Reader Submission #87 – Five Common Mistakes Modders Make

GSC 2016-02-23 19-43-48-84Not too long ago here on PRC.net, we ran a lengthy Reader Submission from an anonymous yet respectable sim racing community member, who kindly took time out of his day to break down the inherent lack of simulation value in most popular third party mods, and why corners are cut during a project’s development. Our resident science teacher has returned on this Wednesday afternoonwith another reader submission, this time detailing the five most prominent mistakes sim racing mod teams make. While his explanations may fly over the heads of some individuals, consider this a rare glimpse into an element of sim racing that most people will never think twice about.


screenhunter_11Hello, readers of PRC.net! I’d like to follow up on my previous submission with a significantly more technical piece. I wanted to list five of the most common mistakes sim racing mod teams make in regards to how car physics are handled within the isiMotor engine. Beware, this might get a little complicated.

  • Weird Tire Values – This one isn’t a surprise. Tires are still somewhat of a mystery to even the most advanced engineers among us. If you’re able to get most numbers within a reasonable range, that might not necessarily yield the expected results when in-game. Yes, there are a few tire characteristics that are scientific fact, such as slip angles or load sensitivities, but it’s not uncommon to see totally wild values for these elements, due to the fact that they “feel better” on a subjective basis. What I’m saying is that tire values end up being fudged to appeal to those beta testing the mod, and what they prefer from their sim racing experience.
  • Botched Suspension – Suspensions are hard, don’t kid yourselves. If it wasn’t for carFactory or other modding tools, it would be almost impossible to get right if you don’t have experience in designing them in the first place. Now, the safe modding teams either copy/paste some suspension designs, or make a generic suspension in carFactory. This isn’t a very realistic depiction of the particular car, but if they copy something that was done right the first time, at least it will generate proper suspension behavior. However, most people don’t even know, for example, that a BMW M3 E30 has semi-trailing arms in the back, and behaves completely different compared to most modern cars. This is a mistake many mod teams, and even probably some first party sim racing developers, have made. The brave modding teams that try to design an entire suspension from the ground up, usually don’t have enough knowledge to get it right, so we end up with suspensions featuring massive bump steer, totally wrong ackerman components, too small travel values, weird camber variations, and other suspension oddities.
  • Ultra Stiff Suspension Settings – This here is what most modders use to hide what I just stated above in the previous element. If you make the suspension hard as a rock, it won’t move much, and therefore it won’t show all those nasty characteristics of wrong suspension geometry. It also “feels more responsive” to beta testers who have no clue that even a Group 6 Porsche 936 actually exhibited a bit of body roll while cornering. This, coupled with the wrong tire values, turns a car into what I like to call “skateboarding physics”, where your steering wheels seems to control how much the rear of your giant go-kart is sliding.
  • Low Inertia or Center of Gravity Height Values – Copied once, twice, maybe even three times from the very first Image Space Incorporated vanilla content, all the way back from the days of F1 Challenge 99-02. It’s not uncommon to see anything with wheels slapped together with values better suited to a Formula Ford. This helps to make the skateboard effect described above even worse, making cars that roll much less and therefore more twitchy, but your average sim racer will rave how “responsive” and “dangerous” these cars are to drive.
  • Excessive Downforce – Cars with strange tires, rock solid suspension components, and low inertia values can be a bit tricky in high speed corners. So the solution for most modders is to just add more downforce. If we’re comparing building a car in an isiMotor title to baking a cake, Downforce is comparable to sugar. It makes everything “good” again, and your giant go-kart will now blast through medium and fast corners as if it’s on rails, but it will be twitchy and nervous in slow bends. Sound familiar? In the world of sim racing, some modders believe even touring cars or 1970’s Formula One entries must behave like some high downforce ground effects prototype.

These are the most common issues off the top of my head, and in all my years of working with the isiMotor engine, I can safely say that only a handful of mods got every aspect right.


acs 2015-07-11 14-48-25-39I’ve probably told this story before, but I’m going to tell it again because it’s relevant to this submission.

I work for an extremely large rental car company as what’s essentially an inventory guy, and we have pretty much every North American car you can think of, from lowly shitboxes to high-end muscle cars reserved for longtime customers. During the summer months last year, we received a fleet of convertible Chevrolet Camaro’s, as well as a huge array of brand new 2015 Ford Mustangs. At the time, the rental car company were a bit short staffed on the customer service side, so us inventory guys got to take on the role of shuttle drivers – a position traditionally reserved for the older gentlemen. Basically, there were days at work where I didn’t do a whole lot other than drive new Mustangs and Camaro’s from our lot to the airport and back again. Sure, it isn’t track day experience aside from that one time I took a rental Charger to the drag strip, but I sort of know what both of these cars feel like.

The Camaro has shitty visibility & heavy steering, while the Mustang is pretty much perfect. The end.

Anyways, for those who have been around this place for more than a few months, at one point I uploaded a revised 2015 Ford Mustang for Assetto Corsa. The source material came from a shady Russian modder, and I claimed that I’d made a few adjustments to the car based on my own personal experience driving it around in traffic.acs 2015-07-11 14-50-41-10To my surprise, a lot of people praised the revised car, including a certain Assetto Corsa shill with over 8,500 posts on the official forum. I have no modding experience – none – so it was pretty hilarious to receive any positive feedback at all. Yet the general consensus among those who downloaded it, was that whatever I did under the hood, it was pretty damn good.

Would you like to know what I did?

total massI adjusted the mass of the Ford Mustang to be 2015 kilograms, for no other reason other than the fact that it was currently the 2015 calendar year.

This was enough to earn the approval of those who downloaded the revised mod, so it’s no surprise that extremely unrealistic car physics are flying over the heads of other sim racing community members.

Advertisements

32 thoughts on “Reader Submission #87 – Five Common Mistakes Modders Make

  1. For the inertia and CoG in particular it’d be nice if the expert had provided some rules of thumb. For example, within road cars, a typical CoG value is 35-40% of the height of the roofline – you should only stray outside that range if you’ve got hard facts on your side, and you won’t be too far wrong if you just pick 38%.

    Within race cars it’s a little less strictly true because they tend to exploit the bounding box – for example, almost all GT3 wings are at the extreme top/back of the allowed space, because the farther they get from the body the more effective they are. Going from the top of the roll cage or crash structure is probably the safest bet.

    Like

    1. A few words about Assetto Corsa’s physics limitations.

      Modern openwheelers can be very stiff. i mean extremely stiff, springs and ARBs (dampers as well)! and Assetto Corsa has engine limitations that doesn’t allow using high suspension rates (causes weird bugs)!

      Interactive aerodynamics physics is a must for an openwheeler, rear aero depends on what’s happening at the front and the rear not only the rear! and vice versa (interaction between front wing and diffuser for example)

      Tyre model is not well explained and needs more variables to be added, to get realistic grip and temperatures/pressure behaviour! You can’t cook accurate slicks with it.

      Lack of setup options possibilities in general

      Like

      1. The physics of Assetto Corsa are not of simulation value, when you compare it to the way a racing driver and racing team and even vehicle manufacturers can use rFactor and rFactor 2’s physics to develop their cars and their drivers in a way that provides accurate data which they can rely on to use at races.

        To give you an example. All our race team clients and even our series clients, they don’t only use rFactor for track familiarization. They also use it to test setup data before they go to an event, they use it to test potential new part data, by developing new brakes or new engine performances and testing it in the game before they actually commit to building it in real life. rF1 and rF2 provide far higher accuracy for those things than any other simulation on the market. When it comes to Assetto Corsa, their game is marketed for a set audience which covers the Semi-Arcade side. When I drive rF1 or rF2, I’m always in a serious state of mind, I can’t just go out on the track and run a few laps. It is too simulation based and I treat it as such. I work on my setups, I push to find every tenth I can throughout a lap. I analyze too much. When I play Assetto Corsa, I can easily go in, pick a car and track, chill back and drive When using my Xbox360 controller and not give a care. The physics seem very much like Forza Motorsport 5, a console game physics HOWEVER, Kunos has done a great job for the simcade fans, but cut down to the wire, rFactor 2’s physics engine is by far the most simulation based physics engine using real life aero and physical data that no other title has come close to.

        Like

      2. When you enter real data to rFactor 2, it works like real life, it’s not the case for Assetto Corsa, you always have to use workarounds and substitute models 🙂

        Like

  2. @ author, “I can safely say that only a handful of mods got every aspect right.”

    Can you at least tells us which ones that is?, give us all a reference point.

    Liked by 2 people

    1. That’s true. Problem is, many pretend otherwise (and even paywall), then flip out when people point out obvious discrepancies.

      If you’re pushing out your first mods and add in a note about not having a whole lot of previous experience or lack of data to work with, no problem. Keep going, I will not complain and I’m far more likely to check it out.

      Liked by 2 people

  3. I’ve enjoyed these last couple reader submissions. Question though – what are some *good* mods? i.e. a whitelist of mods that did get the physics right.

    Liked by 1 person

  4. Not only mods, most sims people often easily praised games like ProjectCars having great cars, even if they are just like mods in essence, they are riddled with problems from suspension, tire, brake, lsd etc. SMS can’t even do drift and road cars right, where is the simulation value when they got suspension setup wrong, fudged tire, and camber do not interact with temperature across tire. Look at the new Ford Fusion, compare SMS made nascar to other sims. Hell, try the radbull MX5, that car has weird tire made from NITTO NT01 ( supposedly more grippy tire in real than the NT05 ) NT01 is DOT comp tire, a tier higher than NT05.

    What the physics man from SMS says :

    “New tires for it (Nitto NT05) are a modified version of our Nitto NT01 road tire with better behavior ‘post-peak’ to aid control when in a big slide. It works on some ideas which became apparent to our tire guru after getting to do some ride-alongs with Formula D drivers. Basically it adds lateral grip when sliding as a result of the tire surface turning into a gooey layer with slower response and balances with how it fades away from heat buildup.”

    So, a road tire is made from dot comp tire for drifting and it still sucks to drift with odd behavior. SMS took the wrong turn doing the tires, drift or grip, it’s on the car setup and driver, not the tires fudged to work in specific situation.

    Like

    1. Try some math next time. The higher-spec mustangs are heavy as fuck and reach just below 4k lbs. Add in your standard fat man-bitch mustang driver and you’re getting pretty damned closed.

      Liked by 1 person

  5. The first part of the story was actually quite good reading. As well as the previous reader submission.

    Second part is there propably just because there was no bad things said about AC in the first part and yeah, we know you have your personal crusade..

    Liked by 1 person

      1. >Get mod
        >Tweak mod (2015, kek)
        >Release mod with “I drive these everyday” validation, consciously knowing shit’s rigged
        >Wait for reviews, interestingly get positive reception
        >Wait for the right news/submission/topic to come up
        >?????
        >”lol u guys so dumb”

        It’s more making fun of Kunos/AC fans (and their willingness to swallow anything ‘good’), but still

        Could’ve also mentioned the Flat6 tires debacle for the umpteenth time, or gone the iceRacing route, you know – add some variety

        Instead you go right to the AC route, and only AC because SIMULATION VALUE

        And that’s profiling mister :B

        Like

    1. A few words about Assetto Corsa’s physics limitations.

      Modern openwheelers can be very stiff.. i mean extremely stiff, springs and ARBs (dampers as well)! and Assetto Corsa has engine limitations that doesn’t allow using high suspension rates (causes weird bugs)!

      Interactive aerodynamics physics is a must for an openwheeler, rear aero depends on what’s happening at the front and the rear not only the rear! and vice versa (interaction between front wing and diffuser for example)

      Tyre model is not well explained and needs more variables to be added, to get realistic grip and temperatures/pressure behaviour! You can’t cook accurate slicks with it.

      Lack of setup options possibilities in general

      Like

  6. Interesting. Just 2 hours ago, some of my friends were (again) arguing about R3E, AMS, RF2, and AC, and they actually did mention your AC Mustang mod. Although they disagreed with each other on many, many things, they all thought your Mustang was bad.

    Liked by 1 person

  7. LOL, that’s some funny stuff.

    Really though, ~4.4k Lbs. isn’t all that far off. Also, basically the exact weight of amazingly shitty high-spec option ’15 challenger.

    Like

  8. Formula Armaroli
    Formula Truck 2013
    Game Stock Car 2011
    Game Stock Car 2012
    Game Stock Car 2013
    Game Stock Car Extreme
    Copa Petrobras de Marcas
    Automobilista

    Reiza Studios has released EIGHT FUCKING rFactor 1-based titles.

    Liked by 1 person

  9. DRM physics suck balls, yet people praise it because it looks shiny. Can’t wait to see what kind of special garbage they’ll manage to release for AC.

    Like

  10. Assetto Corsa is developed by Italian studio Kunos Simulazion – Released last year for Windows, the game will be coming to PS4 and Xbox One next year, it does cede a little more ground to the notion that ‘game’ isn’t such a dirty word.

    According to James Dover, the game can be happily played with a controller. But while the game does deliver in terms of graphics, James reckons its physics engine “lacks seriousness”.

    Like

  11. oh these rfactrolls. You’re mad at AC and mad at DRM mods. In reality you’re just upset that DRM doesn’t want to take any more bullshit from rfactrolls community, so they stopped modding there.
    There is rfactrolls and rfactor community, in case I’m being misunderstood.

    Like

  12. When you enter real data to rFactor 2, it works like real life, it’s not the case for Assetto Corsa, you always have to use workarounds and substitute models 🙂

    Like

Ratio of vowels to consonants will be monitored. Post at your own discretion.

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s