The Guide X: A Tale of the A.I. Trail

by Adrian Hon

The Wrap-up Report

See the newest updates. Lost? Want to read this from the beginning? Go to the 'Before 24th April' page. Want the full table of contents? The reason they're not here is because this isn't really part of the Guide, it's the Wrap-up Report.

Index

Introduction

This Wrap-up Report is intended to be my look at how this AI game went, how enjoyable it was, what it did right (and wrong), and just various personal insights and thought on the game. I'll also be looking at similar web-based 'immersive' games that are currently running, most notably Majestic but also other games that manage to consume our time on a basis equivalent to this game. I'll follow that up with my wish-list for the next web game like this one, and then finish with a few words on the experiences I've had during writing the Guide and some advice for anyone foolish enough to want to try and write something like this in the future.

This isn't supposed to be an exhaustive analysis - if it was, I would have done more research and you'd see a bibliography and list of references at the end. It's just my thoughts on the game, as the author of the Guide and someone who holds a strong interest in online gaming and the Internet.

[From hereon, the AI web game will be referred to as the 'game', except where specified otherwise. Just so you don't get confused]

Game Analysis

It's always difficult to make a good analysis of a game which has created what is arguably a new genre in computer entertainment; there are few points of reference and any comparisons to elements of games in other genres will not necessarily be fair due to the different gaming environment. But these difficulties are only present because this is such an innovative game, so whatever follows, at least it'll be interesting.

Overview

Why is the game so innovative? It's not the first immersive web-based game; we've all seen the success of Massively Multiplayer Online Role Playing Games, as well as First Person Shooters which at least consume up as much time and are arguably just as immersive to some gamers. It certainly isn't the first web-based narrative, what with the numerous online soaps and stories that litter the web - in fact, it's far less interactive than most cutting edge online narratives. And with hard-core puzzler games such as The Stone, the puzzles aren't the best.

No, the reason why this game excels is because it has created something new; a primarily web-based game that isn't afraid to intrude into the 'real world' and is based on the premise that a strong story coupled with high production values and basic levels of interactivity will attract gamers and perhaps most importantly make them work together.

I stress the teamwork aspect because this game is possibly the first that has required its participants to work together in order to advance the game storyline. The other elements of the game should certainly not be overlooked, but if I had to name a single thing about the game that was new, it's the teamwork. I won't go so far as to say that before this game there has never been one that has properly realised the abilities of the Internet, but it is up there.

In this game analysis I'll be examining each of the elements of the game in turn; puzzles, storyline, websites, continuity, multimedia, offline interaction, recommendations and a conclusion.

Puzzles

Surprisingly enough, the most memorable part of the game for most players is the puzzles. The principal purpose of the puzzles in the game was to provide a higher level of interactivity to the game above that of simple web-browsing. Completion of a puzzle essentially allowed access to new game content, almost all of which was necessary to understand and advance the game.

The puzzles within the game can be split into two broad categories; original and routine puzzles. Original puzzles have several characteristics:

It goes without saying that the original puzzles were by far more enjoyable, in that they provided stimulating mental challenges and were generally unique every week. They were sometimes about interesting subjects of which most players would have very little knowledge (e.g. peculiarly named British food, cryptography, mathematics) and offered opportunities for learning. The use of different languages certainly endeared the game to a large international audience and such was the interest in Japan that a large portion of this Guide was translated into Japanese. Evidently a small gesture of writing in another language can go a long way.

It's a tribute to the creativity of the game creators that a simple binary classification such as above does little justice to the puzzles within the game; while there are certainly many routine puzzles which are easily categorized, there is a wide spectrum of original puzzles. Some, like the routine puzzles, were simply text based and perhaps involved word association or spotting details in a photo. Others required offline interaction such as the ARM rally puzzles that involved co-ordination across three different time zones.

There are a few universal characteristics of puzzles that were praised by players. Good puzzles involved a large amount player interaction and took time to complete. However, they could be tackled incrementally and players knew when they were on the correct track. They didn't require multimedia, but offline interaction was a bonus. Also, they didn't immediately appear as puzzles.

Examples of good puzzles include the Statue Security puzzle in which players had to convince Mike Royal to save Dwayne (investigation of Mike Royal and the exchanging of tips between callers was vital and enjoyable) and the epic Enigma puzzle that required the identification of visual clues in photos, the recognition of a Morse code sequence and interpretation of clues into rotor settings. In the latter case, players knew that it was an Enigma code and they knew what they were looking for - the machine settings. They knew where they could find the clues, hence they knew that the puzzle wasn't insurmountable and that the end was near.

It's very interesting to note that the puzzles that weren't obviously puzzles were also lauded. A number of puzzles, such as the RUR-14 puzzle and really any puzzle that involved raw code had players simply greeted by an empty text and they were expected to do something with it; it's not as if people were irritated by these, but they weren't particularly enthused. However, the most mentioned puzzles were those that fitted seamlessly into the game, such as the Kent Whitman PDA puzzle where players had to apply a green filter to the image; it was a very elegant and natural puzzle with a logical but elusive solution. The same applies for 'hidden' puzzles such as the Brutus image/Kate Nei word-number puzzle, which players actively had to discover. In these cases, the sense of achievement on completion of the puzzles is much higher.

[The Mowz puzzles are exempt from these generalisations because they were quite obviously supposed to be puzzles set by a hacker - people viewed them as a personal challenge and tried to get them solved as soon as possible.]

As for offline puzzles (outside of the rallies), such as those that appeared in newspapers and on the TV, these were not especially well received. Perhaps it was due to the limiting nature of paper media that a chess puzzle was chosen for the newspapers but I didn't feel many people were interested by them; neither were they by the confusing words and signs that flashed up on the TV adverts. I did hear that the game creators were considering putting a fake job advert in one newspaper, which I'm sure would have pleased players. The lesson here is simple: a dull or boring puzzle in the 'real world' is just as bad as one online, even if it did cost more to put it there.

You might wonder why the game creators even bothered with the routine puzzles such as the Martin Swinton diary entry puzzle - I certainly did, since they were so simple to solve and proved to be irritating obstacles in some cases. Well, the only real (and perfectly acceptable) solution is that they provide a sense of continuity and personality to the characters within the game. Martin Swinton has a false sense of security which is exploited in the game when he uses his laughably simple entry codes. The Red King, as a matter of principle, encodes everything he writes - not to stop people from reading it, but because he's a hacker.

However, why would Martin protect his diary so poorly when he's an expert at house AI security and shows that he's more than competent in his dealings with Diane Fletcher, especially when he hired her? What would have made more sense is if he had one single hard puzzle to find the location of his diaries, and then once that was broken players would have access to all of his diaries without the puzzle being updated afterwards.

It's a small inconsistency, but they add up.

In general, the puzzles were of quite high quality and the quality remained relatively independent of the media used or even the production values. Offline interactive puzzles such as the rallies and Statue Security were received extremely well (as they should, considering their cost). Puzzles that involved programming, notably the Eliza puzzle, were liable to breaking down and completely infuriated players when they did. While testing might not be feasible in such a limited time environment as this game, there is little point in creating a sophisticated puzzle that doesn't work. There are a few solutions: get better programmers, create the puzzles beforehand so that proper testing can be conducted and cut down on the amount of puzzles that require programming.

The status of the 7000-strong Cloudmakers community as being a mean, lean puzzle-solving machine has been stressed many times but observant players will have noted that 60-80% of all puzzles are solved by the same dozen or so hard-core players. Occasionally they'll need to consult others on a point of obscure information, and occasionally puzzles will be solved by unknowns, but these are exceptions. So, you might ask, why do you keep on saying that puzzles require the efforts of the entire community when they quite evidently do not?

There's a simple answer - in order to sustain this hard-core of puzzle solvers, you need the 7000 other lurkers, speculators and sometime puzzle-solvers. You can't just have the puzzle solvers on their own, and you can't expect every single member of the community to have the time to invest in looking up things on Google or thinking up rotor settings for Enigma machines.

In any case, I suppose it's not only the teamwork that counts, but the illusion of teamwork. When you're a member of the Cloudmakers, and you've made a few speculative posts and suggestions about puzzles, you feel like you're contributing and that you're making a difference, even if you're not. So the difficult puzzles might not require 7000 people, but they do draw people together and provide a common point of reference, e.g. 'You mean they still haven't solved the Mowz Business puzzle?!'

Digression: I suspect that if you plotted a graph of players' involvement and investment of time in puzzles, it'd follow a Gaussian distribution (that's a normal or bell curve to you and me). You have a minority of people who do a near-infinite amount of work and a large number of people who chip in with the occasional suggestion which, who knows, might help the hard workers on their way to solving the puzzles.

Storyline

The storyline of the game was extremely highly regarded by players and was reflected by the depth of speculation and discussion about the game and the personal attachment that players had with game characters (case in point: the condolence letters to Laia about Jeanine's death). There's little that I can say about what makes a story good, but the AI game to me excelled by having numerous different writing styles and set-pieces with strong, identifiable characters.

Perhaps the old adage that you don't miss something until it's gone is not out of place here; very few people commented on the literally vast range of writing styles employed within the game: the diary entries, the cyberpunk ministory of Martin and Diane's first encounter, the magazines that differed wildly in tone (and that a 'real' magazine commended on its perfect writing, in Metropolitan Living Homes' case). It's a shame that it went unappreciated but imagine the outcry there would have been if the game had been written in the same style all the way through. I think the success here can be attributed to a strong principal writer or team that creates the framework of all the different writing that has to be done, and either a wide range of contract writers or a small number of talented writers who can employ different styles.

"Evan Chan was murdered, and we want to find out why," - could there be a better hook into the game? From our first explorations into the websites of Evan Chan and the Salla Family, we were drawn into the storyline quickly yet it didn't limit itself to one thread, and instead developed numerous subplots and conspiracies that were drawn together with remarkably little mess at the end of the game. The story was well conceived and was probably outlined well in advance of the start of the game - which was a good thing.

It was also adamantly non-interactive. Despite what people might tell you, this is without doubt a good thing. Allowing players to determine the course of the story could have been a disaster and ended up in people simply choosing the most 'sexy' paths. Then there are the problems of how it would have been implemented - who gets to determine the path of the story? The 7000 players? The Cloudmakers? The people who solve the puzzles? Are we supposed to take a vote on every time Laia is choosing between staying at home or visiting Dwayne? A good storyline should not be sacrificed for the sake of a buzzword. I could go on about how people want to be entertained in games like this, they don't want to have to make decisions, but I strongly suspect you've heard it all before.

The fact that there were very few fan-created websites that mimicked the game websites should not be taken as evidence that players were not interested in the storyline. As someone who was involved at ground-zero with the issue of fan-created websites, it was decided by most Cloudmakers very early in the game that they would confuse players and especially beginners who might not be able to tell a game site from a fan site. Even worse, it was feared that a fan site might disrupt the flow of the story, something which most people were against.

We know that the game creators would have liked to have seen more fan created content, if only to save them the trouble of making it themselves. We even saw an early effort of theirs that had middling success - the Metropolitan Living Homes Debutante Ball, which was some excellent entries - with the emphasis being on 'some'. Perhaps if the game went on longer, we might even have seen more websites like the nigh-well perfect Shipbrook.com fan site - but there just wasn't enough time...

Websites

...What's more, there wasn't enough topics to write about. A cursory examination of the Trail reveals 36 separate websites that probably hold well over a thousand individual pages of information and images. If a company was mentioned in one website, you could bet that that company would appear with its own website soon enough - which pretty much limited the scope for fan created websites. The problem was, the game authors worked so well that there just wasn't anything that you could create a fan site about.

Why is fan created content so important? It helps out the writers. In a game where the players have an insatiable diet for new content, it's simply not possible for there to be too much content, and there will always be players who are complaining that they are bored and they want more updates. There are few solutions to this problem, and most aren't satisifactory - you could spend more money on writers, or you could perhaps try and squeeze more out of them. In both cases, something has to give - money and/or quality.

It could be argued that you could simply develop more content in advance (entire magazines, for example, could be done in advance), but that merely shifts the problem elsewhere - the content still has to be written and money still has to be spent. The 'only' thing you have solved is the matter of timeliness.

So really, fan created content is the cheap and cheerful answer.

I'm not really being very helpful here, am I? However, I do think I have a partial solution. Simply because the majority of fan created content is not up to the required standard doesn't mean you have to reject it all; there just has to be a screening procedure. In order to have an effective screening procedure, all game-creater endorsed fan created content must reside on websites that are controlled by the game-creators, in order to instill confidence in that fan content.

So far, so good. Of course, you can't have your fan writers finding out about what's happening next in the story, so they can't be allowed to write storyline-critical content. Instead, you have them writing peripheral material. For example, they could write editorial pieces on the Mann Act, or reviews of fictional Gladmech battles.

Inevitably though, some fan writers will be dissatisfied by the lack of control that they hold. At this point it really becomes an issue for the game-creators as to the amount of control they are willing to surrender to the fan writers in order to get 'free' content. What might work well is if you had an entire website that was controlled by fans - fan graphics, fan writing - but it would all have to be vetted by the game-creators beforehand. Whether or not this would be worth the hassle for the game-creators is up to them, but I imagine a standardised and semi-automated process could be used.

(i.e. It's not my problem, it's theirs. I just think up the ideas.)

The websites were generally of a good quality. That's not to say that they all looked good, but they generally all looked the part, so cheap and nasty companies like Electric Toyland would have cheap and nasty websites. In some cases however, the websites were a bit too cheap and nasty - you would have thought that AIT, being the best university in the world, would have had a website more sophisticated than the one it has right now (and less so for BWU). Admittedly most of the sub-standard websites were created early in the game (AIT, BWU, SPCB) but that's not really a valid excuse - there's no reason why they couldn't all have been as good as the websites that we saw later on, such as Cybertronics or 7to1.

But hold on a second here - I'm not advocating the use of Flash. The reason I liked Cybertronics and 7to1 was because they looked good, not because they used Flash - they could have been constructed in HTML (granted, it would have been sophisticated HTML, but even so). In most cases, I dislike Flash - it's not good for the disabled, you can't control font size (which is irritating because usually the text is way too small on Flash sites) and it sucks up bandwidth like there's no end. I know that the guys at Microsoft might all have broadband connections but they should try accessing Cybertronics from a dial-up connection on a slow computer one of these days. It's a nightmare, nothing less.

So. The websites were in most cases suitable but my main criticism is that they invariably looked too bad instead of looking too good.

That's enough with the appearance. The content was pretty much faultless and went a huge way to creating the illusion that the game world was 'real' - apparently useless articles that you could see on AIT and MLH nevertheless were still essential so that you could understand the game universe and really be drawn into it. One piece of advice though: the saying that 'It's not a game site if there isn't a typo' is sadly true, and even though you might argue that misspellings and misquotations are in character, it doesn't convince me. And for God's sake, would you please spell Allen Hobby's name correctly!

Continuity

Setting an update time of Tuesdays seems horribly artificial and ungenerous, but it turned out quite well for a number of reasons.

• It reinforces the community: Every week on Tuesdays, the excitement and tension building up in the IRC chat rooms and on the messageboards was palpable. You could literally sense players straining at the bit to find the first updates and then unleash themselves on the new puzzles. And since the updates were at a regular time, you could always rely on seeing your friends on IRC at that time and chat about the latest speculation. The Tuesday Updates were a common point of reference for all game players, facilitating communication.

• It reduces the strain on players: As much as some players would have liked to have seen constant updates, very few players had the time or energy to keep up with such a frenetic and unpredictable pace. However, since they knew that 90% of the week's updates would be on a Tuesday, they could make time in their schedule to be free and participate in the fun.

• It reduces the strain on the programmers: Okay, I don't know this for sure, but I imagine that having regular bursts of stress is preferable for the programmers to being 'on call' permanently for constant updates. At least this way they could have a rest on Wednesdays and Thursdays, safe in the knowledge that they wouldn't have to make another big push until next Monday.

The Tuesday had a few staples that people could rely on. You always knew that Laia would write an email and a Meditations piece. Martin would always write something for his diary and most weeks the Red King would write something new in his hacker page. Beyond that, there would normally be some news updates at BWU, ARM and AIT but everything else would be seemingly random. This way, casual players could always check out the new diary entry or Laia's Meditations if they wanted new content, while leaving the 'professional' players to scour the websites for new puzzles.

By and large, the game websites changed relatively well with time, adding new news updates and information (e.g. about the Mann Act). Some fared less well, such as MLH, which rarely had new articles.

Sometimes you could feel the pressure of time on you as a player, especially when the Cloudmakers were stuck at a puzzle and had to be helped along by increasingly obvious clues from the game creaters (e.g. Mowz Business puzzle, Cloudmaker Founder puzzle); I imagine that in mid-game it didn't matter too much if people took a while longer with a puzzle, but as you reach the end-game and events are more scripted and time-critical, things get tense.

To be honest, things will be time-critical for as long as you have scripted events such as TV adverts and large real-life events. I suppose the only way to combat this is to keep things as flexible as possible, since you can never rely on the Cloudmakers or indeed any other large group of players to solve things too slowly or quickly.

I feel that we were pretty much going as fast as possible with the main flow of the storyline in this game; any faster and people would have had real difficult keeping track of what was going on. Please note that I say 'main flow of the storyline' - it's entirely possible that peripheral content could have been added constantly, such as magazine articles or news stories about the game universe that weren't related to the puzzles. This would have kept interest up during the times when people were waiting for 'proper' updates that would relate to the main storyline and puzzles, and also wouldn't have scared away people who had less time. Of course, this would require more writers and/or more time, and alas, more cash.

Multimedia

I still remember the time when we found the first video in the game, from the inside of Enrico Basta's house. The quality wasn't perfect, the acting could've been better, but even so the Cloudmakers were completely stunned and loved it. I remember telling my friends, 'This game is so amazing, they even made a video for it,' and they looked suitably impressed (mind you, they could have just been humouring me).

Was it worth it, though? Having videos produced by outside agencies (as this and the others were made) is notoriously pricey and could probably pay for an entire new game magazine issue to be produced. The game-creators didn't have to use a video, they could have just as easily (and cheaply) made a sound recording or a transcript. However, neither of them would have created quite the sensation that the video produced and it's undeniable that seeing our characters 'in the flesh' added to the atmosphere of the game.

The video was a success, partly for the reasons above and partly because it was so novel. I imagine that eventually players would have become used to seeing the videos and would have started griping about the acting and dialogue. Clearly it's best to use multimedia such as videos in moderation, to keep such events special, to keep costs down and also to ease the suffering of the poor souls who don't have broadband connection - they're still the vast majority of players.

The audio in the game was pretty well received and most players agree that the voice actors were very competent; that the audio was all produced in-house by Microsoft (and so was cheap!) makes this even more of a success.

To be honest, there isn't much point having too much audio and video anyway. Put too much in and it doesn't feel like a web-based game any more, it feels too artificial. We don't have that much audio or video on the 'normal' Internet and we don't expect to see any more - or less - in the datasphere.

Offline Interaction

If a random Cloudmaker is asked about the game, you can be sure that seconds into his explanation he'll start talking about how it tied into newspaper adverts, the TV, phone calls in the middle of the night, voicemails, everything, oh my.

Even though the vast majority of the game was based on websites, players tend to exaggerate the role that offline interaction had. Perhaps it's just because it's more memorable or they want people to think that they game is particularly cool. Whatever the explanation is, player loved the offline aspects.

When playing any computer game, you have a feeling of total control - when you turn off the computer, you know that you won't have to worry about the game. You can choose to visit the websites you want, and you know that it doesn't really matter if you don't solve this puzzle by tomorrow since it's not as if anyone is going to die. But the game turns all of this on its head.

Not only is the game partially time-critical, but it also insidiously invades your life. You receive phone calls in the middle of the night from crazed robots, you become emotionally attached to security guards at the Statue of Liberty and for some reason begin to impersonate pro-AI activist leaders, even though you're the wrong sex for the job. You even get dressed up especially for anti-AI rallies and fish pieces of paper out of toilets.

The offline interaction lifts the game above and beyond any other game before it - and it didn't even take much to do it! Voicemails and phone calls aren't expensive or too difficult to produce, and the Statue Security puzzle just involved one guy at a computer with a phone - hell, pay me £10 an hour and I'll answer the phone myself!

True, the rallies must have cost quite a lot to hold - Microsoft would have had to expedite production of all of the puzzles, posters and banners for the three sites, and also write dialogue for the actors. Then there's the hiring costs for the club and promotion of the rallies online. It's difficult and expensive. But the sheer amount of publicity produced made it more than worthwhile - practically every newspaper report after the rallies would mention them, and it's difficult to name an event within the game that excited the players any more. I was three thousand miles away from the nearest rally and I still stayed up until 4:30AM in the morning to follow them.

It wasn't all good. The TV adverts could have been done better and I've already mentioned that I didn't find the newspaper advert chess puzzles to be very exciting. I suppose the moral of the story here is that static puzzles are the same anywhere (by and large - I still think that the job advert in newspaper puzzle would've been great) but true player-game interaction that is other than passive browsing of websites is highly desired by all players.

Recommendations

• More quality control. From faulty puzzles to spelling mistakes, a bit more testing and proof-reading could have improved the game without too much effort. Time-constraints are a problem, but if there's even one error in a puzzle that annoys players, or even worse stops it from being solved, the puzzle has done far more harm than good. In a way, the casual typos and mistakes (e.g. Svetlana Cellini, Allen/Allan Hobby) ended up creating more work for the game-creators.

• Increased fan content. I've already talked at length about this, but working out a system where good fan content can be integrated into the game will please players and creators alike, and wouldn't require anyone going out of character.

• Better use of offline content. Have people hunt through newspapers for personal adverts or jobs - don't waste your money on obvious chess puzzles.

• Puzzles that aren't puzzles. This isn't applicable in all cases, but puzzles that 'make sense' in all cases are preferred by players, as well as puzzles that aren't obviously puzzles. Consider that the Kent Whitman PDA and the Statue Security puzzles are regarded as being ideal puzzles.

• Keep the weekly updates. They brought players together, helped them organise their time and created a common point of reference. Who knows, they might even have helped the creators?

• In general, create more content in advance. I wasn't the first person to say this (in fact, the first people to say this were the creators themselves) but it helps quality control and eases the strain.

• Don't make the storyline interactive. I have nightmares thinking about exactly how difficult it would be to do this. The result of the Mann Act was determined by players - it was perhaps the only truly interactive segment of the game, and even at the end necessitated the filming of two separate video sequences. Imagine if the Mann Act was placed in the middle of the game; while there'd be some redundancy in content, you're making things extremely difficult for yourself with regards to producing enough content in time, even in advance. And the Mann Act isn't a good example - you can't put every interactive decision to a simple vote.

Conclusion

Ask any player of the game whether they thought the game was a success, and they'll tell you that of course it was. The game creators got all of the fundamental elements correct: the puzzles were enjoyable, the storyline was compelling and strong and the presence of 'extra' features such as offline interaction and continuous development made it something special.

The game wasn't perfect though, as reflected by my recommendations above, and there is a great deal of scope for new elements beyond simple improvements; I believe that this game is merely the first of a new genre of games that will fully utilise the world-spanning communication and team-working abilities of the Internet - and grow beyond the Internet to penetrate into the 'real world' with offline interaction.

Could the AI game, in a second incarnation, make any money as a subscriber based game or does it require sponsorship? I'll be considering this point after the next section, in my wish-list.

Back to Index >>

Other Games

With the success of the AI game, it is inevitable that similar games will be produced both by Microsoft and other games companies in the near future. Electronic Art's Majestic project, which is particularly reminiscent of the AI game, has already begun (and is erroneously believed to have been inspired by the AI game by some). Despite the fact that other than Majestic, there are no other truly similar games to the AI game, there are still many lessons to be learned from the more popular online multiplayer games such as MMORPGs and First Person Shooters.

Introduction

It's always dangerous to talk about 'progress' in any subject, but over the last few years there's been a marked increase in the number of people playing online games. Among the vast array of multiplayer online games that exists, the most popular include First Person Shooters such as Quake and Unreal Tournament, and MMORPGs. Some take this increase to mean that single-player games are gradually being superseded by online multiplayer games; to me this seems like a foolish assumption based on the belief that online games are somehow 'superior' and preferred by gamers to single-player games, when in fact they 'merely' offer new options.

The more popular online games are not particularly innovative, and generally involve the transplantation of a well-used and beloved genre into the Internet, with the replacement of NPCs (Non-Playable Characters) with live individuals; so now in Quake, you can kill your friends, and in Everquest, you can talk to them, barter with them and then kill them, instead of having to deal with pesky AI bots.

That's not to say that these online games aren't entertaining or fun due to lack of innovation, or else no-one would play them. What it does mean is that there is room for innovation in wholly new types of online games that truly utilise the communication and teamworking abilities of the Internet. With the development of games like the AI game and The Sims, which emphasises social interaction above all else and replaces the traditional purpose of online games (killing/making money) with increasing your social standing and impressing your neighbours with your immaculately decorated home, we'll be seeing some interesting changes in online gaming.

In this section I'll be examining MMORPGs, First Person Shooters, Miscellaneous Online Games and finally Majestic to see the current state of online gaming and how these games reflect on the future of immersive games such as the AI game.

MMORPGs (Massively Multiplayer Online Role Playing Games)

Much to the irritation of the rest of the online gaming community, MMORPGs tend to monopolise the press (along with FPSs) for Internet gaming. I can't say that this is particularly surprising considering that MMORPGs have two newsworthy defining characteristics: firstly, they have a lot of players. Secondly, they make a lot of money.

Case in point: As of mid-July 2001, Everquest, the most popular North American-based MMORPG, had 658,504 total subscribers and 405,522 active subscribers. The cost of buying the Everquest game is $30 and monthly subscription rates are up to $10. My back of the napkin calculations reveal that Everquest has made nearly $20 million from software sales (this figure is misleading, however, since EA will only see a small part of it with the rest going to stores and distributors), but the real cash comes from the $4 million they make off subscriptions each month.

All of that adds up to $48 million per year. Apart from a very few exceptions, most popular computer games don't come anywhere close to spending that much on their budgets (Final Fantasy VIII is one of those exceptions with a budget of $50 million). While much of that cash will be eaten up by administration costs, bandwidth and server charges, overheads and wages, there'll still be a hefty chunk of change left and what's more, the subscription costs don't even have to pay for the development of sequels - the box sales can pay for that! So really, a $48 million budget per year could probably fund over 100 full time programmers and associated employees.

Yet even Everquest pales in comparison to the Korean-based behemoth of the MMORPG world, Lineage: The Blood Pledge. Depending on which reports you look at, Lineage has from a mere 2 million to 8 million active subscribers paying $15 a month for the privilege of playing (it costs nothing to download the game). Most reports that I've read place total active subscribers at 2.5 million as of the end of 2000, so I wouldn't be surprised if that's increased to 3 million by now - which is larger than the combined subscriber base of all the North American MMORPGs.

This of course means that Lineage makes in a month what Everquest makes in a year, and that its total income tops half a billion dollars annually. It's almost unbelievable (if you believe their figures, that is). Assuming that economies of scale apply to MMORPGs, someone is making a lot of money from Lineage.

The point of throwing all these figures at you is to impress the fact that MMORPGs are a big business and they make serious money. The question is, why are they so popular? Ask any MMORPG enthusiast and they'll tell you about the socialisation and the various adventures they can go on at any time; they'll tell you about the complex social structures and the conventions that they hold in real life, and the economy that underpins the game. All of these are valid points.

However, the real attraction of MMORPGs is their persistency. For $10 to $15 per month, you have effectively unlimited entertainment; if you wanted to, you could spend every waking hour playing an MMORPG (some players practically do) and there's a good chance that you wouldn't get bored, since you could spend your time chatting in a virtual pub or loping around from city to city or gawping at the new monsters of the day. Granted, most players won't spend quite that much time - last year, Everquest players spent an average of four hours per day (still a huge amount) on the game - but however much they play, they have the knowledge that whenever and for however long they choose to log on, there will always be a game waiting for them with new quests and friends waiting.

In comparison, the AI game's weekly updates seem positively ungenerous; I expect that you could probably read the total amount of content produced per week by the game in two to three hours, if you knew where all of it was. It's true that more time will be spent on speculation and solving puzzles, but even that grows boring and repetitive. The point is, you can't just log on to the AI game at any time and expect to find new updates - you know that you have to wait until next Tuesday for that. What's more, it's possible that the cerebral nature of the story and puzzles put some players off. Couple that with the relatively small (well, zero) marketing that the game had, and it's a wonder it managed to get 7000 Cloudmakers.

This doesn't mean that games of AIs ilk are doomed to an eternity of having a small amount of subscribers with big brains. Improved marketing and more content (fuelled by bigger budgets and fan created content) will be sure to attract more players, and as the public gets used to the concept of games such as this and Majestic, player numbers will increase.

We know at the very least that there are several million gamers out there who are willing to invest time and money in multiplayer online games; it's not known whether they'd be interested in games other than hack'n'slash RPGs, but we'll find out soon enough.

But before we move off the subject, there is one more fundamental difference between MMORPGs and the AI game; MMORPGs are based on player-player competition at their heart, no matter what they might preach about teamwork. Being in an MMORPG might involve conducting quests in teams, but at the end of the day it's about the guy with the best equipment, highest experience levels and most money. Why else would there be player-killing? I'm not saying that all MMORPG players are bloodthirsty killers but in comparison to games such as the AI game where teamwork is required and there's such a wide community spirit dedicated to beating the 'bad guys' and finding out who killed Evan Chan, they might as well be.

[I know that Everquest is game in which you have to work together in small groups to succeed, but what about Lineage and Ultima Online? I feel that heartfelt co-operation is the exception rather than the rule in MMORPGs]

Concommitant is the lack of strong storyline in MMORPGs - but that, along with player-player competition isn't a disadvantage of MMORPGs, it's just a difference. Some players enjoy competition and they enjoy the independance that a lack of storyline gives them. Others will enjoy co-operation and the sense of excitement of the unknown that a strong storyline imbues. I'm not sure how you could integrate competition and co-operation into an AI game but it's probably possible (see my wishlist for more on this, through puzzle competitions and rewards).

As for the lack of a storyline, you can look at it in two ways. Obviously any game like the AI game will have to have a strong storyline which pretty much destroys any player independance. However, that only goes for storyline-critical elements; there's still a huge game universe out there that has nothing to do with the storyline that players should be allowed to explore and add to; if they can be encouraged and helped in created fan websites and articles, they'll entertain themselves and others. It makes sense.

MMORPGs are a completely different genre to the AI game genre, but through them we can see that you can't place too high a value on persistency, competition and independance.

First Person Shooters

I don't really want to spend much time talking about FPSs (e.g. Quake 3, Counterstrike, Unreal Tournament) since they're so alien compared to the AI game genre. However, I feel that many of the aspects that make FPSs so popular are shared with MMORPGs - and are absent from the AI game. Namely, the ability to dip into the game at any time (not really persistency but instant-gratification, I suppose). Then there's player-player competition, which probably is more prominent in FPSs than in any other game - you have the constant striving to be the best Quake 3 player, or get the most frags and reach the top of the ladder. That striving can sometimes push players to become far more committed to the game than co-operation ever could.

Miscellaneous Online Games

Not all popular online games fall into the twin categories of FPSs and MMORPGs; The Stone and the upcoming The Sims Online certainly won't.

The Stone is a puzzle-solving game in which players have to purchase a physical stone (well, you can get an e-stone but you still have to pay) and then can participate in the online Stone game where you can solve puzzles and also earn rewards and prizes. The puzzles are quite difficult and very well made, and a large community has grown up around the game to give each other hints and exchange theories about the elusive backstory of the game.

As you might have guessed, there isn't really a very strong storyline behind The Stone; as far as I know there isn't much text about or characters - the best description is to imagine a game that had all of the very best web-only puzzles in the AI game and the plonk them on the game. All of The Stone puzzles are supposed to interlink together into something called the Enigma, which is a 'metapuzzle' and is a grand mystery. Or something like that.

The point is, The Stone seems to make money - it's still around, isn't it? But then, they don't have to worry about developing thousands of web pages of content, and videos, and offline events. I think that what The Stone shows us is that there are thousands of people out there willing to pay to play a solely puzzle-based game; and if there are thousands who'll do that, how many more might be willing to pay for a story-based game with puzzles?

The Sims Online is by far the better known of these two games and is set to become an international hit when it releases, judging from the huge popularity of The Sims (around 4 million copies sold). The Sims Online could be said to be the closest we've gotten to a simulation of 'real life' on the Internet - I'm not talking about role playing games, but 'mundane' reality. In the game you'll be able to build a house in a town of 50,000 and your avatar will be able to decorate it, interact with other avatars, go out to other houses and clubs, run businesses, gamble at the casino, everything. There's no real goal - you could conceivably play it to become the most famous or richest Sim, although I suspect most will play it just for the social interaction and minigames.

Like MMORPGs, The Sims Online will be persistent and players will be able to dip in whenever they want. It will require the purchase of the software and then a monthly subscription charge which is likely to be around $10 to $15. There will be no strong storyline although I'm sure the game creators will occasionally hold events in towns or unleash natural disasters and space monsters; universal events, in other words.

So far, so good. But the thing is, The Sims Online will appeal to a whole new demographic of game players; while 'traditional' players will find the game fun, many other computer users who don't play Quake or Everquest will sign up, as well as a higher percentage of females. This, incidentally, also occurred with the AI game.

The Sims Online will have no puzzles and no storyline; it has no defined goals. What it does have is a shared universe which the players inhabit and can alter to their choosing; in other words, the storyline is sacrificed for maximum malleability of the world and player independance.

Can a strong storyline and independence be reconciled? We've examined this problem before with the MMORPGs - to an extent, yes.

Majestic

The name of Electronic Arts' new flagship online product should be well known to any player of the AI game and indeed anyone with an interest in computer games. Majestic can be said to be the twin game of the AI game, involving a strong storyline and puzzles, with lots of content and a game community. The only difference is, you have to pay to play.

Majestic has been in development by EA for over a year and they've spent about $10 million in developing the game. Launched in June 2001, Majestic will consist of 8 monthly episodes costing $10 each (it might be continued longer if successful). The $10 is in fact for a subscription to the EA Platinum Service which also offers other games.

According to EA, around 100,000 people signed up for the free pilot episode of Majestic. In order to break even, Majestic will require something in the region of 125,000 subscribers on average for each episode, just to get the $10 million back. Okay, that accounting isn't particularly fair considering that some of that $10 million was probably spent on backend technology that is reusable. However, in the Industry Standard's report, EA still believes it will need 110,000 subscribers.

[For some reason, the Industry Standard thinks that Majestic only costs $5 a month. They are wrong - it is $9.99, and it says so on EA's website. But even so, the calculation of 110,000 still works out. Looks like the Industry Standard needs better fact checkers.]

I'll comment on the game itself soon, but looking at these numbers I'm wondering how Majestic will muster up the 110,000 subscribers. Bearing in mind that less than that number signed up for the free pilot episode, and that they will be sure to lose a hefty percentage when cash is required, I'm thinking that they probably only have 60,000 active players at the most. I've said 60,000 because the retention rate on Everquest, for people who bought the game and then decided to keep on subscribing, was 61.58%. That was for the people who had already bought the game - if anything, it's only going to be lower when you're dealing with people who signed up for a free product.

The moral of the story here is that with a game costing $10 million (how on Earth did they manage to spend so much money?) that doesn't seem to be much more complex or wide-ranging than the AI game, you're going to have a real problem finding enough subscribers to make the game economically viable. I think EA made a mistake here - they specifically went out to match and beat their other online title, Ultima Online. But it doesn't work the same way. Majestic shouldn't require a huge backend operation on the scale of UO and it shouldn't require the level of constant administration either. In other words, it shouldn't cost as much. Concommitantly, players shouldn't be charged as much as they are being now, because they don't have constant entertainment.

Much as I would like to hate Majestic (for how could there be anything superior to the AI game?), all the reports that I've seen have been at least middling to positive. It seems that most people playing the game are enjoying themselves; I haven't had the opportunity to interview anyone who is directly playing the game but even with the positive EA spin, things are probably doing fairly well.

But is 'fairly well' good enough for a $10 million venture? And exactly how similar is it to the AI game - is it fair to draw direct comparisons?

To the second question, yes. While Majestic goes about things in a more heavy-handed manner than the AI game (which admittedly may be necessary, since people are paying money and things have to be right), with its special Majestic Web Browser and episodic downloads of audio and video required, it's business as usual, broadly speaking. The game is more individualistic than the AI game as you're expected to solve puzzles yourself and you're emailed, messaged and called personally, but it's the same old web puzzles and surfing for clues. You do have teammates of a sort who will assist you, and there are many Majestic fan sites and official messageboards. However, I don't get the impression that there is the same kind of collective and united effort as there was with the AI game - maybe people prefer it that way, I don't know.

Also, what with the episodic downloads, you know it's a game - as if you didn't already, from the looks of your bank balance. The AI game never properly admitted it was a game. A small distinction, I'll admit.

The storyline of Majestic is quite different than that of the AI game - both involve conspiracies, but the similarities end there. Majestic is set in a contemporary setting, which allows it to get away with linking to contemporary UFO-related sites, thus partially solving the problem of developing enough content. Not that this hasn't caused problems in its own right, but it makes sense for a game set in the present-day to make reference to present-day events and so link to related websites.

Finally - and to me, crucially - the game is only for people in North America, which means that I can't play it, and neither can Western Europe or Asia. In strictly objective terms, it probably makes sense to do this since it saves time on support issues and other backend problems (phone calls, for example). However, to players it seems remarkably insensitive compared to the AI game which didn't limit its scope to North America and had websites and features all over the world, in Japan, Germany and the UK.

Majestic is much like the AI game, with few major differences. What differences there are are mainly due to the commercial nature of the game and the consequences that entail from that, notably the requirement for a markedly more individual-based game.

It's a good game. Is it good enough?

Back to Index >>

Adrian's Wishlist

There are many things that I'd like to see in the sequel to the AI game, whoever makes it. Some of these things are wildly overreaching and would cost large amounts of money. Others are more realistic and wouldn't require much more money, if any - they could even save money. All, in my opinion, would improve the game.

I've split this section up accordingly into three categories; realistic suggestions for the cash-strapped developer, advanced suggestions for those who want to go beyond the AI game and utterly implausible suggestions for the billionaires among us. Some suggestions will straddle categories.

After the suggestions will be a short discussion about the viability of a subscription based AI sequel.

Realistic suggestions for the cash-strapped developer

• Creator-controlled fan content. I can't stress this much more than I already have; the players of these immersive games just can't get enough content, and unless you have unlimited legions of writers and lots of cash (and of course you don't, because you're cash strapped) this is a real problem for you. Either you just tell your players to put up with it - and they might - or you enlist them to create content for you. You ensure that quality is high by vetting all fan created content and keep the storyline under wraps by having them create peripheral material within the game universe.

  Going further on this path, you could have entire magazines and company websites created by fans based on brief in-game descriptions of them by other websites or characters. All would have to be vetted, but this process could be streamlined and a single dedicated community-watcher employee could probably vet several websites per day. Now that's value for money.

• Create content in advance, and conduct quality control. As far as possible, all content should be created in advance. This might end up increasing costs (no!) but in the end it will promote player loyalty by increasing standards. If you're thinking of using complex puzzles that involve programming, test them thoroughly across a range of platforms and browsers.

• Multimedia isn't everything. It costs a lot and in most cases isn't worth the hassle or money, unless you do it in house and can guarantee quality and timeliness (which probably isn't the case for complex multimedia like videos). If you are going to create some videos, and you should since it'll be expected, only produce a small amount but make sure that they're of high quality.

• Don't forget the offline world. Yes, it does cost a lot to do things off the Internet but as has been said before, a little can go a long way. If hyped up properly, even a single ARM-rally type event can have huge dividends and keep players happy for months.

• Disguise puzzles and keep them logical. Make puzzles appear naturally and don't give them simple number-crunching solutions. You might be of the attitude that a puzzle is a puzzle and nothing else matters, but the sense of pride and achievement that players gain when they've discovered and bested a well-crafted puzzle is far more than you'll see with your simple 'Here is a puzzle - here is a text dialogue box - please figure out the answer'. Okay, this isn't applicable in all circumstances but it's important to bear it in mind.

• Storyline is crucial. Even if you have to pay top dollar to get a good author, it'll be worth it. Without a good story, not even the best puzzles can save your game, when your players aren't emotionally involved or interested in your characters and universe. Don't make the storyline interactive, and if you do, keep it to a minimum.

• Not everyone has broadband. By all means create pretty websites. But don't forget that broadband penetration is still pitifully low and that players aren't happy waiting around for huge Flash websites to load, even if they do have fancy animations.

• Somehow, try and recognise individuals. I can't overstate the value of teamwork in this new game genre, but you can't forget the efforts of individuals. Even a simple table of people who solve puzzles the quickest (as was used on the Mowz puzzles) can do wonders for the pride of puzzle-solvers and it's not unreasonable to reward good puzzle-solvers with physical goods. This is a risky course of action though, because as soon as you start rewarding individuals they will start holding back on information to other players and the community may well start crashing down around you. Hey - this is a wishlist, remember?

Advanced suggestions for the brave

• Create a space for players. As a moderator of the Cloudmakers, there wasn't a day that went by without us worrying about bandwidth problems or whether the YahooGroups server was going to crash right at the Tuesday updates. Giving players an empty space in which to discuss the game with the functionality of YahooGroups plus a good messageboard system will immediately put you up several notches in the esteem of your patrons.

  It will cause you grief and it will cost you money; the bandwidth charges will be high and you'll have to make sure that the system stays online without crashing unless you enjoy being insulted by your players. In the end though, it's worth it because not only will your players appreciate it but it will also give you increased access and possibly an element of control over the community. Not too much control though, since dedicated game communities can get touchy about that sort of thing.

• More innovative content. I'm not just saying 'more content' here - you should already have more with the fan created content. I'm talking about content that will stimulate the production of fan created content; create fictional sports such as the Gladiator Robots that will build up a following and thus fan websites. Try and make players interact within the game, not only outside of it.

• Experiment with direct player interaction. Characters could chat on player IRC channels or answer emails from players (the latter is less risky as it stays in the game universe and isn't prone to mistakes). Requires more manpower but would be enjoyable for players.

• More offline interaction. Not just obvious stuff like ARM rallies but Statue Security stunts where people can call an in-game character. Players loved the chance to actually talk to someone who was actually part of the game and have the illusion of directly influencing game events.

• Experiment with offline events. From the reports I've read about the ARM rallies, they weren't that world shattering; a group of players were herded into a room with some jigsaw puzzles and torn up books and had to solve them. Yes, it was enjoyable and innovative, but you can do better. Hire actors to stage set pieces; imagine if in the middle of an ARM rally, a BIOS activist burst in and kidnapped a fake player or held the group hostage and would have to be outwitted - exciting stuff, huh? Even better, there could be more of a tie-in with the online world so that it would be absolutely necessary for online players to do research that would help the offline players in their predicament.

  In a similar theme, don't just limit yourself to static media for taking the game offline. Imagine having a game character phone into a real radio talk show with a fictional problem, or having a game company sponsor a sporting event.

• Consider community-community competition. If you can't work out player-player competition, and providing that you have a large enough player base, it's entirely possible for you to encourage the creation of separate communities that will compete in solving the game; at a stroke, you've introduced the element of competition while retaining co-operation on a good scale. However, traitors flourish on the Internet and it's too easy for people to pass information between different communities - but that's their problem, not yours, and you can be sure that there'll be plenty of hijinks and fun before the game is over - and all of the players' own creation.

Utterly implausible suggestions for the froody

• Ridiculous amounts of offline interaction. You want your game to get in the newspapers and you have more cash to burn than any of the failed dot.coms? Hold offline events every week across the world! Place them so that 70-80% of your players will have the opportunity to reach one, and ensure that each contributes to the story. Place webcams and Internet terminals at each event so that online players can also interact - you could even rig up some kind of teleoperated robot to zoom around the place! (not nearly as difficult as you'd think).

  In addition, contact people via SMS messages - the Europeans will love that.

• Super-size it. Damn the fan-created content! Just hire legions of the best writers in the world and have them churn out content constantly. Increase the length of the story and create concurrent storyline tracks that different player communities could work on separately - occasionally there would be crossovers between the tracks and as the game comes to a close, it would all be wrapped up together.

• Encourage worldwide interaction. Continuing on from the concurrent tracks idea previously, why not have the concurrent tracks take place in different countries and languages? Japan could have its own track, America could have its own, Western Europe has its own; this would increase the audience and make it a properly localised yet also globalised game. Co-ordination is obviously the main problem here as you'd have to have several different teams, ideally scattered across the globe. However, don't make the mistake of forcing players from one country to take part in their local game - if they don't want to, who are you to disagree?

• Sponsor offline community events. Outwardly the community might not like you meddling in their affairs, but if you promoted offline community events such as regular meetings in a bar or club (the game creators would pay to hire the place out or sponsor the event somehow) they'll love you for it, and it'll cement the community further.

Spending money to make money - is AI version 2 economically viable?

I admit that this is probably not the best topic for me to tackle when I don't know how much the game cost in the first place. Still, I think I can make a few broad insights and suggestions based on a few assumptions about how much hardware and manpower costs, and that the cost of Majestic is already known.

Now, clearly AI version 1 was economically viable; it's been branded a success by all and didn't abruptly end before the story was finished. However, AI version 1 wasn't supposed to make money in a direct sense; it was funded by some unholy combination of Microsoft, Warner Brothers and Dreamworks SKG and its goal was to promote the film to punters. Also, since the game was produced by the Microsoft Concept Development Group, Entertainment Business Division, AI version 1 was at least partially a proof-of-concept and/or loss-leader. All of this is fair enough, since after all it is the first game of its kind for Microsoft.

Is it possible that all future versions of the AI game will be funded in the same way? I suppose it's possible but very unlikely. I can see other film studios eyeing up the success of the AI marketing campaign and wondering if they might be able to replicate and improve upon it, and I'm sure they'd be willing to spend a bit of cash on a similar game. Whether that happens remains to be seen.

The real promise of AI version 2 isn't in the film industry though, which we all know now is purportedly smaller than the computer games industry (only in the strictest of senses though, via box office takings). It's in the business of making money, just like its big brother Majestic and all the other MMORPGs and subscription-based online games. We know that people enjoy playing the game, and even that people are willing to pay for it - Majestic must have some subscribers at least.

We'll soon find out if Majestic succeeds or fails through its subscription-based revenue stream, but I'm resigned to the fact that it will not - the game simply cost too much to produce and it may prove to be too novel and alien to game players to work well just yet. Plus, there are the problems of persistency that I keep on mentioning - Majestic doesn't give you the same return that Everquest or FPSs do, when it can keep you waiting for days for a new game update. All of this will impact on its subscriber base.

So here's a thought - aren't there other ways of generating revenue than through subscriptions and single sponsors? Could you perhaps use a multitude of sponsors? Let's have a gedanken: imagine if you let slip in the game that a major character will be talking on a certain radio show that day. Immediately, you've made thousands if not more players tune into that radio station for the entire day. That's got to be worth some cash. What if you have one exclusive radio station that characters say things on periodically? Then your players will be listening to that radio station constantly.

Or what if a character places a personal ad in a newspaper? The newspaper is bought by thousands. What if a huge data file is deposited on a cover-DVD of a computer games magazine by a character (huge, so that people won't be able to pass it around the Internet) - then everyone buys the magazine. Stunts like this can make people money. Couple it with a gold-level subscription service for the hard-core players that offers a slightly superior game experience (don't ask me how it does it, I haven't thought that far ahead yet) and you have the makings of an alternative funding model that isn't totally reliant on subscribers.

I am not advocating the use of adverts within the game - unless it feels like they should be there. Yes, if you have a game set in the present day and you have a fictional magazine, there's no real reason why you couldn't put banner adverts or popups or sponsored links on it. People expect that sort of thing - not that you're likely to make much money from adverts or sponsored links in the current economic climate, but that's besides the point. If a game character had a personal homepage, they might want to put it on Geocities and use a weblogging software like Blogger, both of whom might pay for the good publicity. But on the AI game, set in the future, any sort of banner advert would have stuck out like a sore thumb, so when dealing with in-game adverts, use common sense.

I do think that AI version 2 has a future, but I don't think it has to be through single sponsors like film studios, and it doesn't have to be through subscriptions. The AI game pioneered a new gameplaying genre. There's no reason why it shouldn't also pioneer a new revenue-stream model.

Back to Index >>

MetaGuide

You're still here, after all of that? Congratulations! Now you get to listen to me drone on about my experiences in writing this 600+ KB monster that I've called the Guide. This section will basically consist of ramblings from myself about what writing the Guide was like and some advice for people who might want to do the same for Majestic or AI version 2 - because I sure as hell won't do it again!

Writing the Guide

I've always enjoyed reading well-written walkthroughs for games and general Internet commentaries, and with my discovery of the game there wasn't a better opportunity for me to give it a go myself. When I first started the Guide, I remember it took several days of solid writing just for me to reach the cutting-edge of the game and even by then it was quite a large document, but I had an immense amount of fun doing it and inserting jokes in the commentary. Back in those days, when the Guide was updated as soon as the game was updated, I was possibly one of the most well-informed individuals about the game and I really did enjoy taking part in the speculation.

As the weeks went on and my end of year university exams approached, time was beginning to run short for me and I found I really didn't have the time to spend hours in the IRC chat channel. The Guide updates began to slip and most weeks I wouldn't have too much of an idea of what on Earth was going on. I began to dread making updates to the Guide and often spent more time worrying about how long it would take me than actually on writing it.

The emails I received thanking me for writing the Guide and making suggestions for the next update were unbelievably good at spurring me on and convinced me that it was worth writing it. I wouldn't say that I didn't enjoy writing the Guide at this point (maybe around May/June) but it was feeling like a chore. The funny thing was that it didn't actually take that long to write the Guide - if I got my head down and just wrote it, I could have probably done it on time every week even with the added pressure of revision.

Of course, during the run-up to my exams and the exams themselves I made zero updates. After that, I gave myself a bit of a rest and then I ended up with something like four weeks of updates that had to be added to the Guide. Let's just say that the prospect of writing four times what I was used to made me blanch and then the whole thing turned into a cycle of despair with me constantly putting off updating the Guide and thus the work piling up. It didn't help that I hadn't the slightest clue of what was going on the game and that to my bemusement there appeared to be Londoners with cockney rhyming slang who'd entered the game.

Eventually the Guide was updated in what seemed to a Herculean effort, and then I went off on another break (my summer holidays) and I forget exactly what happened but I recall there was another Herculean effort just before I went off on my trip to Seattle to see Microsoft. After that, the game had ended and I honestly didn't have the inclination to finish off the Guide. However, the constant barrage of emails that I received from loyal readers (who I sadly have mistreated) ensured this last ever update of the Guide.

You might be wondering how my interaction with Microsoft during the game affected my writing the Guide. In a word, it didn't. They never gave away what was to happen next in the game and I had no more idea of the storyline or eventual ending than anyone else. Every so often they might drop in to the IRC channel and complain about long hours and lack of sleep and other inconsequentials, but that was about it really. If you want to know more about it, you should read my trip report to Seattle (link above).

I don't want to give people the impression that I didn't enjoy updating the Guide. I did, especially in the beginning. However, an undertaking like the Guide is not something to be approached lightly - it's not the amount of work that's the problem. I estimate that I spent somewhere in the region of 3-4 hours every week solely on updating the Guide. Some weeks it was less than that - it was rarely more. No, it's not the work but it's the regularity and unpredictability of the content of the updates that's the problem. Every week I'd have to make an update and if I didn't, there'd be two updates to make next week. It came to be intimidating sometimes.

However, the fact that the storyline of the game was excellent helped no end and whenever I began to feel overwhelmed I'd just read Laia's latest Meditations or Martin's diary. I often received offers of help from Cloudmakers which believe me did not go unappreciated but I felt that I'd come this far on my own and I'd make it to the end on my own as well - besides, the Guide is a narrative, not a reference book and it wouldn't have read quite as well with another author joining in halfway through.

I did have this crazy idea that I would also manage to write a weekly column about the game (I still have a long list of article ideas squirreled away somewhere on my computer) but that obviously didn't work too well.

Advice for aspiring Guide-writers

I generally refer to the Guide as a walkthrough to the game, but it's not the same as writing a walkthrough to any old adventure or puzzle game. There are some key differences - you have to make regular updates and you don't have the advantage of hindsight when writing. You're doing everything in realtime. Also, the sheer scale is completely different; there aren't any other computer games (with perhaps the exception of the Final Fantasies) that are quite as complex or long as this one.

Your job as a writer will be made more difficult by the fact that the game will be constantly evolving - the game websites will change from week to week and you will always get emails from people who haven't read through to the end of your guide telling you 'Hey, there's a new news article at XYZ.com' when you've clearly written about it several pages down. But you'll cope.

If the game is anything like the AI game, you'll want to write it as a narrative - there really isn't any other good way of making it readable and accessible. However, you might want to consider using a different method of navigation within your guide - I have the big table of contents at the top of each page but if I had to do all of this again I'd use some kind of javascript popup or DHTML trickery to reduce the space taken up by it.

I'm going to maintain that having a single author is the way to go. It'll give you complete editorial control and you don't have to worry about pleasing anyone else or dividing up what has to be done. It'll also give a pleasant unique tone to your walkthrough and allow you to make jokes and digressions; your readers will identify with you, hopefully, and you should have some fun in the process.

If you do want to split the workload, what I suggest you do is to take alternate updates. There's no point trying to split the work of a single update because you'll end up missing out some things and duplicating others and it'll degenerate into a mess. The best way is for each person to handle entire updates.

Whatever you do, try and keep your updates prompt; it shouldn't take too long to write an update and if you let the work pile up things will only get worse, not better, and you'll just begin to feel bad about yourself. Don't forgot that in writing your guide, you're doing it for yourself and for your readers; people out there really do appreciate what you write. Also, don't be afraid to ask the game community for help on sticking points - there was many a time I'd enter the IRC channel to ask who a certain character was or when we first found a puzzle.

If you bear all of this in mind, you'll do fine. It's not an easy job being the author of what some considered to be the Cloudmaker's Saga, but if you love writing and love the game, it's worth the effort.

Thanks to...

...All the guys at the Microsoft game team, Warner Brothers and Dreamworks SKG for creating such a wonderful game and the movie that made it possible.

...All of the Cloudmakers and other players of the game, without whom the game would not have been possible and the Guide would have been rudderless.

...All of my friends and family who've put up with my incessant babbling about some strange person called Evan Chan who's going to do in 141 years.

There are far too many of you to name individually here, but if you're reading this then you'll know who you are.

Signing off

This is really the last update of the Guide - if I do add to it, it'll only be some small link corrections or addendums. I have had fun playing the game and writing about it, and I hope you've enjoyed reading this Guide. When AI version 2 appears on the Internet, you can be sure that I'll be watching - and also that I won't be writing another Guide. Once is enough :)

If, for some wild reason, you want to read more of my writing you should check out my website, Vavatch Orbital, which has a large collection of embarrassingly old essays of mine and in particular my regularly updated weblog.

That's all folks,

Adrian Hon
Author of the Guide, Cloudmakers co-moderator
2/9/2001

Back to Index >>