Release Date

 By Shamus Feb 7, 2008 35 comments

article_calendar.jpg
It’s the start of February, and yet right now the people at Columbia pictures know that the movie Hancock will hit theaters on July 2nd, five months from now. Given that the advertising campaign has launched, they have most likely known this for some time. This is no big deal. They always know when their movie will be releasing. Once in a long while you’ll hear about a movie which was delayed in production but for the most part people who make movies know how long it will take before they even start.

So why isn’t this true for videogames? Movies have to cope with larger crews, complex scheduling challenges, and logistics problems in moving the actors, equipment, props, costumes, and crew all over the world. A software developer just needs to get their small group of people into the office every day until the game is done. That’s far simpler, yet they can never reliably predict how long it will take. And when they’re wrong, they’re always guilty of underestimating. Games which don’t come out “late” are usually ones that were allowed to develop “until it’s done”, rather than shooting for a fixed date at the start of a project.

Part of the reason is (and if you’re a longtime reader you probably saw this coming) the constantly evolving world of realtime graphics. Movies sometimes will innovate a little or change technology. New cameras. New special effects. The move to digital. But these advances aren’t very common, are limited in scope, and are probably not rolled out in the middle of a project. In computer games this evolution is constant and effects every part of the production pipeline. Imagine how smooth filming a movie would go if every eighteen months there were all new cameras, all new lights, all new editing tools, all new sound equipment, and new ways of producing special effects. Everyone would be not only struggling to learn to use their own tools, but figuring out how they are affected by the changes everyone else is going through. It would be chaos. And that’s not too different from what game developers seem to be going through.

Exacerbating the problem is that they often do this innovation during development. They are often trying to do production and R&D at the same time. It’s easy to know how long it will take to make something. It’s much harder to know how long it will take to invent something.

But as much as I love to blame problems on the endless process of graphics one-upsmanship, that can’t be the whole reason. Certainly there are games which build on a stable, established platform and end up being delayed anyway.

I keep expecting videogames to grow out of this, but there’s no end in sight. If anything, games are more notoriously and regularly late now than they were in the 90′s. Maybe this isn’t just a problem with videogames. Maybe this sort of thing is just pandemic within the software industry in general.

I don’t really mind that games are delayed – I’m not the sort to pick up games on release day – but I do wonder why it happens so often.

201535 comments. Hurry up and add yours before it becomes passé.


  1. Vyolynce says:

    There are a myriad of reasons, I’m sure.

    Translating from Japanese into English is usually a big hurdle, especially for the longer RPGs, but the major difference between movies and games is interactivity. A movie just has to play; a game has to be played, which involves input and reactions to that input. A movie doesn’t need AI or physics engines (depending on its level of CGI), and a movie won’t crash and freeze up thanks to some bug in the code. A movie also doesn’t have to deal with any sort of online connectivity, which I’m sure introduces its own unique set of problems to coding. There’s also the matter of scale: a good movie will last anywhere from 1.5 to 3 hours; a good game should run at least eight or ten if they’re going to charge full price for it at launch.

    Movies only need X number of copies available on release (assuming one large release and not a slow-rolling build up from Limited to Wide like a lot of smaller indie movies) that are going to play on known devices to dozens of people at a time over multiple showings; they can worry about DVD releases later. Games, on the other hand, need Y number of DVD-equivalent copies available at launch that will play on as many as seven or eight different systems (X360, PS3, Wii, DS, PSP, and PC, with PS2 still receiving support and — at least in theory — a Mac version also possible) each one requiring its own unique coding (especially the Wii) if they want the “ported” versions to be worth playing; each of these individual discs/carts will be enjoyed, on average, by one or two people, so just on a per-customer basis the difference between Y and X is probably a couple orders of magnitude. Even exclusive titles need to meet a certain level of production that I’m sure outweighs the number of movie theaters playing a given film on release.

    Frankly, sometimes I’m a little amazed that any games get released “on time”.

  2. houser2112 says:

    I think it has a lot to do with the fact that games are a physical product (thankfully, this appears to be changing). A product that is commonly given as a gift on Spendmas. You can’t really give someone a specific movie as a gift, so there’s less pressure to push a movie out by that time.

  3. Dana says:

    I see it as being (at least partly) something to do with the inherently fragile nature of software design. In contrast, if you have one bad scene (or FRAME!) in a movie, it won’t “crash” the film. And if you make a minor change to one scene of the movie, it likely won’t affect the other scenes much. Film is a lot more “fault-tolerant” and (effectively) “object-oriented” than code.

  4. Phlux says:

    Part of this is money also. In hollywood, if your movie sucks, editing can only fix so much and re-shoots are very expensive. If your lead actor sucks the whole way through, you can’t really do anything about it unless the studio gives you another 50 million to redo your whole shoot with another actor. I’ve never heard of this happening.

    In a game, if you find your fundamental mechanics are flawed, unfun or the game is loaded with bugs, it’s a lot easier to ask for another 5-10% more money and 3 extra months than it is to ask for a do-over on a whole movie.

    Some games do get do-overs, but they are almost always self-financed projects (Duke Nukem Forever pops to mind).

  5. Kleedrac says:

    I also think you’re missing a large factor in the sense that when a movie’s release date is announced they are in post-production. The filming is done. The actors have already gone home. It’s just the director, producers and editors cutting the film together and adding music. Less scheduling involved. More akin to a game being in beta and setting a release date (for which they are still a bit notorious for being late) than from concept to alpha.

  6. Henebry says:

    My bet is that the deciding factor is the movie theaters, with which the studios need to make arrangements almost a year in advance. A movie multiplex has 12 screens to fill, and if it books 3 of them for the upcoming SuperMegaMolluscMan movie and SuperMegaMolluscMan is delayed by even a week there will be hell to pay, because by then it’ll be too late to bring in a decent substitute.

    By contrast, the gaming shelves at WallMart and BestBuy aren’t a fixed commodity. If DoomLegacyGlaiveHammer gets delayed, no one except the fans will complain. No suits for breach of contract, etc.

    In my experience, a hard deadline can make the impossible happen. As Samuel Johnson once observed in regard to his own tendency to procrastination as a writer, “When a man knows he is to be hanged in a fortnight, it concentrates his mind wonderfully.”

    The movie industry has developed an elaborate system of procedures (and the job positions to go with them) dedicated to keeping movies on track once they’ve entered a certain phase of development. I doubt that game companies have anything like this sort of mechanism—though one exception may be games developed for simultaneous release with a feature film.

    One interesting test of this theory of mine might be movies that are released straight to video, a segment of the industry which is more closely parallel to the game industry in its marketing.

  7. rflrob says:

    I’m not sure how much I buy into the Hard Deadline theory Henebry suggested. The theatres must have some flexibility in their lineup to account for the fact that, say, SuperMegaMolluscMan is crap and nobody wants to see it, while some documentary is wildly popular and is still selling out shows a month after its release.

    Late software products have been around almost as long as the entire industry. I believe the classic treatise on the subject is Fred Brooks’ The Mythical Man Month, one of the primary conclusions of which is that adding people to a late product only makes it later.

  8. cypher says:

    I think the main reason is that games are interactive and movies are pretty static. A director can simply look at the storyboard for a shot and then shoot it that way. And the consumer will always see it exactly how the director shot it.

    Games, on the other hand, allow you to make your own decisions. Not only does the game designer have to keep the story going (and make sure that the player actually follows the story), he has to make sure it still works in a consistent way with the consequences the player expects from his actions.

    Take, for example, the classic “maiden in distress”-scenario. Most movies handle that by having the hero shoot himself through hordes of bad guys, then saving the maiden. The director can create a storyboard for the whole movie, and decide how he wants a specific shot to look like, all before actually shooting the movie. So when the movie gets shot, everyone pretty much knows what to do.
    The game designer, on the other hand, might plan for the player to shoot through the bad guys as well. But what if the player decides to sneak through instead, without killing anyone? What if the player shoots the maiden instead of saving her?

    A game designer has to account for all those possibilities, and make sure the game still works (or at least gives an appropriate response). A movie director does not.

  9. Shinjin says:

    Maybe this sort of thing is just pandemic within the software industry in general.

    I’m in the software industry (non-game related) and this is so.

    I’m sure there are some management types out there who will argue with me on this, but there is no practical way that someone can sit down and state

    - I have X hours of design work,
    - Y hours code to write
    - which will generate Z bugs
    - which will take N hours to correct

    Sure, with experience you can have a rough idea. But there are simply too many variables to account for. You can try to budget time for each of these steps, but you will always be wrong.

    And I’m leaving out the beta stage of development here, since it gets even worse if the designers don’t accurately predict what it is the users actually want. (No, user interviews aren’t always involved in the design process.)

  10. Jansolo says:

    Here applies the common problems for…

    …software engineering.

    Where there is few engineering, some of software, and a lot of guys selling the project at half the cost and half the time.

    I presume it doesn’t happen in making movies (nor in rocket science, bridge construction and so on)

  11. Dave says:

    Really, I’d guess it’s because there’s usually a considerable delay — months, at minimum — between the last post-production work on a movie and its opening date. Whereas a video game will often be in stores within weeks of ‘going gold’. So there’s far less of a buffer if development bogs down.

  12. Hal says:

    All I know is that I was supposed to have a copy of Super Smash Bros. Brawl in my hands in 2007.

    *Sigh*

  13. Macguffin says:

    “I think the main reason is that games are interactive and movies are pretty static.”

    I think that’s a huge chunk, along with the relative extra complexity of software development and the relative youth of software and game creation as disciplines. We spend a lot of time reinventing wheels because we are still focused on essentially creating our own movie cameras and lighting rigs, instead of all just buying them. And even when we buy them, we still have to crawl around in the innards to make sure it’s working the way we need it to… because while a movie can choose from (I assume) piles of cameras and lights, AAA game makers only have a handful of models they can seriously consider.

  14. Deoxy says:

    I work in the software industry (not games), and the problems are almost as bad here as well, leading me to believe that graphics issues are only one problem among many (though still significant).

    Problems I see:

    1. Bad specs. (“User input? What’s that?” In fairness, getting users to tell you their needs, as opposed to what they think will fill those needs, is usually really quite hard.) This leads to large delays in the Beta phase, as the large portions of the product have to be re-written because they just do the wrong thing.
    2. Bad management. The biggest and most common sin here is delivery dates promised to the client before so much as mentioning the concept to the developers – I’ve had 6 month projects handed to me with 1-month contractually-obligated delivery dates. Fun…
    3. Bad sales incentives. Salesman who are given no penalty when they promise stuff we can’t deliver. I was on a project where a salesman sold an add-on to our product to the client… an add-on that DID NOT EXIST. Not “wasn’t finished”, not even “wasn’t started” – was not even considered or put on the drawing board yet. THAT was a fun project, oh yeah. Salesman got his bonus for a big sale, the rest of us busted our butts and got chewed out by the client. Good times…

    There’s probably more, but I need to get back to work.

    I don’t see games being immune to any of these.

    Edit: Oh, and what other people have said about movies is right, too: orders of magnitude less complexity, relatively stable tool base, less actual content, etc, etc.

  15. Ingvar says:

    rflrob:

    I don’t know the exact innards of multiplex cinemas outside Sweden, but there they usually have film transport between different projectors, so they can show one individual copy on 2-3 projectors at one time. This requires a bit of set-up and break-down, but is comparatiely quick (compared with getting a new copy delivered from the distributor). That way, if you have SuperDuperMan XIV and HyperBomberBoy XXIII opening on the same week, both estimated to take three screens and SDM14 only fills one, whereas HBB23 is sold out for three weeks solid, you may be able to take one or two SDM screens and turn them into HBB screens.

  16. Robert says:

    I’ve worked extensively in software at little companies (Optika) and big ones (MS). Although it was never my job to be the project manager, for which I daily give thanks to the gods of silicon at a small altar of my own construction, I did see the decision process. The reason why movies can do it and that software can’t is that the movies have a schedule that involves a physical facility. If you say the print will be there on Monday and it isn’t, they don’t just shrug and move the existing titles over a little bit to make the shelf look full, they have an empty theater full of non-revenue-generating seats. And then heads roll, and people get fired.

    Software is a physical product still, at least sometimes, but it’s a flexible one. EB can put more copies of Madden on the shelf if ShamusWorldIV: The Blood Reckoning slips, but you can’t really do that easily at a theater. For one thing, the print you’ve got is already in demand somewhere else – the theaters rent them and when their time is up, it’s up. Since movies are less flexible, the people in the industry are less flexible, too. This creates a requirement for timeliness that just doesn’t come into play on the software side.

    This does mean, by the way, that a software company that wanted to be serious about its ship dates could do so – the Hollywood idiots can manage it, after all. It’s just that the benefits of a slip (saving a million dollars in tech support calls) far outweigh the benefits of hewing tightly to the schedule (pleasing a few gamers who can’t wait until 2/11 for something that should have been out on 2/1, but who are going to buy it either way).

  17. onosson says:

    “In computer games this evolution is constant and effects every part of the production pipeline.”

    I hate to be a grammar nerd, but shouldn’t that be “affects”?

    Nah, actually I like being a grammar nerd :P

  18. WysiWyg says:

    Sounds to me that the common denominator here is people, more importantly people who underestimate the time it takes to create something.

    I’m thinking it can have to do with costs too, if the people with the wallet saw how long it would take to make a piece of program (and thus the cost in “manhours”), they wouldn’t “greenlight” it?

    Oh, and Deoxy; you REALLY need to get yourself a union. If someone tried to do that in Sweden there would be hell to pay. Well, at least it would have been a couple of years back. But that’s a WHOLE other story.

  19. houser2112 says:

    “Oh, and Deoxy; you REALLY need to get yourself a union. If someone tried to do that in Sweden there would be hell to pay. Well, at least it would have been a couple of years back. But that’s a WHOLE other story.”

    Ever read Dilbert? What he describes is standard operating procedure, or at least not uncommon. Also, I’ve never heard of a union for software programmers.

    I’m not sure all of his points apply to games, though, since the revenue streams for games vs other software are different. Games are typically made for consumers, other software is typically made for internal clients or other companies. (and there are actual contracts signed with financial penalties for the developers when things aren’t done on time, and for users when they want to change the specs)

    It’s rare for end users of games to have formal input into a game’s design, and their effect on a company for being late is indirect.

  20. Jacob says:

    In addition to the things other software developers have mentioned, let me point out that games have a lot of variables that movies don’t. A movie knows it is going to be roughly 2 hours long, that theaters will need to receive a specific format, that actors will be played by humans and have to conform to human capabilities and limitations, that certain special effects are possible and certain others can be worked in with CGI, etc. Games, otoh, are wide-open. Genre expectations can narrow scope a little, but even there, you have a lot that you can do and most of it is unexplored territory.

  21. One Word:

    Technology.

    To explain:

    Movies are often “in the can” long before they are released. They are held back, maneuvered, manipulated, marketed, and everything else. Release is based on marketing, not completion. And there HAVE been cases where movies have been delayed… but usually there’s a wide enough margin for between completion and release that they can be off by a bit and still make it.

    Games have very little of this take place. One of the reasons is technology. Video games have been chasing technology around since before Pong. While SOME of this takes place with games, now, too (and most games DO ship “on time” – or close enough to it – it’s only a few high-profile, cutting-edge games that have the notorious delays), for the most part you DO NOT WANT to hold back a game for six months for marketing. The problem is that the cutting-edge tech becomes stale, and the competition catches up.

    The other factor is risk associated with technology. Making a movie is a well-understood science. How you make games changes with EVERY generation, to some extent. This year, it’s all about… I dunno… the physics engine. There are huge risks riding on the back of technology, and people are re-inventing the engines constantly.

    I imagine if they were trying to re-create the movie theaters, camera technology, and experimenting with new ideas every single year in film, you’d get a lot more delays, too.

  22. Mark says:

    I think a couple of people have touched on the fact that the release dates of movies are easier to keep to because they set the dates when the film is in post-production.

    Also worth considering is that film releases are dependent on a limited resource. There are only so many theaters, and when you’re planning a release of your film, you want to find a time when there’s not a lot of competition. So Hancock is set for July 2, but I’m betting that it was planned for that date more because the summer is prime spending time for audiences and the studio thinks it can get the biggest bang for their buck. Indeed, you see all sorts of small delays or advances based on that sort of thing. It’s only a week or two, usually, so it’s not all that noticeable… Anyway, the film can be in the can long before the release dates for this reason.

    Software, on the other hand, is notoriously difficult to guage, especially when you’re starting out on the project… other comments have hit this subject well…

  23. Nathan says:

    The main difference between movies and videogames in terms of release dates is basically just a matter of the difference of when the marketing starts.

    In movies, marketing starts up and release dates get named when the movie is pretty much complete already. A movie can be fully ready for showing months ahead of the release date.

    In the videogame market, a game will be hyped, marketed, and given a release date while it is still in development and there are still potential hurdles and problems facing the game. Once the game is complete though, it will probably be on store shelves within a few weeks at most.

    Right now, comparing the ability of the two mediums to stick to release dates is the same as comparing apples to oranges. A lot of the factors you are guessing at are irrelevant, actually. However, there are a lot of people in the videogame business who are arguing that videogames should move over to the same model as movies, where the hype for a game is only started once the game has already gone gold, and a release date is made only then. Of course, moving over to that system from the current system could be tricky.

  24. Taellosse says:

    I would imagine there’s another significant factor, and that’s simply the length of the respective industries’ existence: the film industry has been around for the better part of a century, and the gaming industry as currently conceived is not even 20 years old. Hollywood has had the time to refine it’s production process over the course of several generations by now, with, as you say, relatively incremental improvements to techniques, whereas Silicon Valley is still trying to figure out how to make the process work efficiently. While the technology involved has gone through dozens of generations, the people haven’t yet–the organizational systems are still being developed.

  25. Don’t forget that they usually don’t advertise a release date for a movie until ALL of the filming is done (or for a book until it’s actually written). IIRC release dates for computer games are usually set LONG before the game itself is finished.

  26. Deoxy says:

    “writing” software is really more akin to construction (that’s one of the best analogies, anyway – I use it all the time) – how often are construction projects behind schedule? About as often as not.

  27. Dev Null says:

    I’ve got to go with Dana and the fragility of the medium on this one, though obviously a lot of good points being made. But when it comes down to 2 weeks before you’re meant to have the final edit done, and you just can’t make that scene work? Cut the damn scene, or print it with the flawed version; either way its out the door on time. With software, if its two weeks to release and your graphics engine crashes every 5 minutes, what’re you going to do? Cut the graphics engine?

    Of course, that comes back to Shamus’ point about simultaneous development of medium and message. Games built on existing engines _should_ be able to make it out the door with a little more reliability… anyone know if thats actually true? It seems like I’m seeing more-and-more engine re-use in the industry, and I would hope that trend would continue. Seems dumb to reinvent the camera for every film.

  28. Burning says:

    My software experience is not with new products but updates of existing products. That being said, it is not true that it is impossible to have a software project keep to a hard deadline. It does however, require good management, organization, and planning. Set multiple clearly defined interim deadlines. Limit the projects each programmer works on at once. Limit the features you commit to in advance. Keep open lines of communication between all people involved in any stage of the development process. And be prepared to say “this feature won’t be ready on time. Leave it out.”

    Now I’m not dealing with entertainment software. I acknowledge that introduces extra complications. However, I have seen that a well run project can produce a stable, useful, usable application on a fixed schedule.

  29. Nothing says:

    I’m with Deoxy about why software in general tends to be behind schedule, but the video games industry has several things going for it that the rest of the software industry doesn’t.
    1.The changing standards- (i.e. ‘technology’ which we’ve already talked about enough.)
    2.Competition- Competition tends to affect gaming companies more. If a website has a feature it’s competitors don’t then it’s competitors can just add the feature. (The exception being something like Google’s search algorithm.) While if a game has a significantly better graphics engine the companies competitors may have to throw out the engine they have and start over or take a marketing hit.
    3.The people- (For this part I mean the EAs of the industry and not the Blizzards.) The gaming industry often hires young coders who they can overwork burn out and then replace cheaply rather than keeping their experienced coders who will bring out better code faster but at a higher salary per worker. I think this may be one of the biggest factors to differentiate the gaming industry from the software industry at large. I don’t mean that all the gaming companies do this but this is why companies like Blizzard stand out.

  30. Davesnot says:

    Movies have a much stronger relationship between producer, director and .. for lack of a better word.. publisher… hell.. they are done, for the most part, with most movies long before release date.. which allows them to maneuver their product to compete with the titles they want to compete with at the time they want to compete..

    Games seem much more like a rich guy telling someone to mow his lawn and have it done by noon… the guy says, “ok” and then realizes the lawn is a sprawling 18-acre estate.

    Plus.. a movie has (usually) one ending.. always one path to that ending.. that makes it much much smaller in scope.. and it takes very little time to get Brad Pitt to say, “Hi there” than it does to create a Brad Pytt and make him look real.. and talk.. and make his hair move right.. and then make him move right.. and move when you want.. but not move if you .. cripes.. all you have to do with the real Brad is say.. walk to your mark.. say your damn line.. lets do it 100 times… still takes less time.

  31. ngthagg says:

    Shamus, I don’t buy the “innovation during development” idea. If that were true, I think we’d see a lot more delays on special effects laden movies. But a company like Pixar, for example, can tell you exactly what month their next movie is coming out at least a year in advance.

    My best guess is that movies are simply much more traditional. Movies are released on time because delaying a movie just isn’t done. On the other hand, we almost expect a game to be delayed. If it isn’t, we wonder what’s wrong with it.

    I also suspect that the movie industry is helped by the easy breakdown of tasks. Scenes, shots, takes, etc. make for very straightforward scheduling. Games lack this kind of compartmentability.

  32. Tirgayon says:

    Don’t forget that they’ve been making movies for more than a century. After 115 years an industry starts to have “institutional experience” they can draw on.

    Games have been made for just about 30 years, and most of that it has been with continuously changing technology. Making a game for a PS3 is a bit different than making one for the Atari 2600.

    In fact writing code for the PS3 is pretty different from writing code for the PS2.

    Film makers are continually inventing new ways to create imagery, but usually we invent the various techniques before we start making movies that require them. In the rare circumstances when a script calls for something that can’t be done, the DP, VFX Super and TD’s have years to work on them from when the issue is spotted in pre-production.

  33. guy says:

    remember, the question is not, “why do games take longer to make than movies?” it’s, “why are movie release dates more predictable?” the fact that movie release dates are made in post-production doesn’t account for stuff like the Lord of The Rings trilogy, in which they knew the release date for the later movies before major filming had started on them. mind, the fact that a major redo in post-production is pretty much impossible does matter.

    @burning

    wow, that sounds like the plan for… EVERY SINGLE SOFTWARE PROJECT EVER! you usually just miss the interim deadlines, or accept feature creep, or can’t even do what you want the project to do with the budget and time you have

  34. Bizarre says:

    I’m personally one who isn’t all that bothered by late games.

    I can name 3 companies off the top of my head who routinely show they aren’t afraid to delay a game – Nintendo, Maxis and Blizzard.

    I can also name 3 companies who make what I consider some of the best games in the industry – Nintendo, Maxis and Blizzard.

    Games have fewer consequences for being late (is anyone NOT buying Smash Bros. or Spore because of the delay who wasn’t already not planning to?) and so they can be kept back until they are polished and working correctly. Particularly with console games, which are only now becoming feasible to patch.

  35. FakeFrenchie says:

    Maybe it’s because riding herd on software programmers is a lot like herding cats. ;-)

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!