Sim to Asphalt: The Case For Simulation In Race Prep
The first corner of an unfamiliar track doesn't politely introduce itself. It just appears in the windshield: blind, off-camber, wider than expected, tighter than expected, or somehow both….all while your brain tries to reconstruct a track map built from YouTube onboards, a PDF, and the written testimony of the one guy online who swore "you can take that flat" with the confidence of someone who isn't paying for our brake pads.
That was us at NOLA in March and CMP in April. First races as a team, first laps at both tracks, and the learning curve showed up on the stopwatch. We've already published the data from both weekends (New Orleans, by the Numbers and Carolina, by the Numbers) so we won't drag the evidence locker back out here.
This article is about what comes next.
There's a sim rig in the livingroom now. Assetto Corsa is installed, NCM Motorsports Park is loaded, and soon we will be logging laps. The next race, NCM in Bowling Green, Kentucky, September 19–20, will be the first time we arrive at a track we've already driven a few hundred times without leaving Virginia.
We haven't raced on sim-prepped knowledge yet, so this isn't a victory lap. It's the experiment before the experiment: we're writing down what we expect to transfer, so after the race we can grade the theory against the stopwatch and the driver notes. Public homework, public grading.
Why We're Doing It
The financial argument writes itself. A real test day costs entry fees, tow fuel, race fuel, tires, pads, lodging, time off work, and a quiet withdrawal from the useful life of the car. A sim session costs electricity.
We race a 2007 Honda Accord because the math made sense. This is the same math.
The sim won't replace seat time, nothing does, and the car and stopwatch alway gets the final vote. But real laps are expensive, limited, and arrive late in the process. By the time the trailer rolls through the gate, every unknown corner is charging us for fuel, brakes, tires, and attention. Our goal isn't chasing lap records in a digital car with suspiciously cooperative tires. It's arriving at NCM with fewer blank spaces in the our' heads, because blank spaces make drivers late with their eyes, late with their hands, and early on the brake pedal. In a race where time is position, that's the gap we're trying to close.
The Rig
The rig didn't arrive in a carbon-fiber crate with a financing plan. It started with 2x4s, measurements adapted from a proven DIY layout, and the standard Black Flag question: how close can we get with what we already have?
Close enough to get to work.
The frame is ordinary lumber, built to hold the wheel, pedals, and screen in the same position every session so every driver trains from the same reference points. Perfect realism would be nice; repeatability matters more. The control end is where the money went: a Moza Racing steering wheel with haptic feedback, which means drivers get real force cues (cornering load, slip, curb strikes) instead of guessing what the front tires are doing. The seat, for now, is an office chair standing in until that budget line clears, which keeps spending pointed where it belongs: the race car, safety gear, and the parts that actually have to survive NCM.
That's the whole program in miniature. Any team can spend money. The useful ones turn limited resources into forward motion. The rig is one more piece of a system — car, drivers, crew, data, fabrication — all pointed in the same direction, and built to upgrade as the program grows.
WHAT WE THINK SHOULD TRANSFER
There is some skepticism on the team. "It's a video game," someone said, in the same tone usually reserved for miracle fuel additives. The truth is they're not entirely wrong. A home rig doesn't smell like hot brakes or turn a fire suit into a slow cooker. But feel isn't what we're importing. We think four things should survive the trip from screen to pavement.
Layout. By lap thirty in the sim, a driver should know that the fast section feeds a slower one, that the exit curb matters more than the entry curb, and that Turn 7 has a personality disorder. The real track will correct the details, but correction is faster than discovery. We're arriving with a sketch instead of a blank page.
Vision. Possibly the biggest gain. A track map shows where the corners are; sim laps teach where to look. Most new-track mistakes start before the braking zone (eyes too short, processing too late, turn-in based on panic instead of reference). The sim drills the rhythm of looking farther ahead and filing away the markers that matter.
Sequence. The sim will lie about exact braking distances, every sim does, especially when the real car is a front-drive Accord with endurance tires and the aerodynamic subtlety of a mailbox (we did track down a 2006 Accord Euro model for Assetto Corsa, which closes some of the gap), but the choreography transfers: lift, brake, release, turn, wait, feed throttle, open the hands. That smooth, repeatable rhythm is the same discipline we covered in Fuel Strategy and Hero Laps, and its easier on fuel, brakes, tires, and everyone leaning over the pit wall. The sim lets us rehearse it without turning consumables into smoke.
Traffic. Lemons traffic is a living weather system of ambition, rust, and wildly mismatched performance envelopes. That's not a complaint; that's the sport. Sim traffic won't replicate it exactly, but it rehearses the judgment: closing speeds, corner compromises, and the moments when the fastest move is waiting one more corner.
What Won't Transfer
We're not treating the sim like a magic box. It won't teach discomfort, No software models the moment a driver's inner monologue switches from "focus on the apex" to "why can I hear my own pulse?" It won't teach consequences either. In the sim, a wall is a reset button; at the track, the best case scenerio is bodywork, alignment, and a conversation conducted in very short sentences. The sim also won't teach our exact car, the gearing, weight transfer, and brakes all differ, and some of our Accord's habits aren't available as downloadable content.
That's fine. We're not trying to win the sim. We're trying to learn the track before the track starts charging us by the mistake.
The Plan: Learn the Track Three Times
First, the sim, weeks out. The goal is familiarity so deep it becomes boring. Boring is useful . Boring means the hands aren't waiting for the eyes to finish reading the road. Once everyone has laps, we compare notes: where are you lifting, which curbs are useful, is Turn 7 flat? It isn't. Someone will insist it is. Someone will get extra homework.
Second, the track walk. This is the reality check the monitor can't provide: curb heights, pavement seams, surface changes, and runoff that looked generous on screen but in person reads as a polite invitation to ruin your weekend. Your ankles are surprisingly good data loggers for elevation and camber.
Third, Friday test laps. The sim proposes, the track walk edits, Friday tells the truth. Braking points will move. Lines will adjust for our actual car, tires, and traffic. That's not failure, that's the process working. The win condition is that Friday becomes refinement instead of discovery: learning how the #86 behaves at NCM, not learning which way the next corner goes while another car fills the mirror.
The Rule: Believe the Shapes, Not the Numbers
We're writing this down now so future us can't pretend we didn't see it coming: don't believe the sim's lap time. The sim may say 2:14 and the real car may say 2:26, and neither number is the product. The shape is the product: the sequence, the flow, the reference points, the driver showing up with a working theory.
A working theory can be wrong. That's allowed. "We thought Turn 4 was flat and it isn't" gets fixed in one lap. "We have no idea what Turn 4 wants from us" is slower, more expensive, and usually followed by the smell of brakes.
Being specifically wrong beats being generally lost.
How We'll Grade It
Not by lap-time prediction. By the quality of the first real sessions: do drivers settle in faster, are radio calls more specific, are mistakes smaller, does the first stint look like calibration instead of exploration?
If it works, sim prep becomes a permanent part of the race cycle, its a driver-development tool that costs less than a set of tires. If it doesn't, we'll publish that too. That's how this team operates: run the experiment, keep the data, print the results either way. It's the same approach behind the engine program, the aero program, and everything else on the build log. If backing a team that works that way sounds appealing, our sponsorship page is always open.
Verdict Pending
Sim laps are underway. The track walk is on the schedule. The Accord, for its part, has no opinion. It's never seen the rig and wouldn't be impressed by our lap count. It only cares what happens when the green flag drops.
Fair enough. Our job is to show up at NCM knowing more than we knew at NOLA and learning faster than we did at CMP with drivers already oriented before the track starts asking real questions.
NCM Motorsports Park, Bowling Green, Kentucky, September 19–20. The experiment is live.
Turn 7 still isn't flat.
Probably.
