The perfect racing simulation oriented game?

In the real world, there are forces that a driver feels from the acceleration,
braking, cornering, or yawing (oversteer) that aren't present in a racing game.

Even without assists in the real world, a driver always has those forces that
can be felt and provide useful feedback.

However in a game, these forces don't exist, so having some level of assist
to compensate for the lack of forces might end up being more realistic than
no level of assists.

My experience with racing games is that a player mostly memorizes the control
inputs for a particular car and track combination, much more so that would be
done in real life. At the same time the real world car's response to control
inputs is changing (due to tire wear, heat, ...) much more so than any car in
a racing game.

The "perfect" simulator would probably be a remotely controlled race car.
The remote view would be the normal in car view. However I would expect that
that an external view might provide better feedback and result in better lap
times. At the extreme, a very remote external view, one that overlooks an entire
track, would might be better still, assuming headphone playback of the on board
sounds were used to eliminate any delay in audio feedback.

OK, now for a PC based racing game simulator? What, if any, compromises should be
made to make the racing game more realistic? The normal assists in current games
would always be available as an option.

In my opinion, modifying the physics to make the cars more forgiving when pushed
past the limits would help compensate for the lack of real world force feedbak.
This would include changes such as: more forgiving tire physics, reduce the rate
of yaw response (in most games, the rate of yaw response is faster than real life,
which is slower, even if not recoverable), modified audible feedback (tire scrub
sounds), and exaggerated visual roll and yaw response (some games include
adjustements for the visual stuff).

You pretty much just described the 'arcade' genre.

-Larry

"jeffareid" > wrote in message
...
> In the real world, there are forces that a driver feels from the
> acceleration,
> braking, cornering, or yawing (oversteer) that aren't present in a racing
> game.
>
> Even without assists in the real world, a driver always has those forces
> that
> can be felt and provide useful feedback.
>
> However in a game, these forces don't exist, so having some level of
> assist
> to compensate for the lack of forces might end up being more realistic
> than
> no level of assists.
>
> My experience with racing games is that a player mostly memorizes the
> control
> inputs for a particular car and track combination, much more so that would
> be
> done in real life. At the same time the real world car's response to
> control
> inputs is changing (due to tire wear, heat, ...) much more so than any car
> in
> a racing game.
>
> The "perfect" simulator would probably be a remotely controlled race car.
> The remote view would be the normal in car view. However I would expect
> that
> that an external view might provide better feedback and result in better
> lap
> times. At the extreme, a very remote external view, one that overlooks an
> entire
> track, would might be better still, assuming headphone playback of the on
> board
> sounds were used to eliminate any delay in audio feedback.
>
> OK, now for a PC based racing game simulator? What, if any, compromises
> should be
> made to make the racing game more realistic? The normal assists in current
> games
> would always be available as an option.
>
> In my opinion, modifying the physics to make the cars more forgiving when
> pushed
> past the limits would help compensate for the lack of real world force
> feedbak.
> This would include changes such as: more forgiving tire physics, reduce
> the rate
> of yaw response (in most games, the rate of yaw response is faster than
> real life,
> which is slower, even if not recoverable), modified audible feedback (tire
> scrub
> sounds), and exaggerated visual roll and yaw response (some games include
> adjustements for the visual stuff).
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>

On Jul 23, 8:14*am, "jeffareid" > wrote:
> However in a game, these forces don't exist, so having some level of assist
> to compensate for the lack of forces might end up being more realistic than
> no level of assists.

I'm with you as long as the assists relate to visual and audio
feedback. Enhancing audio or visual cues is reasonable. If you are
talking about true driving aids - steering help, braking help, etc.,
then I don't think that contributes to realism at all. Same thing
goes for ABS, TC, or stability control that isn't available in the
real vehicle that is being modeled. Don't mess with the vehicle
dynamics.


> My experience with racing games is that a player mostly memorizes the control
> inputs for a particular car and track combination, much more so that would be
> done in real life.

I agree, and it's readily apparent every time I switch FFB off and
on. I can drive without FFB. It's actually easier when hotlapping.
Once you add good FFB into the mix you can no longer memorize control
inputs. Driving becomes a significantly more interactive experience.


> At the same time the real world car's response to control
> inputs is changing (due to tire wear, heat, ...) much more so than any car in
> a racing game.

I don't agree with that. Changes in tire wear or temp in a sim has a
lot more impact on my driving response than I ever experienced in
limited real life racing experience. With the infinitely higher
feedback resolution you get in real life, responding to those sorts of
changes in the real world is easier, or even sub-concious.


> In my opinion, modifying the physics to make the cars more forgiving when pushed
> past the limits would help compensate for the lack of real world force feedbak.
> This would include changes such as: more forgiving tire physics, reduce the rate
> of yaw response (in most games, the rate of yaw response is faster than real life,
> which is slower, even if not recoverable),

I think that properly modeled vehicle dynamics will yield a more
controllable vehicle than we've seen in almost every attempt at a
racing simulator. For an example of where it's been done wrong, I
point to the complete lack of pnuematic trail modeled in almost every
rFactor mod. Pneumatic trail accounts for a huge part of the inherent
stability in a real car. Then there is the poorly implemented front
end geometry in so many rFactor mods. The original F3 car is
terrible. When the car suspension and inertia is modeled properly,
and the tires behave correctly, car simulations aren't difficult to
drive.

IMO, if you start messing with the vehicle dynamics to make it more
driveable you've left field of simulation.

> modified audible feedback (tire scrub
> sounds), and exaggerated visual roll and yaw response (some games include
> adjustements for the visual stuff).

Yep. All proper for a simulation IMO. You aren't making the car
easier to drive. Ideally you are helping the driver understand what a
realistic car simulation is doing.

--
Pat Dotson
UltraForce Simulations LLC
http://ultraforcesim.com

> FFB - Driving becomes a significantly more interactive experience.

> > At the same time the real world car's response to control inputs is
> > changing (due to tire wear, heat, ...) much more so than any car in
> > a racing game.

> I don't agree with that. Changes in tire wear or temp in a sim has a
> lot more impact on my driving response than I ever experienced in
> limited real life racing experience. With the infinitely higher
> feedback resolution you get in real life, responding to those sorts of
> changes in the real world is easier, or even sub-concious.

I wasn't referring to the impact on the player / driver, just the fact that
a real world car has more dynamic and subtle changes in it's response to
wear and setup than the current racing games offer. Although I'm not a Nascar
fan, Nascar does have the advantage that these small changes show up as tenths
of a second difference in lap times, and that the drivers are aware of these
small changes. Maybe it's a skill thing, but in a racing game, I'm not sure
if I could optimize a setup to the nearest tenth of a second on even an
oval track without wondering how much of this was due to driving variations.

> I think that properly modeled vehicle dynamics will yield a more
> controllable vehicle than we've seen in almost every attempt at a
> racing simulator.

Maybe that's the issue I have with a lot of the current racing games. I still
like GPL's physics, maybe not the greatest, but the high degree of yaw
(slip angle) involved with those old bias ply tires, gave a lot of visual
feedback, and I became a fan of lift throttle induced oversteer (as opposed
to trail braking) for racing games. The only real car I ever experienced
lift throttle oversteer was in a Caterham that I used to own, a 1350lb
car with 250hp (a special engine mod at the time, now a standard option).
The first time it happened, the car recovered quite nicely, and much
faster than the recovery time for my heart rate.

> > modified audible feedback (tire scrub sounds), and exaggerated visual
> > roll and yaw response

> All proper for a simulation IMO. You aren't making the car
> easier to drive. Ideally you are helping the driver understand what a
> realistic car simulation is doing.

I'll keep waiting for this sim then. Thanks for the response.

Heh. Not the Skippy. It NEVER does the same thing twice

-Larry

> My experience with racing games is that a player mostly memorizes the
> control
> inputs for a particular car and track combination, much more so that would
> be
> done in real life.

You can also add some physics, see for example the I-Way building in
Lyon. ;-) But those are expensive puppies.
Still, playing with cues too much leaves simulation and enters arcade
indeed. The question is always if your goal is 'fun' or 'realism'. More
fun often means adding some non-realistic stuff.

Ruud

jeffareid wrote:
> In the real world, there are forces that a driver feels from the acceleration,
> braking, cornering, or yawing (oversteer) that aren't present in a racing game.
>
> Even without assists in the real world, a driver always has those forces that
> can be felt and provide useful feedback.
>
> However in a game, these forces don't exist, so having some level of assist
> to compensate for the lack of forces might end up being more realistic than
> no level of assists.
>
> My experience with racing games is that a player mostly memorizes the control
> inputs for a particular car and track combination, much more so that would be
> done in real life. At the same time the real world car's response to control
> inputs is changing (due to tire wear, heat, ...) much more so than any car in
> a racing game.
>
> The "perfect" simulator would probably be a remotely controlled race car.
> The remote view would be the normal in car view. However I would expect that
> that an external view might provide better feedback and result in better lap
> times. At the extreme, a very remote external view, one that overlooks an entire
> track, would might be better still, assuming headphone playback of the on board
> sounds were used to eliminate any delay in audio feedback.
>
> OK, now for a PC based racing game simulator? What, if any, compromises should be
> made to make the racing game more realistic? The normal assists in current games
> would always be available as an option.
>
> In my opinion, modifying the physics to make the cars more forgiving when pushed
> past the limits would help compensate for the lack of real world force feedbak.
> This would include changes such as: more forgiving tire physics, reduce the rate
> of yaw response (in most games, the rate of yaw response is faster than real life,
> which is slower, even if not recoverable), modified audible feedback (tire scrub
> sounds), and exaggerated visual roll and yaw response (some games include
> adjustements for the visual stuff).