To Reiza: What physics do you actually want?

And it's on you to assume the knowledge of the physics department of Reiza Studios?

15 years of modding certainly helps but is a different foundation than working on the especially unique driveline and tire model components directly and on this particular engine and knowing its advantages and disadvantages or special needs. You aren't the only person here absorbing information (and putting it into practice) about the topic at hand but you act right now as if you would be.

 

Just so you can rest assured: The team is aware that this is a known quirk and it will be explored further, there are already considerations. But this really isn't a thing to do "by a fingersnip".

 

Sticky diff was deeper than anything else physics wise when it came to problems in the past (a true behemoth of a quirk that wasn't even remotely to be expected to be that deep, i was one of the guys reporting and revealing the issues with driveline back then, i remember well how i was also not believing it really). It was a code issue and additionally falsely set properties actually (also surrounding components).

This is with very high likelyhood not even remotely the case with any other quirk right now due to things being far more explored than back then.

 

It's important to mention that cause and effect don't need to be necessarily of identical magnitude - example: something like faulty F-Trainer tires a few builds ago wasn't anything "deeper" and more a lack of proper fine tuning (emphazis on "fine") but resulting in a big difference.

I can't tell you at this point. I'm hesistant to give any promises wether something is easy to do. Also no timelines.

Due to the nature of the way the engine was adopted by the studio it's also a learning process for the guys directly involved in the process of creation.

The point i start to react allergic is when an issue is associated with lack of knowledge (how it should behave) or ignorance towards feedback.
When there is nothing to say yet, nothing will be said - this doesn't mean that it's ignored or dismissed.

 

To All:

**warning, this is a very long post** so grab your popcorn.

To the degree that I may have contributed to any misunderstanding, behaved in a unreasonably condescending tone I apologize. There is a certain degree of tension in the face of criticism and frustration can come across that way.

One of the most difficult things to do on an internet forum is to maintain a topic when statements are misinterpreted, assumptions/opinions are written to sound like facts and respondents literally put their own words in "quotes" alleging that that is what others have stated. With this context in mind, I stand behind what I have actually stated with regards to default setups and their role to play in the direction of the sim and how cars respond on track physics wise. Replies to any of my statements by others, with interpretations, assumptions or made up quotes or otherwise, are not my statements.

I was asked to review this thread from another forum member and as a Physics Consultant (official title) and proud member of the physics team, (with a not-insignificant role as a *ahem* developer) I felt an obligation in some way to address the claims in as best a way as possible.

First, one must admit that the question of "what physics do you actually want" is as enigmatic as they come, and to be honest I don't think anyone can view this question as anything more than rhetorical.

However, in an attempt to answer this question, I will state again, that this sim is still under development. Characteristics of car behavior are under consistent and steady review. Insofar as any issues can be identified (either from the player base or internally) that can be objectively confirmed internally, (be it a global issue or based on individual car design) once authorized for further review, it is subject for future revision.

It has been observed that many of the cars exhibit similar characteristics ('saminess' was the term?). It has also been stated or implied that the cars are directionless which I take to mean 'all over the place' from a physics perspective. These two observations are polar opposites to one another.

From an internal physics perspective, each and every car is modeled with their RL counterparts in mind. Weight distributions, CoGs, inertias, suspension geometries, front/rear tracks, wheelbases, aero capabilities, CoPs, HP, torque curves etc. etc These characteristics in sum, are the 'natural' characteristics of a car. They have a huge impact on how each car handles for better or for worse.

As far as the default setups are concerned, having one setup for each car that needs to be sufficiently capable for all tracks, from Oulton Park to Spa, from Azure Circuit to the Daytona Road Course, logically and factually presents its own degree of challenges and limitations. When I say that the default setups are created to be as accessible to as many players as possible, from novice to expert, and yet have the capability to still be as quick and competitive as possible it must be viewed in this context, it does NOT mean we are watering down the cars or creating illusions or attempting to force car behavior that is contrary to their nature. It means the cars require to be setup for the trickiest of braking zones, the most difficult of corners, as capable in flowing tracks as they are with tracks that feature sharp chicanes, capable of handling relatively flat tracks as well as those that feature undulations or sausage kerbs, under various track conditions and weather states. This means, conservative/controllable braking, adequate rotation but at the same time maintaining a degree of aggressiveness that can give all players the on throttle satisfaction and perception of edge of grip feel and speed they desire when done properly. So no, the setups are not directionless, and to the degree that there is a "saminess" to them, this should be a sufficient explanation as to why one might feel that way.

With this in mind, for those of you who have the ability to optimize setups to suit their own driving style overall or even fine tune each car for particular tracks, the capability is there. On the other hand, for those of you who wish to sit and drive, rest assured we've done everything possible to make sure you are not faced with a unreasonably large degree of a disadvantage, no matter what track you decide to drive on. Will it be the fastest the car can be, no. Can it get you through a race in a competitive fashion, within a reasonable degree of difficulty depending on the type and nature of the car itself, we certainly hope so.

With regards to specific car examples, I will start with the FUSAs at Indianapolis. It has been observed that the cars exhibit potentially excessive (and recoverable) yaw rates mid-corner and a reluctance to spin. So I return to my original statement. "Insofar as any issues can be identified (either from the player base or internally) that can be objectively confirmed internally, (be it a global issue or based on individual car design), once authorized for further review, it is subject for future revision."

Moving on the the Lotus 98T as an example of an alleged global tire wheelslip/slide characteristic issue. There was a bit of misunderstanding along this threads timeline where the term 'wandering' may have been used to describe different distinct issues and at some point the Super V8 was mentioned, as well as FFB issues.

I have not observed a 'wandering' or swim or sway in a straight line with the Lotus 98T. However, this has been observed as an issue with several other cars. I attempted to describe the cause and effect of this sensation and attribute it to several factors such as FFB, underdamping and excessive unlocking of the clutch plate differential. Grip loss, wheel inputs and load transfers all contribute.

On the other hand I have indeed observed with the Lotus 98T and many other cars, an on power traction loss "kick" or "snap" or "wander", however you choose to describe it. For reasons described below, this characteristic was perhaps a bit too obvious to me and may be the reason this thread hasn't been as clear as it otherwise could have been.

To me, these two scenarios are distinct. The first being a red flag for a type of issues as described above and a potential physics or FFB issue.

The second being a side-effect of torque based traction loss. I think it is rather clear that this traction loss snap is related to the degree which torque is applied to the drive wheels, due to the fact that this characteristic is more and more pronounced the torquier the car you drive. It can be observed in some cars that have a low RPM torque curve, such as the BMW M6 and Nissan GTR GT3, which makes initial throttle inputs snappier but the most explicit examples are FRetros, FClassics and Gr Cs, particular those with turbo lag.

So why do the cars snap laterally, requiring the driver to drop throttle? Well, the tires themselves have longitudinal and lateral grip limits. But you cannot treat them in a bubble in each direction. The tires in AMS2 share a similar characteristics in slip tolerance. Generally speaking this is between 5-20%. Radials being on the low end of the spectrum and Bias Plys being on the more forgiving end of the spectrum. The Lotus 98T carcass, as constructed in the physical tire model is based off of the Goodyear radial provided to cars of this era. The tires have to be relatively stiff in comparison to a Bias Ply type you may see in earlier generations due not only to the vertical load requirements, but lateral load and flex, given the level of grip, so in essence the sidewalls are stronger to prevent lateral collapse. This presents a degree of hardness or sharpness, and limiting the amount of generous on power slide characteristics that a softer radial might provide.

That being said, and as I tried to describe earlier, the degree to which the tires exceed their slip limits in the longitudinal direction, limits their grip in the lateral direction for two reasons. First, as the SETA model simulates it, the rubber itself has some flex in both directions but the kernel of this stretch is limited to the effect that the tire cannot have max rubber stretch in both laterally and longitudinally at the same time. This is again subject to the hardness or stiffness of the tires, as mentioned above with sidewalls being a contributory factor. Second, as the grip limits are exceeded, 10% and beyond, the rubber itself looses natural adhesiveness, as simulated, the rubber no longer has the capability to sink into the natural asperities of the track surface.

Therefore the sharpness of this snap is a direct consequence of throttle inputs and torque exceeding the capabilities of the tire quicker than the driver can react.

This, I believe explains why the cars are doing what they are doing and the inherent difficulties in coming to terms with cars with upwards of 1200hp at peak turbo.

The tires indeed have limits that can easily be exceeded, very quickly with these types of outputs. The engines and drivelines in AMS2, with their torque and HP curves are also simulated rather precisely, this includes realistic levels of turbo pressure, fuel consumption and other factors. Peak torque for cars of this era are sometimes reliably documented (sometimes not) as well as peak HP, and both of those peaks are set to their documented RPMs. What cannot be exactly determined are set to realistic levels based on other factors such as documented engine capacity.

This is not a failure in the physics or a bug. In fact it may be an example of the physics working as intended. In other words you have a static tire, with its defined limits and you have the engine and its defined limits. The only other X-Factor is power application.

Certainly, drivability of the cars is and was a concern in real life, however they did not have telemetry until the late 80's I believe, so race teams had to make due by adjusting gearing, boost pressure, fuel consumption, packing as much downforce as possible on the cars and good old fashioned driver skill. And they certainly didn't run the car at max dyno levels for various reasons. The teams also would adjust their car for every track and to suit the needs of each of the two drivers.

There a few onboard resources for the Lotus 98T, the best one being Johnny Dumfries at Adelade circuit. One of the things that stands out the most is how much time is spend coasting through corners and how little Dumfries spends at 100% throttle and if he is at 100% its in a straight line. You can literally see the front tires sliding across the track at mid-corner to exit. This indicates a rather stiff front rebound damping. Also the gearing seems rather suited particularly for Adelade, (Dumfries is using the standard 6 speed transmission, ours features Senna's 5). What cannot be determined is the boost pressure, fuel mode or any other factors that would keep the car within the tire limits.

So how do we balance these realistic factors along with their real life risk scenarios (engine failure, crashes and death) with a sim racer's expectation of difficulty and balls to the wall, no risk/no concern approach, where all that matters is a TT leaderboard time. Should we limit the car's capabilities to fall squarely within the grip limits of the tire or present the car raw, with all its inherent difficulties and challenges with all modes set for challenging levels of performance and let the players decide how they want to drive or tune the car?

This last paragraph is an example of the design choices that have to be made when developing cars. And I agree with it. Present the car in all its glory and let you guys decide. Give you guys a decent default setup that you can drive with the car's own inherent sim racing type challenges, skill floor and ceiling. You can decide to max the engine out to it theoretical capabilities, detune it, under drive it, over drive, use the defaults, or tinker. Its all part of the experience IMO. Another option is to drive the defaults, overdrive the tires, compare the car to that of other games or mods and criticize the physics of the game or the developers themselves.

When I drive the Lotus 98T, I choose to adapt the car to my capabilities as a driver and also use it to learn throttle patience. Learn how the car responds to different throttle inputs at various levels of lateral g.

You may disagree with this approach, you may not. But ultimately the Lotus 98T has gone through a rigorous review process, including being driven by all levels within the Reiza organization. Certainly there is a range of opinions about the car internally as well, but this representation of it has reached a degree of acceptable consensus within Reiza with respect to the car's natural capabilities, and respective of its competitiveness within the racing category. Furthermore, if there are marginal improvements that can be implemented, via new discoveries or physics philosophies in the future, the Lotus 98T, among all the other cars in the game will certainly become a benefactor at the discretion Reiza.