World's Fastest V-8 Wagons and Sedans! - Jason Cammisa's Ultimate Drag Race Replay [11:21]

And for the record, the SF90 ran what it ran. We did THAT race a dozen times minimum, and the thing just could NOT figure out the surface at Willow. I explained that, too, both here and on IG: it was surprising because the 296 (the other really powerful Ferrari we've tested there recently) learned quickly to ignore the wheelspin that you get from the tar strips and patches, but the SF90 just couldn't deal. It would launch hard, spin the rears, cut power, and fade it in slowly. It did it every single time — never once getting a clean launch the whole day. Randy (Pobst) drove that SF90 for that segment. If only y'all could hear what he had to say during filming... lol. The man is \*not\* used to losing on track.

They're not fake. At all. What you're seeing is multiple takes — we probably launched those cars 50 times to get every conceivable angle without seeing another camera.. and to get the launches as close as humanly possible. When you're talking about cars this fast and this closely matched, launch reaction time makes a huge difference. I've explained this many times — but the results you see at the end of the race, and what you hear in my commentary — are ALL based on instrumented testing I perform with a data logger before we start filming, with a fully warmed up powertrain and a full tank of whatever gas the manufacturer recommends. This is scientific, SAE J1349 weather-corrected (where applicable) instrumented testing that's the same as I did when I tested cars for Automobile, Motor Trend, Car and Driver, and Road & Track. We'd need a near-unlimited production budget to film long enough to get every shot launch down to the millisecond and still wouldn't get it perfect. And then, we'd still run into some cars derating at different times due to the back-to-back launches. And we'd need to top up all the cars every few launches because they use fuel at different rates and therefore their weight changes. Oh, and then there's tires: we'd be changing them multiple times, too. And having to cool them down continually. It's a losing battle. And in the case of these two cars, loko how close the actual test numbers are. The cars are within a HUNDREDTH of a second (which is, for the record, WAY below the resolution of any test gear) on the launch and less than a tenth the whole way through 1st gear. The BMW's slow 1-2 shift lets the Audi pull almost 0.2s ahead, but the BMW is ultimately faster. They're about dead even at the 1/8 mile, and then the BMW pulls on. Test Audi BMW 0-10 0.39 0.40 0-20 0.81 0.82 0-30 1.28 1.35 0-40 1.82 1.99 0-50 2.47 2.59 0-60 3.22 3.30 0-60\* 3.01 3.11 (\*1-foot rollout subtracted) 1/8 mi 7.18 @ 97.36 7.20 @ 99.89 1000' 9.37 @ 111.47 9.31 @ 115.87 1/4 mi 11.22 @ 121.90 11.09 @ 127.12 Happy to lift the veil, but look at those numbers. These cars are literally separated by HALF of the time it takes to blink your eyes — at most — during a full race, so think about the difference a tenth of a second of RT would make either direction. And then go look at RTs at your local drag strip. It would change the finishing order between these cars dramatically. There's a reason CarWow's results are different every time they race: RT counts for so much, even when they get a good launch. And that's precisely the reason I use instrumented testing: it pulls the RT and the driver out of the launch and explains what the car is capable of, best case. We could just as easily do the CarWow thing, but instead we make 100x the work for ourselves doing it the scientific way. On the minus, you might see a couple frames used from a less-than perfect run. On the plus side, you know which car is \*actually\* quicker. I want the car to be the star. No fake race: the Audi is quicker to 60, the BMW is quicker through the quarter-mile.