badge system sucks

[ju1cydrag0nfru1t]

i absolutely hate the badge system. i did about 6 races and got crashed into by only other cars(non avoidable) and got from clean driver to !. and now i only get in with people who are even worse and i cant even try too defend or i am going to die. its unbelieveble.

 

[ecknine]

You say 'non-avoidable' but accident avoidance is a thing. Would love to see a replay of all 6 races.

 

[Shador]

I started as a green badge, platin 3 player 4 Races in Atlanta. I'm on the way to S2. At least half of the grid has a red badge. For what is this badge system when i has to race with them???
All 4 Races a red badge player wants to go through me in the first corner, so we crashed. I lost no safety rating, i also keep my green badge, but S2 rating isnt possible with this system.
I would love to race with other green badge players, doesnt matter the speed of them. Please change this.

 

[Nectaria Coutayar]

i absolutely hate the badge system. i did about 6 races and got crashed into by only other cars(non avoidable) and got from clean driver to !. and now i only get in with people who are even worse and i cant even try too defend or i am going to die. its unbelieveble.

6 incidents that were non-avoidable, just means you're lacking basic driving and situational awareness skills.

 

[SaltyDog]

Though I am sure there is room for improvement and fine tuning, the badge system seems to have worked remarkably well through my progression from rookie to S1 safety rating. Almost all of this time I have held the green trusted badge and most of my races since acquiring it have been solid green badge races. The racing has been far safer than the racing I have experienced in iRacing, and I suppose this could be coincidental, but so far the badge system seems to work better than I had expected.

I don't know if your 6 races were truly unavoidable contact, a bit of bad luck or partly your fault, but I am sure that if luck played a part in it that it will equal out over time. If you have thoughts on how the badge system could be improved I would be interested to hear your ideas. Continuous improvement should be the goal of the developers, especially if this is to be a largely subscription based service, but complaining about how bad it is without also offering thoughts on how to improve it isn't all that helpful.

 

[Felix Perez G]

What’s the point of striving for a Gold Safety Rating if you’re allowed to compete in Gold series with orange and red licenses? This is the result of a race yesterday at Road Atlanta: I went from P20 to finish in P6—out of 54 cars on the grid, only 25 of us finished...

 

[Blue_Monkey]

What’s the point of striving for a Gold Safety Rating if you’re allowed to compete in Gold series with orange and red licenses? This is the result of a race yesterday at Road Atlanta: I went from P20 to finish in P6—out of 54 cars on the grid, only 25 of us finished...

these are 2 different systems for 2 different purposes:
badges track short term
SR tracks long term driving standards

the issue is a different one, the long term one is too easy to gain, SR needs to be stricter

 

[dan1714]

I feel the incident responsibility needs to be attached. My GT3 car gets rear ended off the track by a hypercar, for example, that's the responsibility of the hypercar, so why the hell am I getting hit with responsibility. Contact is NOT legal in sports car racing, why is it acceptable in LeMans Ultimate? It's ridiculous that I have maintained a P3 safety rating for a year and in the last 2 weeks I've gone from a diamond badge to a red badge because drivers are hitting me, because they don't give a damn about their ratings, or it's too easy to get into silver or gold events.

 

[Blue_Monkey]

I feel the incident responsibility needs to be attached. My GT3 car gets rear ended off the track by a hypercar, for example, that's the responsibility of the hypercar, so why the hell am I getting hit with responsibility. Contact is NOT legal in sports car racing, why is it acceptable in LeMans Ultimate? It's ridiculous that I have maintained a P3 safety rating for a year and in the last 2 weeks I've gone from a diamond badge to a red badge because drivers are hitting me, because they don't give a damn about their ratings, or it's too easy to get into silver or gold events.

think a few steps ahead
you can't design a reliable system that determines fault for now ... even the Stewards currently get it wrong a lot of times ...

that said, RED badge is usually a sign that you fight/lap too aggressively!

 

[dan1714]

think a few steps ahead
you can't design a reliable system that determines fault for now ... even the Stewards currently get it wrong a lot of times ...

that said, RED badge is usually a sign that you fight/lap too aggressively!

You really can. They can detect an event that should cause the car to lose traction, and as a result they can attach responsibility to it. The software can determine if a car is going to be able to make a turn based on speed, entry angle, tire and track temps and a plethora of other items. So based on that, if the passing car would not be able to make a turn based on that, without hitting another car, bingo you have instant incident responsibility. It is ALWAYS the passing cars responsibility to complete a clean and safe pass, so I don't see where this is so tricky. "I am faster than you so you need to get out of the way!" Is a gamer attitude. A racer knows that each car has equal right to the racing surface and respects the cars around him.

 

[SaltyDog]

What’s the point of striving for a Gold Safety Rating if you’re allowed to compete in Gold series with orange and red licenses? This is the result of a race yesterday at Road Atlanta: I went from P20 to finish in P6—out of 54 cars on the grid, only 25 of us finished...

As I understand it, and I could be wrong, but the SR measures your historical driving safety over an extended period of time, and is used primarily to determine what car classes and race types you are eligible to drive. DR is used to measure driver skill and the system primarily uses this to sort drivers into splits so that similar skill levels are driving together as much as possible. This is similar to the way iRacing manages SR and IR (DR in LMU), though the math behind them isn't the same. The driver badges look at driver behavior and safety over a shorter window of only 10 races, and are used to sort drivers with similar DR into splits with more similar recent safety history. So for example, if 40 drivers have similar DR, but 10 of those drivers are Green badges, they end up in the same split before filling in the remainder of the field with the next best driver badge (blue/good) and that sorting continues through the badge colors until the field is full.

The problem with SR alone in both games, is a blind spot that exists in both games when judging driver safety using only a long horizon metric; using only long horizon metrics allows the player to game the safety system by using the SR rating like a currency to be spent. When the safety rating is high, a driver can choose to drive much more aggressively, carelessly or even maliciously with little impact or consequences over a race or even a few races. The badge system in LMU accounts for this by also comparing a short horizon metric to expose the aggressive/malicious/careless behavior and tries to cull those drivers out of the best splits. How well this works depends in part on there being enough drivers overall, and enough with higher badge levels to sort effectively into high skill, high safety splits. If there aren't enough registered players with a green badge, then it's not possible to prevent padding out the split with progressively worse badge level drivers.

The system isn't perfect, but it is much better at preventing unsafe drivers from being chronic offenders than the SR alone system used by iRacing. I think I have how the system functions at a high level correct, and the explanations make sense, and I am sure if I have any of the details wrong someone will help with a correction, but maybe this helps a little.

 

[Blue_Monkey]

You really can. They can detect an event that should cause the car to lose traction, and as a result they can attach responsibility to it. The software can determine if a car is going to be able to make a turn based on speed, entry angle, tire and track temps and a plethora of other items. So based on that, if the passing car would not be able to make a turn based on that, without hitting another car, bingo you have instant incident responsibility. It is ALWAYS the passing cars responsibility to complete a clean and safe pass, so I don't see where this is so tricky. "I am faster than you so you need to get out of the way!" Is a gamer attitude. A racer knows that each car has equal right to the racing surface and respects the cars around him.

No, you can't, at least not reliably with conventional coding.

What if the loss of traction happened because the car ahead moved under braking? Weaved? Turned in unexpectedly and forced the passing car to avoid? I came up with 3 valid cases in 5 seconds, and there are many, many more.

The problem is not detecting that a car could not make the corner. The problem is determining why. To handle that reliably, you would need an enormous number of algorithms and rules covering all those edge cases and their interactions.
The required development effort would be completely unreasonable. You could spend 10 years building a traditional rule-based system and it would still get situations wrong. Just look at racing AI today: changing conditions, race starts, unusual situations, mistakes and interactions between cars are still difficult to handle properly.

AI/ML/GenAI-based systems may eventually make this feasible with reasonable development resources, but that remains to be seen. It is nowhere near as simple as "the car couldn't make the corner, therefore the passing car is responsible."

 

[dan1714]

No, you can't, at least not reliably with conventional coding.

What if the loss of traction happened because the car ahead moved under braking? Weaved? Turned in unexpectedly and forced the passing car to avoid? I came up with 3 valid cases in 5 seconds, and there are many, many more.

The problem is not detecting that a car could not make the corner. The problem is determining why. To handle that reliably, you would need an enormous number of algorithms and rules covering all those edge cases and their interactions.
The required development effort would be completely unreasonable. You could spend 10 years building a traditional rule-based system and it would still get situations wrong. Just look at racing AI today: changing conditions, race starts, unusual situations, mistakes and interactions between cars are still difficult to handle properly.

AI/ML/GenAI-based systems may eventually make this feasible with reasonable development resources, but that remains to be seen. It is nowhere near as simple as "the car couldn't make the corner, therefore the passing car is responsible."

Each of your scenarios provides information from both cars to determine loss of traction. How is loss of traction easier to determine than cause of contact. Moved under braking is legal and allowed. That's a gamer term. I am allowed to turn into the corner anytime I want provided you are not significantly along side me. The passing car is responsible for a clean and safe pass. Period. It is not the responsibility of the car in front to move out of the way.

 

[Blue_Monkey]

Each of your scenarios provides information from both cars to determine loss of traction. How is loss of traction easier to determine than cause of contact. Moved under braking is legal and allowed. That's a gamer term. I am allowed to turn into the corner anytime I want provided you are not significantly along side me. The passing car is responsible for a clean and safe pass. Period. It is not the responsibility of the car in front to move out of the way.

you seem to be short of reading accurately and thinking it through before replying

 

[dan1714]

As I understand it, and I could be wrong, but the SR measures your historical driving safety over an extended period of time, and is used primarily to determine what car classes and race types you are eligible to drive. DR is used to measure driver skill and the system primarily uses this to sort drivers into splits so that similar skill levels are driving together as much as possible. This is similar to the way iRacing manages SR and IR (DR in LMU), though the math behind them isn't the same. The driver badges look at driver behavior and safety over a shorter window of only 10 races, and are used to sort drivers with similar DR into splits with more similar recent safety history. So for example, if 40 drivers have similar DR, but 10 of those drivers are Green badges, they end up in the same split before filling in the remainder of the field with the next best driver badge (blue/good) and that sorting continues through the badge colors until the field is full.

The problem with SR alone in both games, is a blind spot that exists in both games when judging driver safety using only a long horizon metric; using only long horizon metrics allows the player to game the safety system by using the SR rating like a currency to be spent. When the safety rating is high, a driver can choose to drive much more aggressively, carelessly or even maliciously with little impact or consequences over a race or even a few races. The badge system in LMU accounts for this by also comparing a short horizon metric to expose the aggressive/malicious/careless behavior and tries to cull those drivers out of the best splits. How well this works depends in part on there being enough drivers overall, and enough with higher badge levels to sort effectively into high skill, high safety splits. If there aren't enough registered players with a green badge, then it's not possible to prevent padding out the split with progressively worse badge level drivers.

The system isn't perfect, but it is much better at preventing unsafe drivers from being chronic offenders than the SR alone system used by iRacing. I think I have how the system functions at a high level correct, and the explanations make sense, and I am sure if I have any of the details wrong someone will help with a correction, but maybe this helps a little.

Although I agree, to some extent, myself running in GT3 and being taken out at Road Atlanta multiple times per race by hypercars (to the tune of 198 seconds each pitstop) is really frustrating. Having to wait for a safe place to pass slower traffic is part of what makes Multiclass racing great. It's using your brain not just raw lap time.

 

[SaltyDog]

Although I agree, to some extent, myself running in GT3 and being taken out at Road Atlanta multiple times per race by hypercars (to the tune of 198 seconds each pitstop) is really frustrating.

As frustrating as that is, its not completely unlike real life sometimes; if you watched the Petit Le mans yesterday you would have seen a lot of the same kind of carnage in real life. What a mess that race was. Good thing they didn't have a SR or driver badge to protect I guess

 

[Babylon]

you can't design a reliable system that determines fault for now ... even the Stewards currently get it wrong a lot of times ...

They could with some, If you get rear ended out in the real world it's not down to you, Even if you braked it's not down to you, The system knows you have been hit from behind and it knows how hard because it gives you damage, So unless I stuck it in reverse, That's not down to me.

It knows you hit a car because you have front damage, And it knows the car you hit had side damage, So it could from that calculate that the fault is really with you, Even if by some freak car play it wasn't your fault, The other car hasn't got away because he has damage.

They have crash investigators to work out who was at fault, Were the damage is on one car, To were it is on the other, Were did car A come from and were was car B going, Most people from just looking at a picture of a car crash can work out fault to a reasonable certainty.
So it wouldn't be that hard for the system to assert blame to a reasonable high accuracy, And even if it did get it wrong sometimes, It would still be better than what we have were you would get punished if a car fell out the air and landed on your roof.

 

[Blue_Monkey]

They could with some, If you get rear ended out in the real world it's not down to you, Even if you braked it's not down to you, The system knows you have been hit from behind and it knows how hard because it gives you damage, So unless I stuck it in reverse, That's not down to me.

by that logic Lewis Hamilton should have gotten the penalty in the Formula 1 2021 Jeddah Grand Prix, instead of the brake checking Max Verstappen ...
the rest of your argumentation implodes thereafter ...

 

[SaltyDog]

 

[Babylon]

by that logic Lewis Hamilton should have gotten the penalty in the Formula 1 2021 Jeddah Grand Prix, instead of the brake checking Max Verstappen ...
the rest of your argumentation implodes thereafter ...

That is such a stupid example, That's in a race setting with live FIA stewards, For every 1 brake test rear end there must be hundreds of normal rear ends, And I would rather an automated system got it wrong once, As opposed to hundreds times.

Probably too much for me to assume you read my comment right to the end were I said.
"And even if it did get it wrong sometimes, It would still be better than what we have were you would get punished if a car fell out the air and landed on your roof"

But yer, Lets not try to come up with suggestions, Lets just keep blaming everybody all of the time, Whatever is easier for S397.