Skip to main content

Site upgrade

Careful readers may have noticed small changes here and there. Today, I performed a somewhat bigger upgrade behind the scenes. Everything seems to be working like before... Still, if you see something fishy, please let me know.

Trailer for new Pixar movie Wall-E revealed

Disney just released a new trailer for Wall-E, Pixar's next movie (which I blogged about earlier). This is definitely looking like the most detailed Pixar movie ever. The light on that dust, wow.

Apparently Wall-E won't contain any intelligible speech! Just R2D2-esque sounds. That sounds like a very interesting creative constraint: it allows the filmmakers to go back to pure animation without appearing to be weird.

There's also a nice Wall-E-related website for a company called Buy n Large, where you can read about stuff like "BnL Annual Report: The Musical".

(Via Kottke.)

Awesome info visualization in The Kingdom

Exposition is always a difficult problem in storytelling. How do you quickly convey crucial background information without boring your audience? "Surely you know, professor Horneman, that one can't just have people tell each other things they already know!"

Well, here you can see the first 4 minutes of The Kingdom, a new movie about the FBI and Saudi Arabia. The credits sequence is a great info visualization that explains the history of the relationship between Saudi Arabia and the U.S. from the 1930s until now. They manage to make it pretty exciting - I am imagining a lecture using this kind of stuff...

Anyway, while not revolutionary, it is a clever idea well executed, and it's not very common to see so much information of this kind in a Hollywood movie, especially at the beginning.

(Via Making Light.)

Is Bungie leaving Microsoft?

Rumor has it that Bungie is leaving Microsoft. The full email from someone who apparently is in the know and whose NDA recently expired:
"So heres my big secret. You should google Bungie + Microsoft + separation this week. You know that big ol BILLION dollar franchise Bungie has created for Microsoft, to show their appreciate Microsoft is letting Bungie leave. Of course Microsoft gets to keep all rights to the Halo franchise, but as today Bungie no longer part of Microsoft. Ask anyone who works there to search the global address book, they're no longer in there. Microsoft was supposed to release the press release today but if they wait till the 10/6 the impact wont effect the quarterly results. However today is the actual official date and the day the NDAs expire, however you still didn't hear this from me." "Apparently MS just wants Bungie to make Halo for the rest of their natural days, and Bungie doesn't like how MS is constantly trying to "handle" everything they do; the way they market their games, the way they interact with their fans (basically the fact that they do appreciate their fans), and how stingie they are with the profits (comparable to the rest of the industry). So as of today they are their own independent entity. They'll probably make Halo 4 for Microsoft, however hey are also free to create new intellectual properties for whatever system they want. (Even though they prefer the xbox platform)"

I don't know if I should believe this. I can see two ways for Bungie to leave Microsoft: they come to an amicable agreement with their parent company, or everyone packs up and leaves. Let's look at each option:

Amicable agreement
Microsoft making an amicable agreement with Bungie to let them go? Highly unlikely. IP is valuable, but I would argue that in the games industry, the capability to execute and exploit an IP is more valuable. Nobody can make a Halo game like Bungie. Look at Take2 and the Max Payne license: they have the IP, but what are they doing with it? Was that worth the $45 million? For Microsoft to let Bungie go would be an awfully nice move, and Microsoft is not known for playing nice, especially when something as crucial as the Halo license is involved.

Everyone packs up and leaves
Sounds like you could do it, right? Well, there's a couple of snags:

  • Employment contracts. The higher the position, the longer your notice period and the longer it takes before you can actually leave and start something new. It gets even more fun with no-compete clauses. Remember Ubisoft Montreal? The higher people are in Bungie, the more they would be driving this transition, and the harder it would be for them to actually pull it off.
  • Company size. How big is Bungie? Probably about 150 people. To get all of those people to quit and re-apply at another company right after a major crunch phase is very, very hard. It's much more likely that only a number of key people would leave. Even then, those people have to make a choice between staying with the status quo at Microsoft and maybe getting a promotion because the guy above them left, or going off to start something new. Admittedly, they could start again under potentially great conditions - the Halo team would have people beating down their doors with suitcases full of money. But how many people at Bungie remember what it is like to start a new company?
  • Operations. Halo 3 was just released. Who handles the operations of the website? Because there's a ton of stuff going on there. Who makes downloadable content or patches? Who makes the PC and Mac ports? The rumor sorta kinda covers this by claiming Bungie agreed to do Halo 4, but again, that only works if this really is an amicable arrangement.

What makes the rumor sound somewhat plausible (which doesn't mean it is plausible) is the whole 'Microsoft is micro-managing us and we don't want to make Halo for the rest of our lives'. Working on a big AAA title like Halo is a double-edged sword: it is exciting as hell and looks great on your resumé, but you're a cog in a very big machine and there will be much more people above you who have very strong ideas about what you should be doing (e.g. 'more Halo' but also 'move that rock to the left'). Also, big and expensive does not guarantee good management, but it does guarantee high pressure.

On the other hand: EA has shown you can create a studio where people just make incremental upgrades year after year after year, and do a very good job of it. The people at some EA studios, notably the EA Sports studio in Canada, have a different mindset and culture from typical developers, but they are great at what they do - just don't ask them to build a new IP from scratch. Who says Microsoft couldn't do the same thing with Bungie? Sure, senior people would leave - they already have - and over time the spark that made Halo special would probably go away. But, sadly, this scenario is way more plausible than an amicable split.

In my opinion this is just a rumor. One that could affect Microsoft's share price. Coincidence? We will see.

(Thanks Chris.)

Update: It's not a rumor anymore.

Why game design is important

A long, long time ago I worked at a company that did not have a game design position, and I wanted to convince the people there that, you know, maybe you should have game designers. Like, one per project, at least. Ah, the good old days.

So with the help of Mark Barrett I wrote a little essay and sent it around. And although I am not claiming a direct causal link, we did get a proper game design position after a while and game design was taken a bit more seriously.

This was a long time ago and I hope nobody nowadays has to convince people that maybe you need someone who is paid to think about exactly how this game will be fun. But you never know...

Here is the essay, unchanged from what I wrote around a decade ago:

Why Game Design is Important

Introduction : What is game design?

The goal of a computer game is to deliver an entertaining interactive experience to the player. If we examine games that have been commercial and/or critical successes, we find that the most common trait of these games is that they are entertaining. Although beautiful graphics, fancy technology, or a big license can contribute to a game's success, none of these, on their own, are sufficient for even minimal success.

If we want to increase the appeal of our games, it is necessary to understand what makes an experience entertaining, and how such an experience can be created. The discipline that tries to answer these questions is called game design.

Game design is not programming or art

The entertainment value of a game is not linked to any well-understood discipline such as graphic art or software engineering. A game can have brilliant graphics, a groundbreaking 3D engine, or fantastic artificial intelligence, while at the same time being frustrating or uninteresting to play. Conversely, a game such as Civilization, with lousy graphics and a lame 2D display engine, can be a classic masterpiece that has become both a critical and a commercial success.

So quality of entertainment is not solely linked to the quality of graphics or the quality of programming, although both certainly have an impact. Instead, successful game developers have found a way to make a game more than the sum of its graphical and programming parts, by focusing on and emphasizing design. Implicit in this is the idea that game design should be recognized as distinct from other aspects of game development.

Game design is hard

A close look at games on the market today clearly shows that making entertaining games is hard. How many mediocre games are released each year? How many games are exciting because of some gosh-wow feature, only to be forgotten a month later? If making software in general is difficult, then making software entertaining should be recognized as being extremely difficult.

Designing games takes time, effort, skill and sensibility. It is important to realize that game design is difficult, and to make sure that appropriate resources are directed at the process.

There is no "official" way to learn game design, therefore it is often ignored

In order to make an interactive experience entertaining, it is necessary to understand what an entertaining experience is and how it can be created. Unfortunately, although it is possible to study various disciplines that are part of the game development process, such as art and programming, these disciplines are not concerned with entertainment.

Worse, there is no formal education that teaches how to make entertaining games. Is there at least an informal way of learning game design? Aren't there theories on game design that are commonly accepted in the industry? The answer is no, and the central reason for this is that the field is new and constantly changing.

Which means that the skills needed to design an entertaining game cannot be easily learned or taught. Just as there are no accepted universal theories, there is no shared vocabulary. "Technical" terms commonly used in the industry, such as "fun" and "cool", are less than helpful. "Playability" and "good gameplay" are a slight improvement, but are far from being basic, well-defined concepts that allow for deeper discussion, or successful implementation. Still, it is possible to learn how to design entertaining games, but it requires a lot of effort and original thought. It cannot be done incidentally.

Game design is hard to identify

Without a commonly accepted theory of game design, it is hard to recognize good game design when it is present, and easy to ignore it during development and critical appraisal. How does a company determine if someone is a good designer? How does it train game designers? How does it prove that bad game design was why its last game was a failure?

The most serious consequence of this lack of common knowledge is that it is easy to underestimate the value and worth of game design. If you can't study game design as an educational discipline, and there are no accepted methods by which to measure the return of investment of a game designer, how can you justify investing time and money in game design? How can you justify paying someone to focus on game design?

Game design is essential

Given that game design is integral to product success, the more relevant question is, how can you justify not paying someone to focus on game design? However vague or ill-defined the process of game design currently is, without someone in the role of designer, the probability of making an entertaining game is largely left to chance. With a talented designer in a well-defined role, the probability of making a great game is greatly increased.

Someone must be responsible for a game's design

Someone must be responsible for the design of a game, and for making sure that the enjoyment of the game survives the production process. If no one is specifically charged with this responsibility, it is too easy to lose track of the end goal while facing the pressures and necessities of production.

Making someone responsible for game design ensures that game design skills are actually present within the team. It also forces people in the team to think more about game design, because they have to deal with the designer, in the same way that a lead programmer cannot make graphical decisions on his own because he knows the lead artist has to be consulted.

Does a game designer make all of the creative decisions on a project? Since game design is about entertainment value, and this is the aspect of a game that has the highest correlation with success, the game designer has a vital position within the team. However, without good graphics and programming, even the most well designed game will not be successful. It is the entire experience, the synergy that makes the whole more than its parts, that makes a game entertaining.

Contrary to popular opinion, one of the most important tasks of the game designer is taking the ideas of other team members and advising on the best way to integrate those ideas into a coherent whole. At other times it is important that the designer be able to explain the consequences of a given design suggestion, and to defend the quality and coherence of the design against suggestions which might damage the project's overall intent. (A lead programmer has the same task.)

Additional benefits of game design

While increasing the chance that a game will turn out entertaining and successful, a dedicated game designer and design process also provide other benefits. Chief among these is the cost savings associated with the streamlining of the production process that a good game design yields. Fewer features end up being implemented on a speculative basis, only to be cut later. Fewer surprises arise if only because someone has considered the ramifications of various design choices beforehand. And fewer morale problems arise because there is, from the beginning, a viable, coherent vision of what the product can and should be.

Conclusion: Design is good

Game design is a vital element of the game development process. Games such as Goldeneye 007, Metal Gear Solid, Starcraft, Resident Evil or Half-Life, to name just a few recent successful games, did not become great games by accident. They did not become hugely entertaining because they had good graphics or fantastic code. They became great because the people who made these games wanted to deliver an entertaining experience to the player. Like any non-trivial activity, achieving that end required careful thought and an intentional focus.

In any act of commercial production, it is important to know what you want to do, and how you are going to do it, before you begin. Although a relatively new medium, interactive entertainment is no different, and without careful game design as a basis for production the success of a given product is left to chance. When taken together with the fact that there is no detrimental effect related to an insistence on good game design, it only makes sense to pursue good design as vigorously as possible on a product-by-product and company-wide basis.

Update: I wrote about what to do if you don't have a game designer here.

More about Yojimbo, the OS X information organizer

A while ago, I wrote a review of Yojimbo, the Mac OS X information organizer from Bare Bones Software. Basically, I didn't feel Yojimbo supported the workflow I wanted.

I sent an email with my feedback to Bare Bones Software (obviously I feel a need to bother others with my opinions, but you already knew that since you are reading this blog).

Some time later (not that much later, I'm a bit behind on posting this), I got a very nice reply from Patrick Woolsey, Director of Technical Services at Bare Bones Software:

Thank you for contacting us. We appreciate your interest in Yojimbo, and I've attempted to provide comments about some of your remarks below.

Apple + I does not open the inspector. I can inspect items in Yojimbo. In other words, I can look at the fields and edit them. I am used to doing this using the Apple + I shortcut from other programs (Finder, iTunes). This doesn't work in Yojimbo. I don't know what Apple's Human Interface Guidelines have to say about this, but I missed it.

Since unlike the Finder and iTunes, Yojimbo supports editing styled text (meaning that Command-I invokes Format -> Style -> Italic), it uses the de facto standard key equivalent of Command-Shift-I to bring up the Inspector instead.

Am I Moving to Trash or Deleting? The icon says one thing, the drop down menu says the other. Is there a difference? I am confused.

For items, the result is always a Move to Trash; collections (only, never their contents) are deleted directly

Tag Collections only use 'and' relationships between tags. You can make tag collections in Yojimbo, kind of like smart playlists in iTunes, only dumber. You can enter a number of tags, but the collection only selects items that have all of these tags (an 'and' relationship). You can't have collections that select items that have any of the tags (an 'or' relationship).

That's correct. We do plan to support user-definable smart collections in a future version, which will allow additional relationships. Tag collections however will remain in their current form, as we may add support for tag-by-dropping in some future version.

Tag entry shortcuts suck. I can't think of a better way to put it. I had the following tags: 'to_blog' and 'to_read'. When I am tagging an item, I start typing 'to_'. As soon as I hit the underscore, Yojimbo selects the first tag that start with 'to_' even though I had several other tags that start with 'to_'. This is about as annoying as it gets. I renamed all my tags to (the inferior) 'blog_it', 'read_it', etc. [...]

Yojimbo is using a standard system facility for tag item creation (as also used for e.g. address entry in Apple Mail).

If you have tags "to_blog" and "to_read", upon typing "to_", you will get a popup list of possible matches from which you can either select directly, or you may continue typing to further narrow.

[I followed up on this as there really was an issue with underscores. This was logged as a new bug report.]

Renaming tags is impossible. You just can't do it in Yojimbo. Or at least I couldn't find out how to do it in less than 30 seconds, which is the same thing. (Remember: I am not arrogant, I read Donald Norman's book.)

There is no facility for renaming tags at present (other than via scripting); this will probably change in some future version. :)

Entering tags is counterintuitive. To me at least. Type 'atag anothertag' and you get ONE tag called 'atag anothertag', instead of two tags. You have to type 'atag, anothertag'.

The ability to use phrases as tags is intended.

Tags that are not used keep hanging around.

This also is by intent, so that (using the autocompletion) you can reapply the same tags you've previously used, even though there may not be any instances at the moment.

There is no purge, you need to delete unusued tags by hand, and you cannot easily see if a tag is in use or not.

If you attempt to delete a tag that's in use, Yojimbo will warn you. Beyond that, I expect that going forward, there will be additional tag management options though I can't say at present what exact forms these might take.

Adding items is inconsistent. In Yojimbo, you can add items using several different methods. Sadly, each method works slightly differently: *
* When you drop a link (a URL) on the drop dock, there is no feedback whatsoever, but the software creates a new item with the correct title and URL. You just have to go in later, find the item, and tag it, which is a) hard (see below) and b) bad workflow.

The intent of the drop dock (and indeed all of Yojimbo's input mechanisms) is to make entering info as easy as possible, so that you're more likely to do so. :)

As we haven't yet been able to identify a sufficiently low-friction way to handle tagging on drop, this doesn't happen. :)

* When you press F8 with a URL in the clipboard, Yojimbo creates an item with the correct title and URL, but you can't enter tags (even though a dialog box pops up).

If you open the tag field by clicking the triangular control next to the Name field, you can indeed do so.

* When you drag a link on to the dock icon, Yojimbo creates a new item that does not have the correct title (it uses the URL instead), but you can enter tags.

For bookmarks, this is a limitation resulting from how the Dock handles dragged info; we have more control over actions performed through Yojimbo's own Drop Dock.

I can't make an 'Untagged' tag collection.

Please see View -> Smart Collections -> Untagged Items.

So, I missed some stuff, I fell on the wrong side of a few design trade-offs, and some features are planned for the future. This was kind of what I expected, but still, I was very pleased by the nice reply I got. Although I still don't think Yojimbo is quite right for me, you might want to check it out.

Surreal Game Design: a group blog on game design

Surreal Game Design is a group blog on game design by the designers - all 17 of them! - at Surreal Software in Seattle. They make me feel inadequate by writing lots of great articles about professional game design, but then again, they're 17 and I am just by myself. Also, I need a lot of time to trawl the web for funny cat pictures and stuff.

Anyway, go check them out if you want to read about game design from people in the trenches. And if you like what you're reading, you can see if they have open vacancies on the same site! Smart.

Interview with Ridley Scott about Blade Runner: The Final Cut

There's an interview with Ridley Scott in Wired magazine about the final cut of Blade Runner that is finally coming to DVD and theaters this year. (You can read about how I've been waiting for this movie for nine frickin' years over here.)

Some excerpts:

Wired: You've called Blade Runner your most complete and personal film. How so?

Scott: I just finished American Gangster. It's about two real guys who are still alive, so you want to make it absolutely accurate. It's not a documentary, but it feels awfully real. For Black Hawk Down, I went to the location and shot it. Legend was more imaginative, but it borrowed from Jean Cocteau's Beauty and the Beast and the best of Disney. Blade Runner involved full-bore imagination. Deckard's universe had to be expanded into credibility. That's probably the hardest thing I've done, because there was nothing to borrow from.


Wired: Is it true that you didn't read Do Androids Dream of Electric Sheep?, the book on which Blade Runner was based, before making the movie? Scott:I honestly couldn't get into it. It's so dense, by page 32 there's about 17 story lines. So one of the problems is distilling it down into a three-act play that can be filmed. Fancher did that with a script he called Dangerous Days. Deeley came to see me when I was mixing Alien and said, "Do you want to do another science fiction?" I said, "I don't really want to go down that route if I can avoid it." But, to cut a long story short, eight months later, the script stayed with me. So I went back to Deeley saying, "You know, we can expand this into something more spectacular if we push it outside onto the street and create a futuristic urban universe." I could never shake loose the fact that I was a designer — which I'm constantly criticized for, and I really don't give a shit. At the end of the day, it has proved to be quite useful.


Wired: Blade Runner was prescient in many ways, anticipating globalization, genetic engineering, biometric security. How do you gauge the movie's influence? Scott: Enormous. One of the top architects in the world told me he used to run it in his office once a month. Architecture would have been my game if I hadn't done movies. Frequently an architect will design a building and then walk, and not care about what's put inside it, which is a pity. If I were an architect I'd say, you know, "You can't have that chair." And I think of Charles Knode's wardrobe, which people don't talk about often enough. Usually you get bad futuristic suits, right? Deckard's was very well done. And Rachael's clothes were stunning. I think there was a lot of influence from the film in that direction. And interiors, definitely. A big clothing designer sent me pictures of the interior of his place, and the factories looked like Blade Runner. Hotels in New York started to look like the movie.

Wired: What have you learned about Blade Runner — the story, the characters, the ideas — that you didn't know when the production began? Scott: I always knew everything! I knew all the characters. Of course, I got more experience as a filmmaker. But the more experience you get, the less you know — because the more you know, the more you know can go wrong. It can make you insecure. But I don't worry much about that.

Also nice to see Mamoru Oshii is a fan - why am I not surprised?

(Via Boing Boing.)

The OLE Coordinate System simulates M.C. Escher in real-time 3D

OLE Coordinate System is a mind-blowing graphics system developed by Jun Fujiki.

Mr. Fujiki has written a 3D display / collision engine that simulates the impossible spatial logic from the work of M.C. Escher. The reality of the autonomous character is affected by how his reality looks to you, the observer. If you don't see a gap, the character doesn't experience one. If you see two platforms touching, the character can cross.

I did not fully grasp it until I saw the video, which I highly recommend. Not only does it show very clearly what is going on, it also is a very nicely made video in itself.

It should be possible to make a puzzle game based on this technology. A constrained set of possibilities to place objects in 3D, combined with choosing the right angle to enable a connection from A to B.

You can even download the software if you have a PC with DirectX 9 installed. There's also a Java applet, but it ran too slow for me. Pity.

(Via Andrew Armstrong.)

Update: As was pointed out in the comments, this technology has been turned into a downloadable PS3 game called Echochrome. You can read a preview from the 2007 Tokyo Game Show here. Thanks Andreas and Nur-ab-sal!