You could also add a collision group that works like this…
If a user is only within the radius of one other car (no walls, etc…) then he purposely crashes, the car he crashes in to will become invincible… Its hard to explain.
So if the user is only within one other user AND NOTHING else then he turns in to the other car this would be considered illegal activity (therefor the other car would go invincible)
Why so the user can rejoin and then troll again? Vote Kicks are typically only good if your game has 1,000’s of people… But then the user will just rejoin and troll a different set of groups…
In NASCAR, you’re constantly up along side people. If I was just going to make it so two cars that are side by side can’t click against one another, then I’d just make all the cars unable to touch each other. This takes out pretty much 90% of the racing aspect, and makes the game essentially just who can go around the track quicker.
That’s why I said you can only collide with other cars if there is more then one car near you. There is no reason for you to turn in to another car if you have all of the open room you need.
It would help if you started logging events in your game. Compare how many crashes this user has compared to actually bad drivers, and the circumstances in which they occur. Once you have that data, you can sift through it for trends that should make it easier to automatically identify malicious drivers.
Except in NASCAR they drive on tight tracks and are always very close to each other. Sometimes accidents happen. This is isn’t supposed to take away accidents. Only purposeful incidents.
Premium servers (pay 100 R$ or so), so if he does buy in and troll on a premium server, you can be compensated for your time spent banning him. Your loyal base could enjoy a little more peace in those premium servers.
Matchmaking so he’d be balanced with poor performing players. Your skilled players would probably have a better experience because of it. Your poor performing players are still learning the game, so they’d probably get less bummed out by a troll.
Shadow bans – let him play the game, but don’t have him interact with other cars, be visible to other players, or even rank on the leaderboards (you can show him his race time/would-be rank, though). The shadow ban would presumably be sticky for a period of time even after showing good behavior. This is hard to circumvent because you don’t get immediate feedback of what is considered “good behavior”.
Good luck! It seems really hard to distinguish accidental bumping from intentional bumping, but there’s probably some other heuristics to use.
They’re overzealous with regard to friends who may occasionally game on each others’ Wi-Fi (also, siblings).
Some people have ISPs who switch their IP often, at which point IP bans are ineffective except for banning random people.
Admittedly, they’re a lot easier than some of the other tricks, but you have to hand hold them when they inevitably start banning people you didn’t intend to ban.
The issue with that is how do you define malicious. Of course driving backwards is easy. But what about people that just rear end people to crash them? Or what about people that brake check people to have other people rear-end them, giving those users a penalty?
That’s what metrics will help find out. Avg crashes will be X, avg crashes for bad drivers will be Y, and number of crashes for this one malicious user will be Z. If Z is significantly higher than X and Y, then that driver can be flagged as malicious. It will never impact the average user because it’s checking for a quantity high above the average.
The issue is that in racing games implementing such a system either makes it impossible to race without constantly being kicked for malicious driving, or allows for so much malicious driving that racing isn’t fun. This is because accidents that can be seen as malicious, such as rear ending someone, spinning out and blocking the track, etc happen fairly regularly even if you are a safe driver. Then one people find out what the limit is, they just have some races where they drive slowly around the track to improve their statistics, just to treat every few races like a demo derby.
I don’t think you understand the concept of average and outliers. This user is so far past the normal that others are recognizing that his behavior is different. If he decreases the number of accidents he causes to avoid the limit, then everyone wins because now he’s only bad enough to potentially pass as a legitimately awful driver.
No, there is a massive difference between being a malicious driver and being a legitimately awful driver. Legitimately awful drivers have a hard time staying on the road, and as a result never are close enough to the front to ruin others races, outside the first couple turns. A good malicious driver can get themselves close enough to the front to take out other legitimately good drivers, and well as cause massive pile-ups that can ruin the race for all of the drivers.
It doesn’t matter how the user decides to drive. If the malicious user lowers their crash rate so they’re no longer an outlier, then they’re less distinguishable from a legitimate driver and their impact on the server will be a lot less painful.