Updated Assetto Corsa EVO Roadmap Shows Drastic Changes

ArchTheSecond · May 8, 2025

Durge Driven said:
The fact any sim needs such a convoluted bugged pay ware manager to be playable speaks volumes.

The physics features were added because they're necessary for elite motorsports modeling. AC + CSP is the go-to atm.

It doesn't need it to be "playable", physics wise AC was already ahead of any alternatives. It's just needed for additional functionality.

eg01st · May 8, 2025

Durge Driven said:
The fact any sim needs such a convoluted bugged pay ware manager to be playable speaks volumes.

Well you miss what I want ACE for and that's online free roam.
imho ISI is still the best driving engine ( warts and all )
Tell me what are all those driving AC doing online, playing tiddly winks ?
Superior handling compared to ACE ? I don't think so.

P.S.
I don't even understand why people want race cars in this.
I thought the main object was to make best free roam / track day sim ever.

To drive from your house to race whatever you like where ever you like is what I want.

MCO II

I think that's the point of needing the best possible physics - you have all those street cars and free roam, so it MUST handle like real life counter part. I think every sim player has searched if there exists an AC mod of his real life car. And if it does, we want to see how close to reality it drives.

imurgodau · May 8, 2025

Chatopex said:
Sometimes I believe the team at Kunos sees our feedback as mere trolling. I understand that the work involved is complex and demanding. However, this is a critique of a flagship title in the world of Sim Racing... it is not just another game, it is Assetto Corsa.
Today is a significant setback for me. A more careful approach (and updates) will be awesome!

I agree. I saw Stefano ban someone from the forums a few years back for politely asking if there were any plans to fix the VR in ACC.

Spinelli · May 9, 2025

ArchTheSecond said:
You're overcomplicating it. Having a suspension kinematics solver with accurate forces will behave worse than a precomputed system, if the parameters in the solver are too simple, and in some cases if it's not running fast enough.

If you want cars to behave properly, you absolutely need comprehensive kinematics, elastokinematics and kinetics. Anything less is increasingly problematic, especially if you use a realtime solver.

More complexity "automatically" meaning more realistic or accurate results is a common misunderstanding in software development, especially in fields like physics engines or simulation.

Realism in physics simulations is not just about the quantity of code or how many systems are modeled - it's about how well the underlying physics principles are implemented and how accurately they align with the behavior being simulated.

In over-simplified terms:

A bloated engine might simulate many niche edge cases, use lots of approximations, or be poorly tuned for real-world behavior. It can be mathematically complex but still behave and/or feel and/or look unnatural in gameplay / end-results.
A leaner engine may focus on fewer, core physical principles but execute them with high fidelity, numerical stability, accuracy, and tight integration which leads to more realistic, believable, and natural results.

A 500 page novel isn't necessarily better than a 200-page one. It depends on the clarity of the writing, the strength of the ideas, it's depth, and the execution, not the length. The same applies to physics engines - more code doesn't necessarily mean better realism.

FrejkiChodak · May 9, 2025

"The latest release, v0.2, has been the most hotly anticipated update for Kunos Simulazioni's big-ticket simulator."
-Because it's the first and only update...???

mantasisg · May 9, 2025

ArchTheSecond said:
physics wise AC was already ahead of any alternatives. It's just needed for additional functionality.

I think you are smart enough to know that it isn't so and never was. It was/is great sim overall, no denying that.

I think even the notorious early access post hotfix ACC totally mogged AC in terms of physics and how everything was configured. Sadly it just wasn't meant to be this way in Kunos metaverse. Mostly because people got so custom to AC and brainwashed themselves to believe that it is "the thing".

I think the date ACC had to reverse from greatness back to gamers marks the pivotal moment in simracing history. When physics in terms of realism globally became no longer a thing of focus, and rather than exciting bit to explore and discuss, it has become a fun spoiling thing that is taboo to talk about. Physics which is matter of facts got replaced by a matter of feelings and emotions.

But I guess in ACE we can still expect to have decent good physics that will be more than enough for majority of people. It is just they will always have mixed feelings and emotions as long as developers will care about them.

There is also no other reasons ACE couldn't be miles ahead AC physics, than just the insane sensitivity of AC userbase to any sort of changes and differences and perhaps extremely tempting potential to tap into Forza market. I think Kunos are capable to do incredible simulation if only there was a market to it. And it still could have free roam and all fun aspects, because physics doesn't prevent that.

I am feeling for simracing. But it was a good run, and achievements are still amazing, I just wish it kept going. And I also wish people wouldn't have degraded so much in what they choose and accept. For example rF2 is amazing simulation, and is has great stuff. But for some reason people keeps picking up trash.

People makes the world go round. Or stall LOL

ArchTheSecond · May 9, 2025

Spinelli said:
More complexity "automatically" meaning more realistic or accurate results is a common misunderstanding in software development, especially in fields like physics engines or simulation.

Realism in physics simulations is not just about the quantity of code or how many systems are modeled - it's about how well the underlying physics principles are implemented and how accurately they align with the behavior being simulated.

In over-simplified terms:

A bloated engine might simulate many niche edge cases, use lots of approximations, or be poorly tuned for real-world behavior. It can be mathematically complex but still behave and/or feel and/or look unnatural in gameplay / end-results.

A leaner engine may focus on fewer, core physical principles but execute them with high fidelity, numerical stability, accuracy, and tight integration which leads to more realistic, believable, and natural results.

A 500 page novel isn't necessarily better than a 200-page one. It depends on the clarity of the writing, the strength of the ideas, it's depth, and the execution, not the length. The same applies to physics engines - more code doesn't necessarily mean better realism.

Just for clarity, I've developed hundreds of simulation vehicles and have been involved as a hobbyist for over a decade and as a paid expert for a bit less than that. I'm talking from actual exact real-world scenarios, to me it seems like you're theorycrafting and overcomplicating it.

ArchTheSecond · May 9, 2025

mantasisg said:
I think you are smart enough to know that it isn't so and never was. It was/is great sim overall, no denying that.

I take it you say this from experience correlating simulation models to use for real-world purposes, right?

Spinelli · May 9, 2025

ArchTheSecond said:
Just for clarity, I've developed hundreds of simulation vehicles and have been involved as a hobbyist for over a decade and as a paid expert for a bit less than that. I'm talking from actual exact real-world scenarios, to me it seems like you're theorycrafting and overcomplicating it.

I respect your experience and I'm not trying to crap on you or anything but rather trying to address a common misconception in, both, development and game physics circles: that more code or system complexity inherently results in more realistic behavior.

In many areas of engineering and simulation, including physics engines, more complexity doesn't inherently lead to better or more realistic results. It may enable more features or simulate a broader range of behaviors, but unless that complexity is being tightly managed, precisely implemented, and thoroughly validated, it can actually degrade the final outcome / end-result.

In practice, I've seen (and many devs will agree) that simpler, well-tuned engines often yield more believable results especially when it comes to player-facing experiences. I know multiple people (including a close family member) who have studied cases where larger physics systems introduced more room for edge-case bugs, instability, and/or visually awkward physics behaviors.

This isn’t "theorycrafting", it’s a fact and one that can be observed across a wide range of systems: clean, focused, well-calibrated models often produce more stable, believable, and realistic behavior than large systems burdened by edge cases, interdependencies, numerical instability, etc., etc..

Sure, more lines of code might reflect capability, but they don't guarantee accuracy or realism without solid fundamentals, "proper" implementation, and careful tuning.

ArchTheSecond · May 9, 2025

Spinelli said:
I respect your experience and I'm not trying to crap on you or anything but rather trying to address a common misconception in, both, development and game physics circles: that more code or system complexity inherently results in more realistic behavior.

In many areas of engineering and simulation, including physics engines, more complexity doesn't inherently lead to better or more realistic results. It may enable more features or simulate a broader range of behaviors, but unless that complexity is being tightly managed, precisely implemented, and thoroughly validated, it can actually degrade the final outcome / end-result.

In practice, I've seen (and many devs will agree) that simpler, well-tuned engines often yield more believable results especially when it comes to player-facing experiences. I know multiple people (including a close family member) who have studied cases where larger physics systems introduced more room for edge-case bugs, instability, and/or visually awkward physics behaviors.

This isn’t "theorycrafting", it’s a fact and one that can be observed across a wide range of systems: clean, focused, well-calibrated models often produce more stable, believable, and realistic behavior than large systems burdened by edge cases, interdependencies, numerical instability, etc., etc..

Sure, more lines of code might reflect capability, but they don't guarantee accuracy or realism without solid fundamentals, "proper" implementation, and careful tuning.

All of this is a given. GIGO. What I am trying to say is that there are *hard prerequisites* if you're going to expect some behavior out of some systems. Like rubber bumpstops will never, ever work correctly if you model them as a linear spring, and so on. Some objective behavior issues can be seen in the suspension and tire behavior in that NetKar video; just because it moves more doesn't mean it's more realistic.

Dpounder420 · May 10, 2025

CobraCat said:
I agree the Miata looked terrible. No suspension movement at all, just flat through the corner like a billboard sign on wheels. Odd because other cars don't exhibit this behavior and look quite convincing while cornering.

the one not too long after (forget what it was) at the imola chicane looked way better and is more rpesentative of what i see and feel when i play ACE. i think the tire flex, both visually and how it feels, is one of the parts they improved the most over the original ac and acc. its on par with rf2 and lmu now for me and might actually feel overall better to me but i cant say for sure yet. feels very controllable on the limit but still with the potential to snap realistically. it feels more solid in a way i cant really explain either. the original ac and acc both felt more solid to me than basically any other sim but then i got into rfactor 2 and started really enjoying the enhanced sense of tire flex and just the general lack of rigidity. this feels like it took the best of both worlds and improved on them so far. best driving feel of any sim basically except maybe just on par with lmu, im not completely sure yet and think its too early to say. there is an added solidity to lmu compared to rf2 that is the same thing ive always felt with ac, especially compared to project cars 2 when it first came out. like a more solid and real connected feeling. for me that used to be the tradeoff between the original ac and rf2, rf2 felt less like i was driving something with bricks for wheels but also that bit less solid and believable outside of the tire feel (even though i also felt that rf2 was more like driving a real car). these differences are super subtle and probably very subjective though, im not sure how to really put it into words. my point is that the driving feel in ACE feels like the best of my favourite sims combined and with an even higher degree of pure realism to the subjective feel, at least for me. performance hasnt been as bad for me as some say but i also havent been racing ai when i play evo, just hot lapping to feel the driving physics. i really cant wait for the ability to put performance parts in cars. i want a delta hf integrale thats fully upgraded to around 5-600hp and weight-reduced to rip around in

TeddyBearx · May 10, 2025

I want a forza game with good physics and FFB, and AC EVO is what I need and dream, I like so much road cars, to hotlap the Nordschleife track, this game is made for me and those who want a racing game like Forza with so much road cars and different cars but with good graphics, sound and FFB.

Kunos I'm with you, you make me life so happy.

mantasisg · May 10, 2025

@ArchTheSecond I think we are talking about simulations, not long ago it was synonimous with realism. Perhaps not anymore IDK

Can someone help me to find ACE onboard video from someone who is actually capable driver and has proper skills and understanding of 911 ? I want to see whats up. I have a hunch that less than 5% of simracers actually are capable of controling such cars at and over the limit. And most of them aren't upploading videos. And most of them are given up on socialising online with mere gamers and overconfident novices.

Updated Assetto Corsa EVO Roadmap Shows Drastic Changes

The Eifel region free-roam​

Modding tools​

Comments

Article information

Share this article

Sabre

KT Racing

Milestone

The Last Garage

BeamNG

iRacing

Straight4 Studios

Bugbear Entertainment

Motorsport Games / Studio 397

Kunos Simulazioni

Reiza Studios

Other (add in the comments below)

The Eifel region free-roam

Modding tools