Realism, cont'd [long]
Robin J. Lee
Current flight sims are most realistic in the categories at the lowest
level of the spectrum, or those which focus on the operation of a single
aircraft. Thus, we are seeing unparalleled detail in systems realism
(Back to Baghdad, Hornet 3.0, Su-27, Longbow) and aerodynamic-physical
realism (A-10 Cuba, Su-27, among others). The best sims are able to
excel in both of these areas. The last low-level factor,
pilot-perspective realism, is improving slowly, but I suspect that the
most significant limitation is that of hardware.
Excellence in atmospheric realism has been achieved by products which
otherwise do not rank highly on the realism scale. The trouble with
world creation is that it takes up an inordinate amount of resources, and
some development strategies have focused upon that realism aspect at the
expense of others. I think atmospheric realism has a certain mass
appeal, as well; more people can appreciate a beautifully rendered
valley than can appreciate the modeling of the gimbal limits of the
LANTIRN pod. Largely, the focus has been disproportionately on terrain,
although further world development continues to emerge. In many cases
there seems to be a choice between highly developed terrain or a highly
developed active world, but not both.
The most advanced levels of tactical realism currently excel in the realm
of single-ship tactics. Enemy single-ship AI continues to undergo
continuous improvement. However, sims are sorely lacking in simulating
the combined effects of multi-ship, multi-formation tactics. In a
modern tactical environment where the average pilot will be, well,
average, there is no particular need to develop a perfect AI ace
(which may well be impossible). Instead, the focus should shift to the
AI's systematic utilization of multiple assets to achieve victories.
Tomorrow's high-level tactical AI will win the air battle not by
outmaneuvering the player side's aircraft one by one, but by drawing off
player fighters in one direction and punching through the resulting gap;
by targeting AWACS and support aircraft; by presenting the player with
difficult choices involving the distribution of limited resources.
Bracketing and drastic vertical separation to avoid player radar scan
volumes will be a regular part of the tactical repertoire, while rolling
scissors and close-quarters guns defense will become more rare. Missile
warfare is *not* push-button warfare; those who treat it as such will make
an involuntary habit of dying in active-radar missile head-on jousts.
Instead, it requires a whole new set of tactical skills that current AI
schemes do not yet address. In an age of repeating rifles there still is
a disproportionate focus on swordfighting. For the modern combat sim,
everybody's tactical mindset needs to shift upwards from single-ship to
multi-ship engagements.
A similar dynamic will exist for ground-based tactics, with SAM sites
equipped with rudimentary AI that directs them to strategically emit and
go silent, to recognize aircraft types by speed and hold fire for the
high value strike package. This alone will increase the challenge to
the player tenfold. Finally, all mission planners should reflect the
fundamental importance of precise timing and coordination in combat
aviation.
Operational-strategic realism is the current rage among the high-end
high-fidelity sim developers, and perhaps it has been overemphasized. At
the center of all the attention is the much sought-after "dynamic
campaign," which would actively track a wide series of relevant
variables and input them into the campaign algorithm, generating
outcomes "on the fly." A random element would ensure that no two
campaigns were exactly alike, while the algorithms themselves would keep
events within a logically derived range of outcomes.
However, a number of factors militate against the diversion of too many
resources toward that goal. First, striking the balance between
logically determined outcomes (that are predictable) and random events is
going to be difficult. War is full of unusual happenstance that often
only can be explained years after the event. Operational plans succeed
or fail based on an infinite number of variables, all of which cannot be
anticipated and modeled. In this sea of uncertainty, a random result
within certain limiting parameters will serve roughly the same purpose. If
a limited number of key variables are tracked in order to prevent
absurdity, the rest can probably be safely left up to chance.
It is vital to remember that the key to achieving operational-strategic
realism in a sim that is focused on the pilot level is largely one of
perception. Falcon 3.0 used a simplified campaign engine which continues
to fool people into thinking that is faithfully tracking more variables
than it actually is. Since the player, operating at the pilot level,
will only be able to impact the campaign in highly indirect ways, it
will not always be clear how much effect his actions should have on the
campaign at all. For an effective dynamic campaign, maybe a streamlined
model based around a few immutable principles is better than one which
tracks thousands of variables and tries to super-rationalize everything.
Finally, I think that scenario realism is more important than many people
may realize. Most sims today cover high-end warfare: this year's jets
against next year's SAMs. The single adversary is enough of a handful
that the rest of the threat environment is ignored. But real threat
scenarios are not on the WWIII high-end scale. The reality of the matter
is more complex. The "hard threats" (in the form of SAMs, AAA, etc.) are
considerably less lethal than S-300V/SA-10s planted on every block. But
the real complications come in the nature of the mission. Typical
challenges:
(a) Hide-and-go-seek. Time to put that radar modeling to good use. The
fact of the matter is that strike intelligence has not evolved to a
level where the target will always be under the waypoint. This is
particularly true for mobile targets. (The Great Scud Hunt, Feb 1991)
(b) Interdict enemy air in a zone with friendlies and neutrals, where
IFF isn't enough to guarantee that you're not going to accidentally bag
an airliner. Visual or electro-optical ID is necessary. With graphics
being what they are now, I think the technology has reached a state where
the player can tell aircraft type by looking at it. (Operation DENY
FLIGHT, NATO AFSOUTH, Bosnia-Herzegovina)
(c) Retaliatory Weasel. Anybody can hit an SA-6 site, but what if
you're not authorized to shoot unless the SA-6 locks on? (After all, in
the real world, one doesn't just go shooting at SAM sites for nothing.)
What if you're on Weasel duty in a no-fly-zone where there are quiet (so
far) SA-6s -- but there are friendly CAPs forever overhead? Mission:
make sure that nobody gets shot down, but don't shoot unless you have to.
(Iraq, 1993)
In addition, mission parameters should include factors like fuel
limitations -- well, we're extended hundreds of miles because there's no
base in the area, so do the mission fast. Plug in the burner to chase
the MiG down, or play it safe so you can get home? What about ambient
flying conditions -- "pitch black" never is quite dark enough in most
sims, yet fatigue in the darkness claims losses all the time.
There are lots of things that can kill you on a mission beyond the best
and the baddest SAM in the world. In most cases the best and the baddest
SAM will not realistically be present, but all of the other factors will
be. It'd be nice if future scenario writers would take that into
consideration.
Realism is always a compromise in a world with limited resources.
Developers will find different ways to emphasize the things that they
think are important while simply touching on the things that they do
not. I think that the single-aircraft sim -- in the sense of a sim that
focuses on the operations of the player's aircraft -- has plateaued with
extraordinarily high degrees of systems realism and aerodynamic-physical
realism. The next step is to capture the world outside the cockpit, in
all of its dimensions.
--
____________________________________________________________________________
Robin J. Lee amr...@netcom.com
Vulture's Row Worldwide Web Page URL: http://www.webcom.com/amraam/
Robin G. Kim
Robin J. Lee <amr...@netcom.com> wrote:
Interesting treatise. I'm sorry I didn't catch part 1.
>Thus, we are seeing unparalleled detail in systems realism
>(Back to Baghdad, Hornet 3.0, Su-27, Longbow) and aerodynamic-physical
>realism (A-10 Cuba, Su-27, among others). The best sims are able to
>excel in both of these areas. The last low-level factor,
>pilot-perspective realism, is improving slowly, but I suspect that the
>most significant limitation is that of hardware.
Air Warrior and Warbirds fans would argue that the latter has already
been achieved. :^)
>Excellence in atmospheric realism has been achieved by products which
>otherwise do not rank highly on the realism scale.
I'd argue that some few sims do both well.
>I think atmospheric realism has a certain mass
>appeal, as well; more people can appreciate a beautifully rendered
>valley than can appreciate the modeling of the gimbal limits of the
>LANTIRN pod.
Absolutely.
>The most advanced levels of tactical realism currently excel in the realm
>of single-ship tactics.
As long as you exclude the online/network sims, I'd mostly agree.
>Enemy single-ship AI continues to undergo continuous improvement.
Progress here has been very erratic, IMHO.
>However, sims are sorely lacking in simulating
>the combined effects of multi-ship, multi-formation tactics. In a
>modern tactical environment where the average pilot will be, well,
>average, there is no particular need to develop a perfect AI ace
>(which may well be impossible).
True, but while some sims do very well in this respect, many others are
still inadequate. Don't cut the laggards a break just because the
leaders have (almost) made the grade.
>Instead, the focus should shift to the
>AI's systematic utilization of multiple assets to achieve victories.
To the extent that it's realistic for the enemy forces being represented.
Good coordination only comes from extensive training.
>Tomorrow's high-level tactical AI will win the air battle not by
>outmaneuvering the player side's aircraft one by one, but by drawing off
>player fighters in one direction and punching through the resulting gap;
>by targeting AWACS and support aircraft; by presenting the player with
>difficult choices involving the distribution of limited resources.
IMHO, this type of AI practically necessitates the use of canned missions,
though maybe this will change in the future.
>For the modern combat sim,
>everybody's tactical mindset needs to shift upwards from single-ship to
>multi-ship engagements.
For that, we'll not only need much smarter wingmen, but more flexible
and extensive wingman commands.
>A similar dynamic will exist for ground-based tactics, with SAM sites
>equipped with rudimentary AI that directs them to strategically emit and
>go silent, to recognize aircraft types by speed and hold fire for the
>high value strike package.
They need to get both smarter _and_ stupider. In many sims,
ground-based air defenses are unrealistically alert and efficient.
>It is vital to remember that the key to achieving operational-strategic
>realism in a sim that is focused on the pilot level is largely one of
>perception.
This is true of every aspect of every flight simulation. In fact, I
would almost define the art of simulation as doing as little as
possible while still achieving the desired perception. The problem is
that most players tend to become more observant and demanding over time
as they grow accustomed to a steadily advancing state of the art.
>Finally, I think that scenario realism is more important than many people
>may realize.
While the examples you mention (which I have not quoted) may be more
realistic, in terms of PC flight sims I think it's more a matter of taste.
>Realism is always a compromise in a world with limited resources.
>Developers will find different ways to emphasize the things that they
>think are important while simply touching on the things that they do
>not.
Again, this is true for any successful simulation. Another thing to
keep in mind is that in a fast growing market, a greater and greater
percentage of buyers will be less experienced, so future sims must be
able to cater to a wide variety of skill levels. Most players require
only enough realism to immerse themselves in a sim; after that, they
just want fun gameplay. Of course, the amount of realism (which should
definitely be considered a _weighted_ average) required to achieve
immersion varies considerably from person to person.
>I think that the single-aircraft sim -- in the sense of a sim that
>focuses on the operations of the player's aircraft -- has plateaued with
>extraordinarily high degrees of systems realism and aerodynamic-physical
>realism.
I'll slightly disagree here. I believe these areas are only _starting_
to plateau.
>The next step is to capture the world outside the cockpit, in
>all of its dimensions.
Those are the things people seem to complain about most in today's sims,
so you're probably right.
pap...@ix.netcom.com
>
>Interesting treatise. I'm sorry I didn't catch part 1.
>
I reposted the first part it was so good. Alot of what Robin wants in
multiship may be answered soon by a number of sims...also I agree with
him about missile warfare being extremly interesting and most
difficult. So difficult that most of the pilots I fly with online in
Flanker will not fly BVR with me...they say its boring. I think that
when multiplayer goes online for Jet Combat sims...the number 16 comes
to mind...<VBG> then the fun will truly begin as multiship engagements
with humans doing the flying will be extremly tense and involving
flights. I dont believe that anyone will do in the next couple of
years, a truly convincing complex dynamic campaign without resorting
to mirrors...or cheats ala Falcon. The requirements for that are too
high as I believe to do this correctly you must have a multithreaded
OS with multiple Chip computers...and powerful ones at that...I can
see myself in front of the wifey explaining the need for a 6 chip PPRO
but darling FLANKER 3.5 wont run on anything less that a 6chip 400mhz,
1 terabyte ram puter...
PAPA DOC
pap...@ix.netcom.com
Great reading thanks....!
PAPA DOC
Robin J. Lee
In article <5im4hf$6...@ssbunews.ih.lucent.com>,
Robin G. Kim <opu...@lucent.com> wrote:
>
>>Excellence in atmospheric realism has been achieved by products which
>>otherwise do not rank highly on the realism scale.
>
>I'd argue that some few sims do both well.
Oh, I don't disagree. I didn't mean to imply that excellence has been
achieved *only* by these products; I just wanted to point out that this
is a category in which there is broader competition from sims which are
not targeting the hard-core realism crowd. This is probably less true
for some of the other factors, which tend to have the same (narrower)
audience.
>>The most advanced levels of tactical realism currently excel in the realm
>>of single-ship tactics.
>
>As long as you exclude the online/network sims, I'd mostly agree.
I've never tried the online sims; my comments really are mostly aimed at
the standalones. I think that many of the limitations of the standalone
game engine can (and are) "worked around" in the multiplayer sims by the
use of scenario rules, etc. This reminds me a bit of the TacOps
(land-based tactical wargame) philosophy -- why code it into the engine
if the players can just set down the ground rules?
>>Instead, the focus should shift to the
>>AI's systematic utilization of multiple assets to achieve victories.
>
>To the extent that it's realistic for the enemy forces being represented.
>Good coordination only comes from extensive training.
True. But I'm not sure it's more realistic to achieve play balance
through giving the enemy top of the line next-generation nightmare
equipment that he uses poorly, or through more realistic hardware mixes
used in an intelligent fashion. Certainly the military seems to regard
the latter option as the preferable training strategy, but then you can
argue about whether it's easier to truly train a competent army or just to
buy a lot of fancy weapon.
>>Tomorrow's high-level tactical AI will win the air battle not by
>>outmaneuvering the player side's aircraft one by one, but by drawing off
>>player fighters in one direction and punching through the resulting gap;
>>by targeting AWACS and support aircraft; by presenting the player with
>>difficult choices involving the distribution of limited resources.
>
>IMHO, this type of AI practically necessitates the use of canned missions,
>though maybe this will change in the future.
I was thinking of the AI implementation in tactical wargames, which,
while generally not brilliant, is able to react dynamically to the
player's actions on the slightly higher levels I'm suggesting here.
Granted, the processor has much more to do in EF2000 as opposed to Steel
Panthers, and therefore might not be able to spare the cycles for this
kind of AI, but it's a thought for the future, anyway.
I guess I'm making a statement less about processor load and more about
the direction of AI development. The epitome of air combat AI for a
modern jet sim, IMHO, is not in a perfectly tuned single-aircraft
routine that has every individual jet flown by an ace. It's in having an
AI that may not be as smart in the furball, but is quite competent in
maneuvering groups of planes. Fleet Defender started along this track,
but I fear that sims have since moved back into the "every
*individual* airplane flies intelligently" mindset, which could be a
mistake of emphasis.
>>A similar dynamic will exist for ground-based tactics, with SAM sites
>>equipped with rudimentary AI that directs them to strategically emit and
>>go silent, to recognize aircraft types by speed and hold fire for the
>>high value strike package.
>
>They need to get both smarter _and_ stupider. In many sims,
>ground-based air defenses are unrealistically alert and efficient.
The truly "smart" AI will also have a human error multiplier. :)
>This is true of every aspect of every flight simulation. In fact, I
>would almost define the art of simulation as doing as little as
>possible while still achieving the desired perception. The problem is
>that most players tend to become more observant and demanding over time
>as they grow accustomed to a steadily advancing state of the art.
I think that's a remarkably good way of putting it. In some sense I also
think that simulations are a bit like sausages; you really don't want to
know too much about how they're made. I can't think of a better way to
ruin a marvelous campaign engine than by looking at it as a set of
equations and complex formulae. The enjoyment, after all, is reveling in
the illusion. (Note that the culture of this newsgroup sometimes seems
to ignore this fact.)
>>Finally, I think that scenario realism is more important than many people
>>may realize.
>
>While the examples you mention (which I have not quoted) may be more
>realistic, in terms of PC flight sims I think it's more a matter of taste.
I'd go even further and say that realism itself is a matter of taste. We
may not want certain aspects to be "realistic"; I can think of a series
of areas where I would prefer other considerations to predominate. My
comments were really addressing the aspects of "realism" without making
any normative statement as to whether a "perfectly realistic" sim
would be even desirable. I think the assumption is often made that the
more realistic, the better, but I don't think that assumption should go
unexamined. Exciting and interesting as combat aviation is, there are
some parts of it that just don't make a good game. (And there is, of
course, the learning curve problem that you brought up -- although that
is a product strategy consideration that I've also left out of the
analysis of "realism" per se.)
Robin J. Lee
In article <334ff77d...@nntp.ix.netcom.com>,
<pap...@ix.netcom.com> wrote:
>
>I reposted the first part it was so good. Alot of what Robin wants in
>multiship may be answered soon by a number of sims...also I agree with
>him about missile warfare being extremly interesting and most
>difficult. So difficult that most of the pilots I fly with online in
>Flanker will not fly BVR with me...they say its boring. I think that
>when multiplayer goes online for Jet Combat sims...the number 16 comes
>to mind...<VBG> then the fun will truly begin as multiship engagements
>with humans doing the flying will be extremly tense and involving
>flights.
I suspect we'll all do a lot of learning about the limitations of radar
and become intimately familiar with determining the edges of missile
envelopes, both things which have been modeled with some detail in the
past, but haven't really gotten the emphasis necessary for BVR tactics.
We'll also become familiar with fuel ladders, acceleration, and whether
or not you can catch the bugger on the edge of the scope or not... <g>
>I dont believe that anyone will do in the next couple of
>years, a truly convincing complex dynamic campaign without resorting
>to mirrors...or cheats ala Falcon.
Mirrors and cheats are fine, just as long as you never find out that
they're mirrors or cheats. :)
Robin G. Kim
>>
>>>The most advanced levels of tactical realism currently excel in the realm
>>>of single-ship tactics.
>>
>>As long as you exclude the online/network sims, I'd mostly agree.
>
>I've never tried the online sims; my comments really are mostly aimed at
>the standalones. I think that many of the limitations of the standalone
>game engine can (and are) "worked around" in the multiplayer sims by the
>use of scenario rules, etc.
At a high level, you are partially correct (you're partially incorrect
in that some squadrons or even ad hoc collections of individuals
sometimes collaborate on fairly well defined missions). However, to
even imply that online sims do not provide a host of multi-ship
tactical challenges is just plain wrong (I know I'm probably reading
too much into what you said, but I can't stop now--I'm on a roll).
Opportunities to employ pure single-ship tactics are relatively rare
online, and that has nothing to do with whether scenario rules are in
place.
I know you prefer jets, but I urge you to at least try AW or WB online
some time just so you get a feel for what they're like. I think you would
be impressed.
>I guess I'm making a statement less about processor load and more about
>the direction of AI development. The epitome of air combat AI for a
>modern jet sim, IMHO, is not in a perfectly tuned single-aircraft
>routine that has every individual jet flown by an ace. It's in having an
>AI that may not be as smart in the furball, but is quite competent in
>maneuvering groups of planes. Fleet Defender started along this track,
>but I fear that sims have since moved back into the "every
>*individual* airplane flies intelligently" mindset, which could be a
>mistake of emphasis.
I don't disagree, but I think you're ignoring the fact that individual
plane AI still has a long way to go even if the designers' goal is to
simulate a pilot with _mediocre_ abilities. The most obvious example
is disengagement. Many AI pilots NEVER run away, resulting in a far
higher casualty rate on both sides than is realistic (a pet peeve of
mine). Those who do run away often do so in a stupid manner. It takes
intelligence (artificial or otherwise) to know when and how to disengage
from a fight, or when to avoid one altogether.
I could come up with other examples, but I think you get the idea. In
all fairness to game designers and developers, I should point out that
I do see progress being made in this area--AI is getting better and
better at what I consider a satisfactory rate. I don't expect
perfection overnight.
>>They need to get both smarter _and_ stupider. In many sims,
>>ground-based air defenses are unrealistically alert and efficient.
>
>The truly "smart" AI will also have a human error multiplier. :)
Yes! To be truly intelligent, you must be capable of making mistakes.
>I also think that simulations are a bit like sausages
This statement is even more profound when taken totally out of context,
as I'm doing here. ;^)
>I can't think of a better way to
>ruin a marvelous campaign engine than by looking at it as a set of
>equations and complex formulae. The enjoyment, after all, is reveling in
>the illusion. (Note that the culture of this newsgroup sometimes seems
>to ignore this fact.)
Or to directly contradict it! Some people actually enjoy taking things
apart to find out how they tick. I have a tendency to be that way
myself, but I'll admit that if I analyze a game too much (as my writing
job often requires me to do), it takes away a lot of the fun.
Corsair
In article <amraamE8...@netcom.com> amr...@netcom.com (Robin J. Lee) writes:
>From: amr...@netcom.com (Robin J. Lee)
>Subject: Realism, cont'd [long]
>Date: Fri, 11 Apr 1997 17:46:17 GMT
>The most advanced levels of tactical realism currently excel in the realm
>of single-ship tactics. Enemy single-ship AI continues to undergo
>continuous improvement. However, sims are sorely lacking in simulating
>the combined effects of multi-ship, multi-formation tactics. In a
>modern tactical environment where the average pilot will be, well,
>average, there is no particular need to develop a perfect AI ace
>(which may well be impossible). Instead, the focus should shift to the
>AI's systematic utilization of multiple assets to achieve victories.
This last point I tend to agree with. Rather than have every enemy an ace,
perhaps the AI C&C should be improved to utilize the assets available. In
addition, I think that multi-player will offset some of these AI limitations.
>(a) Hide-and-go-seek. Time to put that radar modeling to good use. The
>fact of the matter is that strike intelligence has not evolved to a
>level where the target will always be under the waypoint. This is
>particularly true for mobile targets. (The Great Scud Hunt, Feb 1991)
GS2000 did a very good job at times of moving the tanks to give you a hell of
a time finding them.
>(b) Interdict enemy air in a zone with friendlies and neutrals, where
>IFF isn't enough to guarantee that you're not going to accidentally bag
>an airliner. Visual or electro-optical ID is necessary. With graphics
>being what they are now, I think the technology has reached a state where
>the player can tell aircraft type by looking at it. (Operation DENY
>FLIGHT, NATO AFSOUTH, Bosnia-Herzegovina)
I don't think the graphics have reached that point yet. I think you'd have to
get very close to visually ID a target in a sim (unrealistically close). Sean
Long mentioned the distance that he could ID a target and I don't think sims
can imitate that yet. Do we know enough about NCTR to model it in a sim yet? I
don't we do to have this help either.
>not. I think that the single-aircraft sim -- in the sense of a sim that
>focuses on the operations of the player's aircraft -- has plateaued with
>extraordinarily high degrees of systems realism and aerodynamic-physical
>realism. The next step is to capture the world outside the cockpit, in
>all of its dimensions.
I tend to disagree. At least in terms of avionics realism I think we've just
started getting there. And as it is, I think that only two released sim has
reached the level that I've wanted for some time: Back to Baghdad (based on
Buzz's comments) and Longbow. In terms of Longbow I don't think it truly
reached that level until Flashpoint: Korea. Don't get me wrong, Longbow was
fantastic. However, there were things about targeting and camera operation
that I knew were off. Those improvements that came about in FPK hit that
bull's-eye that I had been waiting for.
I may agree after I see what all of these new releases scheduled shortly do,
but to be honest I don't think we'll reach that plateau this time either. The
past have seen a lot of avionics shortcuts. Part of this has been because of
the complexity of these systems. I'd rather see the systems get very complex
with options as to the realism so that players of different levels can be
accomodated. This would also allow these people to grow with the sim as well.
Corsair
__________________________________________________________
Web CAG of The Unofficial "Jolly Rogers" Site
http://www-home.calumet.yorku.ca/mdonalds/www/home.htm
__________________________________________________________
CAG of the "VF-84 Jolly Rogers" Simulation Squadron
__________________________________________________________
"Water is composed of two gins, Oxygin and Hydrogin.
Oxygin is pure gin. Hydrogin is gin and water."
__________________________________________________________
Robin J. Lee
In article <5iqthr$f...@ssbunews.ih.lucent.com>,
>>the standalones. I think that many of the limitations of the standalone
>>game engine can (and are) "worked around" in the multiplayer sims by the
>>use of scenario rules, etc.
>
>in that some squadrons or even ad hoc collections of individuals
>sometimes collaborate on fairly well defined missions). However, to
>even imply that online sims do not provide a host of multi-ship
>tactical challenges is just plain wrong (I know I'm probably reading
>too much into what you said, but I can't stop now--I'm on a roll).
>Opportunities to employ pure single-ship tactics are relatively rare
>online, and that has nothing to do with whether scenario rules are in
>place.
Fair enough. Let me amend my previous statement; my comments were
directed *solely* at stand-alone sims. :)
>I know you prefer jets, but I urge you to at least try AW or WB online
>some time just so you get a feel for what they're like. I think you would
>be impressed.
I think I would be too, at least judging from the traffic on this
group. Unfortunately my current schedule doesn't allow for frequent
flight-sim'ing these days, and the I'm sure the lack of practice would
make me meat on the table for the vets out there. ;) (Heh, twenty
hours a day of practice wouldn't likely change things much, but being
"busy" is a comfortable excuse.)
But I'll try to squeeze online sometime and have a look at what all the
fuss is about. :)
>>I can't think of a better way to
>>ruin a marvelous campaign engine than by looking at it as a set of
>>equations and complex formulae. The enjoyment, after all, is reveling in
>>the illusion. (Note that the culture of this newsgroup sometimes seems
>>to ignore this fact.)
>
>Or to directly contradict it! Some people actually enjoy taking things
>apart to find out how they tick. I have a tendency to be that way
>myself, but I'll admit that if I analyze a game too much (as my writing
>job often requires me to do), it takes away a lot of the fun.
Granted. I do wonder if that strictly qualifies as enjoying the
simulation as a simulation, or simulation as thing-with-which-to-tinker,
but that's just putting things into different semantic boxes (which I
think I've done far too much of already. :)
On a somewhat related note, it does strike me that we've become quite
good at tinkering over the years, and that older sims may actually avoid
a lot of the criticism that we level at newer ones simply because at the
time, we weren't quite so discerning.
I remember when F3 first was released, and I read in the back of that
wonderful manual that "terrain masking effects on radar are not
currently modeled" but we should fly low anyway because it was realistic
and it was good practice. I shrugged that off without a thought; today,
things like that would bother me. Now, one may argue that technology
has advanced to a point that the general standards have been raised, but
even adjusting for technology (i.e. considering the lack of a feature
that is every bit as revolutionary today as terrain-masking was in 1991) I
just have a feeling that it's more than that. It's almost that we are
less inclined to exercise our imaginations to fill the gaps in the sim.
Perhaps it's a question of whether we're still ready to approach sims
with a *willing* suspension of disbelief rather than with a skeptical one.
Perhaps the simulation hobby has matured sufficiently to enter the age of
winged grognards?
Robin J. Lee
In article <mdonalds.11...@calumet.yorku.ca>,
Corsair <mdon...@calumet.yorku.ca> wrote:
>In article <amraamE8...@netcom.com> amr...@netcom.com (Robin J.
>Lee) writes:
>
>This last point I tend to agree with. Rather than have every enemy an ace,
>perhaps the AI C&C should be improved to utilize the assets available. In
>addition, I think that multi-player will offset some of these AI limitations.
Something to consider is that it's much more likely to wind up
with a competent general and a poor army than with a poor general and a
competent army, if only because it's easier to find/train one competent
individual and pin stars on him than it is to find/train thirty thousand
competent individuals and fill the ranks. So, you centralize the
thinking at the top and make it as simple as possible on the end-user
side (students of Soviet military thought will find this notion
familiar). A similar model of AI that emphasizes the higher level
thinking over the lower level thinking may be the way to go.
>>(b) Interdict enemy air in a zone with friendlies and neutrals, where
>>IFF isn't enough to guarantee that you're not going to accidentally bag
>>an airliner. Visual or electro-optical ID is necessary. With graphics
>>being what they are now, I think the technology has reached a state where
>>the player can tell aircraft type by looking at it. (Operation DENY
>>FLIGHT, NATO AFSOUTH, Bosnia-Herzegovina)
>
>I don't think the graphics have reached that point yet. I think you'd have to
>get very close to visually ID a target in a sim (unrealistically close). Sean
>Long mentioned the distance that he could ID a target and I don't think sims
>can imitate that yet.
Interface reworking, maybe, with an inconspicuous pop-up at realistic
fighter-pilot VID ranges. But you probably should be forced to get
closer than radar range to ID something; that's a big part of modern air
operations.
>Do we know enough about NCTR to model it in a sim yet? I don't we do to
>have this help either.
Well, spooky technologies that have made it into the press, like
jet-engine-modulation, can be painted in really broad strokes -- the
numbers will be off, but that's true for every classified technology
modeled. I'm also not sure Non-Cooperative Target Recognition has
evolved to the point where it's a big factor in operations -- if only
because it's not well-known enough that the political powers trust it
enough to adapt the ROE appropriately. But that's obviously speculation
on my part.
>>I think that the single-aircraft sim -- in the sense of a sim that
>>focuses on the operations of the player's aircraft -- has plateaued with
>>extraordinarily high degrees of systems realism and aerodynamic-physical
>>realism. The next step is to capture the world outside the cockpit, in
>>all of its dimensions.
>
>I tend to disagree. At least in terms of avionics realism I think we've just
>started getting there. And as it is, I think that only two released sim has
>reached the level that I've wanted for some time: Back to Baghdad (based on
>Buzz's comments) and Longbow. In terms of Longbow I don't think it truly
>reached that level until Flashpoint: Korea. Don't get me wrong, Longbow was
>fantastic. However, there were things about targeting and camera operation
>that I knew were off. Those improvements that came about in FPK hit that
>bull's-eye that I had been waiting for.
Maybe "plateau" was too strong. What I meant to say is that as a general
matter, the improvements we're seeing in avionics are incrementally less
of an improvement now than they were in the past, that the (functional)
jump in realism we saw when sims started modeling a user-directed radar
scan volume and optical sensors with limited field-of-regard/gimbal
limits was far greater than what we're likely to get in the near future
(additional radar modes, brightness wash-out in goggle sensors, more
environmental effects). To me, that leveling-off of the fidelity curve
is an indication that resources *might* be more profitably devoted to
other areas where the incremental benefit may be more. But I could be
wrong; I won't be too heartbroken if I am, because I am first and
foremost a systems realism guy. ;)
I don't mean to let any products "off the hook," by saying that, by the
way. I'm well aware that there are only a few sims that have reached the
levels I'm describing, and that those are not indicative of the general
market. However, the flight sim business seems to be fairly quick to
adopt breakthroughs once they're made -- padlock, the virtual cockpit (to
the chagrin of the AW/WB crowd, I know :), adjustable scan volumes, etc.
The real trick seems to be taking the first step.
>I may agree after I see what all of these new releases scheduled shortly do,
>but to be honest I don't think we'll reach that plateau this time either. The
>past have seen a lot of avionics shortcuts. Part of this has been because of
>the complexity of these systems.
Obviously we have a long way to go until it's perfect, or as near-perfect
as it can be. But I suspect the growth will slow down some, at least for
a while. We'll see.
Have you seen Hornet 3.0? In terms of systems realism (excluding the
FLIR pod and one of the weapons HUD modes), this is comparable to B2B,
IMHO.
Robin G. Kim
>>some time just so you get a feel for what they're like. I think you would
>>be impressed.
>But I'll try to squeeze online sometime and have a look at what all the
>fuss is about. :)
If it's Warbirds, let me know when you'll be up and I'll wing with you.
>>Some people actually enjoy taking things
>>apart to find out how they tick. I have a tendency to be that way
>>myself, but I'll admit that if I analyze a game too much (as my writing
>>job often requires me to do), it takes away a lot of the fun.
>
>Granted. I do wonder if that strictly qualifies as enjoying the
>simulation as a simulation, or simulation as thing-with-which-to-tinker,
>but that's just putting things into different semantic boxes (which I
>think I've done far too much of already. :)
It may be just semantics, but I think it's a valid point.
>On a somewhat related note, it does strike me that we've become quite
>good at tinkering over the years, and that older sims may actually avoid
>a lot of the criticism that we level at newer ones simply because at the
>time, we weren't quite so discerning.
Absolutely. The average hard core flight simmer is getting more and more
sophisticated and demanding over time. I don't think this is the whole
story, though.
>I remember when F3 first was released, and I read in the back of that
>wonderful manual that "terrain masking effects on radar are not
>currently modeled" but we should fly low anyway because it was realistic
>and it was good practice. I shrugged that off without a thought;
It bothered me. ;^)
>today, things like that would bother me. Now, one may argue that technology
>has advanced to a point that the general standards have been raised, but
>even adjusting for technology (i.e. considering the lack of a feature
>that is every bit as revolutionary today as terrain-masking was in 1991) I
>just have a feeling that it's more than that. It's almost that we are
>less inclined to exercise our imaginations to fill the gaps in the sim.
>Perhaps it's a question of whether we're still ready to approach sims
>with a *willing* suspension of disbelief rather than with a skeptical one.
I personally think sim players today are less willing to gloss over
omissions because their standards have been raised by the advance of
flight sim technology. For example, once you've played Tornado, with
its terrain-masking and accurate missile modeling, it becomes harder to
accept sims that are less realistic in these respects. Falcon 3.0 was
the best there was in many areas when it first came out, despite its
imperfections and omissions.
This spiral of rising expectations makes life particularly tough on sim
developers these days. Because the market seems to be growing so
quickly, a higher and higher percentage of sim buyers will be rank
novices, so developers must either pick a niche or expend huge amounts
of resources trying to appeal to both the beginner and the expert
crowds.
>Perhaps the simulation hobby has matured sufficiently to enter the age of
>winged grognards?
I believe expectations from some people are definitely high enough to qualify
them as grognards. I suppose I'm one, but I'm not proud of it. ;^)
Ridder
In the spirit of Robin Lee’s excellent essay on realism, I thought I
would add a few words on the subject. Specifically, this is my wish
list of things that I’d like to see that would help transform today’s
*flight* sims into *modern air combat* sims.
The obvious things we’d all like to see have been discussed in this
newsgroup: accurate flight modeling, avionics, a mission planner,
dynamic campaign, multiplayer, better wingman AI, high framerates,
textures, no textures, 3D support, a perfect view system, etc., etc.,
etc.
I’d like to add the following, which could be characterized as *realism*
enhancements. Yes, I realize that some sims have some of these things.
Yes, I realize that it would be a lot of work to implement all of this.
Yes, I realize that the machines that could run such a thing are
probably *N* years away. Yes, I realize that the audience for this
level of realism would be small. Yes, I realize I’ll never get all of
this. I’m just dreaming here and trying to be constructive.
Here’s what I’d like to see:
1) A true IADS. There are many sims that give us discrete SAMs and
even a certain level of interaction between an EWR and a SAM. But I’d
like to see an honest-to-goodness IADS. That is, early warning/gci
radars should report to a command and control post which designates SAMs
or aircraft for engagement. A real IADS would give me several ways to
interrupt the flow of information to the threat. I could kill the
command and control bunker, I could jam or kill the EW radar, I could
jam the communications links, I could kill or jam the SAM, etc.
2) Realistic response times. A radar requires more than 1 scan to
establish a track. It takes time to pass the track to the command and
control post. It takes time for the command and control post to cue a
SAM. It takes time to...you get the idea. Everything takes
time...let’s see it accounted for properly.
3) Realistic Ground AI. Most current sims assume that a radar is on
continuously as long as there is a target within detection range.
Knowing that there’s somebody with ARMs out there who wants to shoot you
dead, would YOU illuminate continuously? Ground AI should weigh the
factors of detection and effectiveness vs. vulnerability. Ignorance or
extreme gutsiness would lead to more illumination time. A measure of
caution would lead to less. A coward may never light up.
4) Diversity of SAMs. There are more SAM types in the world than just
the SA-6, SA-8, SA-11, and SA-15, yet for many sims that’s all we get.
How about more? More diversity means more uncertainty...and more
challenge.
5) Realistic SAMs. In many sims, the only difference between one SAM
type and the next is the engagement envelope. There’s much more to it
than that. Their engagement sequences are different. Their ECCM’s are
different (Where’s the home-on-jam?). Their employments are different.
Their purposes are different.
6) More realistic SEAD. There’s more to SEAD than just firing an ARM
at a threatening SAM. What about standoff jamming (e.g., EA-6B, EF-111,
EC-130H)? What about being forced to create real target priorities
which require you to suppress the IADS (see 1)? More and better SEAD
assets coupled with more and better IADS modeling require a whole new
and realistic way of thinking. Mission planning may now require waves
of attacks with multiple assets: standoff jamming to protect strikers
(including both radar and comm jamming), waves of ARM launches in
various modes (preemptive and reactive), bomb droppers, CAP, etc.
Oh...almost forgot. How about some air-launched decoys?
7) Uncertainties. This is the big one. What if nothing was perfect?
I’m not talking about systems failures like in Flanker. I’m talking
about imperfections in systems. How about a realistic RWR that provides
ambiguous, or maybe even erroneous ID of a threat? How about
inaccuracies in reporting position? How about missed intercepts? How
about IFF imperfections (threat of blue-on-blue)? How about botched
intel? How about aged intel (you wouldn’t know exactly the number,
types, or locations of threats)? Give the issue of uncertainty some
thought and you’ll begin to see how this one can drastically alter
gameplay (just as it complicates reality).
8) Improved avionics modeling. Don’t get me wrong, I love the avionics
modeling in many of the sims. Longbow’s multi-page MFDs, ORT, etc. are
great. However, I’d like to see imperfect systems. Similar to the
uncertainties I mentioned above, what if the avionics modeling actually
accounted for inadequacies of some systems by using realistic frequency
ranges, sampling rates, receiver sensitivies, azimuth and elevation
coverage? I know they’d be a guess, but ANY guess is better than
assuming perfection.
9) Kill removal. Campaigns are great, but without kill removal they
aren’t really campaigns...just a series of missions that are either
scripted (e.g., Longbow) or dynamically generated (e.g., EF2000). I
want to see my hard work of today pay off tomorrow such that the thing I
destroyed is truly an asset lost to the enemy. Also, I’d like to see
things that are damaged (but repairable) be repaired at an appropriate
rate which accounts for available assets and existence of supply lines
(logistics). For example, if I destroy a radar dish today, that radar
is out of commission for awhile. It may come back online, however, if
the enemy holds the turf long enough, there is a spare part available,
and there is a supply line that can get the part to the site. A
permanent hard kill would have been achieved if I had destroyed the
transmitter rather than just the dish.
Just a few thoughts.
JR
Robin J. Lee wrote:
>
> <An Outstanding Essay>
CJ Martin
In article <335587...@erols.com>, jri...@erols.com wrote:
[snip]
>Here’s what I’d like to see:
>
>1) A true IADS. There are many sims that give us discrete SAMs and
>even a certain level of interaction between an EWR and a SAM. But I’d
>like to see an honest-to-goodness IADS. That is, early warning/gci
>radars should report to a command and control post which designates SAMs
>or aircraft for engagement. A real IADS would give me several ways to
>interrupt the flow of information to the threat. I could kill the
>command and control bunker, I could jam or kill the EW radar, I could
>jam the communications links, I could kill or jam the SAM, etc.
Having learned a thing or two about AI's in general, AI's in flight
sims in particular, and specifically AI controlled ground defenses, I
can tell you that a sim does exist right now that does this:
Longbow.
I know, I know, I'm biased. But these things are modeled in Longbow.
>2) Realistic response times. A radar requires more than 1 scan to
>establish a track. It takes time to pass the track to the command and
>control post. It takes time for the command and control post to cue a
>SAM. It takes time to...you get the idea. Everything takes
>time...let’s see it accounted for properly.
Longbow has this too.
>3) Realistic Ground AI. Most current sims assume that a radar is on
>continuously as long as there is a target within detection range.
>Knowing that there’s somebody with ARMs out there who wants to shoot you
>dead, would YOU illuminate continuously? Ground AI should weigh the
>factors of detection and effectiveness vs. vulnerability. Ignorance or
>extreme gutsiness would lead to more illumination time. A measure of
>caution would lead to less. A coward may never light up.
And this.
>4) Diversity of SAMs. There are more SAM types in the world than just
>the SA-6, SA-8, SA-11, and SA-15, yet for many sims that’s all we get.
>How about more? More diversity means more uncertainty...and more
>challenge.
And this too.
>5) Realistic SAMs. In many sims, the only difference between one SAM
>type and the next is the engagement envelope. There’s much more to it
>than that. Their engagement sequences are different. Their ECCM’s are
>different (Where’s the home-on-jam?). Their employments are different.
>Their purposes are different.
This I honestly am not sure about.
>6) More realistic SEAD. There’s more to SEAD than just firing an ARM
>at a threatening SAM. What about standoff jamming (e.g., EA-6B, EF-111,
>EC-130H)? What about being forced to create real target priorities
>which require you to suppress the IADS (see 1)? More and better SEAD
>assets coupled with more and better IADS modeling require a whole new
>and realistic way of thinking. Mission planning may now require waves
>of attacks with multiple assets: standoff jamming to protect strikers
>(including both radar and comm jamming), waves of ARM launches in
>various modes (preemptive and reactive), bomb droppers, CAP, etc.
>Oh...almost forgot. How about some air-launched decoys?
Coming soon...
>7) Uncertainties. This is the big one. What if nothing was perfect?
>I’m not talking about systems failures like in Flanker. I’m talking
>about imperfections in systems. How about a realistic RWR that provides
>ambiguous, or maybe even erroneous ID of a threat? How about
>inaccuracies in reporting position? How about missed intercepts? How
>about IFF imperfections (threat of blue-on-blue)? How about botched
>intel? How about aged intel (you wouldn’t know exactly the number,
>types, or locations of threats)? Give the issue of uncertainty some
>thought and you’ll begin to see how this one can drastically alter
>gameplay (just as it complicates reality).
We (in the office) go round and round on this. Hard core guys would
love the imperfections of real life modeled, but the fear is that
everyone else would consider it bugs, or bad design. I can see both
sides, and having a 12 year background in combat avionics maintenance
gives me a pretty good basis for knowing what is real, and what is
not. I'm not sure I want my perfectly planned mission to go in the
dumper at the last minute because my radar transmitter croaked...but
that is real life. The simulation is important, however, there is also
a story, a feeling, a...something that is hard to put into words.
After all, these are still games, still entertainment. What if Lukes
X-wing had a weapons system malfunction during "the trench run" and
the proton torpedos 'hung'? Real life, or bad script?
That said, I'm not disagreeing with you...just point out some of what
I've learned in the past six months. Some times you can do both things
through the use of user choice (the "realistic" vs "arcade" settings);
however, there is also a limitation as to how much programming time
(and testing time, perhaps more important) any team can realistically
spend. Do you support 2% of the users, if that support costs you 50%
of your development time? (extreme examples, I know).
>8) Improved avionics modeling. Don’t get me wrong, I love the avionics
>modeling in many of the sims. Longbow’s multi-page MFDs, ORT, etc. are
>great. However, I’d like to see imperfect systems. Similar to the
>uncertainties I mentioned above, what if the avionics modeling actually
>accounted for inadequacies of some systems by using realistic frequency
>ranges, sampling rates, receiver sensitivies, azimuth and elevation
>coverage? I know they’d be a guess, but ANY guess is better than
>assuming perfection.
See above.
>9) Kill removal. Campaigns are great, but without kill removal they
>aren’t really campaigns...just a series of missions that are either
>scripted (e.g., Longbow) or dynamically generated (e.g., EF2000). I
>want to see my hard work of today pay off tomorrow such that the thing I
>destroyed is truly an asset lost to the enemy. Also, I’d like to see
>things that are damaged (but repairable) be repaired at an appropriate
>rate which accounts for available assets and existence of supply lines
>(logistics). For example, if I destroy a radar dish today, that radar
>is out of commission for awhile. It may come back online, however, if
>the enemy holds the turf long enough, there is a spare part available,
>and there is a supply line that can get the part to the site. A
>permanent hard kill would have been achieved if I had destroyed the
>transmitter rather than just the dish.
Hehehe...where *is* that bug???
>Just a few thoughts.
>
>JR
And damn fine ones.
CJ
Robin G. Kim
CJ Martin <cma...@origin.ea.com> wrote:
>We (in the office) go round and round on this. Hard core guys would
>love the imperfections of real life modeled, but the fear is that
>everyone else would consider it bugs, or bad design.
If you document it, it's not a bug. :^) If you decide to implement
random system failures, just list all such failures not related to damage
in the debrief so the player will understand what happened.
Implementing such a feature may be a bit of work, but turning it off
should be easy--just set the failure frequencies to zero.
Ridder
CJ Martin wrote:
>
> In article <335587...@erols.com>, jri...@erols.com wrote:
>
> [snip]
>
> >Here’s what I’d like to see:
> >
> >1) A true IADS. [snip]
>
> Having learned a thing or two about AI's in general, AI's in flight
> sims in particular, and specifically AI controlled ground defenses, I
> can tell you that a sim does exist right now that does this:
>
> Longbow.
>
> I know, I know, I'm biased. But these things are modeled in Longbow.
I agree that Longbow probably does the best job of simulating an IADS of
any current sim. However, the IADS is somewhat degraded by the nature
of the Longbow itself...i.e., to move with stealth in low numbers, evade
the IADS, and attack only those elements that are either part of your
mission or necessary for survival. In effect, the NOE flight
characteristics of the Longbow reduces what could be many-on-many
engagements to few-on-few engagements...meaning that the IADS model
could be greatly simplified (and still be realistic) by dealing with
only a few objects at once. The next question is: is the Longbow IADS
model robust enough to work with a jet sim, with many emitters and SAMs
of many types engaging many aircraft of multiple types?
>
> >2) Realistic response times. [snip]
>
> Longbow has this too.
Agreed. The delay with which enemy helo’s are dispatched is proof that
Longbow does a very good job of this. However, many sims don’t...and
this is a major problem in my view. The timing of information flow,
sensor to shooter, is a critical aspect of realism.
>
> >3) Realistic Ground AI. [snip]
>
> And this.
Yeah...I’ve noticed a bit of such behavior in Longbow. Again, for the
aspects of the IADS that an Apache would have to deal with it does an
excellent job. I just wonder if it would be robust enough to extend it
to a jet sim. For example, create a multi-axis, multi-phase,
multi-platform strike. You’ve got 18’s with HARMs, 15E’s, CAP, jamming,
air launched decoys...the whole works. Can you really simulate the
effects of real-time confusion and decision making going on at the
EW/GCI radar? At the C2 post? At the SAM?
>
> >4) Diversity of SAMs. [snip]
>
> And this too.
I can only partially agree with you here. The Baltic Hammer campaign
may have an appropriate mix of all Russian equipment (although I’d like
to see some SA-10’s, -12’s, and -18’s). But in 3rd world scenarios,
you’re going to see a motley mix of systems...including French, U.S.,
and Chinese. Wouldn't it be great to go into a scenario where your
intel told you the air defenses were mostly Russian and French, only to
find a few U.S. Hawk systems in the mix?
>
>[snip]
>
> >6) More realistic SEAD. [snip]
>
> Coming soon...
Oh? And the name of that sim would be...(I’m not sure, but I might be
getting downright giddy!)
>
> >7) Uncertainties. [snip]
>
> We (in the office) go round and round on this. Hard core guys would
> love the imperfections of real life modeled, but the fear is that
> everyone else would consider it bugs, or bad design. I can see both
> sides, and having a 12 year background in combat avionics maintenance
> gives me a pretty good basis for knowing what is real, and what is
> not. I'm not sure I want my perfectly planned mission to go in the
> dumper at the last minute because my radar transmitter croaked...but
> that is real life. The simulation is important, however, there is also
> a story, a feeling, a...something that is hard to put into words.
> After all, these are still games, still entertainment. What if Lukes
> X-wing had a weapons system malfunction during "the trench run" and
> the proton torpedos 'hung'? Real life, or bad script?
I can almost hear Jack Nicholson..."You want realism?!? You can’t
handle realism!!!" <g>
>
> That said, I'm not disagreeing with you...just point out some of what
> I've learned in the past six months. Some times you can do both things
> through the use of user choice (the "realistic" vs "arcade" settings);
> however, there is also a limitation as to how much programming time
> (and testing time, perhaps more important) any team can realistically
> spend. Do you support 2% of the users, if that support costs you 50%
> of your development time? (extreme examples, I know).
I agree...it is just a game, and the desires of the 2% may not be worthy
of service. That’s the way the market works and I accept that. But it
is a game that pretends to simulate reality to one degree or another.
We all know how the world of warfare works:
1) Somebody develops a system.
2) Somebody else develops a system to defeat that system, rendering it
ineffective.
3) Repeat 2 forever or until world peace breaks out (in which case you
go back to 1).
There’s a reason that the F/A-18E/F’s IDECM system is vastly different
from the F/A-18C/D’s countermeasures suite. However, in the *average*
sim today you would never be able to tell the difference. (Disclaimer:
I have not played Hornet 3.0 so I am not referring in any way to this
sim).
Here’s an example of what I’m talking about: I love Flanker, but it is
very unrealistic for a threat to be *precisely* geolocated by the RWR
the moment it lights up (there’s that response time thing again...).
There’s a physical limitation. The sim developers will need to account
for that limitation properly if they ever want to include GPS weapons
such as JSOW. With JSOW and a sensor that precisely geolocates emitters
instantly, the sim just became an arcade game.
>
>[snip]
>
> >9) Kill removal. [snip]
>
> Hehehe...where *is* that bug???
That settles it...I am definitely giddy!
JR
Aaron S.
>
> >7) Uncertainties. This is the big one. What if nothing was perfect?
> >I’m not talking about systems failures like in Flanker. I’m talking
> >about imperfections in systems. How about a realistic RWR that provides
> >ambiguous, or maybe even erroneous ID of a threat? How about
> >inaccuracies in reporting position? How about missed intercepts? How
> >about IFF imperfections (threat of blue-on-blue)? How about botched
> >intel? How about aged intel (you wouldn’t know exactly the number,
> >types, or locations of threats)? Give the issue of uncertainty some
> >thought and you’ll begin to see how this one can drastically alter
> >gameplay (just as it complicates reality).
>
> We (in the office) go round and round on this. Hard core guys would
> love the imperfections of real life modeled, but the fear is that
> everyone else would consider it bugs, or bad design. I can see both
> sides, and having a 12 year background in combat avionics maintenance
> gives me a pretty good basis for knowing what is real, and what is
> not. I'm not sure I want my perfectly planned mission to go in the
> dumper at the last minute because my radar transmitter croaked...but
> that is real life. The simulation is important, however, there is also
> a story, a feeling, a...something that is hard to put into words.
> After all, these are still games, still entertainment. What if Lukes
> X-wing had a weapons system malfunction during "the trench run" and
> the proton torpedos 'hung'? Real life, or bad script?
>
If the proton torpedos had 'hung' they either wouldn't have made the
movie, or they would have made it from the prospective of the Empire
("Beware of the Light Side of the Force"). Good thing Luke didn't have
a systems failure (even though he didn't use the targeting computer).
Aaron
Andy Hollis
Ridder <jri...@erols.com> wrote in article <33569E...@erols.com>...
[snip]
>
> I agree that Longbow probably does the best job of simulating an IADS of
> any current sim. However, the IADS is somewhat degraded by the nature
> of the Longbow itself...i.e., to move with stealth in low numbers, evade
> the IADS, and attack only those elements that are either part of your
> mission or necessary for survival. In effect, the NOE flight
> characteristics of the Longbow reduces what could be many-on-many
> engagements to few-on-few engagements...meaning that the IADS model
> could be greatly simplified (and still be realistic) by dealing with
> only a few objects at once. The next question is: is the Longbow IADS
> model robust enough to work with a jet sim, with many emitters and SAMs
> of many types engaging many aircraft of multiple types?
The basic system is capable, but needs expansion to handle the larger
coverage area and enumeration. The basic principles still apply, though.
As for the variety of emitters, that's easy. Just have different radar
cross-sections for each (table driven) and keep track of radar on/off, type
and direction. Variety in search/track radars (attached to SAMs) is easy
too. You can table drive most of the important differences (range,
frequency, susceptibility to various jamming, radar type...), and more
importantly, you can table drive the crew reactions and search/track/launch
timing parameters. The only hard part is finding enough cpu cycles to
handle more of everything. That's when you get into "horizons of activity
resolution". In other words, stuff that you interact with more (i.e. near
you) gets more resolution in simulation, while stuff that's out of your
zone of influence is handled in a more coarse fashion.
[snip]
> multi-platform strike. You’ve got 18’s with HARMs, 15E’s, CAP, jamming,
> air launched decoys...the whole works. Can you really simulate the
> effects of real-time confusion and decision making going on at the
> EW/GCI radar? At the C2 post? At the SAM?
You don't have to have to simulate all the actual activity, just get the
effect of the activity correct. In other words, how will the pilot know
this is happening? Simulate the effects. This is the "smoke and mirrors"
part of doing sims. If it was just straightforward engineering, everyone
could do a challenging, fun, and realistic sim.
[snip]
> Oh? And the name of that sim would be...(I’m not sure, but I might be
> getting downright giddy!)
"Big Bird"
[snip]
> I agree...it is just a game, and the desires of the 2% may not be worthy
> of service. That’s the way the market works and I accept that. But it
That's what reality level switches are for. Simple mode for the "fun"
crowd, progressively harder settings for each item so people can get all
the realism they can handle or want.
--Andy
Ridder
Andy Hollis wrote:
>
> Ridder <jri...@erols.com> wrote in article <33569E...@erols.com>...
>
> [snip]
>
> > Oh? And the name of that sim would be...(I’m not sure, but I might be
> > getting downright giddy!)
>
> "Big Bird"
Hmmm...could be a simulation of the domestic SA-10's surveillance radar
<g>. On the other hand, an F-15E is a pretty big bird...
JR
>
> [snip]
>
>
> --Andy
Denny Atkin
Ridder <jri...@erols.com> wrote:
>Andy Hollis wrote:
>> Ridder <jri...@erols.com> wrote in article <33569E...@erols.com>...
>> > Oh? And the name of that sim would be...(I’m not sure, but I might be
>> > getting downright giddy!)
>>
>> "Big Bird"
>
>Hmmm...could be a simulation of the domestic SA-10's surveillance radar
><g>. On the other hand, an F-15E is a pretty big bird...
Okay, I can't handle the pressure of the NDA anymore. I have to spill
the beans! It'll be on our web site next week anyway, so no harm done.
Sure, the F-15E is a pretty big bird. But it's not THAT big. Since when
has Andy Hollis ever thought small?
The "fixed wing sim", due out in two weeks, is Jane's C-5A Galaxy. It's
one of the most innovative sims I've ever seen. Not only does it give
you a chance to fly one of the world's largest aircraft, but there's a
fascinating strategic component as well. You have to properly route your
cargo to the war zone, and deliver it safely and on time, or the dynamic
campaign falls quickly to pot. It's an amazing balance of resource
management and sim action.
Of course, there IS combat. You can choose escorts for your C-5A, and by
putting the plane on autopilot, you can choose the external view and
watch F-15s, F-16s, F/A-18Es, and F-22s jump to your defense. But don't
get too distracted, because you'll often get the opportunity to defend
yourself by releasing chaff and flares.
The graphics are just beautiful. Jane's C-5A has native support for the
S3 Virge and Matrox Mystique 3D cards, and Jane's said they might
support others if they perceive a demand for it.
I know this may not seem as exciting as a jet combat sim, but wait until
you just manage to skirt into a landing at a remote field, maneuver
through tight confines, and then move to the loadmaster position to
discharge your tanks and get back in the air before nightfall. We're
talking major adrenaline rush here!
I managed to actually roll the thing flying the beta last night--pretty
amazing, and pretty scary considering how long it takes. And in the
debrief, my mechanic claimed the maneuver had cracked the wing spar! Now
that's realism.
----------------------------------------------------
Denny Atkin / den...@aol.com / den...@mindspring.com
My new troubleshooting book: "CGW's Why Won't This $@#& Game Work?"
http://www.amazon.com/exec/obidos/ISBN=1566865808
Aaron S.
Denny Atkin wrote:
>
> Ridder <jri...@erols.com> wrote:
> >Andy Hollis wrote:
> >> Ridder <jri...@erols.com> wrote in article <33569E...@erols.com>...
> >> > getting downright giddy!)
> >>
<reads above>
<looks at Calender>
(It's not April 1)
<laughs anyways>
CJ Martin
>Okay, I can't handle the pressure of the NDA anymore. I have to spill
>the beans! It'll be on our web site next week anyway, so no harm done.
>Sure, the F-15E is a pretty big bird. But it's not THAT big. Since when
>has Andy Hollis ever thought small?
>
>The "fixed wing sim", due out in two weeks, is Jane's C-5A Galaxy. It's
>one of the most innovative sims I've ever seen. Not only does it give
>you a chance to fly one of the world's largest aircraft, but there's a
>fascinating strategic component as well. You have to properly route your
>cargo to the war zone, and deliver it safely and on time, or the dynamic
>campaign falls quickly to pot. It's an amazing balance of resource
>management and sim action.
[snip]
Andy, I *swear* I did not make the coffee during Denny's visit...
<g>
CJ
(sorry for the sorta inside joke, folks...some of you will get it though)
enzo
> > The "fixed wing sim", due out in two weeks, is Jane's C-5A Galaxy. It's
> > one of the most innovative sims I've ever seen.
> > Of course, there IS combat. You can choose escorts for your C-5A, and
by
> > putting the plane on autopilot, you can choose the external view and
> > watch F-15s, F-16s, F/A-18Es, and F-22s jump to your defense. But don't
> > get too distracted, because you'll often get the opportunity to defend
> > yourself by releasing chaff and flares.
Cargo sim battles the way they were always meant to be played. Excellent!
> > I know this may not seem as exciting as a jet combat sim, but wait
until
> > you just manage to skirt into a landing at a remote field, maneuver
> > through tight confines, and then move to the loadmaster position to
> > discharge your tanks and get back in the air before nightfall. We're
> > talking major adrenaline rush here!
YES!!! The only that could make it better is if you get to take the role
of loadmaster and supervise dropping Sherridan tanks, HMMMV's,
FAE bombs, and Daisy Cutters!!
> > I managed to actually roll the thing flying the beta last night--pretty
> > amazing, and pretty scary considering how long it takes. And in the
> > debrief, my mechanic claimed the maneuver had cracked the wing spar!
Now
> > that's realism.
What would be even better if they model the sound of the spars groaning and
howling under stress echoing in the cargo hold. Brrrrr!! Talk about eeeerie!
pap...@ix.netcom.com
Now Im really really " DISAPPOINTED" (kevin klien Fish called wanda
voice) I was hoping the rumors I heard about an LA freeway Sniper SIM
were true....
Oh Well...
PAPA DOC
>I know this may not seem as exciting as a jet combat sim, but wait until
>you just manage to skirt into a landing at a remote field, maneuver
>through tight confines, and then move to the loadmaster position to
>discharge your tanks and get back in the air before nightfall. We're
>talking major adrenaline rush here!
>I managed to actually roll the thing flying the beta last night--pretty
>amazing, and pretty scary considering how long it takes. And in the
>debrief, my mechanic claimed the maneuver had cracked the wing spar! Now
>that's realism.
> ----------------------------------------------------
Corsair
>The "fixed wing sim", due out in two weeks, is Jane's C-5A Galaxy. It's
>one of the most innovative sims I've ever seen. Not only does it give
>you a chance to fly one of the world's largest aircraft, but there's a
>fascinating strategic component as well. You have to properly route your
>cargo to the war zone, and deliver it safely and on time, or the dynamic
>campaign falls quickly to pot. It's an amazing balance of resource
>management and sim action.
LOL! "Loadmaster position" <snort>.
Ridder
Sorry to beat a dead horse here, but I noticed today that the current
issue of Aviation Week has an excellent article that illustrates exactly
what the 2 points below talk about. The reference is: Aviation Week &
Space Technology, April 14, 1997, P. 55, Article titled "Missile Warning
Fix May be Needed". This article discusses how current RWRs on many
fighters may not even detect many modern threats. It makes the point
for Uncertainties and Imperfections far better than I could.
JR
Ridder wrote:
>
>[snip]
>
> 7) Uncertainties. This is the big one. What if nothing was perfect?
> I’m not talking about systems failures like in Flanker. I’m talking
> about imperfections in systems. How about a realistic RWR that provides
> ambiguous, or maybe even erroneous ID of a threat? How about
> inaccuracies in reporting position? How about missed intercepts? How
> about IFF imperfections (threat of blue-on-blue)? How about botched
> intel? How about aged intel (you wouldn’t know exactly the number,
> types, or locations of threats)? Give the issue of uncertainty some
> thought and you’ll begin to see how this one can drastically alter
> gameplay (just as it complicates reality).
>
> 8) Improved avionics modeling. Don’t get me wrong, I love the avionics
> modeling in many of the sims. Longbow’s multi-page MFDs, ORT, etc. are
> great. However, I’d like to see imperfect systems. Similar to the
> uncertainties I mentioned above, what if the avionics modeling actually
> accounted for inadequacies of some systems by using realistic frequency
> ranges, sampling rates, receiver sensitivies, azimuth and elevation
> coverage? I know they’d be a guess, but ANY guess is better than
> assuming perfection.
>
>[snip]
FastPants
"enzo" <en...@outwest.com> wrote:
>
>> > I managed to actually roll the thing flying the beta last night--pretty
>> > amazing, and pretty scary considering how long it takes. And in the
>> > debrief, my mechanic claimed the maneuver had cracked the wing spar!
>Now
>> > that's realism.
>
>howling under stress echoing in the cargo hold. Brrrrr!! Talk about eeeerie!
A few years ago I had a go on the DC-10 sim at Gatwick (UK) shared by
American Airlines/British Caledonian(b4 they were taken over by BA) -
I saw someone barrel roll it that day, supposedly within
limits..........you'd be amazed by what a heavy transport a/c can
do.....and what some airline pilots think about doing............
That sim was a-mazing......you could hear and feel the 'thump' made
by the tyres going over expansion gaps in the virtual 'concrete' slabs
of the apron whilst taxying ! Heavy work too.......*lots* of
inertia....great with auto-throttles on the ILS, but without them I
ended up overshooting the approach at Hong Kong - you've heard of
Bird strikes..?.......lets just say it was a Fish-strike...*sploosh*!!
FastPants
Corsair
>Andy, I *swear* I did not make the coffee during Denny's visit...
<snort> I asked myself this question when I *finally* stopped laughing after
reading Denny's post too :)
Corsair
I just came to a realization of one of the things I'm looking for when it
comes to realism. I'd really like to be able to take a book about...
oh I don't know... let's say an F-15E, open it to a page that shows the
cockpit, and be able to name each control and what it does. I've always been
big on avionics.
Probably the only way that you could attain that level of realism is by a
clickable cockpit which I think has been long overdue as a mainstream sim
feature. I'm happy that iF-22 will be incorporating it into theirs.