Here’s Why 75 Years of Performance Testing Means MotorTrend Is Prepared For Mobility’s Future
From the stopwatch/fifth-wheel/radar era to GPS and from 0–60 and the quarter mile to the figure-eight test and more, we’ve come a long way.
MotorTrend didn’t invent automotive performance testing as we know it today. We cede credit for that to Autocar (née The Autocar) in Europe and to Tom McCahill (1907–1975) stateside, for first wielding a handheld stopwatch and clicking off the 0–60 times of cars he bought, begged, or borrowed. He published his results in Mechanix Illustrated magazine beginning in February 1946, and car lovers everywhere were quickly addicted to that performance measure. We may have inherited the metric, but we like to think MotorTrend has helped improve the metrology, advancing the science of objective vehicle testing at numerous points throughout our 75-year history.
0:00 / 0:00
In the Beginning ...
We performed no testing on our first cover subject vehicle, the Kurtis Sport Car—at least not until 2019 when MotorTrend purchased that fully restored car (0–60 in 15.3 seconds!). But we performance-tested three cars in our second issue (October 1949): two 1949 MG TCs (20.3 seconds, 14.2 with Shorrock supercharger) and a 1949 Studebaker (16.7 seconds). We started out using essentially the same “technology” McCahill pioneered: calibrate the vehicle speedometer over a measured mile then launch the car and click a stopwatch as the speedo needle swings past the true 60-mph mark. Because the driver’s hands were usually busy, MotorTrend entrusted stopwatch work to a passenger.
Our techniques improved with an eye to systematically eliminating opportunities for human error. The primary tool of our trade for the first several decades was the mechanical fifth-wheel, with Tracktest of Franklin, Michigan, supplying our first units. This bike-wheel/tire on a swing-arm setup generally mounted to the rear bumper (later to a side door), and initially it just eliminated the inaccuracy of the vehicle speedometer while the passenger still operated a stopwatch.
Gradually, the wheels became more sophisticated and transistorized. Early versions of these measured speed by observing the passing of white and black segments on the wheel rim in front of an optical sensor, but sun glare could make black look white. Subsequent designs shined their own light through slots around the wheel to generate the desired speed frequency data. By 1954 we were feeding that data into a Berkeley electronic counter, and over the next three decades computerization continuously improved the fifth-wheel’s accuracy.
Darker Ages
MotorTrend arguably stuck too long with the fifth-wheel, as superior technology eventually emerged. By the late 1980s, the Germans devised the Correvit “optical fifth-wheel,” which shined light on the ground and registered reflected light from the passing road surface with an optical lens. This instantly eliminated problems like the bike tire bouncing off the ground or expanding at higher speeds and getting tangled if the car needed to reverse or even make a tight U-turn.
While some of our competitors jumped on this bandwagon, MotorTrend instead transitioned to Stalker radar in the 1990s. This also eliminated the above noted fifth-wheel woes, boosted accuracy, and eased the transition from testing one car to another. Setting it up on a tripod behind the car at the starting line allowed multiple cars to be tested with one setup, and it was ideal for testing things fifth-wheels couldn’t practically mount to, like jets. On the downside, the driver didn’t have any onboard readout of performance, so the team at the starting line had to radio info comparing launches, elapsed times, and trap speeds from run to run. Mounting the radar unit inside the windshield allowed the driver to review runs on the computer and permitted higher-speed testing. (The reliable range of a stationary-mounted unit didn’t extend much past the quarter mile.)
Heavenly Accuracy
The biggest leap forward for testing accuracy first appeared in the early 1990s and leveraged Global Positioning System satellites. Most industry and media testers still use this technology today, with the newest data loggers boosting accuracy by also receiving signals from European, Chinese, and Russian satellite constellations.
Myriad satellites overhead (31 operational on the GPS network as of 2023) continually broadcast their number and the precise time of day. An antenna mounted to a vehicle’s roof with a clear line of sight to between three and seven of these satellites can generate velocity and vector data by knowing where each satellite is in the heavens (from their numbers) and reading that time info coming in from each with a Doppler shift. From this velocity info, the computer can integrate position/distance and derive acceleration. One drawback: Bridges and other obstacles can cause a brief loss of signal.
Former MotorTrend tech guru Kim Reynolds brought MT into the satellite era with a Racepak unit in the mid-2000s, but its data management proved a bit clunky, so in 2007 and 2008 we purchased two Vbox units from U.K.-based Racelogic, which was by then the de facto supplier of vehicle-performance testing systems used by the entire industry. In the years since adopting the Vbox equipment, we’ve averaged about 200 full road tests per year.
Weather Correction
In the heart of the muscle car era, the Chrysler/Mopar Hemi engineering crowd conducted extensive environmental tests studying the effects of temperature, humidity, and pressure on the performance of naturally aspirated engines. They eventually developed look-up tables of weather correction factors for both the elapsed time and trap speed that promised to equalize all results to the SAE J607 conditions of 60 degrees F, 29.23 inches mercury (Hg) barometric pressure, and zero percent relative humidity. Sister publication Hot Rod published these tables, and magazine testers began employing them, upgrading to a more formal SAE J1349 weather correction protocol (changing temperature to 77 F) in 2005. For years we’ve measured weather via a Computech RaceAir system and corrected our results accordingly.
1-Foot Rollout
Former editor Griff Borgeson made the case in a June 1952 editorial for triggering our acceleration-testing clocks when a driver’s foot hits the floorboard. That protocol prevailed until the days when test-and-tune nights became the easiest way for our readers to measure 0–60 and quarter-mile times on their own cars. Since then, we’ve aligned our testing practice with the NHRA protocol of starting the clock after the car has rolled roughly (or, nowadays, precisely) 1 foot.
Special Tests
The satellite, cloud-connected age of modern computing and sensing opens vast new testing possibilities. Under the auspices of former testing director Reynolds and his successor, Eric Tingwall, MotorTrend has returned to the vanguard of innovation in performance testing over the past decade and a half.
Perhaps our greatest contribution to the pantheon of performance testing has been our invention of the figure-eight test. After years of frustration trying to quantify vehicle handling via slalom or emergency lane-change tests (both of which reward smooth driver inputs and tidy vehicle dimensions over inherently superior chassis dynamics), Reynolds and road test editor Chris Walton conceived a “racetrack in a bottle” test consisting of two 200-foot-diameter circles with their centers separated by 500 feet. Test results include a lap time, the vehicle’s steady-state maximum lateral grip in both directions, and an overall average g-level encompassing accelerating, braking, and cornering grip. It also provides valuable insights into the vehicle’s behavior transitioning from cornering to acceleration and during hard braking into a corner. The test’s only drawback is that it requires smooth pavement measuring 300 by 800 feet.
We’ve quantified ride quality and measured sound levels. We’ve used our Best Handling and Best Driver’s Car contests to innovate tests measuring everything from traditional chassis-dynamics parameters—pitch, yaw, roll, chassis slip angle, steering input angle, and more during racetrack hot laps—to more esoteric stuff, such as inferring driver stress or joy levels by logging instantaneous heart, breathing, or perspiration rates, and even videoing the driver’s face during a lap and asking AI to infer real-time emotional state.
Going Forward
As the industry evolves into a period of increasing electrification and autonomy, we’ve devised test methodologies for assessing real-world highway EV driving range and fast-charging rates to help buyers assess the road-trip viability of electric vehicles. We’ve integrated our objective testing results into our Ultimate Car Rankings model, which works a bit like an instant comparison test of every vehicle in every segment. We’re hard at work on tests for quantifying the driver workload and degree of distraction prompted by different infotainment user interfaces, and we’re working on new ways to quantify the relative merits and performance of various Level 3 or higher driver assist systems.
Whatever the future brings to the automotive market, you can rely on MotorTrend to work out new and innovative ways to test, assess, and rank it all, and we’ll present our results in a way that makes sense to our audience as it faces the increasingly difficult decision of how to invest its time and money in mobili
1,000-Mile Road Tests of the ’50s
In a piece written for the November 1954 issue titled “The Story Behind MT ’s Road Tests,” associate editor Jim Lodge made an accounting of all the various procedures and testing gear his team employed at the time. The to-do list might make our modern test team blanch—but at least their predecessors got hella cool Motor Trend Test Team swag to wear!
At MT HQ, a fuel flow meter was installed between a car’s fuel pump and carburetor(s), and a tachometer/vacuum gauge unit was hooked up and affixed to the dashboard. A curved tubular bubble-in-oil level device was also mounted for use in quantifying body roll, as was a Perf-O-Meter gizmo that displayed accelerative thrust so the driver knew to upshift when said thrust fell off. The trunk was then loaded with a pair of contact boxes and pneumatic hoses to register when tires hit them at each end of the top-speed timing strip, a roll of wire connecting them with a battery and a bank of electrically actuated stopwatches (seven of them, including a “master,” hundredth-second watch), plus battery cables, connectors, and myriad metal stakes to hold things in place. Last came the Tracktest fifth-wheel, the Weston electric speedometer, the Berkeley electronic counter, and the cabling to connect everything. Those last three items alone were said to cost $1,200 in the day, or $13,700 in 2024 dollars. The 25-hertz Vbox Touch data loggers we use today cost just $4,200 and provide far more precision and accuracy.
The Old Test Regimen
Here’s how a multiday, 1,000-plus-mile Motor Trend test went down in 1954. First, every test car did a few pulls on a Clayton Engineering chassis dynamometer to measure road horsepower at various rpm. The car was then fueled and weighed on a calibrated public scale. Next, the team headed for a test site in the high desert, 190 miles away, leaving from home as early as 2 a.m. on days with notably high temperatures forecast.
Once there, the pressure hoses were positioned over the top-speed section, and several runs were made. High-speed acceleration testing was done to obtain time-to-speed data for speeds higher than we could reach on our quarter-mile-plus airstrip. Next, steady-state fuel consumption tests were conducted at several speeds, and finally the cars were driven up and over a mountain pass to assess handling, climbing, transmission flexibility, braking, and more. A second day of 0–30, 0–60, and quarter-mile acceleration and brake testing then happened at a more local airfield. Lodge’s 1954 story also mentioned a new MT closed-course circuit coming into use for handling evaluations.
Certainly, these tests seem more extensive and/or complicated than the ones we execute today, but consider there were no EPA fuel economy ratings, horsepower claims were more spurious, and the sheer number of car models was small enough in those days to cover practically every one of them with just 50 such tests per year—about a quarter of our annual testing workload today.
