rec.autos.simulators

Car Physics: Arcade mode in car sim games

Mike Stanle

Car Physics: Arcade mode in car sim games

by Mike Stanle » Sat, 18 Aug 2001 19:53:45

Hello everyone,

I've noticed that many car sim games have an arcade mode. This mode usually
has simpler handling than sim mode, so that it's more difficult to lose
control and go into a spin, and it's easier to regain control once it's been
lost. I thought it might be nice to include this in a sim using a difficulty
slider e.g. going from 1 - 10, with 1 being complete novice and 10 being
hardened sim player (rather than the usual beginner/advanced/expert
options).

The problem is, I'm not sure how an arcade mode would be derived from a
simulation. It seems to me that the most obvious way of doing this would be
to artificially limit something in the simulation. My initial thoughts on
this are to limit either the angular velocity in the yaw axis or to limit
the ratio of lateral forces applied to the front and rear tyres. But with
either of these approaches I can see unnatural side effects ocurring (most
arcade modes derived from sims do retain a degree of naturalness).

Has anybody else had any thoughts about this?

Cheers,

Mike.

Gregor Vebl

Car Physics: Arcade mode in car sim games

by Gregor Vebl » Sat, 18 Aug 2001 20:31:00

Hi Mike,

I indeed have thought about this, and my conclusions were surprisingly
similar. If I were to do it, the first thing I would take care of would
be yaw dampening. I would introduce an additional torque on the car that
would be proportional and inverse to the angular velocity vector (sort
of as if rotating the object in a viscous fluid). This should help with
the requirement that arcade players tend to not use smooth inputs, as
they usually play with a digital (style) control.

If that is not enough, then at least the front tyres need to have their
lateral grip vs. slip curve altered; arcade players like their car to go
left when they select left, the control reversals or inefficiency when
past limit isn't quite something that is expected, control should be had
at all times. This means that the grip curve should not reverse after a
certain point, but that it should at least saturate or even have a
constant small positive slope, so that control is maintained at all slip
angles. You probably want spins to happen, but you also want to make
them catchable; I believe this could be the key.

The proportionality constant for yaw dampening and the interpolation
from a proper slip curve to a more controlable one can be done
continuously (the slider idea), and without introducing any hard
constraints, which would probably make the arcade mode still feel quite
natural.

Anyway, I'd love to hear more thoughts on the topic, especially from
people who actually have experience implementing such things.

-Gregor


> Hello everyone,

> I've noticed that many car sim games have an arcade mode. This mode usually
> has simpler handling than sim mode, so that it's more difficult to lose
> control and go into a spin, and it's easier to regain control once it's been
> lost. I thought it might be nice to include this in a sim using a difficulty
> slider e.g. going from 1 - 10, with 1 being complete novice and 10 being
> hardened sim player (rather than the usual beginner/advanced/expert
> options).

> The problem is, I'm not sure how an arcade mode would be derived from a
> simulation. It seems to me that the most obvious way of doing this would be
> to artificially limit something in the simulation. My initial thoughts on
> this are to limit either the angular velocity in the yaw axis or to limit
> the ratio of lateral forces applied to the front and rear tyres. But with
> either of these approaches I can see unnatural side effects ocurring (most
> arcade modes derived from sims do retain a degree of naturalness).

> Has anybody else had any thoughts about this?

> Cheers,

> Mike.

Mike Stanle

Car Physics: Arcade mode in car sim games

by Mike Stanle » Sat, 18 Aug 2001 21:10:37

Hi Gregor,


... which would help to avoid spinouts. Sounds good, think I'll give this
one a try.

Digital controllers are something I have given thought to (since I mostly
use one anyway). I thought a good way to deal with these would be to have a
dynamic maximum turn angle depending upon the slip angle. You don't want the
slip angle to exceed the maximum on the traction curve too much since this
will cause reduced grip. I thought I might calculate a turn angle that would
give a slip angle which is close to this maximum value, ie. try to keep the
front wheel's direction close to its velocity direction. This would mean
that if your backend was sliding out, you could steer the front wheels so
that they were in line with their motion, resulting in more torque coming
from the rear wheels than the front and hopefully stabilising the car into a
powerslide.

Haven't implemented this yet but it's on my todo list.

I've tried altering the traction curve so that it becomes level after it's
maximum. This definitely makes the car more responsive to inputs (perhaps a
little too responsive, but that could easily be tweaked). I intend to spend
more time looking into this.

me too :)

Nick

Car Physics: Arcade mode in car sim games

by Nick » Sun, 19 Aug 2001 03:10:12

a good arcade mode is one where the car easily loses grip and heads off into
broadside land, but is easy to hold or correct at will. in this way the
player thinks that he/she is god because they can hold 80yd powerslides. i
think obtaining this functionality from a real sim is quite difficult, so i
am trying to develop an arcade handling routine to go alongside the
realistic one in my game. if anybody knows a site with some arcade handling
code on then the url would be much appreciated, coz it's turning out harder
to do the arcade handling than the realistic :-)
Kirk Hous

Car Physics: Arcade mode in car sim games

by Kirk Hous » Sun, 19 Aug 2001 05:30:10

I'm a big proponent of easy driving through setup tweaks.  When my friends
try out GPL I usually use my newbie setup which is 59% front brake bias,
85/30/5 diff settings and not much rear roll bar for understeer.  For
instance road cars have lots of understeer built in so that when people
swerve to miss a deer or pedestrian, etc. they can keep the car from
spinning.  This way they can still get a feel for the dynamics of braking
and balance without having to relearn everything after graduating from an
"arcade" physics engine.


Jonny Hodgso

Car Physics: Arcade mode in car sim games

by Jonny Hodgso » Sun, 19 Aug 2001 05:29:16


Valid for learning, but unfortunately (IMO) consistent understeer is
about the most boring trait a car can have.  Not quite the aim for an
'arcadey' driving game.

Jonny

Kirk Hous

Car Physics: Arcade mode in car sim games

by Kirk Hous » Sun, 19 Aug 2001 06:21:56

I'm totally confused.  If you can't spin the car you necessarily have
consistent understeer.  I think that's why most of us consider arcadey
physics engines totally boring.  You propose having a car with lots of
oversteer but that is difficult to spin?


Haqsa

Car Physics: Arcade mode in car sim games

by Haqsa » Sun, 19 Aug 2001 06:59:09

Personally I like the way Papy did it in N4, which is to have not only
an arcade mode, but also separately selectable but realistic driving
aids such as ABS, traction control, and stability control (yes stability
control is a real thing).  The arcade mode then keeps all the normal
physics, but adds a setup designed for control and stability, and could
artificially boost the performance characteristics of the car (grip,
downforce, power).  The reason I suggest that is that I think arcade
players are looking for a thrill rather than realism, but if you don't
still have a realistic physics model underlying the whole thing it won't
respond like they expect it to.


Jonny Hodgso

Car Physics: Arcade mode in car sim games

by Jonny Hodgso » Tue, 21 Aug 2001 03:42:56


> I'm totally confused.  If you can't spin the car you necessarily
have
> consistent understeer.  I think that's why most of us consider
arcadey
> physics engines totally boring.  You propose having a car with lots
of
> oversteer but that is difficult to spin?


> > Valid for learning, but unfortunately (IMO) consistent understeer
is
> > about the most boring trait a car can have.  Not quite the aim for
an
> > 'arcadey' driving game.

My understanding of arcade "physics" is that you want to be able to
hold the car sideways without spinning it - that way it's fun, and you
feel like you're an expert driver.  Just having a car that won't turn
is boring - that's more like the ancient "fly spaceship through
caverns" game than "driving".

Yes, I think I do propose exactly that: lots of *linear region*
oversteer to make the car feel like it drifts and powerslides, but
with the *limit* handling capped so that it'll only spin if you're
extremely silly with it.

Any clearer? ;-)
Jonny

Kirk Hous

Car Physics: Arcade mode in car sim games

by Kirk Hous » Tue, 21 Aug 2001 15:05:47

OK, I get what you're saying now.  Maybe the front roll bar could suddenly
stiffen up when the driver gets sideways.  Probably would have made learning
GPL much easier.  Good idea :)


Mike Stanle

Car Physics: Arcade mode in car sim games

by Mike Stanle » Tue, 21 Aug 2001 21:05:36

Hi Nick,

I disagree that obtaining a good arcade mode from a real sim is difficult.
It just depends how arcadey you want it to be.

This may be true for a game like Mario Kart, where you'd probably just end
up discarding most of the real physics to get the handling you want. But for
a game with a bit more realism I think it's possible.

A good example of this is Driver. It seems to me that behind it there is a
realistic handling model (for the tyre forces at least), but they obtain
quite a good arcade feel by using the auto-handbrake feature. I'm not
completely sure of the mechanics of this, but basically the handbrake is
automatically engaged on hard turns, causing mucho powerslides. The car
rarely looses control due to the slowdown incurred by this, and the result
is handling which is a lot of fun and easy to get in to, while also feeling
pretty realistic. Try disabling the autohandbrake feature and you'll get the
spins at low speed/understeer at high speed behaviour you associate with a
sim.

Like I said though, it really depends on what level of realism you want. But
for anything other than a Mario Kart clone, I reckon that beginning with a
realistic mode then limiting it for an arcade mode is the way to go.

Mike.


Mike Stanle

Car Physics: Arcade mode in car sim games

by Mike Stanle » Tue, 21 Aug 2001 21:09:54

That's really an advanced sim feature though. You're probably quite an
advanced player, so you know what to tweak and how it will affect the
handling.

My point is that most arcade players will not want to (or even know how to)
set up the car in this way. They want to simply pick up a game, get straight
into it and have fun. Messing around with front end options and setups is
not really something they want to do.


Jonny Hodgso

Car Physics: Arcade mode in car sim games

by Jonny Hodgso » Wed, 22 Aug 2001 02:05:09


I think whoever mentioned tyre curves without a peak had the right
idea - possibly with different initial gradients, too, to get
oversteer followed by understeer...


> > Yes, I think I do propose exactly that: lots of *linear region*
> > oversteer to make the car feel like it drifts and powerslides, but
> > with the *limit* handling capped so that it'll only spin if you're
> > extremely silly with it.

> > Any clearer? ;-)
> > Jonny

Kirk Hous

Car Physics: Arcade mode in car sim games

by Kirk Hous » Wed, 22 Aug 2001 07:39:41

I agree but if the default setups are easy then there is no need for
modification unless the user really feels like tweaking.


use EXCITE.COM instead

Car Physics: Arcade mode in car sim games

by use EXCITE.COM instead » Wed, 22 Aug 2001 10:36:22


Driver has a nifty driving model, but its collision model needs MAJOR
work.
Traction is handled quite nicely, and cars in general do behave as they
should...
Until you hit something. Of course, being limited to automatic
transmissions and
top speed of 87 mph (except certain cars) is not exactly realistic
either. :-D

--KC


rec.autos.simulators is a usenet newsgroup formed in December, 1993. As this group was always unmoderated there may be some spam or off topic articles included. Some links do point back to racesimcentral.net as we could not validate the original address. Please report any pages that you believe warrant deletion from this archive (include the link in your email). RaceSimCentral.net is in no way responsible and does not endorse any of the content herein.