The Unrealistically Difficult Real Cars Handling Thread... or just incorrect real life physics

These days it has become very usual to promote opinion that "real cars aren't so difficult", there is no lack of that. Many very respectable and competent people talk about that. I don't disagree with that, in fact I agree with that. I am sure so do you. However, what does "easy" or "hard" tell ? Nothing, absolutely nothing. It is subjective for everybody, especially when it comes down to subtle little things, that might sometimes mean big differences. People has different perceptions, different skills and knowledge and observation levels, different ages, different hardware and performance of software, different controllers, different settings, different motivation levels, finally various different conditions which also has an impact on concentration...

In my opinion at the present it has become very agreeable that cars are and should be easy and easier, and easier, it is most easily perceived as improvement, and making anything worse is extremely hard to perceive as improvement, even though it is possible that simulation can be and is likely to be easier at some variables, it is still extremely unintuitive to wish for something to be "worse". As a rule, and I am sure devs know that, do something likely too hard and it will be very disliked, do something likely too good comparing to reality - no one will complain, probably appreciation will be just stronger. It is just natural, although might not neccesarily be more true, just more welcome.

As a reaction to that, I am suggesting not to only speak how brilliant handling should be, but also how bad it could be as well. This way maintaining healthy balance between believing in brilliance of car handling and critical view at possibilities of it going wrong and understanding reasons. Both perspectives are absolutely necessary for correct, realistic simulation.

I thought that it would be funny way to do it by showing examples of cars "misbehaving" from real life, and pretending it is not real life, but "incorrect simulation". Seriously, how incorrect could real life possibly be ?

 

crash #1: -Unrealistically low friction astro turf.
crash #2: - Tires unrealistically has no grip, urealistically low slip ratio, way too hard.
crash #3: - Again no grip... too little downforce, sliding friction much too low, unrealistic and undrivable.
crash #4: - Tires are way too sharp, bad FFB does not help, kerbs upset car handling too much.
crash #5: - No downforce at all, FFB did not warn about oversteer early enough, way too slippery, there is no play with a car, much too unforgiving, if car would drive like that IRL then driver would just beat the engineers and would refuse to drive it. There is surely no fault by the driver, no one could drive that car.

 

Another example:


Grass is way too bouncy and too slippery. Not enough grip. But then immediatley there is way too much grip and launches car off as driver is not being helped by unrealistically unrealistic FFB to point front wheels where they should be pointing in correct moment. Very inconsistent physics, very unrealistic, iceRacing 2.0.

 

There are good mods and bad mods.

Graphics are the obvious aspect of a mod, we all see the images, and there are some really good tools available to develop with.

Physics are quite a bit more challenging. There are some modders seem to lack the understanding of how suspension geometry works and even fewer that have access to the actual dimensions of the suspension for a car they're building. rFactor has a pretty good tool (PTool) for laying out chassis, suspension, hinges and flex. I've opened a number of chassis.ini files on cars, and some are so far from anything found on a car it sometimes makes me chuckle.

I'm not a modder, my racing background come from actual competition that includes engineering and design of chassis, suspension, tires, engine and gearbox development along with tuning, and I've logged in the area of 100,000 miles as a driver. I took up rFactor in 2007 when I retired from racing. That said I've seen some really great mods and some really terrible mods, sometimes both ways from the same modding sources.

I do tinker with a few of them when I have time, but won't share them with others out of respect to the modding teams.

So indeed some of the mods tend to drive like iRacing or even Mario Karts despite the impressive graphics.

Simracers are a mixed audience. Some are of the old guard preferring H-Pattern shifting and vintage cars while some are of the more contemporary breed wanting traction control and limited tunability. Nothing wrong with either one its all about what entertains folks, but those who prefer the latter tend to let driver aids mask the issues of poor physics.

 

Sorry, nice post, but you missed the aim of this thread a lot. The aim is to show difficult side of real life driving, and comment it in such way like it would be something unrealistic happening.

 

This is one of the big problems with rF2 :

Having both the best physics engine and being a modding platform, which means that a very significant part of the cars available in the game do not fully use real data, the latest rF2 technologies and properly used, possibly some help from real pilots/engineers/brands.

As far as I'm concerned, all of this is essential to exploit the potential of the best physics engine on the market.

So I only use by choice the 40 100% S397 cars from the last 4 years.

Also, because these 6 cars were made with real data, real drivers, real engineers and a very great rigor, I also use the 2 karts from Kartsim and the last 4 cars from Slowmotion.


Realism above all (or if you prefer, I give myself every chance to have only realistic content).

EDIT

I'm responding to John, but I don't think I have much to add to the thread specifically, sorry.

 

Rest in peace, thread. You lived short life.

 

Intresting thread, would I have known there will be a thread like this i would have bookmarked a lot of videos, like this:



only first minute of this video is interesting. But what happened here with his Porsche 996 GT3 at Nordschleife Schwedenkreuz? I guess losing grip cause of prominent hilltop and lifting throttle quick (unbalancing the car) ? Though he did the same lifting throttle at hilltop of Flugplatz few seconds before without a problem (?)

(EDIT: though you can see in slow motion he not only was lifting throttle, but additionally was braking -> i think this was to much unbalancing the car with low grip)

 

I think the cuarantine really gave wind to the whole "race cars shouldn´t be THAT hard", thanks to iRacing


This one is a classic. Also, Grosjean managed to save his car from a massive spin today.
Edit: yeah, F1 cars are not drift machines; the rear should´ve remained planted or the car should´ve spun completely out of control. Real world physics are FLAWED

 

I just came back from a track day, same car, same track as rF2. Friends used to say I have the rumble strip force way too high. After today's session I'm more convinced than ever that real life pianos are the ones that are too rough, I'm just catching up with that.

 

Now that I've been enlightened a few comments to mention.
Passing is the most difficult actions both in real life and in the sim. In real life the senses are far more amplified, particularly when you know your opponent. You can often hear tires of the opponent losing grip and generally have a much better peripheral view reaction. Its taking or losing title to the line where its hardest.

Understeer/Oversteer is a given, its what happens when it leads to an off track excursion from what I gather you're interested in and there are so many dependencies with that.

Surfaces are vastly different; on a summer morning when the grass is wetted with dew it still varies widely. On the infield at Charlotte Motor Speedway the clay isn't too soft so its like being on ice. At Mid-Ohio or Roebling road the sandy soil is soft enough that a bit of sideways action even on slicks is woefully unpredictable; it may slide, it may bite, and on a tube frame open wheeler like a Formula Ford the chassis will often dig in; not so much a problem on a flat bottom or monocoque car. At those two tracks the driver is merely a passenger with driving left to the angels.

Gravel is generally predictable after the first 100 milliseconds or so. At least until you reach any raked furrows. Fine pea gravel or crushed limestone is the worst, you have no grip so its best to just lock the brakes and hope for the best.

Rain isn't so bad once the track is wetted. At first the oils/glycols that soaked into the track begin to lift becoming like patches of black ice, its worst through the center of the racing line in turns, its best to use the inside front wheel as a "feedback sensor" for that. Most of us can feel that on the track, but its much harder to feel on a sim.

Puddles aren't so bad unless they happen to be on the exit of a turn. Most drivers can see or at least prepare for puddles on entry and in the event of a slip can compensate a bit on the exit. Throttling down on the exit and hitting a puddle begins an oversteer that can be worsened with trailing throttle oversteer so it takes a perfect modulation of the throttle, backing off just enough to match torque of the tire drag to catch traction.

We set up our cars to deliver the best lap times with as little reserve as the driver feels necessary.

 

Very interesting to read (though I guess @mantasisg created this thread to show/discuss real life examples that real life racing isn't that easier than rF2 simulates). But I appriciate a lot your input! There are only a few knowing what they are talking about when it comes to realism and real racing vs sim racing.

* According to passing. Passing was the most difficult thing when i was league racing. I just didn't know/feel where my opponent is. This changed a lot by using VR now.
* According to dry racing. I think rF2 physics and tire model is not that far off.
* According to wet racing. There i guess is a lot to do by S397 so this gets realistic. Puddels? We don't have them yet in rF2.

 

@vittorio Yes, thats good example of real life physics being broken. It was truly weird how he lost that Porsche. I think you are right, it was a combination of braking when car was unbalanced, plus it did seem to bounce a bit, also it may have not have any downforce. However true reason is that real life physics are incorrect, car shouldn't do that

@Mitch9 You are 1000% correct, it is iRacing that become simracing anti-hero with its over the top difficult handling. I have played iR last time long time ago, but I agree. However, this quarantine also highlighted that there is next to no tendency to talk that something might be too easy, at least I haven't hear much of such comments from real drivers or so. Nothing is too easy then... yeah right

Nice save by Kimi, now thats a pro, but I think we can agree that real life physics appeared to be incorrect right there, rear of the car should have remained planted, I think we all can agree with that Real life physics incorrect - proved Although we could use such video example in opposite thread to this as well, as it shows good handling (possible to save) too. But opposite thread is not needed because nobody talks of anything beeing too easy anywhere.

@atomed Exactly. Real life is sooooo broken. I totally believe that they have rumble strips settings way too high in real life in many tracks, unrealistically high.

@John R Denman Nice post as always, plenty of information, and things potentially leading to deeper and deeper converstions. To make a short statement, I believe that RL is/would be slightly easier than simulation. But that just proves the point that it can still be pretty difficult and leave even best drivers off-guard as it can be seen in many examples documented in footage and more sources. I don't want to apply anything to any particular sim, but to general understanding as I often see people IMO underestimating challenge of driving at high pace and consistently in real life.

 

Slower modern race cars or "bit faster" like GT3 made with focus on gentlemen drivers are not difficult to have decent lap times but they are difficult to be on the elite and that's very clear to see IRL
Simracing fails at this, either the limit of the car is too short or too long, but I believe simracing fails because damage models are poor which makes creating a limit difficult. In sims damage is basically about crashing. IRL damage is friction, vibration, wear and contacts.
Maybe a new sim need to be made where damage is physics, if that makes any sense my 2 cents
Oh add a proper racing school while you are at it, simply throwing people at a green track is wrong.

 

Maybe we can carry on with this "meme", I don't know.

Here is a video of aquaplaning being unrealisticaly difficult IRL, Hunt commented that they couldn't even see where the puddles are.


Spinning out because of lift off mid turn ? Obviously no reason for that spin, but unrealistic real life physics.


Super unrealistic handling right there. I mean why wouldn't the car turn mid turn, why it would then oversteer so hard ?


Shedden spins out Etype, because of unrealistic "drifting physics", obviously way too little friction and slip angles should be bigger


Look at this twitchy mess, absolutely unrealistic. Look at those unrealistic driver inputs. How is it realistic for a car to have no grip, and then so much grip almost at the same time. Undrivable...

 

In many ways it is, mostly due to the way its delivered to our senses. We will never have the instantaneous feel of motion, and even with VR we're limited to flat screens with a singular focal range so we can't adjust the paracentral portion of our peripheral vision. We're taught we have 5 senses, but within each such as sight there are many more; paracentral zone of vision is one of them. (I use paracentral range as one of the components for the Driver Monitoring Systems I develop).

And than there is the flip side. Racing in a multilayer fire resistant suit, socks, shoes, balclava under a helmet on 100° F days in the baking sun whilst feeling G-Forces where your neck is getting 100+ lb forces applied while driving at breakneck speeds with total focus and anticipation. I once drove a Formula Supervee with a leaking front radiator. At the end of 18 laps my feet had 1st and some 2nd ° burns, yet I never felt it until I started to get out of the car and than it hurt like hell. Literally couldn't walk for a few days due to the blisters on my feet. Yet if the cat claws my foot on the sim while hot lapping in a 962, I feel it right way. That level of immersion will always be missing on a sim, but hey thats OK, its still a lot of fun.

 

True about pace, but also difficult to be precise. Decent laptimes, what are they, 4seconds off pace decent enough for most ? What if elite in simracing is multiple seconds faster than elite in real life ? Does it mean real life drivers are slow, or is it too good in simulation ? Another important point is the learning curve. Should it go steep and should one ordinary person do his/her best (within couple tenths) within a hour of first time driving car, untill realising bit of exploit and climbing higher instantly. Or should the learning curve be flatter, should one ordinary guy practice car handling for 4 hours to get within couple of tenths of his/hers capability ? And then having to repeat that at diminishing rate at each new track to get to personal decent laptime.

I disagree about the damage, maybe slightly, bit of truth, also financial side, health and other worries that puts additional stress. However, physically main things are: tires and aerodynamics.

@John R Denman Interesting stuff, I'll need to look up more about it. Speaking of senses, I am confidentally placing anticipation as 6th sense, I think it is a sense of imagination, knowledge and experience combined. Tough story about radiator leaking, surely thats major bug in real life haha, it reminds me that Fangio had similar race, I wonder if he felt the same.

From wiki:
"He won a particularly brutal race at the Gran Premio de la República Argentina. This race was run in Buenos Aires during a gruelling 40 °C (104 °F) heat wave, and with track temperature of over 57 °C (135 °F) few drivers other than Fangio were able to complete the race. The W196's chassis had heated up and Fangio's right leg rubbed against the chassis structure, but even after receiving severe burns he kept going; it took him three months to recover from his injuries."

 

Fucking noobs ruin online racing....

 

Judging the difficulty, handling IRL vs sim is something i do regularly.
It comes down to 3 basic things

1.Driving condition
This contains everything the surface type,air speed,temperature,rain etc

2.Who is driving
A good driver will always give objectively accurate description of the car behaviour. because a pro driver is as good as a normal human body can get good at throttle,braking modulation.
This is why i only watch onboards video that are driven by good drivers.

3.Car setup
IMO in simulator the setup window of the cars seems way big,it seems
like you can do any mindless changes and the car will still perform quite well, while IRL i think setup plays much more crucial role.

That being said My opinion on RF2 is still the same, the cars are way to progressive both on entry and exit especially the modern aero cars.
Although thanks to new ride heights sensitivity they seem much more realistic that what they used to be.

 

The most basic things are tires and aero. And also most mysterious.

The rest comes after, and I think most sims are pretty good with more obvious stuff.

I don't think a driver has much influence on difficulty, it is just that perception of difficulties depends on skills and attitude. I also prefer watching cars by good drivers, but I like to watch cars being driven up and over the limits by anybody. Perhaps I even like seeing imperfect driving more than perfect driving, because absolute best drivers will make it look easier than it is, and they will not let cars to show much or any misbehaviour. That is just unentertaining.