Jousting in the Dark with Flashlights

As I’ve related before in these pages, I am a casual flight-simulation enthusiast.  In my youth I was a much more devoted aficionado (if not a particularly skilled one), and I spent many happy hours flying (mostly one-way) virtual sorties into scarily dense integrated air defense networks.  But these days, I lack the time and proficiency to survive in that kind of unwelcoming environment, and so I satisfy myself with the prosaic tasks of practicing touch-and-goes, failing at in-flight refueling, and occasionally sparring with computer-controlled enemies at a skill level just low enough to make me the conquering hero.  Another oak leaf cluster for my Distinguished Flying Cross?  Why thank you, don’t mind if I do, you can paint the MiG silhouette right over there, that’s right Iceman, I am dangerous.  (Hey, don’t judge me.)

There I was in the blue F-16, trying to jump a partially-anesthetized, unarmed A-4E in a level turn...and, er, I wind up botching the end game with way too much closure. What looks like a tidy little displacement roll on the tape was really a frenzied attempt to avoid a mid-air collision.

However, it’s interesting after all these years to see how current technology is employed by serious simulator fans who have stuck with the hobby.  There are several different layers of simulation complexity.  The first is mastering the control and management of your own aircraft, which is a nontrivial exercise in the era of 700+ page “game manuals.”  (I am stuck at this level.)  The second is basic combat against a computer-controlled aircraft, the so-called “1-v-1” engagement.  This raises the complexity substantially, as weapons systems and tactics and all of the nastiness of an opponent come into play.  The third level is multiplying the number of aircraft in play, which adds the element of multi-tasking under stress.  And the fourth adds human players into the mix, which increases the chaos by an order of magnitude.

The “fourth-level” organized scenarios that are flown by serious devotees of the hobby are fascinating.  While there is plenty to read in the open literature about Basic Fighter Maneuvers, there is not a lot about how a large air battle involving fourth-generation fighters equipped with missiles of the AMRAAM generation would play out tactically (the seminal open-source work on the subject, Robert Shaw’s Fighter Combat: Tactics and Maneuvering, was published in 1985).  These are the intriguing parts of the mission that Lex could not show us on his helmet-cam — and for good reason, for there lie secrets upon which lives may one day depend.

But within the boundaries of what is publicly known, the experiences of the hardcore simulator crowd provide some fuel for thought.  Not so much as a predictive device — but rather, to get the “feel” of the thing, and in particular how tomorrow’s fighter pilot (or UCAV operator) will need to quickly synthesize all kinds of fragmentary information in a very short time to detect, identify, engage, kill, and withdraw.  If these games are any indication, it will be confusing, fast, violent, and curiously cerebral.

The Youtube video above is a recording of an air-to-air encounter during a Falcon 4.0 airfield strike, flown by a four-ship of F-16s over a simulated Korean peninsula.  (Or, more properly, flown by four civilian hobbyists over the Internet, at least one of whom has melodramatic tastes in background music.)  In this setup, they have no airborne radar controller with a God’s-eye view of the battlespace warning them of incoming threats.  Instead, they have to rely on their own radar warning equipment, their awareness of each other’s position, and the onboard data networking equipment that allows them to construct a common picture of what’s going on.  But they have to work to build that situational awareness, based on little half-second blips and buzzes that pulse on their threat receivers, and there is no time to spare, since modern missiles have the speed and range to kill virtually anything that can be detected, very quickly.  Compared to the WWII experience, it seems strangely abstract.

This second video is taken from a mission analysis tape (mimicking the military’s TACTS/ACMI systems, most current flight simulators have some kind of “flight recorder” capability that allows the mission to be carefully dissected afterward — this is an attractive little presentation package called TacView).  Here we see some of the consequences of chaos, when an F-15 takes a missile shot — and while his intended target exits the missile envelope, a friendly F-15 wanders into it.  With modern fighter aircraft being as agile as they are, and modern missiles having the range, kinematics, and “semi-smart” acquisition mechanics that they do, and everything happening fast fast fast…stuff can happen.

I don’t assert that these commercial entertainment products will predict the outcome of future air battles.  (Even assuming that flight simulations get the performance details within the ballpark, I think they still have a platform bias that undermodels operational-strategic capabilities which can change the battlefield fundamentals.)  But I do think that they do a more creditable job at capturing the flavor of that battlefield than a lot of other media, which rely too heavily on accounts of past air wars that are receding in relevance.

And if nothing else, these mission accounts are very entertaining for an old computer game player to watch.  They can be my wingmen anytime, no I can be theirs.  Er.  Well, something like that, anyway.


Filed under Airplanes, Flight simulation, Wargaming

3 responses to “Jousting in the Dark with Flashlights

  1. First order of business: more elbow, less fingers and/or watch.

    (I know you got that.)

    I don’t purport to be an expert, or even a novice in the field of radar and electronic warfare, but the glaring holes found in the modelling of these systems in commercial products are where the real holes lay with respect to their ability to help the layperson understand long range tactics. And it really doesn’t need to be this way- while 99.9% accuracy is in the realm of OPSEC and a potential prison sentence, 85-90% performance model is well within the ability of anyone with a calculator and knowledge of a system’s power output and operating band.

    Unfortunately, right now not one developer even begins to model anywhere near that close, and all of them admit it right up front under the pretext of “classified” and “accuracy”. So instead, what you get is BVR-Quake, where in some simulations there are only three denial techniques: jammer, chaff, and the notch. Otherwise, if you or your target’s hit box falls within maximum programmed range of the attempting system and its current azimuth/elevation scan, you (or they) are a shown radar return.

    This degrades practicable lessons, such as offensive RWR use. Further, without any sort of radar power range and air density falloff, you get systems that are detecting opponents entirely too far out, causing grossly altered tactical shifts- if your system is detecting me ten miles before it really should, and it’s known that yours is also set to mirror mine in pure range for balance purposes, you’ve got a massive benefit that shouldn’t be there. That’s 40 or so seconds worth of tempo I’ve lost to you and your system- I need to start my notch or denial tactic that much sooner. And if you’ve got the longer ranged BVR weapon to boot (that also doesn’t degrade at range like it should), then why am I even bothering to leave the field? Hell, why are you even bothering with the lengthy startup procedure yourself? Neither of us can learn anything by the exercise, which denigrates the term “study sim”.

    At least if you’re 85-90th percentile in the detection model, then we can sweep the rest under the rug and say the following:

    “Our model’s radar sees the battlefield generally as a perfect/less than perfect day, even with clouds and precipitation- those things are modeled, but there are certain aspects of the specific systems that we don’t know which generate differences in performance. So either every system is performing perfectly, right off the factory line, or are degraded due to wear and tear on the components. Everyone receives the same percentage of bonus or degradation, and its within stated guidelines for all around performance of all the systems on an individual basis.”

    I’d take a 90% performance, 90% detection, 90% systems model simulation every day of the week over any 99.99% performance, 0% detection, 99.99% system version. When someone finally says “to hell with this” and starts looking at the radar/detection game in a realistic fashion and comes up with the bright idea to offload that code onto the GPU (which is where it should be, since that’s the processor which is calculating visual aspect- which is where the best model for true RCS can be derived since radar “sees” in a different wavelength band than the human eye), there will be the consumer end paradigm shift.

    Hell, depending on how government systems do it, that might be a paradigm shift for them, too…


    • RJL

      I agree that these simulations tend to be systems-heavy (part of their design heritage, I would guess), and the physical environment is not as well-modeled as, say, the sound/water environment in a submarine simulation. But any simulator (including the government ones) is a series of compromises, and for a lot of people, modeling an eight-minute startup sequence with all of the switches and knobs is more important than getting the EM spectrum model (and its associated tactics) to the 90% mark. Horses for courses, and all that.

      But given your interests, I’d suggest looking into the Falcon 4 series again. The design team is self-conscious and transparent about the choices they make, and I think their EW approach is more sophisticated than many other offerings. (See p. 289 of the RP5 manual, which strikes me as thoughtful.) It’s also strictly a hobbyist crowd at this point, so commercial pressures in the direction of simplification are absent. They also jumped headfirst into NCTR modeling, so they’re not shy about probing the boundaries of the open literature.

  2. cg23sailor

    I would also suggest looking into DCS (Digital Combat Systems) A-10C and the DCS Blackshark simulators.

    As far as realism and modeling goes. I find the same problems even with visual detections in naval simulations such as the excellent Silent Hunter series. Visually sighting is soooo much more than can light physically travel from the contact to the eye of the lookout. SO MUCH MORE!

    In most Sub games, to include Silent Hunter, if you get within a specified distance of a contact then the contact automatically detects you. this specified distance is variable somewhat due to physical parameters such as day/night, haze, rain, and fog. But so much more is the mental training, physical stamina, and alertness or boredom of the targets lookouts.

    As it stand, real life tactics such as those perfected by Otto Kretchmer and used to devastating effect by the likes of Morton and O’Kane cannot be reproduced or used in the game itself. It will not allow for it.

    I am referring here to night surface attacks and charging right into the middle of a convoy, making the attacks, and escaping on the surface in the ensuing confusion. You will automatically be detected before you can ever get close enough.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s