Exactly why are front-engine RWDs uncomfortable to drive?

Forza Horizon Legacy Horizon FH4 Discussion
1

The steering doesn’t respond to throttle by letting the wheels straighten naturally during corner exit. In some MR cars it’s a non-issue because the front tires are usually smaller and unweight during acceleration.

The steering doesn’t naturally reduce lock or feed opposite lock when necessary, and when it does from player input, it often feeds too much, greatly upsetting the car’s trajectory. The ginger re-application of lock post-counter-steer is a further waste of time.

Slow steering can cause fast crashes. Fast steering in the right moments can ensure steady yaw and predictable total behaviors. No god-mode “yaw dampening” necessary. Does anyone at PG/T10 understand that?

1 Like
2

I don’t find them “uncomfortable.” Some of the rear engine RWD cars, on the other hand, I’ve found to be a bit rough.

3

The steering is geared towards AWDs and leaves many RWDs completely in the dark. Why else would 98% of the shared grip tunes be AWD? The PI system works fine. The steering is ignorant. It needs to stop pointing the car and let it flow. That would make the game less of a soul-suck. Conflict of interest?

Some A800 lap times at Ambleside Village circuit. 1 take per car, best of 2 flying.

49.601 - Audi RS5, AWD
49.082 - Subaru WRX '08, AWD (has aero)
49.307 - Mugen Civic, FWD

Abarth Spider RWD, A800, 384hp, 52%, sport tires, 245/255mm f/r and no aero–best I’ve done so far is 50.8xx. Second best lap of about 10 races now, 51.601. Nothing I do with the tune makes this car fun to drive. - update: decreasing power and changing to rally tires, plus aggressive roll bars biased towards understeer (24/4) helped me reach 50.387

AWDs/FWDs in Forza, you basically beat the **** out of to be fast… same as real life. RWDs? No…

FM7 arbitrarily nerfs AWDs via PI even though the lame steering is less of a problem on their smooth and limited tracks that also let you change cars and tunes between lobbies.

4

I have no problems with RWD cars in this game. Most of them are a little bit too rear heavy when untuned but nothing that can’t be fixed.
The only RWD cars which cause some issues are those with thin rear tires. Then, accelerating out of corners turns into a hassle. The recently added Tuscan with 285s is a prime example of this.

5

285 is thin? My Abarth is on 225s and doing ok.
Sounds like you agree the steering doesn’t respect the layout if anything less than ~305 rear tires isn’t enough to find a good flow with.
RWD should be competitive in MP against clean drivers without arbitrary PI nerfs against AWD and FWD. Until it is, something is lacking in the gamepad control’s approach.

6

I have to comment on this one separate. Tires is one area I think people really can oops up on.

Don’t fall into the taller/wider is better for handling trap!

Tires are a great bit of highly undervalued physics. it is also why I absolutely cannot stand when people do the narrow rubber on a large rim “cool” stance on ultalow profile tires. Ugh! It offends my engineering senses!

So width. Running at same air pressure, you are going to have the same overall area of contact patch in contact with ground… mass times area, etc. Weight of car is not changing, air is doing most of the supporting of the load, very little is actually transferred through the structure of the tire itself. Is why tires go flat and why runflats only can function at lower speeds for a short distance before they start to fail.

So changing width will changes the shape of the tire. A tall overall diameter (which we cannot change) with narrow width will create a longitudinal contact patch. Widening tire will square it out and even make it a transversal shaped contact patch.

This is where traction and physics gets interesting. In terms of the molecular/chemical friction, this will not change regardless of shape. What we are concerned about is the mechanical traction. Typically the easy way to think about it is that your mechanical traction increases proportionally to the perpendicular length… how to phrase that better is a wide tire with a high transversal width, will give a higher mechanical advantage on stuff like accelerating and breaking. A narrow tire will actually hold better sideways, but will be easier to break loose from traction.

The reason we see that tiny bit of green when we change this is because of tire roll for handling and the above for breaking. The penalty to acceleration and launch we can see on some cars is simply due to adding weight without gaining any advantage since the smaller tire size already is enough to meet all traction needs. If having trouble in corners? Try actually going narrower if you are not having troubles with power breaking loose.

Wider and steering. Tire scrub. We have differentials because the wheels inside of a turn are turning a sharper radius than the outside. So they don’t turn as fast. The same is true for tires. This becomes more significant with wider tires. Wider tires hurts turning because the tires resist changing direction much like if you were running a locked differential. Real world example. Big trucks. Years ago, I worked detailing in a ford dealership. Had a 100k ford diesel. These are good trucks to drive. I could turn it around in my shop, no hassle. Came back a week after sold with the typical big offroad tires and wide wheels. I had to do multipoint turns just to back it into a bay due to just getting wider tires. No other change.

The other part. Sidewalls.

Tires are a critical part of your suspension. Your control arms, wheels, brakes, the tire itself are only dampened by your rubber. Higher pressures and lower sidewalls limits the ability for your tire to take up these bumps in a road. If running superlowpro without proper sidewall geometry on massive tires and high pressures? the texture of the road (in real world) alone is enough to cause loss of traction just because your unsprung weight is continually unloading from the road. If it is rough for you, and you are bouncing in the seat, imagine what that wheel is doing?

So in Forza, a wider wheel increases your unsprung weight and as such will also then cause traction loss easier over bumps.

This was discovered and tested when using my 959 rally. I had wide tires in lower PI class and simply couldn’t maintain pace. I was sliding when drivertars and such were not. I called cheat, until I went back to default wheel size and now could win without huge fights.

2 Likes
7

WARNING: A long Vaporisor technical post incoming!

TL:DR version, look at last to pictures showing when you AWD swap Zonda.

There are a couple issues I think at play with rwd. First is acceleration and launch. Now no, I havent driven hyper exotic cars and such, but I grew up and regularily drive multi hundred hp rwd vehicles. The tire spin even on stock tune vehicles and the way the rears kick out just feels off. The partial throttle input is way off. But I suspect much of that is as follows:

Clutch, driveshafts, and flywheels.

Clutch was the first one I noticed. Racing clutch does shift faster, but it also takes slippage away from the clutch. It is seeming the game is simulating that. Downgrade your clutch for lower speed technical tracks. It will give smoother engagements and seems some slippage at launch to keep it from kicking out. Real world comparison, my VR6 GTI has performance build and a high performance clutch. Gotta rev match it well because it doesnt have the smooth slippage of a stock.

Driveshaft and flywheels… to some extent, the wheels. I discovered this one actually with my offroad and rally builds. As bouncing along, any slight slippage causes the revs to bounce all over the place. Made high power cross country a pain at times. I downgraded these to stock and got faster times because I wasnt suddenly doing burnouts on every bump. This is rotational mass and will help keep the revs from instantly kicking out and gives you more control of subtle spin, something we do reactively without thinking in most stock or mild tune rwds. The fail vids we see of cars suddenly kicking out is because of these lighter rotational mass. Once traction is lost, they just hit that top rev faster than can react and out of control. Flywheel/Gyroscopic effect is fastinating.

Lastly, I am not certain on it, but I suspect there could be a bug with the way it calculates the PI on conversions. I had been pondering it because of the sheer advantages AWD gets for its low PI cost vs just tuning a RWD up to that same level.

For that, I am going to turn to something I was working on back in the last fall season. I have been meaning to get around to finally doing up a GTA for forzathons. Just a fun, hard accelerating IROC style RWD. But then got to pondering this exact issue. I swapped from RWD to AWD to see what would happen, but now I see a potential significant error in how it is applying PI to these swaps. I will attach reference pictures at the end. First will show the full tuned RWD and it’s PI. Followed is the AWD swap. The picture shows final results after geabox swap, this is critical and to show that I did not forget to reapply all drivetrain upgrades.

The key aspects to look at are the the top speed and acceleration. Launch is to be expected.

Swapping from AWD causes a significant drop in top speed, while adding to acceleration. I am certain that this relates to the AWD swap changing the transmission. Default, this change puts gearing that is far insufficient for the car which puts a slight change on acceleration due to the first gear being slightly different, but after that would accelerate at same rate. Very minor in the actual values. It might go a bit over ten, but in other ones I have looked at since that don’t hit ten, top speed is always significantly more impacted than the acceleration. AWD cars to rwd get PI from top speed with no effect to accel on some.

What does this mean? Without knowing the code for PI, I have absolutely no way to verify and this will just be a theory. The PI is being significantly reduced due to the default top speed, and only partially countered by the initial launch acceleration and some expected boost from the actual launch value. This is a false value though. Anybody who does basic tuning, first thing they do is they go in and change final drive. Now the cars have the exact same top speed and once out of initial launch, they accelerate exactly the same. However in racing with the added launch on this car, it becomes far superior without the added PI due to the top speed/accel handicapping.

To sort of stress my example, was top speed running on the highway with it. Pictures 3 and 4 taken both around when they stopped pulling (while trying to screenshot at high speed in a curving tunnel) but can you tell me which the AWD and RWD is? Heck. I cannot even remember. But the AWD PI is reduced because it has a far more insignificant top speed.

But if you want a real good example. Check out the Zonda, pictures 5 and six.

Going racing? You have to install AWD, just for that bootiful PI reduction!

1920×1080 3.33 MB

1920×1080 3.37 MB

1920×1080 1.81 MB

1920×1080 1.56 MB

1920×1080 2.75 MB

1920×1080 2.75 MB

8

Sounds like we’re on a similar page in terms of the drivability of RWDs in Horizon. There’s also a huuuge margin before they break loose (laterally), and once they do the grip about disappears completely. At least, the unwillingness of gamepad steering to feed appropriate amounts of opposite lock makes it seem like it. Part of it is due to shortcomings in the specific tire modeling approach, not even the values. They will always feel off until the current method is scrapped wholly and re-thought. Rabbit hole of a subject in it’s own…

The Abarth was actually struggling with the heavier flywheel and driveline. Hit a lightpole? Car slows down 20 mph (just over 2000 pounds). Rear wheels break traction because even with the smallest rim diameter and lightest tier of wheels, there’s more inertia there than the tires can stop. Upgrading to the lightest flywheel and driveline minimized that. All the while, too, it had a stock clutch.

My 2WD builds gain PI from swapping to AWD. I wouldn’t use 2WD if it wasn’t cheaper in that sense.

One issue could be the bias towards 0-60/0-120 times (would 40-80/60-120 be a better metric?) doesn’t account for drivability, especially in context of the gamepad steering. Most of Horizon and Forza players use gamepads. Even wheel users won’t all have perfect tunes. RWDs should have performance ceilings somewhere between equal for TCS users and equal for skilled non-TCS users. Right now I don’t think that’s the case. Is there an outright bug in PI calculation? I don’t think so. I think it’s in the steering, maybe the tires, and a little in map design.

Without quoting the whole bit here, I thought wider tires had wider, shorter patches giving better lateral grip but less longitudinal grip. Sounds like you’re saying something different?
Bit to add here too, smaller sidewalls have less flex, which can be more useful in Forza than the potential versatility and even higher limits of higher-profile tires. Anything 20-series and often 25-series is a no-go regardless (Centenario front wheel upgrade, BMW E92 max size, I think the '05 Ford GT is rubber-band-capable on the front too).

I usually go for 1-2 upgrades below max wheel size. That’s what the Spider has (2 below max, stock). 225s are doing better than 245/255… but they are also rally tires, an upgrade from sports and lose traction more progressively.

9

I’m humbled at your efforts but that’s not what I’m on about.
RWDs are fastest with a light touch. Truly well-tuned FWDs and AWDs benefit from quick precision as well. They tolerate rougher inputs better by nature in most cases.

Nothing about the current gamepad steering is gratifying to me outside of being faster than I used to be or faster than other players. It’s heavy-handed and disrespects the car and my wishes, for what I can see the opportunities within the physics are. I know how to play with a car, trust me. Forza barely reminds me of how that feels. If it wasn’t for the tuning and a dying interest in competing in live multiplayer, I’d have no interest left at all by now.

Everything has to be over-the-top fast and huge in FH4. It’s not the playerbase’s fault at all. It’s the responsibility of the physics department for the way a gamepad thumbstick interacts with the interconnected systems of springs and dampers and tires on road (I guess map design is part of this too) to cater to one or other experience.

Referencing an earlier comment, it’s either out of ignorance they’ve made it this way, which is why I have been trying to express my dismay for months, or… it’s to try to trap the player within their system that they can come back to in the future and say they know who we are more than we do. NSA mass-surveillance type of thinking.

10

Ooooh… the gamepad in particular. Yeah. I can see that. I got a wheel and I wouldn’t play this at all with the way the way steering feels on gamepad/keyboard. Especially for the rwds. Feathering throttle indeed can make for a better track times with RWD, but for a majority of cars, it can be essential to ever be able to pull out of corners with a wheel. Keyboard though it seems to brute force auto stop that?

11

Yeah. It seems like more work to code that effect, than it would be to program the steering to predicatively align to the car’s trajectory–intelligent opposite lock… just “intelligent steering”, period.

The physics aren’t great, but check out CarX on YouTube for an extreme example of what I have in mind for Forza…

12

Double-posting to make sure you see this. Looks like, at least in FM4, they were using data from tires alone, which is what I think they are still using in FM7/FH4. Would like to know for sure.

And, quoting myself in an older post
I think Forza uses a contact-patch centric model, unlike the hub-centric model referred to as “the VMS system” in this presentation.

13

The Abarth 124 at A800 is way out of its league… Bringing up small, slow cars from C500-something all the way to A800 and beyond can have strange effects.

14

I actually managed to get it performing fairly reasonably, but it took quite a bit of work, and my shared tune is AWD. It’s not a contender for most OP A class for online racing, but I got it running Lakehurst Forest Sprint a second quicker than Grandma Driving’s RX-8 tune, for example.

1 Like
15

Yeah AWD tends to mask grip issues in the chassis.

These small cars like Miata, BRZ, etc., they don’t have a lot of downforce when stock. If you use power, tires and weight reduction to bring them up to higher classes without something to add grip, they become unwieldy. The same tends to happen with old cars.

Speaking of chassis stiffening that the OP brought up, in Forza it seems to have similar effect to locking the diff, that is, you have more traction but breakup is sharper when it happens. This is why I almost never add the full rollcage, but how much stiffening you need will depend on the car and how soft the stock chassis is.

16

Oooh! I completely neglected to think as to how much simulation there was in chassis stiffening! I mean, I knew it was a thing, but never thought to what extent it simulated.

IE, the shelby cobra. He laughed when talking about it’s “Great handling” and said it worked because the chassis on the first one was so terrible. It was all the flexing in the chassis that was keeping it on the road. I will have to add this to my consideration when building up my rwd cars. I am doing an old lambo… think is from car pass for the community challenge. A good one to test for me to see how much of a difference it makes. Probably exactly this. With tires and flywheel weighting, perhaps the chassis is just too stiff.

17

I’m starting to think it’s twisty/flexy, even with race bracing (been tuning it with sport) and needs a lower rear ride frequency to keep from oversteering. ALL of my builds get race bracing, and the cars I can’t get into shape with balanced springs, I have until dropped and forgot about. This and the E30 M3 must need a different approach. Still, I feel like I’m tuning around the steering’s habits as much as I am around the car’s. And I question the amount of damping going on in the chassis flex. Seems excessive?

18

@Vaporisor I see way more often that awd swap increase PI instead of decreasing that, for that zonda example, i have tuned that S998 with awd but swapping rwd back PI decrease to 997, okey not big difference there, but for another example A800 amc javelin, swaping rwd back would decrease PI down to 775, koenigsegg one:1 stock PI993 and with awd swap and stock tires you cant even keep that car in S2 class.

19

Well, that camaro one has rwd as a lower PI than the AWD. I thought it odd that stock konigsegg the PI drops. I discovered that one just that bit ago.

Perhaps I am just heavier footed and lean more to grip steer turning and tuning. Reason I started looking at it was my lines and grip driving favour powering out of a turn. Heavier some RWD just cannot do anything with severe wheelspin, excessively so on asphalt at speed. Most higher end cars I leave as is. Specially track toy stuff.

But really gets aggrivating on wet tracks. That is why I run the heavier flyweight on rwd, keeps it from kicking out bad on slippage.

Counterpoint you yours was the pvp rip rod drag challenge. AWD on stock tires a better swap then RWD and drag tires? The weighting of something just always has seemed off. So many tracks, specially higher PI just punishing for RWD, even with foot controls. With suspension and downforce setup, I can tune out most disadvantages to the AWD. Outside of cases like that where AWD causes enough PI to prevent adding another critical part, it seems that ones with a rwd PI saving rarely is offset by the additional fitted parts you can add.

Usually the difference between AWD and RWD on my cars when making one for racing is just power adders. The traction means that out of most turns and definitely off the start, AWD makes a significant time difference. RWDs seem to mostly make up time in the higher speed turns for grip or then the top speed.

So unless the chassis is pretty much excessive and already good on grip, not sure what all else I am missing. I just find at S1 and pretty much anything above S1, the gains feel way off. It is why shared those picks. For that GTA, it was just huge gains from coming out of corners and low to high speed pulls. It isnt all the time, but just many. The gains seem way too much for the PI and so just pondering if there is some conditional mathematical issue.

Course this is also thanks to not having tire wear. Those AWD tunes would be shredding rubber and running way hot.

20

I have an S1 RWD Ford GT '05 (aero, race tires, ~850 hp) build that absolutely dominates in MP, when driven correctly. The only car that came close to my pace in an adventure this morning was a 2019 991. I have not had success or even tried to with an S2 RWD. If MP ever puts me in S2 road again, I’ll take the GT-R LM FE out. It hooks early, drives easy, for a 950hp RWD.

As far as I can tell, PI calculations come from circle skidpad and linear accel-decel “tests”. If the accel tests are only happening as they are indicated, 0-60 and 0-120, ideally, they should be done in an outwards spiral fashion… or, 30-100 linear tests, with carefully calibrated bias towards the lower end and increased overall bias in total PI calculation. That’s the range where 2WD most often suffers leaving corners. Launch is almost meaningless, it always has been. The slowest corner is the most launching that should happen during a race and even then, CG height, track width etc… can be factors in the realistic performance impact of AWD over RWD.

The devs could be way ahead of me and already have done this. dunno. They like to keep secrets.

NightDriver - The sharper loss of traction with bracing usually comes with higher limits and precise response. Like I said, the cars that can’t be good with race bracing and balanced ride frequencies, I have, so far, written off. Tuning and learning to be fast with 6 springs and 4 dampers is about as much as I want to worry about. When the chassis adds an unknown number of springs and dampers that don’t cooperate with the ones I can control, I get confused. One could say I’m a rigid thinker. Lol…