Another Q & A, this one from Mark Shrivastava: "If the car feels fine, how do I know if I'm leaving something on the table? If it does nothing wrong, is predictable, is the only thing left to do just trial and error?"
This one comes up so much. If you find yourself ever saying "that's just how the car is," or have the occasional really fast event and then wonder why you aren't there 100% of the time, this one is for you. We most recently spent some time working with someone who's been developing the same car for 6 or 7 years, and has fought the car dynamically basically the entire time. Still winning a lot, but not always. The inevitable spiral follows: doubting your own abilities, guessing "I just suck," leaving events disappointed. But frequently nothing even "feels" wrong to you. It's very easy to feel "Stockholm Syndrome" towards a mediocre car, and maintain a bias against action to change it, sometimes for fear that you'll make it worse, sometimes just because you're able to do well enough as it is.
We're not saying it's *not* the driver, and that the driver isn't the primary variable, but there's no such thing as a car that does nothing wrong. It simply means you aren't a harsh enough critic, or aren't diving in deep enough, or thinking analytically. You can be down on total grip: are both axles pressure optimized to a point where no matter which direction you go, the car gets slower, or less consistent? Even if the car is balance optimized, are you sure you're optimizing the grip through travel (both camber curves) completely? Are you sure you understand every tradeoff your setup has, and are sure it's one worth making?
This is pragmatically intensely difficult and an endless process, so this analysis frequently stops with some event wins + contentness. The car will always have some failure mode you should be chasing, whether it's mistake management, a demand to drive it in a particular way, or a demand for a narrower input window than ideal. If a car was perfectly predictable and perfectly consistent, you'd be able to intuitively drive it every run without making a mistake. Does it behave the same across a wide variety of grip levels and surfaces, without changes? That's truly impossible, so find the ways in which your car is likely to induce a mistake or inconsistency, then chase down the why. Don't even use "the best setup you've ever driven" as a reference. Use "the best and easiest to drive car anyone could ever have," and don't sell yourself short.
This could very easily be an article where we write up the 946 different ways to use logged data, but truly, just being a harsh critic of your own work is going to be much more effective. This is much more of a philosophical question than a scientific one. And here's a point that's counterculture to a lot of tribal logic: be willing to fix the car alongside fixing your own driving mistakes. It's unpopular to say this in a domain where everyone is convinced that the driving is the *only* difference maker, not just the primary one.
The same holds true for analyzing a ruleset. Have you gone through the rules and made sure you're maximizing each one in the best way you can? Whether or not you specifically pursue each line item isn't the most important part here. There are certainly things where the juice isn't worth the squeeze, but most stop before they even understand the juice:squeeze ratio.
Great question, Mark. If this raised more questions than it answered, that's kind of the point, let us know if there's a piece of this you want us to dig into further.
