Few discussions in sim racing are as universal as those that arise after an accident. Whether we’re talking about a casual race between enthusiasts or a high-level competition, all it takes is two cars making contact for the question that has haunted simulator developers for decades to resurface. Who was at fault?
The question seems simple when we watch a replay in slow motion from the comfort of our chair. Yet anyone who has spent time competing online knows that reality is far more complex. What one driver considers an aggressive move, another may see as a legitimate defense. What one calls an irresponsible divebomb, another interprets as a brave overtake. Between the two accounts there usually exists a grey area where certainties vanish and interpretations begin.
That is precisely why the direction Assetto Corsa EVO is taking with its new Safety Rating system is so interesting. Beyond adding a simple classification to separate clean drivers from troublemakers, Kunos appears to be attempting something far more ambitious: building a system capable of analysing incidents and assigning responsibility. In other words, moving closer to what many consider the true Holy Grail of sim racing.
The Reason iRacing Never Wanted to Point Fingers
The iRacing incident system is frequently criticised for penalising everyone involved in a contact regardless of who caused it. It is an understandable complaint. There is nothing more frustrating than watching your Safety Rating suffer after being the victim of someone else’s mistake. However, behind that philosophy lies a logic that often goes unnoticed.
The iRacing system does not attempt to assign blame because it acknowledges that doing so reliably is extraordinarily difficult. Rather than acting as a judge, it functions as a risk counter. The goal is not to punish the offender but to measure how many incidents surround a driver during their races. From that perspective, individual responsibility takes a back seat, and what matters is incentivising all participants to avoid dangerous situations, even when they have the sporting right of way.
It may seem unfair, and in many cases it is, but it also carries an enormous advantage: consistency. Players know exactly how the system works and what consequences each contact will have. There are no complex interpretations or decisions that are difficult to understand. The rule is the same for everyone.
The Appeal of a System That Punishes the Guilty Party
The proposal from Assetto Corsa EVO is so seductive precisely because it attempts to resolve that sense of injustice so many drivers have experienced at some point. The idea that a system could distinguish between aggressor and victim connects with a very human aspiration: that effort and fair play receive a fitting reward.
On paper, the concept makes a great deal of sense. If a driver miscalculates a braking point and hits the car ahead, it seems reasonable that the greater part of the penalty should fall on them. Likewise, if a driver makes a clearly illegal move and causes an accident, it is logical that the system would identify them as the primary responsible party.
The problem emerges when we leave the obvious examples behind and enter the real situations that arise during a race. That is where the boundaries begin to blur. What happens when two drivers contest the same racing line? What if one leaves insufficient room while the other attempts an optimistic overtake? And what happens when both commit small errors that ultimately lead to contact? In those scenarios, even the most experienced observers tend to hold differing opinions.
When Even Humans Cannot Agree
Real motorsport offers a very valuable lesson for understanding this problem. Sporting stewards have access to cameras from multiple angles, detailed telemetry, radio communications, and in some cases statements from the drivers themselves. Even so, it is not uncommon for a single incident to generate debate for days.
We have seen it across every imaginable category. One group of fans considers the penalty excessive. Another holds the exact opposite view. Even former drivers and experts rarely agree on every incident. Interpretation is an inseparable part of the sport. That is why it is so risky to assume that an algorithm can find a definitive answer where human beings continue to disagree. Computers are extraordinary at identifying patterns and analysing objective data, but races are full of contextual nuance. A driver’s intention, their prior history, the moment in the race, or even the reaction of their rival are all elements that frequently influence how a manoeuvre is assessed.
However, focusing the debate solely on the assignment of blame would be to oversimplify Kunos’s proposal. Perhaps the most promising aspect of the system is not its capacity to punish those responsible for an accident, but rather the way it intends to reward close, clean racing. For years, some reputation systems have allowed drivers to progress simply by avoiding any interaction with others. All it took was staying out of trouble, running alone, and accumulating lap miles without incidents. The result was a curious paradox: some drivers earned excellent ratings without ever having truly demonstrated their ability to race wheel to wheel.
The philosophy inherited from Assetto Corsa Competizione seeks precisely the opposite. Proximity is no longer a risk to be avoided but a skill deserving recognition. In a sense, the system tries to reward what defines the best online drivers: the ability to fight for position over several laps without turning every overtake into an emergency. The rise of artificial intelligence has reignited the debate about whether we are finally close to finding a definitive solution. The short answer is probably not, at least not in the immediate future.
Artificial intelligence can be enormously helpful in detecting aggressive behaviour, identifying dangerous driving patterns, and analysing thousands of incidents in a matter of seconds. It is a powerful tool and will likely play an increasingly important role in online competition systems. However, turning it into an infallible referee remains an extraordinarily ambitious goal.
Racing is, at its core, a constant interaction between human beings. And human beings have an extraordinary ability to generate unpredictable situations that do not fit neatly into any predefined rule. Every time we believe we have found a universal solution, a new exception appears that forces us to reconsider everything.
Perhaps the most common mistake in this debate is thinking that the objective is to achieve perfection. The history of sim racing shows that no system will likely ever be able to correctly determine responsibility in one hundred per cent of cases. The relevant question is not whether an algorithm can be perfect, but whether it can be better than what we currently have.
You can purchase Assetto Corsa EVO from our links for about 20 euros:
See you on the track!
This website uses affiliate links which may earn a commission at no additional cost to you.
