computer gaming world 003

8/12/2019 Computer Gaming World 003

1/44


2/44


3/44

VOL. 2 NO. 2 MAR. - APR. 1982

FeaturesSOUTHERN COMMANDReview of SSI s Yom Kippur Wargameob ProctorSO YOU WANT TO WRITE A COMPUTER GAME0Advice on Game Designinghris CrawfordNAPOLEON'S CAMPAIGNS 1813 & 1815: SOME NOTES2Player s, Designer s, and Strategy Notesoie BillingsSO YOU WANT TO WIN A MILLION6Analysis of Hayden s Blackjack Masterichard McGrathESCAPE FROM WOLFENSTEIN8

Short Stay Based on the Popular Gamed CurtisODE TO JOY, PADDLE AND PORT:9Some Components for Game Playinguther ShawTHE CURRENT STATE OF1COMPUTER GAME DOCUMENTATIONteve Rasnic TernNORDEN+, ROBOT KILLER5Winner of CGW s Robotwar Tournamentichard FowellTIGERS IN THE SNOW: A REVIEW0Richard Charles KarrYOU TOO CAN BE AN ACE2Ground School for the A2-FS1 Flight Simulatorob ProctorBUG ATTACK: A REVIEW4

Cavalier s New Arcade Game Analyzedave JonesPINBALL MANIA5Review of David s Midnight Magictanley GreenlawDepartmentsFrom the EditorHobby Industry NewsInitial ommentsLettersThe Si l icon Cerebrum7Micro Reviews6Reader's Input Device0


4/44

From the Editor...Each issue of CGW brings the magazine closerto the format we plan to achieve. This issue (ourthird) is expanded to forty pages. Future issues

will be larger still.In this issue we add a new regular department,

THE SILICON CEREBRUM. THE SILICONCEREBRUM, written by Bruce Webster whommany of you will recognize from his writings inTHE SPACE GAMER magazine, will concentrateon computer programming as related to gamedesign and artificial intelligence.

You will also note in this issue that most gameswill have a small capsule description called "BasicInformation". This feature was added due to therequest of several readers. Your viewpoints areimportant to us. Speaking of your viewpoints, thelast page of this issue is very important. We wantto know what you think of the articles in CGW,what you think of certain ideas we have, and whatyou think of the computer games being covered inCGW. Please take time to respond to "Reader

Input Device". Not only will it help us to know justwhat you want to read, it will also let the man-ufacturers know just what you think about thegames on the market.Our congratulations go to Richard A. Fowell

whose robot, Norden+, won the Robotwar tourna-ment. Richard shares in this issue the story be-hind Norden+. Thanks Richard. Those readerswho have been with us since 1.1 know that we alsohave a tournament in progress called "The Great-est Baseball Team of All Time (GBTOAT). Readersare encouraged to send in the results of theirtournaments, which will be published in our May-June issue. If you do not have CGW 1.1 and 2.1, butare interested in the tournament send $1.00 forreprints of articles that list the eight teams andground rules of the tournament. You need SSI'sComputer Baseball or AH's Major League Base-ball (modification needed on MLBB) to play thetournament. No awards on this tournament; justthe opportunity to publish your replay results.

HOBBY ANDI N D U S T R Y N E W SAutomated Simulations has an-

nounced plans to offer its EPYXand MIND TOYS products for theIBM personal computer. The awardwinning TEMPLE OF APSHAIwill be the first release (March).Automated Simulation's DRA-GON'S EYE is now available forthe Atari 400 and 800 computers.Computerware (1512 EncinatasBlvd., Encinitas CA 92024) has re-leased two new games for the RadioShack Color Computer, MAZER-ACE (board game on computer),and STORM (arcade game).

Avalon Hill's Jack Dott tells usthat AH will be releasing 6 newarcade games in April for the Atariand Apple computers. Also, accord-ing to Dott, current projects at AHinclude STOCKS & BONDS andMIDWAY CAMPAIGN for the IBMpersonal computer, a naval gamecalled ALL HANDS ON DECK,RICHTHOFFEN'S WAR, andCLOSE ASSAULT (Squad Leadertype game). No firm dates are avail-able for most of these.

In MISSION ESCAPE's rule-book, CE Software offered a prize tothe first player to get through all tenlevels of that game. Steven Allen, aresident of Iowa (CE's home state)won the prize. Don Brown of CEstates that the offer has been ex-tended. Another prize of 5 CE Soft-

ware products will be offered to thefirst player outside Iowa to win thegame. To check on the current statusof the contest contact CE Softwareat 515-224-1995.Volumes 6 and 7 of the Sword-thrust series, entitled "The EternalCurse" and "The Hall of Alchemy"

will be out very soon according toCE Software.Synergistic Software announces

its new game APVENTURE TOATLANTIS. A to A, a sequel totheir popular ODYSSEY adventuregame, is an adventure game withembedded arcade action (Apple 48K,$40). Synergistic also plans twoMarch releases, NIGHTMARE GAL-LERY (arcade with supernaturalcreatures) and U-BOAT COMMAND(action/strategy).


5/44

I N I T I A LCOMMENTSIn addition to games mentionedelsewhere in this issue the followinggames have been received by CGW.Most but not all of these games willreceive more detailed attention infuture issues.

Cavalier ComputerP.O. Box 2032Del Mar, CA 92014THE ASTEROID FIELD: Cav-alier's version of the ASTEROIDSarcade game. An excellent trans-lation of the popular arcade gameto the Apple II computer ($24.95).STAR THIEF: Arcade game inwhich the player (or players)take the role of guards. In yourspaceship you guard against thetheft of the powerpods. Apple II48K, $29.95 .

Automated Simulations1988 Leghorn StreetMountain View, CA 94043TEMPLE OF APSHAI: Awardwinning hi-res fantasy adventuregame.UPPER REACHES OF APSHAI:

Four new levels and 150 newrooms to add on to TEMPLE.HELLFIRE WARRIOR: SequeltoTEMPLE. Four levels, sixtyrooms, new commands.THE KEYS OF ACHERON:An expansion dungeon forHELLFIRE.RICOCHET: Abstract strategy-action game in which you maneu-ver blocks to set up a shot at youropponent's goal and try to protectyour own. (Atari 400/800,JABBERTALKY: A program-mable word game. Four differentprograms are included, Alphagram-mer, Cryptogrammer, Jabbergram-mer, and Free Verse. (Apple II &TRS-80).

BEZ4790 Irvine Blvd. Suite 108BIrvine, CA 92714BEZ-MX: Strategy/arcade gamein which two players, or playerand computer plan and executeattacks on each other's country.Weaponry includes MX missilesand Bez-1 bombers. Targets areairfield, MX sites, factory, cities,agriculture (Apple II, $29.95)

LETTERSDear Editor,

I enjoyed reading Chris Crawford'sview of the future of computer war-gaming in the November issue ofComputer Gaming World. This isnot necessarily because I agree withhis vision, but because Chris tries todeal with most of the importantissues facing computer wargaming.In fact, I disagree with many ofChris' contentions and will discussthis under the two broad categoriesof programmers and hardware.In general Chris seems to be quitedispleased with the quality of pro-

grammers currently writing war-game software. As an example hestates that he had not yet met aprogrammer who had "solved themathematics of a hex grid for com-puter use". This is totally untrue, forhe has met several programmerswho have written games for SSI andall of them have solved the hex gridon the computer. In fact 8 of our 17games use a hex grid and they wereprogrammed by four differentindividuals.

Chris also stated that almost all

computer wargame designers areamateurs. If he intended this to meanthat they are all-designing for thelove of wargaming and not for moneyI would be forced to admit that thisis at least partially true. But I get theimpression his meaning is that mostdesigners lack professional skills.Once again I must disagree. Wouldhe call Charlie Merrow (designer ofComputer Air Combat) an amateurwhen he has over 22 years of com-mercial programming experience onover 15 different computers. He maynever have designed a board war-game before, but he's been workingon computer games for a long time.This kind of experience has beenmore often than not the norm, atleast concerning the designers that Ihave worked with.

As for pay, I agree that payingsomeone what he is worth is veryimportant if you're going to get aquality product produced. As forSSI, our staff programmers receivewages and benefits that are verycompetitive with those received byprogrammers in industry (quite afeat considering that we are locatedin Silicon Valley). As for royalties,Chris paints a dim picture for thepart time game designer. Actually, adesigner stands to make between$10,000 and $20,000 from a game,


6/44

half of which he will probably seewithin six months after the releaseof his game. I feel that this is ampleincentive for a programmer to com-mit his spare time for 4 months to ayear (especially if he considers theproject "fun" for at least the first fewmonths).So, as for programmers I mustdispute Chris's claim that therearen't enough good programmers towrite good wargames. I have dealtwith many good programmers writ-ing wargames, and although I lookforward to working with more in thefuture (and there certainly will bemore) the programmers we have nowprovide a great deal of creativetalent.As for the computers themselves,Computer Gaming World shouldhave mentioned that Chris is anemployee of Atari and thus presentshis view of the world from a some-what distorted vantage point. I donot agree that the Atari is clearly

the better computer available todayfor computer wargaming. The graph-ics used in Southern Command onthe Apple stack up against thegraphics of any computer wargameavailable today. As for the disk I/O,the Apple has the clear advantage,operating at speeds up to 10 timesthat of the Atari. This difference isespecially important when one istrying to design a complex gamethat will maintain a gamer's interestduring repeated plays.Being able to chain programs isessential, and the Atari is currentlyclearly inferior. Also, an Atari diskcan hold only 75% of the informationstored on an Apple disk, forcingsimplification. I have no doubt thatas the Atari comes of age it will havethe potential of becoming a bettermachine for wargaming, but at thistime I feel that the Apple is at leastequal to the Atari, if not considerablybetter.

As for the statement that 16Kgames are good enough I must onceagain disagree. My work at SSIallows me to talk with computerwargamers everyday, and the gen-eral consensus seems to be that no16K game to date (yes Chris thisincludes Eastern Front) maintainsa gamer's interest past the first fewgames. I attribute this to the lack ofvariability in 16K games. With only16K a designer must constantlysimplify his game, taking awayoptions until the game generallyfollows the same course every time itis played with little chance for agamer to encounter dramaticallydifferent situations. On the otherhand a 48K game can allow fordifferent scenarios, more random

events, and most importantly, morerules which cause each game to beconsiderably different from the lastone played. Keeping a 16K gamefrom falling into a rut is in myopinion almost impossible, and ifthe people I have talked to are anyindication, I am constantly beingproven right.

As you have probably determinedby now, I disagree strongly withChris on many issues concerningcomputer wargaming, but I do agreewith him that computer wargameswill continually improve, and thatthe machines we are currently usingwill eventually be replaced by newer,more powerful machines, and I forone can't wait to get my hands onone. Joel BillingsStrategic Simulations

Dear Editor,I was grateful to see a magazinedevoted to the aspect of personal

computing I am most interested in game simulation (particularly sportsgames).Unfortunately, my home computer,the TRS-80 Color Computer, has noAvalon Hill/SSI type software. Theonly obvious solution was to writemy own in particular a baseballand football replay program. Insteadof a batter-by-batter simulation, itmatches statistics for an entire teamto determine the winner of eachgame. This allows replay of a com-plete 8-10 team league in less than 30

minutes, including entry of the teamdata.With my game I replayed every

major league season 1901-1976.Using each year's winner, I organ-ized them into 10-year groups foreach league. I played each of thesegroups through a 162/154 gameseason and these are the decadechampions:National League:1906 Cubs942 Cardinals1920 Giants952 Dodgers1928 Cards963 Giants1935 Giants973 RedsAmerican League:1910 Athletics 1948 Indians1917 White Sox 1957 Yankees1927 Yankees968 Tigers1937 Yankees969 Orioles

Then I normalized the statisticsfor each league's...winners (comparedthem all statistically to the sameyear 1940 then altered theactual team data to get an accurate

replay of teams 67years apart...likethe 1906 Cubs and the 1973 Reds).I replayed these 8 team decadewinners through a complete 154game season with these results:

National League1906 Cubs8-661973 Reds3-711952 Dodgers8-761963 Giants7-771935 Giants6-781942 Cards4-801928 Cards1-831920 Giants9-85American League1948 Indians7-651969 Orioles7-671927 Yankees3-711937 Yankees0-741957 Yankees6-781917 White Sox3-811910 Athletics6-881968 Tigers2-92

I used the 1906 Cubs and the 1948Indians for a 7-game "Super Series"and the Cubs won in 4 straight Ithought you might be interested inthis different approach to determine"Baseball's Greatest Team".

L.V. BeattyNorfork, VA


7/44


8/44


9/44

takes 2 hours or more should have (again, EF doesnot). There's no need to say more, those who ownAtaris will buy one while Apple-owners will buythe other.HISTORICAL BACKGROUND

SOUTHERN COMMAND portrays 3 days offighting between the Israelis and Egyptians duringthe Yom Kippur War of 1973. The war began onOctober 6th when Syria attacked the GolanHeights in the north and Egypt attacked in thesouth. The Egyptian forces crossed the Suez Canalinto the Sinai (which Israel had captured in 1967)and broke through the lightly-manned Israelidefenses. Then, for reasons still unexplained, theystopped. This gave Israel time to defeat the Syriansin the north and transfer the units there back totheir Southern Command.

Even with one front cleared up, Israel could notafford to sit passively on defense. Roughly 10% oftheir total population had been mobilized andthese people were all needed back at the farms andfactories where they worked. Israeli leaders had tofind a way to end the war in a few weeks or thecountry's economy would collapse.

They decided to counter-attack at a weak pointin the Egyptian line and cross the canal. Bythreatening Cairo, they hoped to force the Egyptianunits to pull back on defense, allowing Israel tore-establish a defensive line along the canal. Thisplan, called Operation Gazelle, worked. SOUTH-ERN COMMAND gives you a chance to replaythe initial stages of the Israeli counter-attack.GETTING STARTED

The game may be played by two people but theemphasis seems to be more on solitaire play. Withjust one person, the computer always plays theEgyptian side. The Israeli objective is twofold: toget units onto the west bank of the canal and toreduce the number of enemy units on the eastbank. Victory is determined by a complex pointsystem but this is all handled by the computer.You may, at any point during the game, ask for arecap of the scoring and this includes an estimateof the final outcome (nice touch ).

When you boot the game disk, you see a menuwhich gives you 3 basic choices: start a new game,continue one that had been saved, or play ademonstration game. Demonstration mode is amarketing tool; something you'd use to show offthe program in a store. Besides the number ofplayers, there are several options to choose fromwhen starting a new game. There are normally 20turns in the game but you can choose to play 10instead. The Israelis get 100 points to make up forthe 10 missing turns. Also, the game normally

starts with all forces in their actual historicalpositions, but there is an option to let the Egyptianmove all non-infantry units to be in a betterposition. This option gives the Israelis a 20 pointbonus and can be used with either the 10 or 20 turngame. When playing solitaire, there are 2 levels ofdifficulty and each may be played with or withouthidden units giving 4 variations ofcomputer skill in all.GRAPHICS

SOUTHERN COMMAND uses high-resolutionfull-color graphics to display the map and units.The map uses the hexagonal grid seen in mostboard wargames and compared with these, it's anaverage-sized map. It's 28 by 39 hexes but innormal mode you can only see an 8 x 13 "window"so it seems much larger than it is. You can movethe window around with scrolling commands orby moving a cursor using 6 directional keys. Thecursor is used to select any one hex; as it nears theedge of the map window, the window will movewith it.

The various kinds of terrain are distinguishedmostly by color: blue for water, green for forest,orange for light and heavy rough, white for sandand blank for clear. The pattern of dots whichmark the terrain is distinctive, though, and youcan quickly learn to recognize them in black and


10/44

white; thus, there is absolutely no problem playingthe game on a monochrome monitor. The "B" keyis designed to make colored terrain show up betterby changing it all to white. This might be useful ifit weren't for one error. The unit currently beingmoved normally blinks to help you keep track of it,and this blinking stops if you hit "B". I found thisto be a big nuisance, but fortunately, the gamelooks better in color mode no matter what kind ofmonitor you have so just ignore "B". In color, theblinking unit flashes white, or green or orange soit stands out against any terrain (nice touch ).

The units are represented by symbols whichremind you what type of unit it is . Thus, a soldierwith a rifle is an infantry battalion, motorizedinfantry is represented by an armored personnelcarrier (APC), self-propelled guns are clearlydifferent from tanks, and artillery, rocket, anti-aircraft and recon units have their own symbols.Two units may be "stacked" in a hex and whenthis happens an "S" is displayed there. You canexamine the stack by moving the cursor there.Egyptian stacks have a box around the "S" todistinguish them from Israeli stacks.

There are times during any game when you'dlike to be able to see the Big Picture andSOUTHERN COMMAND lets you. A touch ofthe P key will display the strategic map. This is

the entire map, and while it is too small to showunit types, the positions of all Israeli units areclearly marked by a solid square and Egyptiansby a hollow square. Major terrain types are shown(again, readable in either color or monochrome)but roads are not. This map requires quite a bit ofknowledge to interpret. You must know the mapand remember what kind of units are where to useit well. Beginners may find it confusing, and thefact that the picture of it on the rear of the gamebox is upside down won't help Still, it is aninvaluable feature. Using it, for example I wasable to learn that Egyptian units were behind meon the second day. From then on I paid moreattention to the reinforcement scheduleCOMBAT

Combat is based on a step-reduction system,where units lose combat factors in action andregain them when they have a chance to "reor-ganize." If a unit's combat strength is reduced tozero, it is eliminated. The number of factors lost incombat will depend on the relative strengths of theunits, the type of units, and on the situation. Forexample, bombarding an enemy with artillerybefore you attack will tend to increase his casual-ties and decrease yours. Or, if you can attack anenemy unit which is trying to reorganize, it will beless able to defend itself. The combat system isexplained in an appendix in the rule book. A lookat it will tell you such things as: tanks are the bestunits for attacking other tanks and motorizedinfantry but the Israelis can get into real troubletrying to use tanks to knock out Egyptian infantry.Of course, you could learn this lesson on thebattlefield, too

Much of the game's tactics involves choosingthe proper mode for each unit. Attack mode is thebest for combat, although Defensive mode is muchbetter against artillery or air attack. River mode(for crossing the canal) is particularly weak, withReorganization mode and Transportation modeonly slightly better. Each mode has its use, how-ever; it's up to you to learn the proper time andplace for it.

Besides being rated for combat strength andmovement, each unit is rated for efficiency.Efficiency affects how well the combat strengthmay actually be used in combat, how well the unitcan change from one mode to another, and howquickly it can rebuild once it is pulled from thelines and put in Reorg mode. It is in the area ofefficiency, and only here, that the Israelis have aclear superiority. The Israeli commander who canfind the right time and place to rest and refitdepleted units will greatly increase his effectivestrength. The computer, as an opponent, is par-ticularly vicious about attacking these units with


11/44

The Israelis must get units across the canal 3by turn 3, 7 units by turn 9, 11 units by turn 15 soa few units must use their advantage of surpriseon turn one to run past the Egyptians and crossthe canal. Once over, they can stake out a bridge-head and m aybe wipe out som e ant i-aircraft missi lesites. These are worth extra victory points as anincentive for the Israelis to do something on thewest bank besides camp. In fact, a powerful threatwill force the Egyptian to pull units back from theeast side, giving evidence that the game is success-ful as a simulation.

I haven't mentioned airpower yet. Airstrikes actjust like very powerful artillery barrages. TheIsraeli Air Force can give ground support withincertain limits. It may be called in at most, one onthe first day, twice on the second, and three timeson the third. You may not get that many strikes,however, as there is a random element at work.Also, once during the game the Egyptian AirForce may be ordered into the air and they willdisrupt (prevent) an Israeli strike if one is called inthat turn.CONCLUSIONS

SOUTHERN COMMAND is an outstandinggame and a very realistic historical simulation.The rule book and two reference cards are of SSI'susual (high) quality. The only thing I could wantwould be an option to play the Egyptian sidesolitaire, but programming the computer to playthe Israelis would not be a trivial task. Still, it willkeep me busy trying to crack that Egyptian de-fense at the highest difficulty level. I urge you totry it too.

artillery and they are very susceptible to this formof attack while in Reorg mode.THE PLAY

In regular play, you have complete control overthe cursor and may move your units in any orderyou wish. Once you start to move a unit, however,you must complete it. Just move the cursor to theunit you want to move, hit "M", and then use thedirection keys to move the unit. It will stop movingwhen it's movement allowance is reached, or youmay hit "Q" to end before then.

A very good feature, which speeds play quite abit, is the "auto-move" mode. This simply meansthat after you finish moving a unit, the computerwill automatically move the cursor to the closestunit which has not yet moved. If it happens to pickone which you don't want to move yet, the "L" keywill cause it to skip to the next and come back tothat unit later. It is smart enough to select yourartillery units first so you can use them to softenup this turn's targets. Auto-move can be turned offand on at any time. Another nice touch.

Another innovation inSOUTHERN COM-MAND is delayed orders Instead of moving aunit, you give it orders to move and attack (attackonly for artillery) a given hex IF an enemy unitmoves into it. I confess I haven't really masteredthis yet, but the computer uses it if you play themost difficult level and it has caught me in severalnasty ambushes.The computer plays faster than any other SSI

game I've seen. It is almost too fast to follow andwould be if it didn't pause to let you read thecombat results. Of course, if you like, you can turnthis pausing off so that it will finish its entiremove in 15 or 20 seconds.STRATEGY

There really isn't much choice about where theIsraelis will cross the canal. The Suez runs fromthe lower left to the upper right of the map, but theright half is dominated by the Great Bitter Lakewhich cannot be crossed. Most units can crosswherever the canal is narrow, but Israeli tanksneed a pontoon bridge. There is a group ofEngineers who can build one and their slowprogress toward a crossing sight is the focus ofattention on the first day.

The easiest place to clear a corridor for them is inthe center, right along the shore of the lake. Here,only one Egyptian infantry unit must be pushedback to open the way. Of course, this leaves arather narrow, vulnerable supply line and createsa logical target for a counter-attack. The difficultlevel strategy seems to be characterized by justsuch an attack, delivered with a vengeance


12/44

SO YOU WANT TO WRITEA COMPUTER GAMEby Chris Crawford

Chris Crawford isan employee of Atari

You've had your computer for some time now.You've used it to teach yourself how to programand to handle a few household problems. You'vealso used it to play games. You have noticed thatmany of the games available for your computerare less than perfect. Soon you begin finding thetechnical flaws in them, and modifying some ofthem. The day inevitably comes when you decidethat you could have done a better job than thedesigner. You tell yourself 'Maybe I should write acomputer game.'

There's certainly plenty of motivation to write acomputer game. Perhaps you have heard stories ofthe fabulous wealth awaiting the designer of asuccessful game. Perhaps you have read magazinearticles about those wild and glamorous fellowswho design the games. Or perhaps you aremotivated by the simple pleasure of making peoplehappy with your own creativity.

Whatever your reasons you are certainly notalone. More than a million people own personalcomputers. All over the country, thousands andthousands of people are coming to the sameconclusions that you have come to. There is nothingstopping anybody from doing it; you need nospecial training to design a game, nor any specialequipment other than your personal computer. Noinvestment of money is required, nor will you needother people to help you with the programming.You can do it all by yourself in your spare time. It

should be easy Why then, when so many peoplewant to, have so few people been successfuldesigning games? There are four main reasons.

First, the successful game designers work veryhard at their games. They are perfectionists whospare no effort in the quest for excellence. Theyhave an eye for detail, putting in fine points thatfew customers notice. They polish their programsendlessly, rooting out all the bugs to achieve aflawless product. I know a software evaluator, aman of deep insight, who tells me that he canalways tell top-notch software submissions fromtheir user manuals; they are always carefullytyped and free of spelling errors. Inevitably, thesoftware accompanying such manuals is equallyflawless. I have seen many games by amateurdesigners, some good and some bad, and the onlycommon trait of the bad games was insufficienteffort. There is no excuse for a half-finished game.

The second distinguishing trait of successfulgame designers is their technical expertise. Theyall are completely comfortable with assemblylanguage and they all know their machines well.No great paintings were ever painted withcrayons; no great symphonies were ever composedwith kazoos; do you really think that you can writea great program in BASIC? I have yet to see anexcellent program written in BASIC. I have yet tomeet a good game designer who did not have totalmastery of the machine on which he worked.


13/44

Technical expertise is a means to an end, buttechnical expertise alone does not qualify anyoneto write good games. I know a number of brilliantprogrammers who have written wretched games.The best game designers are really artists at heart.Their games are an extension of their personalities.These people are intensely creative because theyinvest so much of themselves into the designprocess. They disdain copying other people's ideas,for that would detract from the clarity of theirself-expression. My advice to the beginning gamedesigner is similar to the advice a famous writeronce gave an aspiring young author: Write a gameabout what you know, not what you think otherpeople admire this is the third factor in success-ful game design. If all you know is ditch-digging,then design a game about ditch-digging. If you putyour heart and soul into it, if you condense thehuman essence of ditch-digging into that game,people will like it. If you only copy an idea somebodyelse has already used, people will see it for thecheap copy it is.

The fourth factor in success with games ismarketing. Don't try to sell the game yourself;turn it over to a software house for royalties. If youare truly a good programmer then you are almostcertainly a poor businessman. Let a good business-man handle the business aspects of the job.

There is, unfortunately, an element of luck in themarketing sector. Three years ago I pounded thepavement trying to sell my games to any of anumber of software houses. Nobody was interested.Now I have standing offers from a dozen houses. Ihaven't changed that much in three years; whathas changed is my reputation. Three years ago Iwas just another unknown software author. Fameand reputation these things can be arbitrary. Bepatient and make sure your product is good. Keeppounding the pavement and your turn will come.

If you decide to go ahead and write that game, Isuggest that you follow these steps:First, toss the idea around in your head forat least a month before you begin any pro-

gramming work. Think it through first Don'trush in and hack the program together. Makesure you know what you are trying toaccomplish before you begin work. Ask your-self, what will people feel as they play mygame? What will the game teach them? Howwill it challenge them?

Second, plan the game I/O before anythingelse. I/O is the big bottleneck in all games; itdictates what can and cannot be done.Graphics are assuming increasing impor-tance in all games. Carefully plan the graph-ics and animation that your program will

use. Pay special attention to the input struc-ture of your program. Most poor programssuffer badly from poor input structures thatconfuse and frustrate the user. Don't let yourgame be ruined by clumsy or overly delicateinput systems.Third, lay out a memory map and stick to it.Decide how much memory your game will

take. Less memory may reach a largeraudience, but more memory will allow abetter game. Allocate specific chunks ofmemory for specific functions. If any functionneeds more space, re-evaluate your entirememory map before allocating additionalspace to the function.

Fourth, develop the algorithms for criticalfunctions of the game. Algorithms are thecritical intermediate step between broad goalsand specific code. Many beginning program-mers leap all too quickly from concept direct-ly into code without taking the time toproperly frame their algorithms. If you thinkthat you can write code without thinking outthe algorithms behind it then you are either agreat deal smarter or a great deal dumberthan I.

Fifth, write the program. Are you surprisedthat writing the program comes so late? Areyou surprised that I treat it so perfunctorily?If so, you are in for many more surpriseswhen you do the game, all of them rude.

Sixth, playtest the game several hundredor even a thousand times. Enlist the aid ofseveral friends to playtest it for you, and payattention to their criticisms. Some of theircriticisms will be bad; ignore those. Some willbe good; implement them immediately andthank your lucky stars for good playtesters.How do you know the difference betweengood criticisms and bad ones? I don't know; ifyou ever find out, please tell me.

Seventh, polish the game.Eighth, polish it some more.And finally, when you are truly and deeplysick of the game and desperate to get it out of

your hair, polish it for at least another month.Polish is the critical element, the final touchthat turns a good game into a shining beauty.I have programmed four computer gamesnow; every single one gives me pangs of guiltbecause I did not polish it adequately. Don'trepeat my mistakes.Obviously, the process of designing and developinga computer game is a long and painful one. Now doyou see why so few game designers are successful?


14/44

NAPOLEON'S CAMPAIGNS1813 & 1815:Some Notes

by Joel BillingsFor years player's, designer's and strategy notes

have been a standard feature of wargames. Some-times the notes appear in the rulebook itselfothertimes the notes are published elsewhere. JoelBillings, of Strategic Simulations, shares his noteson SSI's new game, NAPOLEON'S CAMPAIGNS1813 1815. The numbers indicate the section ofthe rulebook to which the note relates. Here's tobetter campaigningPLAYER'S NOTES3.1 Setting your Line of Communications can becrucial early in the game, for your corps will try toretreat along the corn line if possible. In this wayyou can ensure that your corps that are far fromthe main army will retreat towards the bulk ofyour forces. An example of this is stretching yourcorn line to the Allied 6th corps in Waterloothrough hexes 8,9 and 3,12. This will allow 6thcorps to retreat towards Brussels in the face of amajor French attack through Tournai. The cornline can also be used to create forces, thus allowingyou to specify the march mode of the corps involved.This is the best and only way of guaranteeing thatyour troops will get to where you want them to go.If for example as Napoleon you wish to pursue thesouthern strategy in Leipzig, a good tactic is totrace your corn line from Leipzig to 9,12 - 9,11 -10,10

- 9,9 - 8,9 - 10,8 - 10,4. This will allow you to giveforce orders to 9 of your corps on turn 1, and willallow a quick movement of troops from 10,4 to 9,12.In this way you will have direct control overwhether your army force marches to a possiblebattle south of Leipzig, or takes its time in order toarrive fully rested.3.4 and 3.5 The starting urgency and contactmode can mean the difference between life anddeath for a corps cut off from the main army.Giving an urgency of 9 and a contact mode ofstand to the Allied 1st corps in Waterloo isequivalent to signing its death certificate. Thesame could also be said for the Allied 1st and 2ndcorps in Leipzig. Be very careful to keep your corpsout of a major battle until your army has beenunited. Also, beware of the sucker strategy oftenused against the Allied 2nd corps in Waterloo. Ifthis corps is placed in initiative or attack mode, allthe French player need do is send a cavalry corpsto hex 16,15 followed two turns later by 50,000 menand suddenly the 2nd corps will find itself about tobe destroyed. Although there is no best initial setup, especially in Waterloo, at least be cognizant ofthe tactics you are set up to defeat, and thosewhich may be used to defeat you.7.2 By moving your HQ as part of a force, you willbe able to keep that force stacked together as long


15/44

as desired. This will allow for rapid movement.Without your HQ, your units will spread out alongthe road, with the last unit moving only after thefirst units have vacated the hex, thus slowing theinevitable desired concentration of your corps.7.4 In effect this means that units in a ZOC maystack as much as desired. This allows you toconcentrate your troops adjacent to the enemy andthen attack with your entire force hitting theenemy at one time. In order to do this, give yourunits stand orders, and only convert to attackorders after your army has concentrated.8.0 (A) Movement is by far the most subtleelement to this game, and the part which givesgarners the most problems. If your units fail tomove as you thought they should check for thefollowing:1) Is it or has it been raining?2 Did you give a low urgency (don't expect a

leader to burn rubber if he has an urgency of 1,in fact be happy if he moves more than oncea day)?3 Is the corps commander unaggressive (givingan urgency of 5 to Bernadotte is not much betterthan giving an urgency of 1 to Napoleon, andgiving an urgency of 1 to Bernadotte will puthim to sleep for the rest of the game)?4 Is the unit trying to cross a river?5 Does the unit have stand or initiative orderswhile constantly bumping into enemy units?6 Is it night or 0500 in Leipzig?7 Did you really give him orders to go somewhere?8 Are you constantly sending him new orderseach turn thus eliminating the old order?9 Did you initially give him an urgency of 1 withhis movement orders, and then send him a neworder to convert to urgency 9, and was that theonly order you sent him? Don't forget that a unitwill lose all previous orders whenever he receivesa new non-immediate order. This is the mostcommon mistake made by players. Wheneveryou send a new non-immediate order, even oneas simple as increasing a corps' urgency,remember to also reiterate the movement ordersyou wish the corps to execute.

8.0 (B) Another problem many people face ishaving their corps suddenly start to march off inan unexpected direction. The usual reason for thisis that a unit moved somewhere due to combat(attacking or withdrawing). Units in combat keeptheir old orders (including orders previously issuedduring combat having to do with combat). Whenthe unit is no longer facing an enemy it reverts tothese old orders. Example: A unit had started toretreat to 3,4. The player gives an RT order for 3,5.When the unit finishes retreating to 3,5 it will

move to 3,4. Use the SC command to check if this ishappening, and if your unit has 'bad' orders clearthem with a non-immediate order.9.6.1 Ferocity is the key to combat in this game. Agood leader can hold off a poor leader despite beingoutnumbered by as much as 10 to 1 simply bykeeping the ferocity low. Rear guards and blockingforces should be given urgency 1 in order toprevent the high percentage casualties required tobe routed. A player should only use a high urgencyin combat when he is willing to accept large losses,or when he outnumbers the enemy sufficiently toreduce his losses.9.6.1.2 Be sure to place your fatigued corps in therear during a battle, for putting them in the frontwill only decrease your odds while increasing thechance that the units will rout.11.5 The loss of a fatigue point from normalmarching at night is often an exceptable penaltyin exchange for speed. The ability to move at nightshould be used especially in Leipzig, due to the0500 restriction on movement (the fatigue gainedat night will be lost during the 0500 turn). Avoidforce marching at night if at all possible.13.2.7 For those who desire to get a corps to forcemarch, simply begin a movement order with astarting hex that is several hexes away from thecurrent location of the corps you wish to move. Forexample, in Leipzig, if you wish a unit in 9,12 tomove quickly to 9,15, enter a movement order with9,15 as the starting hex (and ending hex bypressing the space bar). The unit will force marchuntil it reaches 9,15. This can be a very usefultechnique when used in movement along a straightline for short distances. If, however, the path is notstraight, your corps may end up traversing openground and crossing unbridged rivers while alsoforce marching at night, thus destroying the unit'seffectiveness.14.3 The importance of this section cannot beoverstated. It is essential that players realize thatwhen a force is given orders to move at a set marchrate, the march rate is only used during the turn inwhich the order was given. In order to keep a forcemoving at a set march rate, the player mustcontinue to create the force each turn. In doing thiswithout giving further orders to the corps, a playercan ensure that his corps will move as he wishesthem to.STRATEGY NOTESLeipzig (French)The key to French strategy is to develop an initialplan involving the destruction of a major portionof the Allied forces before the Allied armies areable to unite. Once the plan is decided upon stick to


16/44

Continued on page 39

it, unless the Allied forces you hoped to destroyhave withdrawn in the face of your initial thrust.Be sure to mask your true intentions as long aspossible. The game can be divided into two fronts,the northern front facing Bernadotte and Blucher,and the southern front facing Schwarzenberg.Generally it is best to pick one and force march asmany corps as possible to mass for an attack onthe enemy forces in the front chosen. Leave only asmall rearguard to block the movements of theother Allied armies, but be sure to leave at leastone strong leader (6 or Napoleon) with the rear-guard. If it becomes impossible to destroy the forceyou are attacking before other Allied armies beginto surround you (blocking your retreat routes tothe west), pick a new corn line and escape beforeyour retreat becomes a rout. The French can winby causing heavy Allied losses during the firstthree days and then withdrawing the army to thewest intact.Leipzig (Allied)Withdraw I and II corps toward Altenburg, stand-ing only after being united with Schwarzenberg inhex 9,15. Use III and VII corps to disrupt Frenchmovements west of Leipzig, or move III corps tojoin the Allied corps in 9,15. Give high urgencies(over 6) to corps VI and VIII and order them tomarch towards Schwarzenberg. The key to theAllied plan is the ordering of IX and X corps. If youbelieve the French will attack south, order Bluchertowards Leipzig, but with a low urgency and standor initiative orders until you receive details ofenemy movements near Leipzig. If you believethat the French will strike north, leave Blucher inHalle with stand or withdraw orders, and moveBernadotte to Halle as quickly as possible. Assoon as you feel you know where the French aregoing, begin to attack his rearguard while tryingto attrition Napoleon in the area of his attack (butdon't allow your corps to be sufficiently fatigued torout). On the last two or three days your objectivesare clear, 1) take Leipzig 2) block the Frenchwithdrawal (this can be done if you are able to guesshis planned retreat route early in the game). TheVII corps is ideally suited for disrupting a retreatfrom Leipzig.Waterloo (French)Your only hope is to either crush the Prussianarmy and take Liege, or attack Wellington beforehe has time to concentrate more than three corps.In either case a holding force must be used to blocka union of all Allied corps. Cavalry corps should beused to screen French movements and to disruptAllied communications. They can also be used toblock Allied movements (II and VI corps are theeasiest blocked) and to capture Ghent, Liege, andBrussels if they are ungarisoned. The Frenchplayer must keep his main force together and

move quickly until a decisive battle can be forcedon a portion of the Allied Army.Waterloo (Allied)The Allied player should concentrate corps 1-4 inNamur (alternatively, leave IV corps in Liege toguard your communications) and once there begina move toward Waterloo if the French player isattacking Wellington. Concentrate corps 5-8 andWellington near 11,11 and fall back towardsBrussels if pressed early by the main Frencharmy. If the French attack towards Namur, marchto Ligny and further once your corps have united.It may be necessary to leave VIII corps behind tosecure Brussels and your communications linefrom French cavalry raids. Avoid force marchingif at all possible, for your units have extremely lowmorale and can be easily routed in combat. Thekey is to withdraw early in the game, only accep-ting combat when your forces have united. Also,do not allow your Prussians to be attacked byNapoleon for long, for your leadership disadvan-tage will lead to the destruction of your army.DESIGNER'S NOTES7.0 It was decided that normally a corps wouldoccupy an entire hex of roadway while it wasmoving, thus prohibiting others from movingthrough the hex at the same time. The exceptionsto this involved either the concentration or with-drawal of an army in the face of the enemy, or theuse of the army headquarters (logistics staff) toassist the movement of an army more quicklythan otherwise possible (allowing stackedmovement).8.9 Orders to corps were usually vague, designedoften to be interpreted by the corps commander ashe saw fit. This is why we developed the idea of anorder consisting of a stated path of movement andan urgency setting concerning the importance(and thus necessary speed) of the movement. Theurgency is also affected by the aggressiveness ofthe commander, accounting for the historicaltendencies of the commander's interpretation ofhis orders. The affects of urgency on movementand combat are the core of the game.9.6 The key to the combat system is understand-ing our desire to enable an excellent leader with asmall force to hold off a poor leader with a largeforce. The intent was to allow a well led Frenchcorps to hold for at least one day against a forcemany times its size, and this is why casualtiesare directly proportional to the urgency, allowinga good leader to lower the combat urgency andthus sustain relatively light casualties despitebeing outnumbered. Also, a good leader can force a


17/44


18/44

BASIC INFORMATIONNAME:lackjack MasterTYPE:ard Game and TutorSYSTEM:RS-80 Mod. I, Level II, 16KFORMAT: Tape# PLAYERS: 1AUTHOR:orman J. Wazaney Jr.PRICE:29.95PUBLISHER: Hayden Book Co.

50 Essex St.Rochelle Park, NJ 07662

Calmly, you watch as the dealer flips cards toeach player around the table. Y our table-limit betsof $500 have attracted the attention of severalcasino regulars, and a small ring of admirers hasgathered behind you. You detect a faint scent offlowers as a soft hand gently touches your shoulder.

The dealer displays a nine of spades as hebegins his circle of the table. The three players toyour right each draw cards. Each, in turn, silentlypushes his hand and his bet to the dealer when histotal exceeds the conclusive 21 points. You carefullytilt the corners of your cards upward; just enoughto glimpse the two red eights hiding beneath. Withjust a moment's thought, you flip them over andslowly push an additional $500 to the edge of thegreen felt. The hand on your shoulder squeezesgently.

The first card slides across the table and youquickly lift its edge. Another eight, this time black;not what you expected. Without hesitation, youturn it over and move another $500 stack out fromyour depleting bankroll. The hand on yourshoulder tightens perceptibly. Another card leavesthe dealer s fingers. Closer to the statisticalprobabilities this time: A red king to go with thefirst eight. Quickly, your finger taps the secondeight. It is hit with a black seven. Again, not whatyou expected. You make a quick decision, callingfor another card. This time you get a five, for atotal of 20 on that hand. Moving quickly to yourfinal eight, you motion for one more card. Thistime the expected probabilities come through: aqueen.... of hearts.

Now the dealer glances toward the tight lippedpit boss as he turns over the card hiding beneathhis black nine. The crowd gasps as it becomesvisible: the final eight in the deck, giving thedealer a total of 17.

Y ou breath an audible sigh as you accept the pitsupervisor's chit for $1,500.00 and turn to lookdirectly into the smiling blue eyes behind you.

If this is your fantasy of a weekend in LasVegas, then maybe BLACKJACK MASTERcan help make that fantasy come true. It gives theblackjack devotee a way of testing new playingstrategies and betting systems without risking theproverbial farm. The program consists of threeparts: The GAME; the SIMULATOR; and theTUTOR.The GAME is similar to other blackjack

programs. It plays blackjack with the player,offering all of the options available in a standardcasino game, including doubling, splitting cards,insurance, and that little known possibility surrendering. Bets can be made in any amountfrom $2.00 to $500.00, so money managementstrategies can also be analyzed.I have two immediate objections to the gameprogram. First, the game always starts you withan unrealistic (for me, anyway) bankroll of$100,000. I would quickly alter the program logic toeither provide a less intimidating amount or allowthe player to choose his own bankroll. I'm notcomfortable risking $100,000, and, at a maximumbet of only $500, it just takes too long to feel thatyou're either winning big or really blowing the wad

The second objection is more serious. Thegraphics of the game are dull and uninspired.Even with the somewhat limited capabilities ofthe TRS-80, a graphic representation of playingcards is possible. To limit the depiction of aplaying card to a small number crammed into abox detracts from the visual impact of the game.Additionally, the screen text format requires theplayer's eyes to jump between several points onthe screen; increasing the likelihood of playerfatigue and mistake. Possibly the 16K memorylimit prevented the use of graphic playing cardsdue to program size. But, even staying within


19/44

these limits, I believe that a better screen layout ispossible which would improve the visual appeal ofthe game.One technical hitch was discovered. After

running through a few hands in the GAMEprogram, I was pulled up short by an "out of stringspace" error. I corrected the problem by increasingthe cleared space from 800 to 2,000 in line 20 of theprogram. This worked in my 48K machine butmay not satisfy other models. I suggest you playaround with this if it occurs in your version.

The TUTOR portion of the program is probablythe most useful of the three segments for theaverage player. It continuously deals cards to youand to the dealer and asks for your play on eachhand: hit, stand, double, or split. It then immedi-ately analyzes the correctness of your play basedon the playing strategy imbedded in its programlogic (more on that later). If you are wrong, itdisplays the proper play for the hand. At the end ofany number of hands selected by the player, theTUTOR provides a total statistical analysis ofyour play; presenting percentages of correct andincorrect calls when hitting, standing, doublingand splitting. Thus, the player learns which of hisdecisions are weakest and can concentrate atten-tion on improving those segments of his play.

The only real objection I find in this part of theprogram, is that the player decision is analyzed ononly the first two cards dealt in each hand.Obviously, if I'm dealt a 4 and a 3, I'll call foranother card, no matter what the dealer is showing.But, what if the third card dealt to me then is a 9,giving me a total of 16. Should I hit or stand?That's a far more important (and realistic) decision,but the TUTOR doesn't contain the logic to helpwith that one.

One other minor objection is the use of thenumeral one to represent aces in both the GAMEand the TUTOR portions of the program. It some-times caused me to input wrong decisions duesimply to confusion.The final, and most unique, part of the program

is the SIMULATOR, which is designed to runthrough any selected number of hands using theplaying and betting strategy chosen by the player.It then displays a complete analysis of thestrategy; including the total number of hands wonand lost, the number of busts for both the dealerand player, and the number of blackjacks, splits,doubles and pushes. All in all, a rather impressiveperformance. Unfortunately, the program cananalyze only about 50 hands a minute, while theinstructions indicate that about 5,000 hands shouldbe processed for a statistically valid result, sobe prepared for a long wait.

Two other improvements can be suggested forthe strategy simulation program. First, it analyzesonly zero-memory strategies. That is, strategiesthat do' not make use of any memorization ofpreviously played cards. Experienced blackjackplayers, who utilize different betting and playingstrategies based on what cards are remaining inthe deck, may find this too limiting. Anotherlimitation is imposed by its use of only one deck ofcards. Many casinos use a shoe containing threeor four decks at a time, and, in fact, the GAMEportion of the program uses four decks. This canchange the strategies somewhat from single deckplay.

Now for the major criticisms. Two problemsstand out as almost intolerable in this otherwiseacceptable program: the documentation and themethod of inputting the player's selected strategyfor analysis.

First, the documentation. The instruction book-let is an impressive 60 pages in length with a slickcover and quality art work. Of interest to theneophyte programmer is a complete flow chart, alisting of program logic and a list of all variablesused in the program.

The weakness is found in the complexity ofpreparing the tables necessary for inputting theplayer's selected strategy. After carefully explain-ing how the game is played in terms suitable forrank beginners, the instructions continue on intostatements like: "for every entry next to the ace onthe worksheet, put an X in the ace row on thedouble down table-hard form under the two cardtotal slot". Now that's a mouthful, and enough toconfuse and discourage all but the most determinedblackjack players.

The second major difficulty is in the method ofinputting player strategy. After using the con-fusing tables described above to diagram a strategy,it must be input into program logic by listing outthe program and then inserting data statementscontaining strings of l's and 0's at the properpoints in the program structure. In addition, if aplayer desires to save a particular strategy forfuture use, it must be separated from the rest of theprogram, saved to disk in ASCII format and thenre-entered into the main program through the useof the merge command. Certainly not a job for thetypical non-programmer blackjack player.

If this explanation of the strategy input methodleft you confused, then perhaps blackjack masterisn't for you. But if you're a reasonably capableprogrammer and an avid blackjack player, thenthis program may be your guidebook on the roadto riches.


20/44

Fiction:

ESCAPE FROMWOLFENSTEINby Ed Curtis

There it is...Switzerland Just fifty more yards tothe frontier and those dogs can howl all they wantafter that. Those miserable dogs I haven't had awink of sleep in the past four days because of thoseDOGS Hold it... Get hold of your self Don't letthose moon-cast shadows get the better of you. Ofyou? Now I'm talking to myself. That last grenademust have shook me up more than I thought. Thatlast grenade. It seems like a lifetime ago that I gotout of the Castle. How long ago was it? Letssee...the Schweinfurt raid was on the seventeenth,August seventeenth. Only last Tuesday. Just hitthe ball-bearing plant on the swing back from thefighter factory, and the rest was home free. That isuntil you lose your engines and half yourcrew to fighters that were supposedly drawn southwith the other formation. And then, slowly driftingdown as you watch your B-17 hit the ground in aball of flame. I was sure that they were going toshoot me until one of the officers recognized that Iwas a bombardier. All that I remember after thatwas one word which kept reoccurring in theirconversation--Wolfenstein.

Time seems to lose much of its meaning as thepreliminary interrogation took place. "Tell usabout the Norden bombsight and how it works ""Don't be a fool, captain, you know that you willtell us everything before long." "Why make your-self suffer needlessly?" "Your fellow officers havealready given us the information we need; all weask of you is enough information to check whatthey have given for minor errors which may haveoccurred due to their conditions at the time." Onand on, over and over with what appeared to berandom interruptions. It got to the point that Iwasn't sure anymore about my ability to hold out,when it happened.

That night, a British commando, little morethan a pile of rags after what he had gone through,slowly pushed a pistol, that he had somehowgotten, over in my direction. He knew that he wasthrough, he insisted, and told me of the plans foran Operation Rheingold. I now had a duelmission; to keep my information from the Ger-mans, and getting the plans out. More important-ly, I now had the means to do at least the first half

I was being taken to the interrogation rooms

again when the second miracle happened. As wepassed another prisoner, who was being taken tothe holding cells, my guard's attention was off melong enough for the pistol to do its work. The otherprisoner, hit by a stray bullet, at first seemed dead.I don't know what made me check his body, butturning him over brought a small gasp from theotherwise lifeless corps. Coming slowly toconsciousness, he held on, somehow, long enough togive me a possible escape route.

From that point my escape became one battleafter another, as I slowly worked my way downthe building. Off the bodies of the common soldiersI scavenged bullets and grenades that cleared myway room by room. The chests that I openedhelped even more as they provided me with auniform, food, and other very useful items. Pro-tected by a bullet-proof vest and in German uni-form, my only bane proved to be the equallyprotected and unrelenting SS. I don't know howmany times I ran from one room to the safety ofanother only to be pursued by those devils. Fana-tical, they would walk right into a stream ofbullets, trusting their vests and Adolf Hitler toprotect them.

Throw a grenade and duck around the corner.Before I got out of that death-trap, I had developedthat response to anything...throw and duck. That,I'm sure, is why my head feels like something thatwas borrowed from some poor relation. And, thatlast one If miracles come in threes, then that mustsurely have been the third. I'm still certain that Iwas closer to it, when it went off, than the SSofficer.

I don't remember much of what happened thenext day or two, but it seems that I just kept goingaway from the dogs. It is even possible that theGermans, by heading towards the border, kept mefrom wandering aimlessly until they picked meup. Even with the way that I feel, the thought ofthe Germans making my escape possible brings agrin.

Oh, well I've rested enough, and the howlinggrows nearer. Time to check my pockets and gothose last fifty yards. Lets see, wurst, schnaps,pistol, grenades, map, and...oh yes...the plan.


21/44

TO JOY PADDLE AND PORT:Som e Com ponents for Gam e PlayingLuther Shaw

TG Products of Richardson Texasand The Keyboard Company ofGarden Grove California are prob-ably the best known manufacturersof game input devices for the AppleII computer. Both companies sellgame paddles and joysticks. TG Pro-ducts also sells a switchable gameport.

In the year that I have owned myApple II computer I have gonethrough two sets of the low qualityblack-cased paddles which, at thattime, came with the Apple II com-puter. The dials for the old paddleswere often short on the readings onthe upper end. But the worst com-ponent of the old paddles was thefire button which was designed forlight weight electronic duty, not thehot and heavy pounding that comesfrom constantly saving the earthfrom invading aliens and the like.

Happily TG Products and TheKeyboard Company have madeavailable game paddles that are farsuperior to the old ones. The TGpaddles are encased in rectangularplastic cases color matched to theApple. The fire button is large and ofmuch better quality than the oldblack-cased buttons. The paddle dialcontrol has a substantial feel. If youleave your ship/whatever exposedto enemy fire it will be your ownfault not that of the TG paddle whichretails for 39.95 per pair.

TKC's paddles are shaped to fitcomfortably in your hands. The dialis designed to allow the player to usethe paddle as a small dial device or alarge dial device. It reminds one ofthe small steering wheels of thestreet racers versus the large steer-ing wheels the cruisers used to have(does that date me?). The fire buttonis built flush against the case whichmakes it comfortable to use. It isapparent that TKC has put some

thought into the human engineeringof their paddles which retail for 29.95.Like the TKC paddles, the TKCjoystick fits comfortably in yourhand. It is weighted which gives ahigh quality feel. The stick itselfmoves smoothly providing accurateresponse. Rather than the two firebuttons that you find on most joy-sticks, the TKC joystick has one fire

and one dual action. switch. Thedual action switch has a momentaryON position which automaticallyswitches back OFF (for use as aregular fire button) and a perma-nent ON position (for continualfiring). TKC added this dual switchwith the view that players (as wellas those using the device for non-gaming applications) would like thevariety of the dual switch. It wasalso hoped that some games wouldbe designed to make use of thefeature. However, after using thedual switch I find that it doesn't addmuch and is more difficult to usethan a second fire button. The Key-board Company has been consider-ing making the change back to astick with two fire buttons. Theywant to know what the users think.If you have used TKC's joystick youmight drop them a line letting themknow what you think of their dualswitch.

Like their paddles, The TG Pro-ducts' joystick is encased in a durableplastic case color coded to match theApple II. The device has an impor-tant feature lacking in the TKCstick self centering. Certain com-puter games such as CASTLEWOLFENSTEIN and FLIGHT SIM-ULATOR play much better with aself-centering joystick. But if youprefer a non-self centering stick TGProducts provides simple instructions-

on disabling the self-centeringfeature. The two fire buttons are justlike those on their paddles and willstand up to a lot of punishment.A third item marketed by TG Pro-ducts is a must for any gamer who istired of having to get inside hisApple II to manually switch fromthe paddle plug to the stick plug. Notonly is manually switching a hassle,it can also get expensive if you dowhat I have done break a pin inmaking the switch.

provides switch selectable socketsfor various game devices (such aspaddles and joysticks).port and allows you to plug in up tofive devices which would normallyplug into the game port. Socket #1 isa non-switchable socket which di-rectly parallels the game socket. It isutilized for sensitive devices forwhich diode isolation would inter-fere. Socket #2 is switch selectableand modifies the device plugged intoit to operate as paddle 2 and paddle3. This is useful in applicationsrequiring dual joysticks or four gamepaddles. Sockets #3 through #5areswitch selectable ports which maybe connected or disconnected to theApple II via a switch. At 59.95SELECT-A-PORT is not inexpensive,but is a worthwhile purchase for theserious gaming hobbyist.

After having played games withboth sets of paddles this writer findsboth to be of high quality but prefersthe TKC paddles due to physicaldesign and the cost. The choice ofwhich is the better joystick buy ismore a question of taste. While theTKC stick is of costs less ( 49.95 asopposed to the TG stick at 59.95), itlacks the self-centering feature. Ifself-centering is important to youthen your choice is obvious.

From left to right: TKC joystick, TKC paddles, TG Select-a Port, TG paddles, TG joystick.


22/44


23/44

Computer game players often give short shriftto game documentation. And why not? Too oftenit's a barrier the gamer has to overcome, every bitas difficult as the cleverest monster or the mostingenious trap. Instead of being perceived as anessential part of the game, the manual is seen assomething between the gameplayer and the game,a necessary nuisance.I know many computer garners who glance atthe documentation just enough to discover thebasic keyboard commands, then throw it into thewastecan. I know gamers who copy their friends'

game diskettes, but don't even bother to borrow orcopy the manual. They know there will be commands that aren't in the documentation anyway.They know there will be situations, objects, specialrules and conditions which the manufacturerhasn't even bothered to include in the manual, sousually they're content to just discover for them-selves what the rules and limitations are. It hasbecome a game within a game What was left outof the manual and where was it wrong?

I'm not a programmer; I'm a fiction writer,writing computer manuals on the side. I do there-fore have a bias I want gameplayers to readmanuals. But I don't read most of the manualscurrently on the market as I have a rather lowtolerance for boredom. Putting aside the issues ofincompleteness and technical inaccuracies gamemanuals are usually lacking on the most basiclevel of writing they fail to hold our interest.

Consider the example set by adventure gamemanuals. These are the worst offenders simplybecause they need to be longer and more detailed.Most of us can wade through the page or two theaverage arcade-type game requires. Adventure

game manuals, on the otherhand, will typicallybegin with a lengthy description of its cast ofcharacters. The thief, the fighter, the priest, themage, etc. When have you ever been surprised bythese character descriptions? When has therebeen a characteristic which stood out, which wasunusual enough to mark that character as any-thing other than two-dimensional? Read a charac-ter description in one manual and you've readthem all.

Next come the all-important Attributes. Strength,I.Q., Speed, Luck, and Constitution make up atypical list. Of course, these are merely wordsattached to the numbers in the program whichenable the computer to simulate the variety ofhuman types in the real world. The computerplays with these numbers to figure the outcome ofbattles, theft attempts, etc. But the gameplayermust think of these attributes as real humancharacteristics, strengths and shortcomings whichmake his characters individuals, before he canreally enjoy playing the game. Far too often thedocumentation treats these attributes as if theywere just numbers, providing only the mostcursory examination of what these attributesmean. Often the definitions are fuzzy, overlapwith each other, or at worst are just plain non-sensical. There would seem to be no compellingreason why a character might be stronger in oneattribute than in another. Each should meansomething different within the context of eachcharacter's life: added strength, for instance, willmean something different to the fighter than itdoes to the thief. And the case studies of thecharacters in the manual should ideally provide a

T H E C U R R E N T S T T EOF COMPU TER GA MEDOCUMENTATIONBy Steve Rasnic Tern


24/44

context which will make such differences apparent.A fighter might use added strength in swinging amace; a thief might use it to scale a wall to reach acrown. But perhaps the fighter would not have themotivation to use his added strength in thismanner, and would be unable to scale the wall. Ofcourse, in some cases these differences will not beaccounted for in the game system itself, but theycan make a big difference in a player's feeling forthe characters and the situation they are in. Thegame simulation will seem a bit more real.

But a manual which attempts to work out thequestions of character and attributes more thor-oughly and evocatively may provide an addedbonus. Using the above example, the manual'sfuller exploration of character might demonstrateto the designer that he should change the Strengthformula in the program so that the fighter cannotuse added strength to scale a wall. More completedocumentation may suggest improvements to theway abstract numbers are handled, adding com-plexity and realism to the game.Of course, most of these criticisms apply to the"simulation" part of the game. Gameplayers whocouldn't care less about characters and the fan-

tasized scenario, who are quite comfortable playingat a more abstract level in which these attributesand characters truly are numbers, will have noproblem with second-rate game manuals. Themanuals will be totally irrelevant to them beyonda few basic keyboard instructions.

As a science fiction and fantasy writer, thething I'm most struck by when I read computergame manuals is the number of lost opportunities.Few manufacturers seem to have realized thatmanuals can be more than basic, skeletal instruc-tions. They can be part of the fantasy world itself:a magic book, a historical treatise, perhaps ahandbook for the sorceror's apprentice. A gatewayto another world.

In my desk 1 have a handful of computer gamemanuals. Several of their makers seem to havediscovered some of the potential for the gamemanual as an actual fantasy object.

A relatively straightforward treatment is theStarship Mission Simulator Operator's Hand-book (Harry M. Sweeney, 1979), a manual for anumber of the Trek-type games. It's a long manual,with much discussion of quasi-technical matters,game system material which might normallydetract from the role-playing aspects of the game.But Sweeney treats this material as if this were anactual Federation manual for the Starship MissionSimulator. The reader becomes a cadet or highlytrained and skilled Federation officer seeking toimprove his or her skills. The cadet or officer willbe rated; the simulator gives each mission a

performance rating. The role-playing fantasybegins as soon as the gamer opens the manual.The documentation becomes a true fantasy object.

Similarly, Level-10's Rings Of Saturn manualtakes the guise of briefing papers for a pilotemployed by a corporation exploiting the mineralwealth of the Saturn system. The pilot has beenferrying tourists to view the splendor of Saturn'srings when a distress call comes in a researchship has been disabled and is slowly drifting intothe 'A' series of rings. After a short narrativeplacing the gameplayer into this physical andsocial setting, the rest of the manual briefs him onthe capabilities of his craft and what dangers hemight expect. The keyboard is treated in themanual as if it were an actual command center.

These two manuals represent a type of manualwe're beginning to see more of from computergame manufacturers. The imaginative transfor-mation of the manual is of the most elementarykind, but it can be quite effective in placing thegameplayer directly into the fantasy setting.

Strategic Simulations' manual for ComputerAmbush does something effective with the pro-blem of character sketches. Near the end of themanual the gameplayer will find two "soldierdossiers," one for the U.S. Army and one for theWehrmacht. These dossiers contain profiles ofsoldiers in the squad, including facial sketches.The profiles are well-written, and most impor-tantly give the gameplayer enough of the soldier'sbackground (history, lifestyle, hobbies, etc.) thattheir otherwise abstract attributes (Intelligence,Strength, Dexterity, etc.) take on specific meaning.These dossiers make the characters in ComputerAmbush among the most real figures in computergaming.

The rest of the Computer Ambush documen-tation, although very thorough, underscores theproblems involved in writing a manual for a verycomplex game. As in many boardgame manuals,the long lists of rules, combat formulas, andspecial conditions make for very dull reading. Andthis is perhaps one reason wargame rules andprocedures often require many playing hours tounderstand even at an elementary level. Thesolution is not easy to come by, but I would suggestthat a collection of fake documents containing therules in disguised form such as "The Conduct OfWar" by Baron Reich, "A Hand-to-Hand CombatTraining Manual," or "A Handbook Of Codes FroThe French Campaign" might prove to be partof the solution. Using the fake document formatnot only thrusts the gameplayer immediately intorole-playing, but also produces a manual which ismuch more entertaining, and easier, to read.


25/44

Producing such a manual is not easy whenpoorly executed such fake documents would tendto confuse more than clarify but when wellthought-out the results should be quite impressive.One of the more impressive series of computeradventure manuals is Dunjonquest by Automated

Simulations. These manuals include a basicintroduction to role-playing for those unfamiliarwith the phenomenon, a brief fictional narrative,descriptions of the monsters, and descriptions ofthe rooms and levels within the dungeons.Although simplistic in comparison to the highlydescriptive Dungeons & Dragons manuals, thesemanuals encourage the gameplayer to visualizethe fantasy setting to an extent not found pre-viously in computer gaming. Where graphiclimitations prevent the programmer from illus-trating the scene, these short descriptions prod thegameplayer's imagination into making up thedifference. The Dunjonquest manuals provide ahappy medium between the highly descriptiveand involved D & D manuals which would befar too cumbersome for the average computergame and simple, skeletal manuals which donot engage the player's imagination at all.

When I was asked to write some of the manualsfor the Level-10 series of adventure games thesequestions were very much on my mind. The firsttwo of these games, Dragon Fire (by RodneyNelsen) and Kaves Of Karkhan, involve thesame cast of characters. My goal for these twomanuals was to produce "fantasy packages,"which would encourage the gameplayer's role-playing abilities and use of imagination to a highdegree.The introduction of the Dragon Fire manualdelves into the psychology of role-playing, encour-aging the beginning player to enter the personaof one of the DF characters. This is followed by afull-length short story which brings the characterstogether and explores major aspects of this fantasyworld. I felt that the brief narratives in othermanuals, as rarely as they appear, have a numberof limitations. There really isn't enough room toprovide the specific detail, the gradation in mood,the "color" required to make a fantasy world cometo life. The short story form can also show a lotmore about character than the usually briefcharacter descriptions. Motivations can be ex-plored (an aspect lacking in many computergames). And of course in a short story you canbuild suspense, and involve the reader in thesetting in ways not possible in a narrative frag-ment. What better way to enter a role-playingfantasy universe?

Both Dragon Fire and Kaves include charac-ter profiles and detailed character drawings.

These profiles are meant to make the figures comealive, relating anecdotes concerning their livesbefore this particular adventure.Female warriors are not unheard of inthese regions, but their homelands, theirpurposes, sometimes even their names,have remained secret. Three-finger John,the tavern owner, is convinced they hailfrom wandering "daughter bands" ofamazons who kill their husbands. "Theyeat raw meat and drink the blood fromdying animals," he tells his barmaidwith a wink.Wizards are said to love disguises...Manyhas been the time he's dressed the part ofthe elm, the weeping maiden, the grass-covered hillock, or the deer in flight.You can't find an elf, no matter how hardyou try. They will always find you. Elvesare highly secretive creatures, vanishinghere and reappearing there; they enjoytravel more than most anything.

I also tried to bring some life to the usualstopping places of the adventurer on an importantmission: the ubiquitous weapons shop, armorersand tavern. I was tired of shops with no names, noproprietors, and worst of all no visual picture whichreadily came to mind while the player was visitingthere. At Winsome Willy's Weapons Shop: "Willywill go to any lengths to replenish his diversestock, including raiding the bloody battlefields,ambushing unwary warriors, and has even stoopedto grave robbing upon occasion. Please watchyour head; hand axes, maces and battle axes arehung just above eye level." At Fliver's & Sons,Armorers: "There are no 'sons': Fliver just likesthe way it looks on the sign... Try not to stumbleover the customers piled up outside the entrance."

Level-10's Alkemstone provided a differentkind of challenge. This was essentially a puzzlegame, the clues leading you to a real stone (a glasssphere) buried somewhere in the world, with aprize of $7,500 to the finder (unclaimed as of thiswriting). The task was to create a mythic historyfor the stone. So I tackled this project as I wouldany fiction assignment for The Twilight ZoneMagazine or Weird Tales, attempting to create amythic tale spanning centuries, involving "real"characters in conflict with such primeval forces as"The Spectral Serpent."

A vast creature of stardust andmoonshine...its disturbing and bizarredreams are ultimately responsible formost of the races of people who livetoday. The sun lived within a small

Continued on page 39


26/44


27/44

Congratulations to Richard Fowell and hisrobot Norden+ Winners of CGW's Robot-war Tournament

NORDEN+, ROBOTKILLERby R ichard Fow ell.

ROBOTWAR is MUSE's game for the Applecomputer, in which players write control programsfor robot gladiators. At the UCLA Computer Club,fifteen members have been playing ROBOTWARsince the summer of 1981. As of January 1982,NORDEN+ is one of our top two robots . He is myeighth robot, and the fifth in my current series.

NORDEN+ was designed for multi-robot battles,where avoiding damage is more important thaninflicting it. His intent is to let the others kill eachother, then kill the weakened survivor. To avoiddamage, he moves at top speed in both X and Ydirections. This is fast enough that, even if a robothas him dead in his sights, NORDEN+ is usuallygone by the time the shell gets there. Even robotswith tracking algorithms, such as its arch-rivalSCALE, have difficulty connecting. To date, firingblindly seems to hit him as often as anything.

NORDEN+ fights as well as he runs. His abilityto hit targets while moving at full speed earnedhim the name of the famous World War II "Norden"bombsight. He scans to his right, and when hesees a target, he computes the angle to which hemust swivel the gun so that the shell will hit, basedon how far he will have moved between sightingand firing , and on the range to the target. He useshis knowledge of instruction execution time and asimple arctangent approximation. His offense ifhis weak point, but even in one-on-one combat,NORDEN+ will beat any of the robots about ninetimes out of ten.

Another problem with movement is collisions.To avoid robots, NORDEN+ doesn't start movinguntil damaged, giving corner and edge robots timeto leave the middle. Once moving, he stays awayfrom edges to avoid such robots. After turning, helooks along his new path for robots, and turns ifblocked. While waiting initially, he first checksthe direction he intends to move, then does asimple rotary scan. If he spots a robot, he goes intoa firing routine that includes a simple tracking

algorithm. To avoid walls, NORDEN+ uses theinformation from his scan along his movementpath and dead reckoning.NORDEN+ has lots of room for improvement.Rather than detailing his specific implementa-

tion, I think it is more fruitful to mention some ofthe tools employed. Two primary design consider-ations are general strategy and the specifics of therobot simulator.

In strategy, it is important to concentrate onthings that count. A robot that ignores damagewhile locked on to a robot in multiple combat issilly the others will chew him up while he'sdistracted. In one-on-one, the reverse is probablythe case. Spending a lot of trouble on improvingNORDEN+ performance in his stationary state,where he spends little time, is not as rewarding asimproving his performance while moving, etc. Alltoo often, designers spend a lot of space onfeatures that are rarely exercised.

To implement things, it is frequently importantto know precisely how a program will act. We havedone considerable experimentation with test robotson the testbench and in the arena to learn thesedetails. Here are some. The program seems to go toeach robot in order, execute one "tic" of code, movethe robot, resolve collisions, if any, then go to thenext. A "tic" of code is one expensive instruction (aGOTO, GOSUB, or a reference, direct or indirect,to the "real" world, i.e. referencing AIM, SHOT, X,Y, etc.) or ten successive cheap (other) instructionsreferred to are the compiled or assembled instruc-tions, which, by the way, make a perfectly validsource code. NORDEN+ takes full advantage ofthis by putting as many as nine cheap instruc-tions before each expensive one, since such agroup of instructions takes only one tic. Incident-ally, the X distance moved by a robot in a tic isSPEEDX/50, and similarly for Y. For suchthings as dead reckoning and tracking algorithmsthis information is crucial. It also helps to knowthat the gun barrel cools in 9 tics.


28/44

There are also some typical useful code sequences.If the gun barrel angle is stored in variable A, andthe scan increment in B, the sequenceA + B T O A I M T O A T O R A D A R- R A D A R 0 T O S H O Tis useful. Note the use of A to store the value ofAIM, which is expensive to refer to, and that thetest of RADAR leaves precisely the right value tobe fed to SHOT in the accumulator.

If anyone has a robot that beats NORDEN+, thesample robots, and all other robots they have seenin one-on-one combat consistently (we typicallyuse twenty-game matches as a standard) I wouldbe interested in hearing from them. I may becontacted care of this magazine. If not, I hope thatthis article helps someone produce such a robot.

;ROBOT NORDEN+;BY RICHARD A. FOWELL;WINNER OF COMPUTER;GAMING WORLD'S;ROBOTWAR TOURNAMENT

ADROIT MOD1 TO SPEEDY256 TO U TO V715 TO 0100 TO D81 TO

Y28 -256 TO V-1 135 TO 0

90 TO L99 TO M

X > 128 -256 TO U/ V TO H< 0 90 + 0 TO 0

85 TO N95 TO011 TO INDEX29 / 2 TO J1014 * 4 TO E

SCAN + T + B TO A TOAIM58 TO B T + 13 TO TIF DAMAGE # D GOTO STARTA TO RADAR3 - RADAR TO R < 0 GOTO SCAN

TOSHOTSIMPLE2 TO F850 /R TO

INAMINA - C - C TO A

MOREDAMAGE # D GOTO STARTA C TO A TOAIM TORADAR1 + F TOF > 10 GOTO SCAN3 - RADAR TO P < 0GOTO MORE

F < 0 GOTO INAMINR TO SHOT

GOTO SIMPLESTART-500 TO FU TO SPEEDX TO SPEEDY TO RADARRADAR > 0GOTO MAINE

TURNIF H > 0 - U TO SPEEDXIF H < 0 - V TO V TOSPEEDYINDEX + 1 TO INDEXO - H TO H90 + 0 TO 0 TORADARIF H < 0TO UIF INDEX > 15 11 TO INDEX3 - RADAR TOF > 12 621 + 621

/ F + 0 TO AIM> 12 F TO SHOT> -137 GOTO TURNMAINP

-600 TO BTRACK TO CU TOSPEEDXF > -145 TURN TO C-500 TO F TO T TO DO DATA TO I

MAINOT 29 TO T > -77 GOTO TURN > DAMAGE GOTO TURNT - 29 > B I TO AA TORADART > -99 GOTO TURN < -450 766 TO Z3 - RADAR TO F

< 3 + T > -296 GOTO TURN< 3 GOTO CT -84 GOTO TURN + 50 TO BIF C 0 GOTO TURN

TO TMAINE TO C621 + 621 TO ZDAMAGE TO DGOTO MAINO


29/44

A New Column .. .

THE SILICONCEREBRUMby Bruce Webster

AN INTRODUCTIONArtificial intelligence, games, and computershave been linked together for, well, as long ascomputers have been around...longer, in fact, ifyou count the "chess automatons" (usually worked

by a hidden person) of a few centuries ago. In 1950,only a few years after the advent of the digitalcomputer, Claude Shannon published a paper inScientific American describing an "automaticchess player". Since then, computer game playinghas been a significant area of research in ArtificialIntelligence (AI). Recently, much of that researchhas found commercial application in specializedunits that play chess, backgammon, and reversi.It has been slower, however, in being applied togames for personal computers. There are at leasttwo reasons, among others, for this. First, up untila year or so ago, most personal computers systemsdid not lend themselves to sophisticated softwaredevelopment. Second, there have not been thatmany articles on applying AI techniques to com-puter games (outside of alpha-beta/minimaxtechniques, which are not easily applicable to

most of the comuter games out on the market). Thefirst problem is taking care of itself, but whatabout the second one?

This column, hopefully, will help to take care ofthe second problem. Its purpose is to provide aforum for demonstrating useful AI techniques forcomputer games, with the emphasis on "useful".To do this, Computer Gaming World is going toneed the help of those of you who are actuallyusing such techniques in your games right now. Ifyou are interested, you are invited to write me, c/oCGW, to get information on the format we want tofollow in this column. If you don't feel that you canwrite the column yourself due to time or otherconstraints, you can submit your ideas and we'llwrite the column to present them. The goal is forthe techniques to be understandable and applicable.When necessary and feasible, program outlinesand short sections of code will be used.

Some of you may be wondering just what areasof artificial intelligence are applicable to gaming.The answer is; almost all of them. to give you anidea of some of the areas we're interested in andhow they might apply to your programs, here's apartial list of topics with some ideas as to how theyare used in games:

Making decisions. This is probably the "heart"of computer game playing: choosing the best moveor action. It usually involves two separate problems:generating a set of legal moves, and decidingwhich one is best. The classic technique here isgenerating a game tree of alternating moves (so asto look several moves ahead) and using some sortof evaluation function and alpha-beta/minimaxtechniques to minimize how long you have tosearch to find the best path. Unfortunately, alpha-beta/minimax techniques don't work well in "realtime" games or in games where a given side canmove several or even dozens of pieces each gameturn (e.g., your typical wargame). When you havelots of units to move, how do you decide where tomove them all?

Defining and seeking goals. Typically, eachplayer in a game has some goal or set of goals toaccomplish. When there is a single goal (e.g.,putting your opponent's king in checkmate), thisisn't a problem. But what about when you haveseveral goals to accomplish? How does the programdecide which goal to pursue right now? In somegames, only a few goals out of a given set need tobe reached. How do you decide which goals aremost feasible in a given situation? The programmay also need to be able to break down a goal intoa set of sub-goals, each of which is easily definedand accomplished. Finally, there are situationswhen a given goal is no longer worth the cost of


30/44

achieving it. How does the program recognizethat, and how does it pick a new goal?

Representing the "world". A computer gameis played within its own little "universe", whichmust somehow be represented inside the computer.Modelling a chess board and pieces is not adifficult task. Modelling a typical wargame mapand all the attendant counters is another matterentirely. Adventure and all its many progenyhave generated a lot of commercial work in know-ledge representation; what are some of the cleanerand more powerful techniques in use now?

Talking with the player. This problem istwofold. First, the computer has to somehow conveythe necessary information co

computer gaming world 003

Documents