Why do game companies miss release dates?

[Bee]

I was thinking about this today because I have a deadline tomorrow for one of my projects and I was just stamping my approval. I've worked for years with deadlines that often lie months and years ahead and I can only remember missing 1 deadline and that was by one day because of printing problems. Don't get me wrong, I couldn't care less if a game comes out today or two weeks from now. After all, it's just a game and has very little impact on my life. But I'm just trying to figure out why game design is different than other types of design. Problems arise in both types of design, but a smart manager plans for such things and figures in contingencies. I generally have my projects done before the deadline and then tweak things to my satisfaction in the last couple weeks. Sure, I've known other managers who wait until the last second to try and make major changes, but in my experience they are rare and often don't last long anyplace I've worked. When developing schedules it's important to set realistic deadlines and know you have the abilities to meet them. I don't know, maybe I'm missing something that makes game design different...

[cthomer5000]

My best guess would be: they're setting unrealistic release dates.

You would think after a couple decades of this that people would be able to say to themselves "Well, I think we'll be done in september, but we've always run 2 months late. Let's set the release date as early November."

I'd say the reason it happens so often is there is really no consequence for missing the deadline. It's just more time to hype the game.

[cody8200]

Bee, think about it like this. Imagine every single person that is reading your report or whatever else your making originally read a different language. Not a very good analogy but the best I can do 2 minutes after waking up. There are so many different computer configurations...making a game work for all of them is very difficult and is one of the major hurdles of game design.

Point two is that you said yourself you tweak your projects till 'your' satisfied. Well thats not what a game is about. It's about the customer being satisfied. If by and large they want more features then you have to add them. If they want something to work different then you should do it. This is especially true of text sims where the target audience is much smaller.

[mckerney]

Release dates are set by marketing, programmers make the games, programming is an imperfect science.

[Bee]

Originally posted by cody8200
Bee, think about it like this. Imagine every single person that is reading your report or whatever else your making originally read a different language. Not a very good analogy but the best I can do 2 minutes after waking up. There are so many different computer configurations...making a game work for all of them is very difficult and is one of the major hurdles of game design.

Point two is that you said yourself you tweak your projects till 'your' satisfied. Well thats not what a game is about. It's about the customer being satisfied. If by and large they want more features then you have to add them. If they want something to work different then you should do it. This is especially true of text sims where the target audience is much smaller.

I've done projects that were multilingual and still got them done on time. You figure in time for translations, etc. It's not like computer game companies are "surprised" everytime they release a game that they wil have to make the game compatible with various computer set ups.

I have clients as well and when I say to my satisfaction, that is saying that it is to my satisfaction that the client is getting what he is expecting. Again, I usually figure in contingency time to add things that the client wants during the design process.

I just don't see either of these things being specific to the gaming industry or explaining why deadlines can't be met.

[Bee]

Originally posted by cthomer5000
My best guess would be: they're setting unrealistic release dates.

You would think after a couple decades of this that people would be able to say to themselves "Well, I think we'll be done in september, but we've always run 2 months late. Let's set the release date as early November."

I'd say the reason it happens so often is there is really no consequence for missing the deadline. It's just more time to hype the game.

Now this is a possibility. In my business, if you continually miss deadlines you are fired. In the computer industry it's become normal to miss deadlines and there's really little consequence other than some complaining from the customers (who still will buy the game when it's released).

[Joe Canadian]

With computer games, bugs can pop up and suprise you. Then you have to fix the bug and test it all over again... as someone stated above programming is an imperfect science. No matter how hard you try to meet a realistic release date, things can pop-up at the last minute and delay the release.

[Marc Vaughan]

Basically a game isn't a straight forward definition in a similar manner to business software, I developed Flight Simulators (Visuals and physics models) for around 8 years before I went into the games industry.

In the Flight Sim industry I had a very good reputation for hitting deadlines, I'd like to think I still do in the games industry - but the percentage I hit is lower.

This is simply because with Flight Sims you have a strict criteria, hit that and you're 'done'. In the games industry as well as completing the game it has to be 'playable' which isn't something you can put down on paper - its a 'gut feeling' that a games developer has and requires a lot of tuning and tweaking until you're satisfied ... this can take time.

I'd prefer one of my games to slip a deadline and be a 'great' game when its released rather than be released on time and be 'ok' ...

[Marc Vaughan]

Originally posted by Joe Canadian
With computer games, bugs can pop up and suprise you. Then you have to fix the bug and test it all over again... as someone stated above programming is an imperfect science. No matter how hard you try to meet a realistic release date, things can pop-up at the last minute and delay the release.

Thats not really the case - there is generally a lot riding on deadlines for release as shop-space and advertising is generally booked for the expected dates, miss that and it can cost a company a lot of money.

That being said - if your game does well then you're generally forgiven ...

[cody8200]

Bee you may have done multilingual stuff but they werent for a 1000 different languages. Of course there are even more than a 1000 different pc configurations that may or may not work as well with your game. I find that testing usually makes a game delayed.

[Bee]

Originally posted by Joe Canadian
With computer games, bugs can pop up and suprise you. Then you have to fix the bug and test it all over again... as someone stated above programming is an imperfect science. No matter how hard you try to meet a realistic release date, things can pop-up at the last minute and delay the release.

But again, you can have the same thing in other types of design. If I'm designing an airport and suddenly find out the size of an air conditioning unit they are going to use has doubled, there's quite a bit of redesign that might be required for that section of the airport to allow it to fit. These are contingencies that you allow for when developing your schedule. Your schedule should be set up so that nothing can "pop up at the last minute" because at that point anything that would cause major redesign should already be completed. The finishing touches should be minor things (in my work text fonts, text sizes, line weights, adjusting colors, etc), the major things should be complete.

[Bee]

Originally posted by cody8200
Bee you may have done multilingual stuff but they werent for a 1000 different languages. Of course there are even more than a 1000 different pc configurations that may or may not work as well with your game. I find that testing usually makes a game delayed.

But my point is that's not an unknown factor. They know it's part of the process and sufficient time should be alloted for it.

[Bee]

Originally posted by Marc Vaughan

I'd prefer one of my games to slip a deadline and be a 'great' game when its released rather than be released on time and be 'ok' ...

I think this is also a good point. In most other types of design, it's either a good design or bad design. With gaming there's a wider range of reaction you can get from your client (or customers in this case), so there may be cases where it would be worth delaying a "good design" to make it a "great design".

Edit: But I don't see this as a reason for the developer not to have a "good design" at the initial release date, which is often the case.

[cody8200]

How much time is sufficient 'exactly' for testing a game. Should they make sure they'll be done with the game no matter what? Say they give it 6 months for beta testing? They would be done probably no matter what. Should they then release the game right away? If so they would have had no idea about when to use money on advertising. Also for a small company it;s important to have word of mouth for months in advance with a game. Large companies developers' have hardly any control over when their product is released unless your a huge name in the industry...aka Warren Spector. Large companies like Activision will say...get it done or don't get it done but were going to release it. It's happened before and it will happen again.

[Marc Vaughan]

But my point is that's not an unknown factor. They know it's part of the process and sufficient time should be alloted for it.

Time is allocated to this - however sometimes unforseen circumstances can arise, for instance the release of CM3 was done a few weeks after the release of a major new operating system (one of the windows ones ... ME I think if I remember right) ... we'd managed to obtain a beta of it from Microsoft so we had tested against that, but there were a number of last minute issues which appeared because of changes in the final ME release.

Similar things can happen with regards to video cards and other hardware which is beyond a developers control.

SI avoid as much of this as possible by going through 'safe' methods for accessing hardware (generally we try to mimic what Windows itself does as this is the safest way to do things, if windows can't do it then its likely they'll never get to launch our game ;) ).

This is the reason that CM4/CM03-04 use the GDI for graphics rather than DirectX or another lower level route ...

[cody8200]

Also in the software industry their is an overall acceptance by users that patching software later is ok and normal. Since this will probably not be changed, developers do not feel too much pressure to get everything perfect the first time through.

[Marc Vaughan]

Originally posted by Bee
I Edit: But I don't see this as a reason for the developer not to have a "good design" at the initial release date, which is often the case.

I agree and we spend a large amount of time at the outset of our projects designing modules - but we are pragmatic and realise that a lot of the fine detail is liable to alteration if it doesn't work as far as gameplay goes ...

[Bee]

Marc,

Just to make the statement, I think you are one of the very best at hitting release dates and when you don't you always have a legitimate reason for the delay. I think it's great you have that approach, but I guess my point was it's unusual in the industry as a whole.

[Bee]

Originally posted by Marc Vaughan
Time is allocated to this - however sometimes unforseen circumstances can arise, for instance the release of CM3 was done a few weeks after the release of a major new operating system (one of the windows ones ... ME I think if I remember right) ... we'd managed to obtain a beta of it from Microsoft so we had tested against that, but there were a number of last minute issues which appeared because of changes in the final ME release.

Similar things can happen with regards to video cards and other hardware which is beyond a developers control.

SI avoid as much of this as possible by going through 'safe' methods for accessing hardware (generally we try to mimic what Windows itself does as this is the safest way to do things, if windows can't do it then its likely they'll never get to launch our game ;) ).

This is the reason that CM4/CM03-04 use the GDI for graphics rather than DirectX or another lower level route ...

I would compare this to my printing problem. It's something outside your control and delays for reasons like this will happen occasionally. I can definitely understand these types of delays.

[Bee]

Originally posted by cody8200
Also in the software industry their is an overall acceptance by users that patching software later is ok and normal. Since this will probably not be changed, developers do not feel too much pressure to get everything perfect the first time through.

I agree. This goes along with cthomer's comment about there not being any significant consequences in the gaming industry for missed release dates.

[Bee]

BTW, I also want to make the statement that this thread wasn't directed at .400 studios because of their delay of TPF. I think that delay falls into the category of a "good design" that has been delayed to try for a "great design". From everything I understand, they could have made the release date but chose to try to release a great game and delayed the release. Again, to me that's understandable.

[Killebrew]

Originally posted by mckerney
Release dates are set by marketing, programmers make the games, programming is an imperfect science.
Yup. You always want the job to take as little time as possible, not because fans are desperate for a new application but because the longer it takes to make the more it costs to make.

Most software is developed by a team of people since the job (in this day & age) is usually far too much for one programmer. The lead designer/developer will provide the initial ballpark estimate in a brief that goes to the non-tech management. If the developer requests 5 years to build the best damn curling sim the world has ever seen, company management will say cut game features until devel time is 6-8 months or get the hell outta' here. Once the brief is passed the design doc is written with much more elaborate time estimates and the expected tasks required to complete the project.

The developing team is usually lead by the designer/developer and a team manager who assigns (or offers out) the various programming/graphics tasks as dictated by the lead developer.
Each team member does an estimate of the amount of time they need to complete their assigned tasks and tasks are often divided up as needed if some programmer time estimates are higher than others. Each few days or week (depends on timeline) new time estimates are handed in so the developer and team manager always know the current status of the project. Here is where lots of stuff can go wrong, as more inexperienced programmers will have a much tougher time accurately estimating how much more time is required to complete their tasks. If even one programmer is way off in their estimates that can set devel time back. Note that there is a pressure to finish tasks at deadlines so some programmers do not like to admit they are behind.

The last step before release is testing and here is probably the main culprit for genuinely unexpected release delays. Some amount of testing goes on throughout the development process but there are some things you can't test until all the parts of the software are enabled. Some unforseen bugs can require major code rewrites and delays of many weeks or even months. On some software projects major errors found in final testing can be enough to make the software unreleasable (hmm, is that what happened to Black Ice Hockey?).

Small time developers can skip some of this stuff, they do not have the finances to hire a team to help them and their goals are also less time consuming than say coding EA's NHL 2006 on your own and in your free time. That said (grr), small developers must still follow some of the same logic about release times, and in .400's case you still have a developer interacting with management in a push/pull for time.

So for Bee, game companies miss release dates due to incorrect time estimates by the programmer(s) and unexpected bugs in testing. For the recent TPF delay it appears as if the developer was pressured into adding features at the last moment, generally an unwise thing to do but sometimes necessary if the original design scope was short sighted.

[TroyF]

Bee,

Airports don't get delayed? (people in Denver may argue with that one) :)

To act as though delays don't happen in the business world is a gigantic misnomer. They may not happen where you are, but you would be the exception to the rule. I've worked at a computer hardware company where delayed release dates were the norm. I worked on a 5 million dollar on-line newspaper account where it missed the launch deadline SEVEN times. (none of the delays were caused by me thankfully) While there was some heated conversations between everyone setting this account up, not a single person lost their job over it.

I've seen countless projects delayed or scrapped completely. Of course, many of these are behind the scenes where the end user doesn't know what is happening.

The computer game designer doesn't have that luxury. Everyday the product is delayed is more pressure, both from the public and the marketers who want the product released YESTERDAY.

TroyF

[Bee]

Originally posted by TroyF
Bee,

Airports don't get delayed? (people in Denver may argue with that one) :)

To act as though delays don't happen in the business world is a gigantic misnomer. They may not happen where you are, but you would be the exception to the rule. I've worked at a computer hardware company where delayed release dates were the norm. I worked on a 5 million dollar on-line newspaper account where it missed the launch deadline SEVEN times. (none of the delays were caused by me thankfully) While there was some heated conversations between everyone setting this account up, not a single person lost their job over it.

I've seen countless projects delayed or scrapped completely. Of course, many of these are behind the scenes where the end user doesn't know what is happening.

The computer game designer doesn't have that luxury. Everyday the product is delayed is more pressure, both from the public and the marketers who want the product released YESTERDAY.

TroyF

Please don't take my comments out of context. I was speaking specifically of other design businesses which actually make a good comparison to game development, not the business world as a whole. I've also worked construction and there are too many factors that enter in that are outside your hands to consistently meet deadlines (anything from weather to onsite accidents). But in the design world that I've worked in for 15+ years, it is relatively uncommon to not meet deadlines and the results are usually very costly to your company in liquidated damages when it does happen.

[SackAttack]

All of the above are valid points, and likely the main reasons behind delays, but I imagine some of it has to do with surrounding games, as well.

Think of it from a Hollywood perspective. You know how you'll see major films get the fuck out of the way of a 'giant' release like Star Wars, or Lord of the Rings, because they know they're going to get swamped? Give 'em, y'know, a berth of a week or two on either end?

I suspect the games industry is sometimes the same way. If you have a game that would ordinarily get a lot of attention and sell really well, but there just happens to be a half-dozen major releases during the same week, you might push the release back a little bit, maybe into a slower period, so that the public will be more willing to snap your game up by the cartload.

Heck, Phantasy Star Online on the Xbox was finished for quite some time before it ever saw the light of day. Microsoft, as the publisher, sat on the game until they were ready to ship it to retail - not so much, I suspect, because of bugs or competing software, but rather just because they didn't feel the time was right. That happens sometimes too.

[Bee]

That's an angle I never thought of Sack. Good point IMO.

[Buccaneer]

Josh: The best recent example is "Master and Commander", I think. Was all set for a summer release but was delayed until about now just to find its marketing niche. In the game industry, Gothic II is great example of being delayed until the right time (now, 4Q).

[Mota]

I think it's all about the $$$. And in video games, 70% of yearly sales happen in 3 months. I work for the retail distributor of Xbox in Canada, and it's staggering how much things change from August to October. Who wants to release their game in February when everybody's credit cards are due and Xmas is long gone?

I think that the developers set unrealistic expectations upon themselves and then when it comes time, realize that there's no way in hell that they can meet these deadlines.

[sabotai]

Bee, a couple of other reasons.

1) Each year, games become more and more complex. Not a just a little more complexe, a lot more complex. The same amount of people working for the same amount of time might produce a quality game this year, but in 2 years it's out-dated.

2) Each game has to be innovative in some way. All of the "clone" type games are always (and least it seems) done on time. All they do is mimic what another game does, and do it. But for a game to try to be great, it has to do something different

Add those 2 together, more complexity plus more innovation, and you get no basis for comparison. You have nothing to compare to what it is you have to do. So there's no way for the developer to know how long it is going to take.

Some learn their lesson, others don't. id missed a deadline once, and changed their stance forever to "it'll be out when it's done." Others don't. As stated above, mainly because of marketing non-programmer dickwads (I added the "dickwads") who insist on putting out a release date so they can do what is they do best (lie).

[Mac Howard]

There's also the aspect of competition that doesn't exist to the same extent in the business world.

The game definition isn't static. A game may well take, say 2 years to complete. During that time other games will be published in the same area that will contain features that were not included in the original game plan but are now felt necessary. Or it may be that user criticism of the designers' current game cause them to introduce new features. These new features not only need designing but will also affect current design features and integrity testing. Last minute bugs are often the result of last minute expansion of features.

Another thing I find is that even on the day before release I see better ways of doing things. Sometimes the only way you can see these improvements is when everything is put together and operating. Decisions made while the project was incomplete are sometimes seen later as not the best. Wanting to release the best possible game you take the time to introduce these improvements - and then become embroiled in extensive new testing, new bugs leading to new delays.

The wish to release the best possible product always leaves you unhappy with the final release and wishing to delay it all just another few days to introduce the features that the complete design has revealed will be an improvement :rolleyes:

[Marc Vaughan]

The wish to release the best possible product always leaves you unhappy with the final release and wishing to delay it all just another few days to introduce the features that the complete design has revealed will be an improvement
One of the hardest things with each CM update is deciding what from my huge list of ideas I'll implement and what will have to be left for a later version ... worst thing is that each year the list gets longer rather than shorter, I can add things to it so much faster than I can implement them (currently topping 132 pages of ideas!).