rec.autos.simulators

Oh my goodness, you should publish this and make money or something.

:-)

Lee

Think of yourself this way, and you'll never get crazy enough to post
about how accurate GPL cars handle despite never having seen one in
person, let along driving a few laps at Monza in one.

They're all games.

Tim

At least s/he didn't sign it "Publius," "Junius," or (god forbid) "Cato."

And surely this was much more than halfway decent.

But cheers all the same,
DK

PS  and you *did* respond ;)

Over the years, numerous threads have been devoted here to the sim vs arcade
debate. What is a sim and what is an arcade racer? Obviously these discussions
exist because everybody has different expectations and priorities. However, the
general attitude here seems that the absolute realism of the physics makes or
breaks a sim. That is: for more than half of the frequent posters here, a sim
will be spoiled when a car doesn't react like it would in reality. A sim that,
as a matter of speaking, *only* has correct physics to offer seems to be
regarded higher than one that fails in this respect, but has everything else
(see later). Along the same lines, a wheel and pedals are considered
*absolutely* required to fully get the experience right.

I've always had the feeling that this "hardcore" approach misses the point
somewhere. The thinking has gone too much in one direction. Because it seems so
obvious and because nobody really challenged it  :-)  So that's why I did some
hard thinking to rationalize this rather "undefined" feeling that I always had.
When I started, I didn't know where it would end. Turns out like some
philosophical voyage of discovery of simulation, hehe :-)  It's just MY view,
which I do not consider better or worse than other views. All that I want is
that the hardcore guys give it an honest read. And of course discuss it... throw
it all out if the logic fails!!    :-)

Ok, here goes...

I'd like to start with a simple question: WHY do racing simulators exist at all?

Well, millions of people would like to experience racing. Unfortunately, for
several reasons, only a few hundred worldwide are able to *actually* experience
it. Therefore, a racingsim offers ordinary people the best possible illusion of
racing. And to achieve this, it uses the virtual reality capacities of a
computer.

Right. If this would be my ultimate answer, then I could just as well stop here
because with this definition, the what I call "hardcore approach" would be
unchallengable <G>

HOWEVER, we have peeled only the first layer of the onion :-)  The key here is
that, to get to the essence, we also have to answer the question why realworld
racing ITSELF (thinking about F1 in particular) exists. Why people want to be a
racecar driver and why those who already are, find it enjoyable?

I've put together a list of reasons here (would be interested to hear whether
anyone can add things?). They are in no particular order, and most are
interconnected in some way.

1) glamour; the 'happy few', exotic places and flashy cars
2) getting respect, BEING someone; because of 1)
3) competition; measuring yourself against others
4) the challenge,getting the best out of yourself; a battle with your own
capacities
5) *physical* sensations: G-forces, arms hurt, bumps...; these give you an
adrenalin rush because you are well aware that you're "taming" a powerful beast
6) *sound* and *smell* of racecars; (HOW it sounds and smells is secondary to
the fact that these particular sounds and smells are associated with racing)
7) *visual* sensations; the impression of moving very very fast
8) the danger involved; another source of adrenalin
(9) commercials, pit girls,groupies, your own Jet and champagne or Guinness<G>

Apart from the smell and the danger, pretty much everything in this list can be
simulated to some degree. The physical sensations are difficult, the online
community gives respect and, together with AI, competition. Licenses give you an
idea of taking part in the real glamour.

Item (4), the challenge, can be broken down further. The challenge consists of
constantly pushing yourself to improve these particular skills. The constant
battle to improve these skills *IS* racing.

1) "feeling" the car with that part of your body under your back <G>
2) "listening" to the car, in order to preserve it to get it over the finishline
3) understanding how to tweak car parameters to make the car faster/better
4) braveness; not being afraid of taking risks
5) accuracy; how precise is your control over the car
6) reaction-speed; responding fast to the inputs of your senses
7) concentration; staying focused over longer periods
8) physical condition and stamina; to maintain concentration during the course
of a race
9) regularity; the ability to do it over and over again (laps/races)
10) tactics; making the right decisions even in the heat of the action
11) mental strength; controlling nerves and recovering from bad luck

You also need a rich dad and social "connections" or the ability to sell
yourself <G>

Apart from 4), every challenge in this second list is, or can be, simulated to
some degree. Simulating "feeling" the car is difficult once again. Complex car
tweaking is already available. Simulation of tactics and "listening" to the car
will undoubtedly get better and better over time.

**How well everything in this list is implemented, is what really sets a sim
apart from an arcade game!**

I think we pretty much all agree that a simulator which offers all of this,
could be called the PERFECT simulator, not?

Now comes the "mental leap": there is a subtle difference between simulating
autoracing, and simulating the enumerated aspects that make autoracing exciting
and enjoyable!

Some aspects of real autoracing as we know it now, may not be essential at all.
That is: they are what they are, but if they would be different, they would not
make the racing less enjoyable and less of a challenge. The reasons that they
"are what they are" could be historical or practical or even just random. As a
result, they aren't mentioned in the lists above!

Some examples:

- turning a steeringwheel and pushing pedals:

In the real world, they are _practical_: the easiest and most efficient way to
change direction or to produce a mechanical pressure to slow down the vehicle.
But they are also _historical_ in a way. In the past there was no other way to
control a car, but now or in the future it could be possible (and IS even
reality sometimes!!) to use joysticks, paddles and buttons to achieve the same
or better accuracy! In the real world, they have the added advantage of
feedback: a whole pallette of forces is transferred through wheel and pedals to
the pilot.

- real-world physics:

It has been shown that real-world physics seem to be finetuned to make our
Universe possible. If physical constants would be just a little bit different,
the Universe would pretty much fall apart  :-)   However, if we ignore this, we
could come up with some pretty interesting and complex alternative physics.
Physics that are JUST as challenging to handle as the physics that we know.
Springs and dampers acting totally different... Air-resistance increasing when
you brake instead of decreasing... Lateral forces the faster you go in a
straight line.... Current top pilots would STILL be top pilots under those
circumstances (if they were born in that world..) Because the really important
skill is to handle "a" complex environment.

- "trivial" things that are pretty much random: the whole points system,
technical regulations, car-liveries, the F1 circus as a whole...

We could change ALL of these things, and still end up with a great and
fantastically challenging sim. The only reason why a PERFECT simulation of these
particular aspects adds to the essence of a racingsim, is the _association_ with
real world racing. For example the "trivial" aspects mentioned above make it
easy to identify with the "real" thing, so by getting them right the sim tries
to offer you a little of the "glamour". The physics will never be TOO obviously
unreal, because that would also immediately kill the whole identification. And
we like to hold a wheel because the real drivers do so too, and because that's
what we see on TV all the time...

Take special hardware: do a wheel and pedals, as symbols for a more perfect
true-to-life simulation, significantly improve any of the enumerated
"essentials"? Force feedback gives some physical sensations, but nothing like
the real thing. They provide better accuracy, but with keyboard control that is
carefully tuned, you can also get a very challenging and complex
experience.(remember that "accuracy" was just one of the 11 required skills,
BTW) In GP2 for example, it is obviously very easy to drive slow to moderately
fast with the keyboard. But once you go to the limit, lap after lap, it becomes
very challenging and engrossing indeed! The details become important. Admittedly
not quite as difficult as with a wheel, but sufficiently exciting nevertheless!

A quality sim can obviously not get away with _totally_ neglecting all of these
non-essential aspects. The more a player knows about real racing, the more
sensitive he will be for these aspects that mainly improve the identification
with the real thing. As an example, for people who regularly drive real cars to
the limit, it might be harder to ignore the imperfections in the physics model.
(But why do they need a sim anyway?  ;-)  ) They will notice when the car reacts
funny. They will notice when a change in setup does not have the real-world
effect. Those who do NOT regularly drive real cars to the limit, will NOT be
distracted by this, and can nevertheless enjoy a slightly different model that
is just as complex. Kids or people who only occasionally watch races and are
totally unaware of what being a racepilot really takes, will even be satisfied
with an arcade game that offers barely more than the visual experience. They
will not feel the same excitement as a real racer though!

To put the whole idea in a nutshell: Imperfections in the physics model, and
keyboard or joystick control do NOT devaluate the sim-experience as profoundly
as is commonly proclaimed. When a physics model is not TOO obviously unreal, and
challenging enough, and when the control method is well executed and
challenging, ...

read more »

As it is, I'm just sad a fellow countryman posted a half way decent
argument, but was too chicken to really stand up for his opinion.

Jan.
=---

--
regards

Shane Lowry

Well, I like to study race car driving as an "armchair hobbyist."
Only one book out so far (?) on Pod Racing ;)

For me, there is a big difference between just "directing" a car model's motion
and actually "controlling" it. When controlling it I have to worry if I might
get it out of shape and lose it. If all I have to do is throw the stick over
to direct it into a 2G turn, it's not that interesting. I analyzed optimal
trajectories
for many years. Just picking a line isn't that interesting to me. Particularly in a
sterile sim environment.

So I evaluate a model based on what my controls do. If the programmer
can convince me that I'm theta-double-dotting then I like it.
If all I get is psi-dot then I'm bored.  ;)

--
Matthew V. Jessick
WB: para

Hehe...  

I could give you my bank account number, but then I would reveal my identity.
Since I'm chicken, I can't do that.   ;-)

Le Prof
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
A friend is one to whom you can pour out all the contents of your heart;
chaff and grain together.  Knowing that the gentlest of hands will take
it and sift it. Keep what's worth keeping and with the breath of kindness
blow the rest away.
                                YLM
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=

Thanks David. Although I did not consciously reanimate this old tradition. It
was more a matter of re-inventing. <G>

Yes, I admit that I spent some time deciding whether I was going to use my
(kn)own name or not. I admit that I am chicken and that I decided not to.

But Jan, aren't you disguising yourself also (what's this "no-spam" in your
mail-address   ;-)  )? I also find it funny that you call my post a "half decent
argument" without explaining. Maybe you just didn't respond because it's too
difficult to explain WHY? (just trying to trick you into a reply here,hehe )

Why a "nom-de-plume"? First of all an argument should be considered seperate
from the person that presents it. The argument counts, not the man behind it.

And secondly: it seems quite easy to burn yourself here (cfr "Frank"  ;-)  ). I
wasn't sure whether the *** scene wouldn't regard it as pure blasphemy and
"expell" me, hehe (I admit it wouldn't be half as bad as burning the witches in
the middle-ages, but anyway...  :-)  ) People with unusual opinions get "marked"
pretty easily. You end up in a little niche before you know it.

I have been wandering undisturbed around this newsgroup for quite some time now,
and I intend to keep it that way    :-)

Anyway: I'm quite pleased that already a few people have crawled through my
monster-post. When I actually looked back after completing it, and noticed how
impossibly long and precise it all was, I never hoped more than 5-10 people
would bother reading it   :-)

Le Prof
"If you can't laugh at yourself you may be missing the joke of the
century." - Edna Everidge

However, in these days of free internet accounts thrown at you from every
direction (been ducking and weaving for months now), and with the miriad of
free E-mail account providers I would have expected a valid e-mail address
(albeit using anti-spam).

Also, I can't help but feel it devaluates the reasoning a little when the
poster is not really prepared to own up to what he says. But that's probably
just me and my warped sense of values.

Jan.
=---

a truly great post! While I do not agree with what you say, it did make
me think.

I think the physics is something that needs to be modelled properly in
order to make us believe we are racing, asthis is what sims are about.
It's not only about handling a complex environment, but also handling a
somewhat familiar environment, too. While most of us have no idea how it
is to really drive a racecar, even less how to program real world
physics, we all posess very good intuition in how all things behave in a
real world where gravity and drag apply. I think we are quite sensitive
regarding this, and the margin of tolerance of how much off the basic
behaviour in a sim can be is really low. I could get convinced to some
extent by GP2, but F1RS for example really left a lot to be desired.

Basic Newtonian physics has to be applied to feed our intuition. This
cosists of forces and torques that produce lateral and angular
accelerations. It is the less well defined things like specific tyre
behaviour or aerodynamic lift and drag coefficients that need not be
perfectly modelled in order to not shatter the illusion. This is where
sim designers have some playing ground regarding ultimate realism.

-Gregor

No, we need some slack, we need also discuss pointless things from time to time, like
"arcade versus sim".

---Asgeir---

With all this secrecy, it is nice to see you left some clues on your identity,
though :-)

On topic, I have to agree with what Gregor Veble posted a little further in this
thread. I think people can be very sensitive regarding how objects in the real world
behave, and therefore, should behave in a simulation. Canned spins can and will be
spotted, even by non-*** racers. Of course, it all depends on the ideas people
have on the world that is simulated. When I first tried out GP500, I could not
believe my bike had so much understeer (reality is tough :-) ). Also, the amount of
real or virtual racing experience can make or break a sim. People that have tried a
great range of simulations will have a better idea of types of physics models, and
will be able to tell which feels 'more real' or convincing.

I agree with you that a perfect physics model alone does not constitute a great
simulation. 'The feel' of the game is made up from a lot of stuff, ranging from
graphics, AI and setup-options to a nice intro, menu navigation and other small
details. Read the post by Todd Dry ("F1 2000 = disappointing, long") a little
further in this NG, and you see what I mean.

There will still be a large group of gamers in the world, though, that just wants a
simulation in which it can recognize the visual stuff from the real world, and don't
care about physics. This group is not necessarily arcade, because accurate sims can
deliver an immersive visit to a racetrack too, but they will be just as satisfied
with an arcade sim as with a *** sim (provided they can use a lot of help :-) )
as long as the experience of 'driving' in a world they recognize is fulfilled.

Regards.
--

Seriously, your post was a good introspective into our hobby and was a
joy to read, so it surprised me when you said you are posting under a
different name. I saw nothing in your post that would offend anyone or
in any other way be inappropriate in its theme to not be accepted here.

Whoever would have been jumping at your throat because you expressed and
argumented your ideas, well, I think they would have shown where they
stand in this community themselves.

-Gregor

As for where I sit on this, I'm with "Le Professeur".  I love the complexity
of GPL, but for me, if a sim "feels" right, then for me it *is* right.  GPL
manages to be both an incredibly complex dynamic sim, and something that
"feels" right.  GP2 was built up from a lot of tweaking during the testing
(and it sounds like GP3 is the same).  Geoff Crammond and his brother drive
it, Geoff tweaks a few variables in some tables (ok, I simplify) and they do
it again until they think it "feels" right.  And damned it more than 50% of
us here still think it has a nice "feel" for a sim.

I love complex code and accurate mathematical simulations, and in theory
pretty much everything that is happening on this planet, from the cellular
level up to the geological, can be accurately simulated.  It's a question of
computing resources and time.  Corners have to be cut in any simulation to
make it run in real time, so there is nothing inherently wrong with throwing
in a few lookup tables to cram some empirical data in the back door of an
"accurate" mathematical simulation.

I have limited time to enjoy sims, much less than I thought I would have
when I finally finished up my school, so my requirements of a sim are
probably different than others.  I want a good, yet believable race.  I want
something approaching the real-world behaviour of the thing being simulated,
but I don't care how they do it.  For me, if it looks like a duck, walks
like a duck, and quacks like a duck, then for my purposes it is a duck.

So, for me:

GP2, therefore, is still a decent F1 sim in my books.  F1RS and MGPRS2 are
not.
RC2000 is a decent rally sim.  Sega isn't.
GPL is a superb classic F1 sim.  er.... SOS isn't
N3 is a decent Nascar sim.  Nascar Revolution isn't.

These are purely subjective.  I don't pick into the code to decide if it's
good or not.  My day to day job has enough of that.  These sims just feel
right.

I can hop in, the car behaves enough like the real thing that I can put
myself into the***pit and suspend disbelief, and the racing experience in
each of these sims is enough to get me involved to the point that I actually
care how things turn out, I get tense if I***up, and I feel good when I
put in a good lap / stage time.

Stephen