Professional platforms isolate and replicate exact rigid body dynamics. SimCraft’s 6-axis simulators utilize independent degrees of freedom—yaw, pitch, roll, heave, sway, and surge—aligned precisely at the cockpit's center of mass.28 This localized rotation allows drivers to accurately sense early oversteer, weight transfer, and grip limits. Research conducted by the Spartan Motorsport Performance Lab at Michigan State University verified via EEG-based measurements that this specific combination of physical, auditory, and visual stimuli generates the correct cognitive feedback to improve real-world driving focus and performance.28
The Dallara simulator in Speedway, Indiana, represents the apex of current motion capabilities. Developed initially to support aerospace contracts with Raytheon, the $12 million machine mounts a monocoque cockpit atop six independently actuating hydraulic arms.41 Professional IndyCar and IMSA teams rent the facility for $12,000 per day to simulate specific aerodynamic setups, spring rates, and shock absorbers.42 The system includes high-fidelity dome projection, steering wheel control loading, and real-time Moog cueing software, operating with ultra-low latency to eliminate the delay between driver input and chassis response.45
[2] 1. Proprioception and Reaction Differentials [source]
The human vestibular system, which detects spatial orientation, processes physical vehicle sliding faster than the visual system.26 In simulators lacking 6-axis motion, drivers rely entirely on visual and steering force-feedback cues, limiting their ability to catch rapid yaw state changes.
A study comparing professional racing drivers and non-racing drivers in simulators recorded their responses to induced road departures. Non-racing drivers experienced a significantly higher absolute vehicle yaw rate (272 deg/s, SD = 172 deg/s) compared to professional racing drivers (124 deg/s, SD = 67 deg/s).27 The professional drivers naturally minimized the yaw rate despite the lack of physical motion, indicating a highly developed cognitive compensation mechanism for missing vestibular cues.27
Braking mechanics represent the most critical hardware crossover between simulation and the physical track. Load cell pedals—which measure applied force rather than physical position—replicate the pressure-to-deceleration relationship of a real brake pedal.15 Sim Coaches replaces standard strain gauges with proprietary hydraulic pedals, directly mirroring the fluid displacement and pressure buildup required to trail-brake effectively and manage forward pitch in GT3 and formula cars.15
[3] Economics of Virtual Homologation and DIL Systems
The deployment of driver-in-the-loop (DIL) simulators generates a massive return on investment (ROI) by offsetting the severe capital requirements of physical testing.
Formula 1 teams are restricted by FIA regulations to just three days of pre-season testing and two 200-kilometer filming days annually.108 Furthermore, the FIA imposes strict limits on computational fluid dynamics (CFD) and wind tunnel sessions, measured in Mega Allocation Unit hours (MAUh), allocated on a sliding scale based on championship position.103 Physical track time is entirely reserved for validation; all experimental setup exploration occurs in the simulator.
Entry-level professional F1 simulator systems begin at approximately $3.2 million (£2.5 million), with bespoke systems reaching tens of millions of pounds.108 Companies like Vesaro provide commercial crossover units ranging from £12,000 to £30,000, while full-chassis replicas scale to £150,000.112 Despite these capital costs, DIL simulations allow vehicle constructors to evaluate thousands of parameters—suspension geometries, aerodynamic maps, and hybrid deployment strategies—reducing hundreds of potential test candidates by an order of magnitude before a single physical component is manufactured.46
[3] 1. Market Growth and Infrastructure
The global racing simulator market was valued at $0.98 billion in 2025 and is projected to reach $2.04 billion by 2030, representing a compound annual growth rate (CAGR) of 15.78%.50 Research and Development organizations—including universities, national laboratories, and autonomous vehicle (AV) startups—account for 21.7% of total market revenue, utilizing simulators for sensor fusion research and vehicle control algorithms.48
The commercialization of sim racing centers further democratizes access to this technology. Operating an 8-to-13 rig commercial venue requires $3,000–$8,000 for commercial general liability insurance, $500–$2,000 per simulator annually for software licensing (e.g., iRacing, Assetto Corsa Competizione), and $20–$50 per square foot for facility fit-outs.109
[3] 2. Motorsport Workforce Restructuring
The economic efficiency of simulation extends to team personnel. The modern motorsport workforce has restructured, heavily prioritizing virtual engineering and data science over traditional mechanical roles.
| Engineering Role | Salary Range (UK) | Salary Range (US) | Primary Responsibilities |
| Race / Track Engineer | £41,800 - £146,200 | N/A | Track-side execution, homologation compliance, driver communication. |
| Aerodynamicist | £35,000 - £162,000 | $85,000 - $125,000 | Wind tunnel testing, CFD design, aerodynamic surface generation. |
| Simulation / Data Engineer | £32,000 - £86,000 | $103,500 - $150,000 | Virtual testing, mathematical modeling, algorithm development, DIL operation. |
| Performance Engineer | £35,000 - £90,000 | $85,000 - $130,000 | Vehicle optimization, dynamics analysis, setup generation. |
| Race Mechanic | £28,000 - £95,000 | N/A | Physical assembly, teardown, and vehicle maintenance. |
Table 1: Salary benchmarking for modern motorsport roles, demonstrating the premium placed on simulation and aerodynamic expertise.56
Race weekend roles command a 15-20% premium over factory-based counterparts.56 Furthermore, contract Simulation Development Engineers working on specialized programs command day rates equivalent to annual salaries of £150,000 to £200,000.56 These roles demand deep expertise in multi-body dynamics, 1D modeling, and the translation of complex tire coefficients into real-time executable code.60
[4] Digital Twins and Synthetic Environments
The precision of the virtual environment strictly limits the effectiveness of the physical hardware. Simulation software provider rFpro builds engineering-grade digital twins using survey-grade LiDAR scan data, creating track and road surface models accurate to within 1 millimeter.13
This 1mm precision captures exact track imperfections, camber changes, and drain covers. For the 2024 Las Vegas Grand Prix, rFpro supplied a digital twin of the 3.8-mile, 17-corner street circuit to the majority of the Formula 1 grid.14 The model replicates dynamic night-race lighting conditions, including simulated gantry lighting, illuminated billboards, and precise sun and moon positioning.14 These exact lighting and shadow dynamics allow drivers to optimize braking and clipping points under specific low-light conditions long before arriving at the physical circuit.10
The application of digital twins extends beyond motorsport into autonomous vehicle development. rFpro maintains a library of over 130 circuits and testing facilities, including a 36km highly accurate network of Los Angeles roads, the Mcity testing facility at the University of Michigan, the Applus IDIADA proving ground, and the Nardò handling circuit.13 These models permit repeatable, deterministic safety modeling for advanced driver-assistance systems (ADAS), allowing engineers to test subtle parameter refinements—such as fluctuating road surface reflections or changes in pedestrian clothing color—without introducing real-world inconsistencies.11
[5] Physics Engines: CFD Pipelines and the Magic Formula
Aerodynamic physics are ported into simulators directly from Computational Fluid Dynamics (CFD) pipelines. Platforms including Cadence ANSA, AutoSeal, and Fidelity automate the conversion of raw CAD geometry into watertight, simulation-ready meshes.76 This workflow addresses common CAD issues—gaps, overlaps, and intersections—using Boolean operations and intelligent topology handling.76 By automating these repairs, Fidelity AutoSeal reduces geometry preparation from days to minutes; a Honda cabin-space sealing task was compressed from one week to roughly one hour.76
Modern CFD tools utilize GPU-resident solver performance to increase throughput. These systems achieve up to 9X higher throughput and 17X lower energy consumption compared to CPU-based approaches.76 When deployed on platforms like the Millennium M2000 Supercomputer, simulation workloads achieve 80X faster turnaround times.76 AI-enabled simulation intelligence further accelerates this process by utilizing surrogate-based prediction to assess drag and lift in near-real-time (under one second), reserving full high-fidelity computational solves exclusively for the most promising design candidates.76
[5] 1. Tire Dynamics and the Pacejka Magic Formula
Vehicle handling simulation is bound entirely by the mathematical modeling of the tire-road contact patch. Professional simulation relies heavily on Hans B. Pacejka’s "Magic Formula" (MF), specifically iterations like MF5.2 and PAC2002.64 The Magic Formula operates as a curve fit utilizing 15-20 specific coefficients to calculate longitudinal force (Fx), lateral force (Fy), and aligning torque (Mz) based on vertical load (Fz), slip angle (α), and wheel angular velocity (Ω).64
Standard tire models suffer from inaccuracies when tire operating conditions—such as tread depth, inflation pressure, and surface temperature—fluctuate during a stint.61 Advanced thermal tire models, developed by institutions like the University of Surrey, integrate these variables directly into the Pacejka coefficients.61 In these advanced models, both lateral and longitudinal stiffness, alongside peak grip, scale as a linear function of tire surface temperature.61 This mathematical adaptation allows the simulator to accurately predict handling degradation and thermal fall-off over an endurance stint. By applying these adaptation equations to vehicle stability control systems, engineers compensate for grip loss before the physical tire ever touches the asphalt.61
[6] Telemetry Correlation and Setup Optimization [source]
Simulators operate as live engineering tools for optimizing suspension geometry and aerodynamic maps. Physical telemetry data is continuously fed back into the simulator to verify correlation and improve future predictions.
A technical case study utilizing ChassisSim on a Formula 3 vehicle demonstrated this direct correlation mechanism. During a wet qualifying session, the physical F3 car suffered from severe resonant mode oversteer in high-speed corners.94 Engineers loaded the baseline vehicle parameters into ChassisSim and successfully replicated the precise damper oscillations and steering resonance recorded in the real-world data.94 By doubling the rear damping in the simulation, the mathematical model stabilized both front and rear oscillations, reducing the simulated oversteer and increasing corner speed from 210 km/h to 219 km/h.94 This single virtual adjustment yielded a localized lap time gain of 0.5 seconds through Goodwood corner, confirming the exact setup change required for the physical car.94
Aerodynamic optimization follows a similar virtual protocol. In a Formula Student case study utilizing IPG CarMaker 11.0, teams ran 3D steady CFD simulations to generate drag and lift coefficients across various crosswind directions.93 By modeling the physical mass and activation delay of a Drag Reduction System (DRS), the team simulated a 25% reduction in total aerodynamic drag.93 Lap-time simulations identified that increasing the base negative lift force by 15% yielded the largest overall lap time reduction, specifically maximizing gains in Sector 1 without compromising traction in slower sectors.93
[6] 1. Driver-Specific Mechanics and Telemetry Patterns [source]
Telemetry exposes the distinct mechanical behaviors of individual drivers, proving that top-tier sim racers and F1 champions manipulate physics similarly. Analysis of Max Verstappen’s steering inputs and throttle traces reveals a driving profile distinct from the rest of the Formula 1 grid.6 Verstappen consistently brakes later, applies throttle earlier, and manages weight transfer at the absolute limit of adhesion, entirely eliminating the standard safety margins retained by other professional drivers.7 Because simulators output identical data channels to real cars, Red Bull Racing engineers can build physical vehicle setups tailored exclusively to these specific, highly aggressive virtual telemetry patterns.6
Professional drivers employ structured methodologies to maximize simulator utility. Driver James Baldwin utilizes telemetry tools like Coach Dave Delta to overlay his simulator data against reference laps set by other professionals.22 By matching the exact virtual track state, air temperature, and wind speed of the reference lap, Baldwin isolates setup variables, allowing for direct comparison of braking points and throttle application between his sim runs and physical track data.22
BMW M Motorsport utilized a hybrid data approach to finalize the BMW M4 GT3. The manufacturer combined pure engineering data with subjective feedback from works driver Bruno Spengler—an experienced sim racer—to refine the virtual model on iRacing.113 Spengler’s feedback dictated chassis and aerodynamic optimizations, ensuring the virtual car precisely mirrored the physical handling characteristics prior to its physical race debut in the Nürburgring Endurance Series.113
[7] The Operational Role of the Simulator Driver [source]
The severe reduction in real-world F1 testing has elevated the simulator driver from a developmental role to an indispensable operational requirement. Simulator drivers, typically professional racers in other categories, conduct exhaustive setup parameter testing that race drivers lack the time to execute.
Nick Yelloly, a BMW factory driver and simulator driver for Aston Martin F1, logs 40 to 50 days per year in the F1 simulator, alongside 20 days in the BMW M Motorsport simulator in Munich.103 During race weekends, simulator drivers perform live race support, completing three-to-four hour continuous stints to test setup options requested by trackside engineers.103
Crucially, simulator drivers must possess extreme adaptability. They are required to mimic the highly specific driving styles of the team's primary race drivers (e.g., Fernando Alonso or Lance Stroll).103 If a simulator driver evaluates a setup change using a generic driving style, the resulting data will not correlate with the primary driver's physical inputs, leading trackside engineers down incorrect developmental paths.103
Teams maintain deep rosters of reserve and simulator drivers to manage this workload. For the 2025 season, Aston Martin retains Felipe Drugovich, Stoffel Vandoorne, and Jak Crawford.105 Alpine employs Jack Doohan, Ryo Hirakawa, Paul Aron, and Kush Maini, utilizing them for Testing of Previous Cars (TPC) programs and Driver-in-the-Loop simulator development at their Enstone headquarters.102 These roles frequently translate to physical race seats; Ollie Bearman substituted for Carlos Sainz at the 2024 Saudi Arabian Grand Prix, and Franco Colapinto secured a reserve role at Alpine for 2025 following nine races for Williams in 2024.102
[8] Psychological Transfer and Risk Propensity
The transferability of skill from sim racing to physical tracks relies on the brain's visual-kinesthetic processing. Motor learning research confirms that the human brain stores spatial memory regardless of whether the environment is virtual or physical.15 Cranfield University studies demonstrate that drivers who log significant simulator hours on a specific circuit arrive at the real track with materially faster initial lap times due to ingrained spatial memory regarding apex geometry, exit width, and braking markers.15
However, simulator training inherently occurs in a zero-risk environment, which fundamentally alters driver aggression and risk assessment. The lack of physical consequence allows sim racers to endlessly probe the absolute limit of adhesion without the financial or physical penalty of a crash. This builds a hyper-aggressive driving profile that translates effectively when properly managed, but requires calibration for real-world safety.
To counteract the zero-risk origin, Risk Awareness and Perception Training (RAPT) programs are utilized to calibrate risk assessment.72 RAPT teaches drivers to identify risky situations based on roadway geometry and peripheral threats. Studies tracking eye-movements reveal that drivers subjected to targeted simulator training correctly fixated on risk-relevant areas in the driving environment at significantly higher rates than untrained control groups, successfully transferring threat-detection skills from the screen to physical driving.72
[9] Insurance Underwriting and Telematics Integration
The influx of sim-trained drivers into physical race cars, combined with vehicles traveling at speeds up to 233 mph, has forced the motorsport insurance industry to overhaul its risk assessment models. Insurers such as Miller, K&K Insurance, and Safehold Motorsports underwrite physical damage, event cancellation, and participant liability.30 Standard policies covering physical damage for track days, High Performance Driver Education (HPDE), and competition events account for the immense capital risk of operating vehicles valued over $100,000.33
Historically, insurers relied on static variables—age, vehicle type, and basic track parameters—to calculate premiums.87 This legacy model is being replaced by Agentic AI systems and telematics-based driver risk propensity modeling. The global driver risk propensity modeling market reached $2.7 billion in 2025 and is scaling to $6.8 billion by 2033, driven by the integration of real-time behavioral analytics.88
Insurers now utilize usage-based insurance (UBI) models—such as Pay How You Drive (PHYD)—that parse high-frequency dynamic data generated directly from vehicle telemetry.86 Swiss Re’s ADAS Risk Score evaluates the exact performance of vehicle safety features, while advanced telematics platforms analyze braking severity, acceleration patterns, and speed management to generate granular individual risk profiles.88
To validate occupant safety in extreme physical crashes, computational human body models (HBMs) are simulated in crash environments. A Wake Forest study integrated the Global Human Body Models Consortium (GHBMC) 50th percentile male simplified occupant (M50-OS v2.2) into a simulated NASCAR environment featuring a poured-foam seat, 7-point safety belt, and head and neck restraint (HNR).89 Engineers conducted 45 simulations of real-world crashes (10 kph ≤ ΔV ≤ 100 kph) using LS-Dyna R. 9.1 to calculate AIS 1+, 2+, and 3+ injury risks for the head, neck, thorax, and lower extremities.89 These simulations provide underwriters with exact mechanical loading values and injury probabilities, ensuring that physical track limits are accurately priced by insurance actuaries.89
[10] The 2030 Regulatory Horizon [source]
The global governing bodies of motorsport are structurally mandating the continued reliance on simulation, moving the industry toward a permanently virtual-first development model.
The FIA Foundation’s "Strategy 2030" explicitly ties the future of motorsport to decarbonization, sustainable practices, and zero-risk environments.59 By setting a mandate for all FIA World Championships to achieve carbon neutrality by 2030, the governing body inherently restricts the logistical footprint of real-world testing.70 Digital Motor Sport is recognized as a core pillar to achieve this sustainability, transferring the carbon-heavy R&D process—shipping chassis, personnel, and fuel to global tracks—entirely into the digital realm.
The incoming 2030 top-class prototype regulations for the IMSA WeatherTech SportsCar Championship (GTP class) and the FIA World Endurance Championship (WEC Hypercar class) cement this methodology. Beginning in 2030, the regulations mandate a shift to a single, two-wheel drive platform utilizing internal combustion engines paired with hybrid systems, eliminating current all-wheel-drive variations.68
Crucially, the homologation for these new prototypes will be strictly enforced and valid for at least five years with no permitted performance evolutions.68 Because manufacturers will be locked into their fundamental vehicle architecture for half a decade, physical trial-and-error testing is no longer financially or competitively viable. Teams must utilize DIL simulators and 1mm-accurate digital twins to virtually homologate the chassis, aerodynamics, and hybrid powertrain integration long before a physical prototype is milled. In modern motorsport, the simulator is no longer a training accessory; it is the definitive engineering blueprint for real-world execution.
References
[2] IMSA. (2022). "Seven automakers that are serious about online racing." IMSA News. 6: YouTube. (2026). "Why Telemetry Proves Nobody Drives Like Max Verstappen." YouTube. 7: YouTube. (2026). "Max Verstappen Is NOT Just Fast — The Telemetry Proves It." YouTube. 10: rFpro. (2024). "Las Vegas F1 Circuit Digital Twin." F1 Simulator Maniac. 11: AVSandbox. (2023). "The Digital Twin." AVSandbox. 13: rFpro. (2024). "Digital Road Models." rFpro News. 14: rFpro. (2024). "rFpro Unveils Hyper-Realistic Digital Twin of Las Vegas F1 Circuit." rFpro News. 15: Sim Coaches. (2026). "Sim Racing vs Real Racing: Does Simulator Training Work." Sim Coaches. 21: Baldwin, J. (2026). "James Baldwin podcast." YouTube. 22: Coach Dave Academy. (2026). "Learn a New Track Faster Using Coach Dave Delta." Coach Dave. 26: Cruden. (2021). "Motion for immersion vs motion for vehicle dynamics." Cruden Articles. 27: NCBI. (2017). "Reaction time and visual-motor performance in racing drivers." PMC. 28: SimCraft. (2025). "How Realistic is Sim Racing Compared to Real Racing?" SimCraft Blog. 30: Miller Insurance. (2025). "Motorsport Insurance." Miller. 33: Hagerty. (2025). "Motorsports Insurance." Hagerty. 36: Porsche Motorsport. (2022). "Porsche and Coanda Esports Will Start a Long-Term Partnership." Coanda Esports. 37: Porsche Motorsport. (2022). "Porsche unveils first factory esports team." Racer. 38: Porsche Motorsport. (2023). "Porsche Esports Performance Center." Porsche Racing. 41: The Drive. (2017). "Seat Time In This Racing Simulator Costs $12,000 Per Day." The Drive. 42: PCA. (2025). "Taste of the Sim - Dallara Sim Experience." MOR PCA. 45: Moog. (2013). "Moog Driving Simulator Speeds Dallara's Race Car Development." Moog News. 46: Ansible Motion. (2015). "Costs and ROI Benefits of Driving Simulators." Ansible Motion. 48: Market Intelo. (2026). "Driver-in-the-Loop Simulation Market." Market Intelo. 50: Mordor Intelligence. (2025). "Racing Simulator Market." Mordor Intelligence. 51: Los Angeles Times. (2023). "Sim racing to NASCAR: Rajah Caruth's rise." LA Times. 53: Speedway Motorsports. (2024). "Rajah Caruth sim to reality transition." Performance Racing Network. 56: Fluid Jobs. (2024). "Motorsport Salary Survey." Fluid Jobs. 60: Motorsport Engineer. (2025). "How to become a Formula 1 Simulation Engineer." Motorsport Engineer. 61: University of Surrey. (2021). "Thermal tire model for racing applications." Arxiv. 64: FST Lisboa. (2023). "Tyre Models Extended Abstract." Fenix Tecnico. 68: IMSA. (2026). "IMSA, ACO and FIA outline plans for 2030 top prototype class." IMSA News. 70: FIA University. (2023). "Accelerating a Just Transition." FIA. 72: University of Iowa. (2023). "Risk Awareness and Perception Training Program (RAPT)." Pubs Lib. 76: Cadence. (2026). "Scaling Automotive CFD." Cadence Blogs. 86: Swiss Re. (2022). "Future of Motor Risk." Swiss Re. 87: ODSC. (2025). "Transform Insurance Risk Assessment with Agentic Systems." Medium. 88: Data Intelo. (2026). "Driver Risk Propensity Modeling Market." Data Intelo. 89: NCBI. (2020). "Computational human body models in motorsport." PubMed. 93: MDPI. (2024). "IPG CarMaker lap-time simulation and optimization." MDPI 2673-4591. 94: ChassisSim. (2025). "ChassisSim F3 Case Study." Scribd. 102: Formula 1. (2025). "All the 2025 reserve drivers for every F1 team." F1 News. 103: Racecar Engineering. (2024). "Inside the role of an F1 simulator driver." Racecar Engineering. 105: RacingNews365. (2025). "F1 reserve drivers 2025." RacingNews365. 108: F1 Briefing. (2026). "Why Teams Spend Millions on Simulators." F1 Briefing. 109: Sim Coaches. (2026). "How Much Does a Sim Racing Center Cost." Sim Coaches. 112: Reddit. (2023). "High tech sim units, what do they cost?" r/F1Technical. 113: BMW Group. (2021). "As realistic as it gets: iRacing and BMW works driver Bruno Spengler advance the BMW M4 GT3." BMW Press [source]