Skip to content

Your Favorite Sims Could Be Fixed in an Afternoon; Why Aren’t They?

I was one of many who gave Automobilista 2 the benefit of the doubt, and was pleasantly surprised by my first few hours with it. The off-brand Holden Commodore was immediately up there with one of the best sim cars I’d ever driven, and I promptly spent a giant chunk of my evening blasting around Adelaide with a handful of other sim racers who seemed quite satisfied with their impulse purchase.

There were doubts among the more critical portions of the community as to whether or not the Madness Engine could support a truly hardcore sim, after the mass market Project CARS series received mixed reviews. Reiza were demonstrably on the right track.

A year later, they were somehow back at square one.

The Porsche 962c and Sauber C9, released as part of a free update to the sim, exhibited all of the classic traits associated with titles that have tried, and largely failed, to provide a believable driving experience. Sim racers have even created their own derogatory term to describe sims that claim to offer realistic handling cars, but let you get away with sloppy driving fundamentals behind the wheel.

That label, is Sim-cade. A hybrid of realistic auto racing elements – pit stops, tire wear, and fuel consumption – with handling traits that while appearing realistic at first, allow for undisciplined wheel or pedal inputs beyond what you could get away with in a real car.

Automobilista 2 now fell into this category.

Compared to real-world qualification times at Spa-Francorchamps, the Group C Prototypes in AMS2 were four seconds faster than their physical counterparts, a symptom that should have raised red flags over at Reiza HQ, but for whatever reason, didn’t, and is usually a hallmark of simcade titles.

More importantly, there is one specific symptom of a racer in this category that began rearing it’s ugly head as I chased the top leaderboard time at Spa, and it’s what a lot of content creators have both tried and failed to describe when criticizing these games. When people talk about a game being simcade, this is the specific handling trait they’re describing:

Just putting down a single lap, was oddly tiring, even with the Force Feedback turned down to a murmur. It seemed like the front end of the car had infinite mechanical grip, and no matter how quickly you moved your hands on the wheel, you came away from the previous set of corners believing you could always move your hands a little bit faster. Once approaching a world record lap time, your wheel inputs had morphed into erratic, last-second jerking motions that you had sort of memorized for each corner, as if you were playing Guitar Hero.

My lap at @LogitechGC Logitech G Challenge Week 3 at Sugo, Record Lap at Asia Pacific Region. still mistake here and there especially Last Turn, but whocares! 2 week in a row Record Holder! Push for Willow [email protected] #lor #pcars2 #logitechG

— Andika Rama Maulana (@RamStig) June 23, 2020

This is a symptom that I’ve seen across multiple games, ironically each of them being deemed simcade by avid sim racers. The result of this infinite front grip sensation creates a butterfly effect that spirals into the game requiring obscure setup tricks not rooted in conventional setup techniques to be fast, or driving styles that bore no resemblance to reality once at the top of the eSports ladder.

What I saw in AMS2 with the Group C cars, was the same symptom I’d experienced in iRacing’s Super Late Model a few years ago, along with the Dallara IndyCar in Project CARS 2 and pretty much every car featured in Gran Turismo Sport. This was a universal problem that spanned multiple physics engines, racing disciplines, and dev teams. The laptimes were insane compared to real world pace, and I couldn’t move my hands fast enough to make use of front grip that seemingly didn’t end so long as I wasn’t overly egregious in my approach.

By comparison, this has gone against everything I’d learned driving multiple classes of sportsman pavement cars.

Real life teams are constantly worried about over-driving the front tires due to their lack of front grip, which is why their setup values, as well as onboard footage, come across as so drastically different when compared with certain sims. Every adjustment is based around making the car turn better, not expecting the car to turn on a dime by default.

It also explains why, in the eSports kingdom, the competitive scene is dominated by younger kids, whereas professional race car drivers can remain successful well into their 40’s and 50’s. They do not need fighter pilot-like reflexes and arm movements to extract front grip out of their race car; they are instead on a relatively equal playing field as kids half their age, attempting to not over-drive the front tires and produce excessive understeer, just like everybody else is.

Yet if you boot up any of the following games, you receive a driving experience that demands you to play Guitar Hero with a race car; technical corner complexes like Long Beach’s fountain section demanding your jerky, erratic inputs to extract every last percent of the car’s limitless capabilities, become muscle memory.

  • Automobilista 2
  • Project CARS 2
  • Gran Turismo Sport
  • iRacing

These symptoms have resulted in some sim racers swearing off some or all of the above mentioned games, even if the games include content they actively want to drive, or features that are objectively exactly what they want from a racing sim. They might end up in the situation I’m currently in, where I have no interest whatsoever in sports car racing, yet have been forced to become a road racing connoisseur and binge on Assetto Corsa, simply because it featured the vehicle dynamics that felt the most natural to me – warts and all.

What if I told you, that you – yes, you – could theoretically fix these games in an afternoon, and the quality of multiple sims could be improved practically overnight?

What if I told you that developers are equally capable of doing this, but for whatever reason choose not to?

Chevrolet Corvette C8.R - | Motorsport Simulations

The engine powering each racing sim, really isn’t the problem. This is a common misconception perpetuated by content creators or hardcore sim racers who probably mean well, but are trying to punch a bit above their weight in their criticisms of certain games.

Racing simulators, really aren’t the most complex games in the world when it comes to sitting down and editing vehicle dynamics. Everybody who’s owned a copy of Assetto Corsa, rFactor, or Automobilista, have undoubtedly at one point or another poked through the contents of a mod they’ve downloaded, and discovered that most of this intricate product is just the result of fill-in-the-blank text files that you can open with notepad.

Files governing everything from engine, chassis, and transmission properties, as well as more intricate stuff like aerodynamics and tires, are just sitting out there in the open. If you’ve got even a passing interest in cars, a solid chunk of it will make sense to you.

This essentially puts modders and developers on the same level playing field; while one group is getting paid for the work and another is doing it for free as a passion project, they are both building their cars in the exact same manner. It’s just data entry.

In fact, they’re very likely grabbing the same data and plunking it into the same notepad documents, from the same set of reference books available on eBay. And sure, while developers may be granted special access to sensitive values governing the development of a yet-to-be unveiled McLaren supercar or Ferrari F1 racer, or tire suppliers like Pirelli or Michelin offer up data straight from the lab, modders can simply decrypt the officially released content once it’s in the hands of the general public and immediately be brought up to speed.

However, this development process can still result in cars that just don’tfeel right.

As mentioned above, for whatever reason, the car might be four seconds faster than it’s real world counterpart once in the hands of an experienced sim racer, who claims his stint in the virtual car in no way resembles the copious amounts of on-board clips that have been graciously uploaded to YouTube

The problem is, a lot of developers opt to… leave it at that. They might have their reasons – time constraints, underlying design choices, petty stubbornness, or the inability to push a car to its’ limit and discover the edge case oddities typically reserved for eSports competitors – but as a sim racer, you’ve paid for a game that offered unparalleled authenticity, and you didn’t exactly get that. You got a product that doesn’t match the marketing claims, doesn’t match on-board footage, hell it doesn’t even match something as basic as lap times.

That’s not cool.

It’s not exactly false advertising, but you now don’t trust the developers to move your favorite hobby forward, and you grow increasingly irritated that not only are these claims of accuracy slowly getting dismantled by awkward hotlap videos like the one inserted above, but their inability to address these issues after several months or even years indicate you’ve wasted your money.

And you’re actually right to feel that way. Don’t let the fanboys gaslight you into believing they’re trying really hard and there are all of these other factors like temperature or asphalt age that play into a four second difference.

They don’t.

The process of ironing these quirks out isn’t exactly time consuming, is something all developers are easily capable of doing, and is something you yourself could accomplish in an afternoon by just running test laps and cross-referencing the tremendous amount of material on YouTube.

Hypothetically, if developers provided easy access to just a handful of tire values and some sort of on the fly via debug menu in which to change them immediately, an experienced sim racer could fix long-standing issues that have plagued their favorite sim quite efficiently. Amateur and professional drivers, such as outspoken iRacing critic Nicki Thiim, could also sit down and over the course of a single livestream session, improve a game he’s complained about for years via simple trial and error.

This is what that process could potentially entail.

I’ve enlisted the help of veteran isiMotor modder Richard Wilks – who has produced a set of fantastic mods for Automobilista under the International Sim Racing Organization banner – to help explain this process to me in detail, so I can translate this over to you in a much more digestible format. Richard’s physics work includes full fields of Can-Am, USAC, and CART entries, essentially the most detailed and demanding mods available for a modern racing sim.

Developers, theoretically, would first need to provide us with access to just four basic sliders relating to tire load. Each sim usually has several lines and attributes that could be adjusted, making tire files seem like convoluted black magic for the uninitiated, but for the sake of user-friendliness many of these would have to be condensed into just one slider affecting a group of values very similar to what’s done with Force Feedback sliders.

You’ll go out for a few laps and make note of the car’s overall sense of front grip. If it feels like you can aggressively attack corner entry and seemingly can’t move your hands fast enough when cranking the wheel towards the apex, you’ll want to lower a slider that would be named something like normalized front load value incrementally, from whatever it’s default value is. The goal is to lower it to a point where you’re rewarded for a fundamentally sound corner entry and smooth, controlled wheel inputs.

The next area of concern happens to be, as predicted, an equivalent normalized rear load value. You want the car’s rear end to be as neutral as possible – extra rotation is something car setup should dictate. The higher this value, the more the car’s rear will rotate, so if it’s too tight on entry, increase it just a cunthair, but any sensation like the car is being steered from the back end, you need to lower it.

Phase three is adjusting how much throttle control you need off the corner, which could be dictated by something like a group of values operating under the moniker of load dependency factor. A lower value would generally be used in cars that can quite easily overpower the rear tires with a heavy right foot – NASCAR, IndyCar, and historic sports cars immediately come to mind. For cars with more mild relationships between the throttle pedal and wheelspin, think mid-range LMP prototypes or GT cars, you’d slightly increase this number until it generally matched the telemetry available online. Ideally, the developer would provide you with both a minimum and maximum window for this value to prevent you from royally fucking up their game.

Some values would be “set it and forget it.” An adjustment like load dependency linearity, would generally be governed by what kind of tire you’re dealing with on the car itself. A street car, supercar, or historic sports car, to convey that “floaty, four-wheel drift” feeling that street tires, grooved slicks, or historic racing tires supply, this value would probably be a low number. As bias ply technology improved in the 80’s for certain racing slicks, you’d slightly increase this number for 80’s or 90’s race cars, and then for modern day radial slicks, you’d increase this number again for that planted, firm feeling you’d get from a modern GT3 or open wheel ride.

However, in fucking with these numbers, what will often happen is that this will slow the car substantially, since in a lot of cases your adjustments will result in noticeably lower cornering speeds. Cars in iRacing, GT Sport, pCars, and AMS 2 have been built to produce accurate lap times and speeds with the infinite front grip sensation, so by removing this and implementing a bit of throttle control on top, you’re guaranteed to get a bit slower.

To counter this, developers would need to provide a fifth slider dictating an overall tire grip multiplier. This would be your last step, increasing in the smallest of increments until you got within a second of your proposed benchmark laptime – intentionally leaving about 1.5 seconds on the table to account for qualifying trim and an eSports meta’d setup.

Once all five sliders were set, you’d go for another stint of test laps and start all over at the top again to double-check your work.

The entire process relies not on real data, but rather driving ability, subjective feel, and cross-referencing a shitload of YouTube footage and driver interviews. Thankfully, in 2021, finding the latter is piss easy and allows you to binge on a ton of racing coverage that would otherwise be relegated to mindless shitter reading.

Therefore, in the hands of a very competent sim racer, or one of several amateur/professional drivers we’ve got hanging out in our community, many of the titles deemed simcade could immediately enter the discussion as industry leaders with just a couple hours of tweaking.

The hobbyist GT4 drivers, IndyCar pros, or journeymen NASCAR drivers, could sit down for an afternoon and basically pump out the most realistic sim versions of every single car they’ve driven in their respective careers.

But this also begs the question, if changing an entire handling model is this easy, why aren’t developers doing it? Why do tire model updates on a handful of cars take months, if not years to manifest for just a handful of cars at a time? Why are cars constantly being pushed out across a multitude of sims that blatantly miss the mark, time and time again?

The answer is going to hurt some feelings.

Because sim racing as an industry has opted to cozy up with the eSports fad, the top 1% of sim racers who could be used as testers, are now largely off limits. When you’re offering a six-figure prize for some of these championships, lawsuits would undoubtedly manifest if you gave one specific competitor beta access and told him to change the handling model to a style that suits him. This is precisely where eSports integration fucked us as an industry. The most qualified people to assist with making the games better, can’t really help you here without spawning justifiable outrage from other competitors and teams.

Developers are therefore boxed in to a corner, forced to make subjective handling changes via driving skills and feedback that might not be up to the task, or recruit community super-fans based on their availability rather than their prowess.

It’s hard not to imagine nepotism comes into play as well. The top guy running the show may have never driven a car on track in anger, or it’s their buddy doing the physics legwork, with no real skilled driver to back any of this up or give it their stamp of approval. People may have the perception that these are big corporations churning out these games, but sometimes the reality is quite the opposite.

Think of the last online race you ran, point out the guy who finished in 14th, look up who he is whether it be on iRacing or a number of other competitive platforms, and give him the keys to completely change the handling model.

How do you think that goes?

That’s a lot closer to to reality than most might think.

Then you have external factors, such as the underlying design of the game. When you’re selling a product like Gran Turismo to millions upon millions of players, you have to worry about abstract concepts like player retention. A game that is deemed by the userbase to be too difficult, even if pro drivers are singing its’ praises, can adversely affect your bottom line.

The idea isn’t for Gran Turismo to be a real driving simulator. The idea is to convince the Average Joe they’re a competent race car driver. If you actually forced the Average Joe to back up his corner entry and smoothly roll on the throttle, there is a fear Joe would fuck off and play Forza rather than sit down and understand the nuances of the game.

Lastly, the developers might just be a bit stubborn. We’re in an industry where multiple companies will send angry emails or thinly veiled legal threats to professional drivers if they don’t shower their product with praise. I’m writing this article on a PC that was, for all intents and purposes, a bribe to pull down certain articles. Some developers might not even be willing to concede to the idea that their game is less than perfect, a stance that I’m told is surprisingly common over in the iRacing realm.

And it’s a shame, because multiple sims that aren’t perfect, very well could be whipped into that state by talented drivers with an afternoon to spare.

It’s essentially what modders do, for fun.

4/5 - (1 vote)