rFactor 2: BMW M6 GT3 Latest DLC Car Announcement

[Paul Jeffrey]

Following on from the recent Porsche 911 GT3 R announcement, Studio 397 have now confirmed the second of four new cars... the BMW M6 GT3!

Set to form part of a four car #rFchallenges DLC offering for the simulation, the next car confirmed to be heading to the rFactor 2 is the stunning BMW M6 GT3 - due to be added alongside the 911 GT3 R alongside two further as yet unannounced vehicles.

At this very early stage of proceedings we are yet to learn when the new pack is expected to land, nor how much it will cost or even if the two mystery cars will even be from the GT3 category, but one thing is for sure - the next few weeks should most certainly be exciting ones for rFactor 2 players!

rFactor 2 is available exclusively for PC from Steam now.

Check out the rFactor 2 sub forum here at RaceDepartment for all the latest news and discussion with regards to the simulation. You can take part in lively debates with fellow rFactor 2 fans and take part in some great Club and League racing events..! Head over to the forum now!

Like what you see here at RaceDepartment? Don't forget to like, subscribe and follow us on social media!

 ​

 

[Marcel Offermans]

Realistically speaking though, is this even something you can pursue? The original creators who designed and implemented AI logic were unable to address these issues, what suggests that you can? This is in no means malicious either, I would just be interested to hear your perspective on things.

That's a fair question. There is a generic and a specific answer to that question.

The generic one is that "this is software" and in software you can pretty much do anything you like. There might be practical or commercial limits that prevent us from actually doing that, but nothing is impossible. I also don't believe that ISI did not know how to fix this. It is much more likely they never got to it (priority wise).

The specific one is to look at the two things you mentioned:

1. AI not overtaking in the lap they pit. That is probably a very deliberate choice in the code and we would have to investigate why that choice was made. I can think of one thing, and that is that maybe they wanted to prevent a scenario where the AI is in the middle of an overtake when it was supposed to enter pit lane and the car it's overtaking is blocking the entry. A solution to this would be to have the AI be more careful on their "inlap" for this reason. Note that I have not done the investigation yet, I'm just making this up as a plausible possible reason. We could look at then either saying, ah, we don't care, we'll let them overtake and then just have them run out of fuel if they make this mistake, or we can limit the "no overtake" to just the last sector maybe of the lap.
2. AI not using the slipstream. This might be more difficult, as I am not sure how aware the AI is of the different slipstream effects of cars. As you know there are several effects that fight each other when you're following a car. A slipstream effect that will make you faster, but also a dirty air effect that will cause a loss of downforce that will make it harder to closely follow a car. I can certainly imagine that the AI is not aware of the exact effects and how they pan out for different (groups of) cars in front. Again, I have not looked into this.

 

[Marcel Offermans]

NO! But i guess they know what they're doing right? What tells you that the previous team were unable to address these issues also?

If Marcel already said that it makes part of the huge list of things "to do" why are you putting in doubt their capacity to do it?

Just to inform you, we most certainly have developers that know how to program the AI and without getting into much detail (because I'm not allowed to) they were involved a lot in another project we are working on.

 

[Anomalistics]

NO! But i guess they know what they're doing right? What tells you that the previous team were unable to address these issues also?

If Marcel already said that it makes part of the huge list of things "to do" why are you putting in doubt their capacity to do it?

I think you're missing the point, my friend. The original developers of this game were Image Space Incorporated. They released both rFactor 1 & 2. The AI logic on rFactor 2 has always had the slipstream issue, one which was raised to the initial developers and now Studio 397. If ISI were unable to resolve this, I'm almost certain Studio 397 will follow the same path as well.

Marcel is just being polite. If you have your customers best interests at heart, you do not dismiss ideas but instead provide the "possibility" of looking into it, despite it being HIGHLY unlikely.

 

[muzikant]

Marcel is always on the forum to answer question for cummunity people. That says a lot of him
Looking forward to see this sim in furture

 

[Joe Stricklin]

Any chance of getting historic CanAm cars? Also, optimize VR and surround sound in VR.
Thanks

 

[Anomalistics]

That's a fair question. There is a generic and a specific answer to that question.

The generic one is that "this is software" and in software you can pretty much do anything you like. There might be practical or commercial limits that prevent us from actually doing that, but nothing is impossible. I also don't believe that ISI did not know how to fix this. It is much more likely they never got to it (priority wise).

The specific one is to look at the two things you mentioned:

1. AI not overtaking in the lap they pit. That is probably a very deliberate choice in the code and we would have to investigate why that choice was made. I can think of one thing, and that is that maybe they wanted to prevent a scenario where the AI is in the middle of an overtake when it was supposed to enter pit lane and the car it's overtaking is blocking the entry. A solution to this would be to have the AI be more careful on their "inlap" for this reason. Note that I have not done the investigation yet, I'm just making this up as a plausible possible reason. We could look at then either saying, ah, we don't care, we'll let them overtake and then just have them run out of fuel if they make this mistake, or we can limit the "no overtake" to just the last sector maybe of the lap.
2. AI not using the slipstream. This might be more difficult, as I am not sure how aware the AI is of the different slipstream effects of cars. As you know there are several effects that fight each other when you're following a car. A slipstream effect that will make you faster, but also a dirty air effect that will cause a loss of downforce that will make it harder to closely follow a car. I can certainly imagine that the AI is not aware of the exact effects and how they pan out for different (groups of) cars in front. Again, I have not looked into this.

Marcel, thank you for your comments.

Just to enlighten you further regarding the "AI SlipStream". The AI do in fact utilize slipstream, however, if they are behind a car on a long or medium straight, they will effectively park behind the car in front, breaking every 0.2. sec rather than pull out and go for the overtake. I am sure you have seen various demonstrations on the official forums and it is pretty deflating. The above example will be replicated easy for any car with similar speed.

If you do LMP1 vs GTE, however, the AI logic can predict in advanced to take over and they do pull out. So it seems the logic is there, but we just need to incorporate that for when similar pace cars use a slip stream.

 

[Will Mazeo]

https://twitter.com/rFactor2/status/1073627566278434818

 

[Anomalistics]

https://twitter.com/rFactor2/status/1073627566278434818

Now that is impressive.

 

[Emery]

Just to enlighten you further regarding the "AI SlipStream". The AI do in fact utilize slipstream, however, if they are behind a car on a long or medium straight, they will effectively park behind the car in front, breaking every 0.2. sec rather than pull out and go for the overtake.

You do know that there are a several factors that influence this behavior, right? There are track parameters, in the track's AIW, in addition to the AI code. You also have the AI aggressiveness slider in game which is modified by the talent file aggressiveness.

 

[McKiernan]

https://twitter.com/rFactor2/status/1073627566278434818

My fav GT3, was hoping to see this. Would've been happy with other GT3's (aside from the newer Bentley & McLaren) also. Nice to see more variety coming to the class.

 

[Anomalistics]

You do know that there are a several factors that influence this behavior, right? There are track parameters, in the track's AIW, in addition to the AI code. You also have the AI aggressiveness slider in game which is modified by the talent file aggressiveness.

Yes, I am well aware of this. It’s an AI logic issue, not configurational. I have done extensive testing.

 

[Dragon60]

https://twitter.com/rFactor2/status/1073627566278434818

View attachment 281542

 

[tlsmikey]

are you kidding? i have raceroom too, is a very good sim with state of the art sounds, but it lacks several things. weather, rain, day/night cycle, modern tire model, real road, dx11...finished experience??

It blows my mind that people think rF2 is anything close to resembling a finished experience. Come on guys. The frame rates are better than they've ever been but they're still a long ways from good. The content is hit and miss. Some cars are amazing....others are a total embarassment (merc and radical....listen to people like Billy Strange who have driven radicals and how bad it is).

The UI is a mess. The graphics are average at best and the track/car selection is not great (stock content only).

rF2 has a ways to go. I've really enjoyed the GTE cars, but i'm not buying anything further from them until they prove that they are committed to high quality content. If you're going to charge high prices, I expect high quality.

 

[AyrtonForever]

It blows my mind that people think rF2 is anything close to resembling a finished experience. Come on guys. The frame rates are better than they've ever been but they're still a long ways from good. The content is hit and miss. Some cars are amazing....others are a total embarassment (merc and radical....listen to people like Billy Strange who have driven radicals and how bad it is).

The UI is a mess. The graphics are average at best and the track/car selection is not great (stock content only).

rF2 has a ways to go. I've really enjoyed the GTE cars, but i'm not buying anything further from them until they prove that they are committed to high quality content. If you're going to charge high prices, I expect high quality.

post useless, I would reply to those who said that raceroom is a finished product compared to rf2. you will agree with me that this is a ridiculous statement

 

[muzikant]

[QUOTE = "tlsmikey, bericht: 2879155, lid: 238368"] Ik ben ervan overtuigd dat mensen denken dat rF2 allesbehalve lijkt op een voltooide ervaring. Kom op jongens. De framesnelheden zijn beter dan ze ooit zijn geweest, maar ze zijn nog steeds een lange weg van het goede. De inhoud is wisselvallig. Sommige auto's zijn geweldig .... anderen zijn een totale schaamte (merc and radical .... luister naar mensen zoals Billy Strange die radicalen hebben gereden en hoe erg het is).

De gebruikersinterface is een puinhoop. De graphics zijn op zijn best gemiddeld en de selectie van tracks / auto's is niet geweldig (alleen voorraadinhoud).

rF2 heeft een manier om te gaan. Ik heb echt genoten van de GTE-auto's, maar ik koop niets verder dan dat ze bewijzen dat ze zich inzetten voor inhoud van hoge kwaliteit. Als u hoge prijzen gaat rekenen, verwacht ik hoge kwaliteit. [/ QUOTE]

 

[9kret9]

Can't believe Rf2 looks that good now, nice screenshots!

 

[Steve Worrell]

It blows my mind that people think rF2 is anything close to resembling a finished experience. Come on guys. The frame rates are better than they've ever been but they're still a long ways from good. The content is hit and miss. Some cars are amazing....others are a total embarassment (merc and radical....listen to people like Billy Strange who have driven radicals and how bad it is).

The UI is a mess. The graphics are average at best and the track/car selection is not great (stock content only).

rF2 has a ways to go. I've really enjoyed the GTE cars, but i'm not buying anything further from them until they prove that they are committed to high quality content. If you're going to charge high prices, I expect high quality.

I'm surprised you don't think they are producing quality content. The GTE, GT3 and lmp2 have been awesome. I think the sounds (especially exterior) could be improved but other than that I think they are great.

 

[Mankiewicz]

so we'll have the porsche, the BMW, a Ferrari and the audi R8 LMS is that correct?

 

[Anomalistics]

so we'll have the porsche, the BMW, a Ferrari and the audi R8 LMS is that correct?

Where did you see the Ferrari announcement?

 

[Steve Worrell]

Marcel, thank you for your comments.

Just to enlighten you further regarding the "AI SlipStream". The AI do in fact utilize slipstream, however, if they are behind a car on a long or medium straight, they will effectively park behind the car in front, breaking every 0.2. sec rather than pull out and go for the overtake. I am sure you have seen various demonstrations on the official forums and it is pretty deflating. The above example will be replicated easy for any car with similar speed.

If you do LMP1 vs GTE, however, the AI logic can predict in advanced to take over and they do pull out. So it seems the logic is there, but we just need to incorporate that for when similar pace cars use a slip stream.

Marcel - this summary is absolutely spot on. If I could rate this comment agree twice I would.

It's interesting that automobilista suffers with exactly the same issue. It's such an immersion killer when you have been caught by the AI qnd you can artificially hold them off when in reality you would have been passed. Would love this issue to move up the priority list but I appreciate you have a lot on with partners etc like Vive and McLaren which I'm sure are critical business projects.