All But War is Simulation

Thomas L. Clarke

Dennis K. McBride

Institute for Simulation and Training

University of Central Florida

3280 Progress Drive

Orlando, FL  32826

tclarke@ist.ucf.edu, (407)658-5030, FAX:  (407)658-5059

Douglas Reece

SAIC

12479 Research Parkway

Orlando, FL 328263248



 

ABSTRACT

This article discusses the resistance of practitioners to virtual training simulators. Parallels with the philosophy of sports and games are discussed using the history of flight training simulators and other military training simulators as examples..  Similar resistance has been exhibited in some fields of athletics but not in others.   It is argued that the major factor leading to training simulation acceptance is danger.  The article concludes with a discussion of projects at the Institute for Simulation and training to apply military training simulation technology to athletic training that are encountering resistance and for philosophical reasons may be doomed to failure

 

 

INTRODUCTION

            Within the military training simulation community there are three types of training simulations.  These are constructive, virtual and live.  Constructive simulations are essentially elaborate chess-like games that involve the manipulation of tokens representing forces in stylized environments.   These have little use in sports and will not be discussed further.  The issue addressed here is the tension between virtual training simulations, simulations that involve the use of computers or other advanced technology to present a simulation of reality for training purposes, versus live simulations, which are essentially just practice games.   Sports practitioners tend to resist the use of virtual training simulations and in what follows the philosophical reasons for this resistance will be discussed.

While some aspects of the concepts of simulation and virtual reality can be found in the ancient world, e.g. Plato's cave and the use of training devices for gladiators, true simulation and virtual reality are modern.  From Descartes' philosophical cogito to the contemporary motion picture, The Matrix, the possibility of simulation has cast doubt on sensory data while the core self has remained undoubtedly real.   Computer technology has advanced to the point where Descartes' skepticism is almost as factual as portrayed in Wachowskis' (1999) fictional The Matrix.  

Both the philosophical and the fictional connotations of virtual reality tend to be dark.  Descartes contemplated a malign creator who has designed flaws into human faculties.  The Matrix is a world in which computers have put humans into virtual slavery.   This dark side of simulation may account for some of the resistance to virtual training simulation often found within sports.   This connection is made particularly clear in the short story "Spectator Sport" by John D. MacDonald (1952).  In this story a time traveler finds a future in which virtual entertainment has been perfected.  All ambition has been lost except for earning enough money for total and irrevocable immersion into three-dimensional virtual reality.   The perfection of virtual technology has reduced everyone to ultimate couch potatoes. 

A more recent fictional warning of the dark side of virtual reality is Ender's Game (Card, 1991) in which a military cadet thinks he is training on a virtual simulation only to discover that the simulation is in fact controlling a distant battle fleet.  After successfully completing the training task he is horrified to discover that he has in fact just destroyed an entire alien civilization. 

While fear of the dark side of simulation may be a factor in resistance to virtual training simulators for sports, it is probably not the primary factor.  Virtual games such as Doom and Tomb Raider as well as fantasy football leagues are very popular despite the darkness associated with virtual reality; however, the darkness may account for the dark visual tone of many of these games.  

The more likely reason for rejection of virtual training is belief in the inferiority of virtual training to real training.  Similar to this is the fact that sports and games are played for pleasure, so that practicing virtually would give up pleasure that could be obtained in real practice.   A corollary of this is that participants in dangerous sports should have a greater tendency to accept virtual training since practicing a dangerous activity is likely to cut short one's pleasure of participation in a learning accident.

The snobbish attitude toward virtual training is perhaps best captured by Hemingway's "There are only three sports. Bull fighting, mountain climbing and motor racing. All the rest are games." (1954)   These real sports could merit virtual simulation for training for the sake of the sport, but real men don't play games.   A man might frivolously play a game but he would certainly never participate in a virtual simulation of a game for training.  The proliferation of professional sports belies Hemingway's extreme attitude, but a vestige remains in the resistance of many sports practitioners to virtual training. 

Bernard Suits (1978) has taken Wittgenstein's aphorism 66:

Consider for example the proceedings that we call "games". I

 mean board-games, card-games, ball-games, Olympic games, and

 so on. What is common to them all? -- Don't say: "There must be

 something common, or they would not be called 'games' "-but look

 and see whether there is anything common to all. -- For if you look

 at them you will not see something that is common to all, but

 similarities, relationships, and a whole series of them at that. To

repeat: don't think, but look! --  (1968)

And developed the idea that essentially all human activity can be regarded as a game, that is

to engage in activity directed towards bringing about a specific

state of affairs, using only means permitted by rules, where the rules

prohibit more efficient in favor of less efficient means, and where such

rules are accepted just because they make possible such activity.

 

To play a game (indeed much of life) is to accept limitative rules for the sake of the activity.  Following the rules, which dictate inefficiency, is to a large extent what gives pleasure in the game.  To participate in virtual training, that is non-game training, is thus to attempt to become more efficient and thus to detract from the pleasure of the game.

            Somewhat counter-intuitively this attitude persists even in professional sports.  Sartre provides a rationale for this in Being and Nothingness where he argues that the aim of play and sport is to get out of the world, whereas serious and material activities aim to get in the world.  As an entertainment, a professional sport needs to aim out of the world and thus must maintain a core of play.  Were professional athletes to strive for ultimate efficiency they would be leaving the game and thus destroying the entertainment value they create.  Hence philosophically it makes sense for professional athletes to resist efficient training such as virtual simulation.

            In the next sections some parallels will be drawn with the history of flight and military training simulation.

 

ACCEPTANCE OF FLIGHT SIMULATION

Edwin Link invented the first modern flight-training simulator in 1927.  Link intended his simulator, the pilot-maker, to be a trainer, but aviators considered Link's simulator a toy and initial sales were largely to amusement arcades.  Science and Invention magazine noted "such devices would make a valuable adjunct to the multitude of miniature gold courses that now dot the country."  (Killgore, 1989)

In 1930 Link formed a flight training school based upon the pilot-maker and struggled to stay in business during the Depression.   Then in 1934 ten Army airmail pilots were lost within a few weeks flying under instrument conditions.  A group of forward-looking Army officers saw the answer in Link's pilot-maker.  The Army order for six simulators provided necessary airmail pilot training and served to launch the modern flight-training industry.  By WWII German and Japanese as well as American and British pilots had received training on Link simulators.

But even today pilots resist simulator training, claiming it is not as good as real flying.  As discussed in the last section this attitude has partly a basis in fear of the dark side of simulation and partly in snobbery.  While the concept of real would provide a philosophical basis for distinguishing real from virtual and which is superior, it is doubtful if pilots make  philosophical distinctions deeper than discussed above.  Some Top-Gun attitudes look down on training in simulators as not macho, and many pilots, like JFK, Jr., think raw skill and intelligence can overcome the lack of formal training.   Research shows that good training can be had with simple training devices such as Link's wooden WWII era Blue-Box with its pneumatic controls.  Nevertheless military aviators insist on the latest technology so that their simulators become very expensive virtual airplanes. While grudgingly accepting simulation training, pilots demand the best simulator that money can buy.   

As a side effect this has fueled the development of the 3D-computer graphics hardware needed to provide ever more realistic renditions of aerial scenery.   Today this hardware advancement has made the desktop computer a viable training device.  Ensign Herb Lacy, a student aviator at the Naval Air Station in Corpus Christi Texas, Enhanced Microsoft Flight Simulator 98 with add-ons from the Internet so the program had the look and feel of a T-34 Mentor training aircraft.  He added Corpus Christi landmarks and visual references.  After using this standard simulator program on his PC, Lacy "pegged the top of the grading scale" during intermediate flight training.  As a result today even the most conservative aviators are forced to admit the utility of flight simulation training. (Peterson, 1999)

Another factor in the resistance to simulator training has been the phenomenon of simulator sickness.  (Kennedy et al, 1992.)  Even the most experienced pilot can become nauseous in a simulator due to the time delays in the simulation computation.  If the visual scene presented by the computer is delayed by more than 30 milliseconds from the other haptic and kinesthetic cues of the simulation, the disparity quickly leads to nausea.  The increasing speed of computers is reducing this problem, but demands for maximum fidelity simulation always push the capabilities of available computers.  Designers of rides for theme parks have to deal with simulator sickness as well.

Military training by simulation has come a long way from the sword-fighting practice dummies adopted by the Roman army in 105 BC after gladiator training had proved their value.   The US Army is just completing procurement of the C2T2 (Close Combat Tactical Trainer) training system which enables teams of soldiers to drive M1 Abrams tank simulators, M2 Bradley fighting vehicle simulators and other simulated vehicles in a virtual simulated battle.  These simulators are networked using the DIS (Distributed Interactive Simulation) protocol and its derivative HLA (High Level Architecture) so that all participants can see and interact with all other participants. (Clarke, 1995)

Soldiers, like aviators, would rather drive real noisy, smelly tanks around for training than use a mere video game-like simulator.  However, unlike aviators who fight in single combat, tank battles can involve hundreds of vehicles, so that similar numbers of simulators must be networked with DIS for training.  The Army thus cannot afford the state-of-the-art fidelity that the Air Force lavishes on its few multi-million dollar flight-training simulators.  It is thus unlikely that the Army would have ever adopted a system like C2T2 had there not been a push from the top.

A far-sighted officer, Jack Thorpe, at the Defense Advanced Research Projects Agency (DARPA) saw the training value in lots of 70% realistic simulators networked together.  If only medium fidelity is sought, large numbers of simulators become affordable. These first tank simulators were a step or two behind the art a decade ago, but such is the pace of technology that today's Sony Playstation 2 or Microsoft's X-Box has more computing power than these early simulators .  In a pilot project at DARPA, Thorpe built a training facility at Ft. Knox, the heart of the US Army tank command, with 75 of his 70% simulators networked together.  With such close exposure to a real system, the Army was won over and is making C2T2 a standard training system.  No doubt the winning of the Canada Cup, a prestigious NATO tank maneuver competition, by a team which trained with networked simulators helped make the case for C2T2.

The title of this paper "All but war is simulation" is also the motto of the US Army Simulation Training and Instrumentation Command (STRICOM) which designed C2T2 after DARPA has proved its utility.  This motto cuts two ways, however.  While it points to the fact that everything done in preparation for the reality of war (or sport) is in some sense simulation, it also carries the sense that simulation is less than reality with the implication that only non-lethal war is truly training. 

War, or reality, is sometimes inherently dangerous and cannot easily be made non-lethal to the novice.  The Roman gladiators whose training techniques the Roman army adopted, had to know how to fight upon entering the arena the first time or else die.  The airmail pilots in 1934 (and JFK Jr. in 1999) found out that instrument flying is often lethal to the novice.  In situations like these the practitioner naturally adopts training through simulation.  In other cases simulation-based training may have benefits but tends to be resisted unless imposed from above as in the case of the DIS tank simulators.  The same situation seems to apply with sport simulators as will be discussed in the next section.

RESISTANCE TO ATHLETIC TRAINING

In 1984 here at the University of Central Florida (UCF) Dr. Wayne Burroughs developed a simulation-based trainer for baseball batters and proved its positive training value. (Burroughs, 1984) He evaluated the effectiveness of visual-simulation-training film approaches to enhancing the visual pitch recognition and pitch location skills of collegiate baseball batters. Visual-simulation training films were produced by filming collegiate baseball pitchers throwing fastballs and curves using a camera located in the right-handed batter’s box.  These films were edited into a series of learning trials in which subjects were required to identify the type of pitch and its location in the hitting area.  Experiment groups (with controls) were given 1 hour of film training or combinations of slow-motion and/or real-time film training before or after the trials. Group comparisons indicated significant gains in location scores but nonsignificant changes in pitch recognition scores; Dr. Burroughs thinks the lack of change in recognition is probably due to initially high pretest levels. These gains occurred with use of slow-motion or real-time films, and most of the gains were maintained in a 6-week follow-up post-test.

Despite these positive results, Dr. Burroughs was unable to convince the major leagues to adopt his training device.  Part of this is due to not-the-real-thing factor; it's not real training if you can't smell the glove leather.  We argue that much is due to the non-lethal nature of batting.  Aviators and gladiators were forced to adopt simulated training or die, whereas a ball player will just have a lower batting average if he rejects training simulation.

A survey of the Internet and bibliographic indices finds little active use of training simulation by serious athletes.  Recreational sports such as golf have developed training simulators that have good commercial success, but professional and serious amateur level sports do not seem to use simulation as much is possible.  There are exceptions and these support the thesis that danger is a factor.   Rough water sports such as kayaking do make use of simulators.  This sport puts an individual into a situation where drowning is a real possibility so that pre-training the novice in a simulator is a real danger reduction factor. 

The case of Jacques Villeneuve, the Canadian Formula One driver who used a racing simulation game to practice for and win the Belgian Grand Prix is also instructive. (Chidley, 1997)  Auto racing is of course one of Hemingway's dangerous sports.  The same article also reports how golfers use simulations to pre-familiarize with the Banff Springs golf course.  This remote Canadian Rocky Mountain course is only open briefly during the year so that practice, while not dangerous, is difficult.

A possible exception is the video-based quarterback training system reported in Ergonomics in Design. (Walker and Fiske, 1995)  This system, developed at Georgia Tech, was reported to be received enthusiastically by the Georgia Tech football team.  Unlike a pure computer-generated simulation it requires a team to role play all of the desired offense-defense combinations and record them with a video camera located some distance behind the quarterback.  This has the drawback that the plays are not programmable;  in order to try a new offense or defense or combination, it would be necessary to get players to act out the play and make a new video disk or CD-ROM.  This would be impractical on a week-by-week basis.  Also, it would not be possible to vary the speed, behavior or other characteristics of the players.  Nor is it possible for the quarterback trainee to move around (to drop back, move around linemen, etc.) on the virtual football field.

When we first heard of this system we thought we might have been scooped as we are working toward developing a purely computer generated simulation for training football quarterbacks.  The next section discusses our conception of a quarterback training simulator.  This training simulator is one of many that could be used to illustrate how virtual technologies can be used for sport training.   It is philosophically interesting because the proposed training modality is alike yet different from the actually activity of quarterbacking.  It thus affords a means to analyze and identify likely sources of resistance to simulation-based sports training.

SPORTS SIMULATORS AT UCF

IST (Institute for Simulation and Training) has worked primarily in the area of military training simulators.  Recently we (Doug Reece is now at SAIC) have realized that many of the training technologies developed to hone soldiers cognitive and perceptual abilities could be applied to sport.  What in military parlance is called “situational awareness” becomes in football the task of “reading the defense” or "reading the offense."

So far we have focused on the quarterback-training problem, as it is definitely a cognitive/perceptual task.  Lots of people can throw the ball, but only a few can see where to throw it.  Computer graphics have reached the point where affordable hardware can generate images of players on the field with sufficient fidelity to produce positive training transfer.  We envision the QB viewing a projection display of simulated play action to hone his defense reading skills.  Questions remain about the best interface, for example, would pointing with a 3D mouse result in good training or is a more sophisticated instrumented football interface needed to avoid negative training?

One of the sports activities for which simulation training has a high potential payoff is quarterbacking.  Football players at the college and professional level have already developed excellent physical skills through years of practice.  However, players often have to learn new offensive systems and plays, and have to learn their opponent’s formations and plays every week. The defenses used by each opponent may be significantly different.  Quarterbacks in particular are responsible for recognizing the defense and directing plays in response to the defense’s general formation, individual player locations, play.

The quarterback’s defense reading task is as follows: before the snap, 1) recognize the defensive play by watching how the defensive players line up as the offensive players take their positions (or go in motion), and 2) change the play if necessary.  After the snap, A) look at the appropriate defensive player and recognize the defensive play by his initial movements; B) based on (A), watch the correct receiver; C) based on defensive play, throw to the correct receiver, or look for the appropriate secondary or tertiary receiver.  The quarterback has about 3 seconds to make a decision; during this time, the quarterback drops back several yards, and onrushing linemen threaten him, obscure his vision, and may force him to move around.

In order to give the offensive squad practice in playing against a new defense, the coaching staff of a football team watches tapes of the opponent and teaches the third-string defensive squad to play as the opponent does.  While a simulator cannot train all aspects of a quarterback’s job, it does have some significant potential advantages over training against the third-string defensive squad.  In particular, it would not require the defensive squad.  Its advantages would include allowing training in bad weather and allowing the quarterback to get more extensive training without requiring the time of the rest of the team. In addition it would give reserve quarterbacks more opportunities to become familiar with the opponents defense, and it could simulate the speed and other capabilities of opponent players that the third-string defensive players could not duplicate

Our prototype VR training system for quarterbacks would operate as follows:

Offensive and defensive formations and plays would be designed with a simplified user interface before a training session and then selected by a coach before each play.  Then the quarterback would view a display showing the football field and the teams.  He would be immersed in the virtual environment as a player.  At play start he would see the players moving to their positions, and could also move behind the center.  The quarterback could call signals and move back in the pocket when the play started.  When he selected a receiver, the quarterback would signal that he was passing the ball and the play would end.  Afterwards, the play could be reviewed from the quarterback’s perspective or from a stadium-top perspective.

The experimental VR training system for quarterbacks can be constructed mostly from existing hardware and software.  IST has considerable experience in relevant areas.  A football field model would be constructed and then displayed on a computer workstation.  The workstation would also drive a head-mounted display (HMD), which the quarterback would wear.  A computer generated forces (CGF) program would control the football players in the virtual world.  One of the key technologies needed for the success of the trainer is virtual football players.  While IST has been developing a CGF program as part of interactive simulation projects for several years, this program is not capable of modeling all of the necessary motions of football players and cannot model the physical interactions (for example collisions in various poses, pushing and tackling) in football.  The CGF program cannot generate football player behavior.  Finally, there is no method for describing a football play in terms of player motions and behaviors.  These capabilities push the state of the art not only in IST’s software but also in intelligent agent design in general.  The goal of this project is to research and develop these capabilities. 

CONCLUSION

This resistance to simulation training when there is a viable non-simulation alternative, even if the non-simulation approach is inferior or more costly, does not bode well for sports simulation-training programs like the nascent one at UCF.  However, some hope is offered by the experience of the military.  Whereas just the act of instrument flying can be lethal, driving tanks and running through the woods for war gaming is not inherently dangerous.  Nevertheless, the US and other armies are adopting computer-simulated training as cost cutting measures despite resistance by the troops.  It is almost trivially easy to computer generate images with sufficient fidelity for training; individuals have adapted game software to training purposes.  The barrier to good sports training simulation is software development time not hardware cost.  IST intends to develop suitable software and the UCF coaching staff will help determine the utility of sport training simulation. 

Forward-looking coaches like Brian Billick of the Baltimore Ravens are beginning to see the usefulness of simulation for developing strategy.  Studies (Lohr and Scogin, 1998, and Farrow et al, 1998) are appearing that demonstrate the value of simulation-type training in sports.   In the near future the true value of simulation for sports training as well should be realized.  The skills of high-level athletes cluster tightly and simulation-based training can provide the small, but crucial winning edge.  But in view of the introductory discussion, there are strong philosophical reasons for resisting the injection of virtual training into sports.  Sports must retain an element of play or gamefulness if they are to retain what makes them attractive.  Reizler (1941) makes this clear in analyzing what distinguishes a game from serious life.  Briefly, something is serious if it contains "a cause you would fight for".  Despite the fights that sometimes erupt at World Cup matches, it is doubtful that any sport truly contains a cause worth fighting for.  Thus it seems, in a twist on the title of this paper and the STRICOM motto, only war or other life risking activity is worth virtually training for.

Thus the final conclusion is less optimistic than that expressed above.  There seem to be fundamental philosophical reasons for not using virtual simulation training in sport and efforts like the one the Institute for Simulation and Training has been conducting in promoting quarterback training simulation may be doomed to failure.

 

REFERENCES

Burroughs, Wayne A., 1984. "Visual simulation training of baseball batters."  International Journal of Sport Psychology; Vol 15(2), p 117-126

Chidley, Joe, 1997. "Wired for thrills: some sports simulations are startlingly realistic.  Maclean's; July,  p41.

Card, Orson Scott, 1991, Ender's game, Tor, New York.

Clarke, Thomas L., 1995.  Distributed Interactive Simulation Systems for Simulation and Training in the Aerospace Environment; SPIE Press.

Farrow, Damian, Paula Chivers, Carl Hardingham, and Shane Sachse, 1998. "The effect of video-based perceptual training on the tennis return of serve." International Journal of Sport Psychology; Vol 29(3), p 231-242.

Hemingway, Ernest, 1954, The Sun Also Rises, Scribner, New York.

Kennedy, Robert S, Norman E. Lane, Michael G. Lilienthal, Kevin S. Berbaum, and Lawrence J. Hettinger, 1992. "Profile Analysis of Simulator Sickness Symptoms:  Application to Virtual Environment Systems",  Presence; v1, p295. 

Killgore, J.I., 1989. "The Planes that Never Leave the Ground." American Heritage of Invention & Technology; pp. 56-63.

Lohr, Bethany A. and Forrest Scogin, 1998. "Effects of self-administered visuo-motor behavioral rehearsal on sport performance of collegiate athletes."  Journal of Sport Behavior; Vol 21(2), p 206-218.

MacDonald, John D., 1952, "Spectator Sport",  in The Omnibus of Science Fiction, edited by Groff Conklin, Crown, Chatham River, pp. 350-354.

 

Peterson, Eric L., 1999.  "He flies by the seat of his PC",  Government Computer News; May 3, v18, p40.

Sartre, Jean-Paul, 1964, Being and Nothingness, Translated by Hazel E. Barnes, Citadel, New York, Citadel.

Suits, Bernard, 1978, The Grasshopper: Games, Life and Utopia, University of Toronto Press, Toronto.

Riezler, Kurt, 1941, "Play and Seriousness", Journal of Philosophy, XXXVIII, pp505-517.

Wachowski, Andy  and Larry Wachowski, 1991, The Matrix, Warner Bros, Hollywood.

Walker, N. and A. Fisk, 1995. “Human Factors Goes to the Gridiron.”  In Ergonomics in Design; July.

Wittgenstein, Ludwig,  1968,  Philosophical Investigations, Translated by G. E. M. Anscombe,  Macmillan, New York.