About the Author
Mass Effect
Final Fantasy X
Batman:Arkham City
Borderlands Series
Weekly Column
Champions Online
World of Warcraft
DM of the Rings
Good Robot
Project Frontier

Grand Theft Chronology

By Shamus
on Sunday Apr 15, 2007
Filed under:
Video Games


I’m writing about Grand Theft Auto this week. I’ve remarked on this series before.

The series turns ten this year. The game has an odd history. Most of the press surrounding the game focuses on the graphic content and outrageous gameplay, but the thing that has amazed me the most about the game is the technology.

The original Grand Theft Auto games looked primitive for their day.
The original Grand Theft Auto games looked primitive for their day.
The original GTA and the sequel GTA II were both top-down scrollers. GTA came out in 1997. The graphics were fairly dated, but it featured some very shocking gameplay (car theft, murder, and rampant destruction) and a sense of humor, not to mention a refreshing open-ended approach. GTA II came out in 1999 and used more or less the same graphics. Neither game was particularly compelling to me. I found them to be amusing but frustrating.

Then in 2001 Rockstar came out with GTA III and more or less conquered the PS2 with it. I still can’t believe the technological jump they took from the 1999 GTA II to the 2001 GTA III. It was like they got hold of some strange alien technology. The game went from being a simplistic, lo-tech excuse to blow up cars to one of the most cutting edge 3d worlds ever. Even six years later I don’t think anyone else has really matched them for spacious, realistic urban landscapes. I don’t know what happened that enabled them to take such a huge leap forward. Did they harvest some of John Carmack’s blood and make themselves a clone? Or an army of clones?

I find this jump to be astounding. They didn’t just come up with one of the most impressive engines to grace the PS2, but they must have invented a bunch of other in-house tools as well. They went from wiggling 2d sprites to motion-captured animations on 3d figures. They went from pictures of cars to realistic cars with interiors, doors, hoods, and trunks that would open and close, individual damage zones, believable physics and probably a dozen other details I’ve overlooked. They went from all-text to quality voice acting. They went from a bland overhead view of the world to a huge 3d city that used one of the most impressive level-of-detail systems I’ve ever seen. GTA III had nothing in common with its predecessor, except for its premise.

Since then they have churned out games at an amazing pace. Five titles in six years. Putting out new GTA titles differs from printing money only in the fact that it’s legal, so I can understand the motivation. I just don’t understand how they’re doing it. Compare this with the hit Half-Life, which had six years between the first title and the sequel, or with Tomb Raider, which managed to release countless sequels of rapidly declining quality. GTA managed to release new games nearly annually while at the same time improving both the game and the technology that drove it. Again, I have no idea how they are pulling this off. The clone army explanation is starting to sound plausible.

What is interesting is that the games are not sequels in a storytelling sense. The chronology of the series is all backwards and sideways, with the timeline jumping backwards and forwards and following intersecting narratives like a Quintin Tarantino movie, only on the scale of a couple of decades. What makes it really crazy is how the various stories overlap. Characters appear in multiple games at appropriate ages. The main character from one game will be a minor character in another.

GTA III: Takes place right around the year 1999.
GTA Vice City: Takes place in “1986”, although it draws from iconic music and television from all over the decade like an 80’s highlight reel.
GTA San Andreas: Takes place in the early 90’s. 1992-ish.
GTA Liberty City Stories: Takes place in (I think) 1996.
GTA Vice City Stories: Takes place in 1984.

I’m currently working on GTA: Liberty City Stories. The protagonist in this one is not very compelling so far. We’ll see how the game itself shapes up as I get a little further into it.

Comments (17)

  1. Jeremy says:

    If you’re interested, there are a number of storyline guides that attempt to outline the stories of each game and how they all interrelate, they’re pretty surprising in how thourough they are.


  2. Robert says:

    Hey, my antispam word was “d20”. Intentional, or delightful coincidence? Who knows.

    They are able to release a game every year for a very simple reason. They have a design and development team which is being managed with reasonable competence. The reason all those other publishers have such slow and/or low-quality releases is that they are being badly managed. Yes, all of ’em. “Reasonable competence” turns out to be a very high bar to hit, for reasons I only partially understand.

    I can’t make an argument from authority, here; I’m not a software development manager. I can only make an argument from experience. I’ve worked on commercial software releases ranging from OS expansions (Microsoft Bob, anyone?) to games (Dreamworks’ original entertainment software line) to databases (SQL Server). Managing a software development cycle is apparently very hard.

    It seems to me that the people who do it well seem to have one set of traits in common: an understanding of the contingent nature of events (X has to happen so that Y can happen or else Z, which is bad, will happen – and nobody can predict when, where or if Q, which screws everything up, will happen). They have an understanding that, literally, nobody has a good ability to forecast how long a particular feature or design element will take*, and so schedules must be dynamic. And finally, they have an ability to tolerate ambiguity and areas of no or limited information in their domain.

    People who lack these traits tend to manage to a schedule that they believe to be a document that controls reality. People with these traits understand that the schedule is a collective best-guess, and that the damn thing will be done when it’s done. Somewhat paradoxically, the latter interpretation tends to lead to cycles that run faster and get everything done in a timely fashion.

    *Purportedly true story: a developer was going on vacation for a few weeks, but knew that his contribution to the upcoming product would be scheduled while he was gone. So he put a tray outside his office holding two six-sided dice and a note reading “To estimate the development time required for any features I am developing, please roll the dice. The total is the number of weeks the feature will take to code and test.” The project manager went ballistic and called him in and said “you can’t do this!” The developer replied, “this is what I’ve been doing for the past five years when people asked me to predict development times, and my estimates have been just as good as anyone else’s.”

  3. Hmmm… The antispam isn’t being very creative today, for some reason.

    Well now, I AM a software development manager at a medium sized IT firm. Your description of what it takes to be successful (the ability to manage the chaos of a creative environment with wildly varying quality of information) is correct as far as it goes. There are a few other things it takes as well, however.

    You need an intimate understanding of the challenges of development. I got mine by actually doing development work for a while before realizing I was better suited to development management. I know others who have gotten theirs by LONG exposure in low level software development management. I’m sure there are other ways to get that particular skill, but the point is without it you can’t tell the difference between a software engineer who is having a legit problem and just needs some time and one who is having an entirely different kind of problem and needs you to call in some assistance.

    You need to communicate with various types of people well enough for them to respect you. If your software engineers respect you and your skills, they’ll give you their honest opinions of things. While that won’t be ‘hard’ data, it will be as accurate as can be generated. If your clients / superiors respect you, they’ll accept that when you say ‘ten days’, you mean ten days, not five days, and not twenty. Of course, the flip side of that is you’ve got to respect your colleagues. The Engineers might not have people skills (some do, some don’t) but they DO have skills. The clients may not have IT skills, but they obviously have some money making talent, and they’ve worked hard to make money with that talent.

    Those are just two things (three if you count mutual respect), I’m sure there are others.

    Thing is, I consider myself minimally competent, but still competent. My Software Engineers have been looking at me like I’m some kind of gift from a kindly deity, as they have had a series of no less than three less than competent people in my position. I like my company, because they tend to shuffle the less than competent sorts off to areas they’re better suited to as quickly as possible, but it’s still an indicator that you need to go through three ‘good on paper, bad in person’ before you find one ‘good on paper, trainable in person’.

  4. Andre says:

    Shamus, we’re all getting the same antispam word… and honestly I don’t think it’s obscured enough to stop a spammer from getting it using OCR.

  5. josh says:

    Clone armies are notorious for their incompetence.

    I don’t think there’s any mileage to be had from the idea that the GTA II team “morphed” into the GTA III team. It simply doesn’t happen. There was most likely enough new blood and talent for the GTA III team that they should be considered totally different entities.

    Anyway, the more interesting question is: what makes the GTA team so good right now, and how can we make more teams like them? I fear that the answer is more along the lines of find the right people (good managers and good developers, a rare combination) and get the hell out of their way. Changing bad teams into good ones, or ineffective people into productive ones, is far more difficult.

  6. Thad says:

    I think it might be interesting to compare when American companies starting using 3D environments vs when Japanese companies were using it. It might be a biased opinion here, but I suspect the Japanese lead the way…


  7. Sauron says:

    If anybody read the comments yesterday, you would note that Shamus only made one CAPTCHA word for now.

  8. evilmrhenry says:

    BTW, GTA and GTA2 were released by Rockstar a while back for free:
    There is a form to fill out, though.

  9. Raka says:

    “one of the most cutting edge 3d worlds ever”

    Wha? Seriously? The graphics and physics were nothing to write home about. Granted, the level-of-detail system was nice. But the models, the physics, the textures… nothing was even remotely pushing the envelope of PS2 games at the time GTA III was released.

    You are completely right about the “spacious, realistic urban landscapes”. But that has a lot less to do with their engine and a lot more to do with a mind-boggling amount of work that’s unimpressive on a technical level, but astounding on a management level. Somehow they found a sweet-spot for their art, design, and implementation teams; an enormous environment was laid out, the artists put simplistic but individual detail into every single component, and the code-monkeys put it all together with a huge number of interactive elements which balanced well but weren’t innovative in and of themselves.

    The gameplay is also impressive throughout. There are better driving games, and better shooting games, and better track-your-progress-as-you-try-to-do-everything-in-the-game-world games (well… maybe not), but the mix was compelling and the naughty aspects just added to the flavor. The enormous number of mini-games and cute little interactive elements tells me that someone in charge was very good at listening to suggestions that had to be coming from all over the team throughout the building process; the fact that they all work well together tells me that someone was even better at controlling feature creep. Add in the soundtrack(s) and multiple coherent (if not profound) storylines and some ever-addictive stat-building in the later games, and their success isn’t hard to explain.

    None of the games ever struck me as a notable technical achievement. All of them struck me as a damn fine use of a giant budget and the storage space of DVD media, and a textbook example of design management done right. The sequels are straight linear progressions from the first game, with minor fleshing out of existing elements and numerous (but individually trivial) additional elements that are a logical progression of others. New environments and themes (all with the same level of attention to detail) are just enough to make each iteration fresh.

    Valve was trying to re-define and the first-person shooter and deliver a salable and innovative new engine with each game, both of which are targets that keep running away while you’re chasing them. Tomb Raider could’ve deteriorated less quickly than it did, but it didn’t have a lot of ways to build on previous success without losing what made the original fun or rehashing it so completely that it was a glorified map-pack.

  10. Shamus says:

    Raka: I agree that the world itself isn’t the most graphically pleasing. (I have other posts coming that nitpick it a bit.) But I really do think their LOD system is remarkable. To date, nobody else has pulled it off, and I’m sure that’s not for lack of wanting to. San Andreas is the most amazing of the bunch. I’m still astounded at how well the roads work: In SA you can fly over the dense part of the city and watch the LOD do its thing on all those buildings and roads, without any load screens. I still can’t come up with how you would approach such a challange given the low memory and glacial (to a graphics engine) DVD access speed. The engine needs to do a tremendous amount of planning ahead. Given the choice between a huge world with a generous view distance and (say) bump-mapping and realtime shadows, I’ll take the big world. Seeing to the horizon shouldn’t be a special effect. :)

    And what you said about Valve was spot-on.

  11. empty_other says:

    GOURANGA! (and not GAAARUNGA!, patrick :) )
    Havent played GTA:London, nor any of the games which is console only. But i love the series. And its always like the gameplay is so… floating. Nice. Clean. Playable. Yes, i am too amazed at their skills in using the 3d-engine and art toghether (in GTA 3

  12. empty_other says:

    …and onwards ), making the game seem more graphically impressive than it really is. Of course, i slightly disliked GTA 2 (but loved the multiplayer in it), and in GTA 1 i was impressed by…. freedom! And a kind of Hollywood movie realism, like a cool action movie.

    But my one most wanted feature in the game: flashing indicators (or AI who “uses the mirrors”). I HATE when i drive past another car, and all of the sudden, most often perfect timed to when its worst, take a turn, straight into my car. I lose control, hits two pedestrians, and crashes into a lightpole. And fail the mission.

    Hmm… i used a letter in my text that made the blog ignore the rest of my post…

  13. Shamus says:

    empty_other: Yes, I really, really hate the way other cars cut you off with nonsense left turns and the like. I even suspect this may be done on purpose, only because it is so common I can’t believe it is simple chance.

  14. Deoxy says:

    They are able to release a game every year for a very simple reason. They have a design and development team which is being managed with reasonable competence. The reason all those other publishers have such slow and/or low-quality releases is that they are being badly managed. Yes, all of “˜em. “Reasonable competence” turns out to be a very high bar to hit, for reasons I only partially understand.


    My personal belief in the hierarchy of management (in ascending order as compared to value to the company – note that most of these values ar NEGATIVE):

    Boneheadedly stupid management
    Criminal management
    Boneheadedly stupid criminal management
    Dumb bad management
    bad management
    dumb management
    mediocre management
    passable management
    somewhat good management
    good management
    NO management
    GREAT management

    I think most companies would benefit from a lot less management; in fact, I think most companies would benefit from having random developers assigned half their time to management (on a rotating basis) over the current system.

    Yes, truly excellent management is a wonderful thing that benefits the company greatly… which tells you what I think of most management.

    “Reasonable competence” turns out to be a very high bar to hit, for reasons I only partially understand.

  15. phlux says:

    I guess I’m less impressed by their technology expansion than Shamus is. Sure they’ve produced 5 titles in 6 years, but they’ve done it with 1 engine, two upgrades, one downgrade and 3 modeled cities. Here’s how I see the timeline.

    Release GTA1. Start work on GTA 3D for the PC. Run out of money developing massive 3d urban environment. Split development team and crank out GTA2 to pay bills. Merge and enlarge development teams. Switch from PC platform to PS2. Release. Conquer globe.

    Massive amount of money means now they can hire all the developers and artists they want. They do so and begin sculpting the much larger environments of Vice City and possibly San Andreas simultaneously.

    Moderate upgrades are done to the GTA3 engine for Vice City to optimize load times, add new gameplay features and further improve graphics. Celebrity voice talent from Ray Liotta and Jenna Jameson create buzz and genuinely improve the immersiveness of the game.

    Work is continued on San Andreas models. Engine is further optimized for the large scale of the terrain and urban environments. Much work is done improving indoor areas. Flight engine from Vice City is upgraded to include air combat and greater subtlty in control as well as increased variety of craft.

    The PSP games are both very similar in terms of engine, and were probably developed simultaneously at the behest of a money-hungry publisher. I’m sure the development team would rather have focused energy on GTA4, but the team was likely split, and a few unfortunate souls had to go back and dumb down the gamplay mechanics, graphics, and make it all fit into a PSP.

    After the PSP thing, possibly the same unfortunate souls then had to port that thing back to the PS2…only from what I hear they did it without upgrading the engine or textures…probably because they were desperate to get back to work on GTA4.

  16. sleepyfoo says:

    it wouldn’t surprise me if they were able to do it using the procedural generation idea that shamus has talked about before.

Leave a Reply

Comments are moderated and may not be posted immediately. Required fields are marked *


Thanks for joining the discussion. Be nice, don't post angry, and enjoy yourself. This is supposed to be fun.

You can enclose spoilers in <strike> tags like so:
<strike>Darth Vader is Luke's father!</strike>

You can make things italics like this:
Can you imagine having Darth Vader as your <i>father</i>?

You can make things bold like this:
I'm <b>very</b> glad Darth Vader isn't my father.

You can make links like this:
I'm reading about <a href="http://en.wikipedia.org/wiki/Darth_Vader">Darth Vader</a> on Wikipedia!

You can quote someone like this:
Darth Vader said <blockquote>Luke, I am your father.</blockquote>