Blog

• Irrevocable Player Choices

Sat, 27 Mar 2010 12:11:19 -0700

Fair warning: this blog post talks about choices players can make while playing a game that block off or change subsequent game choices or content. It’s impossible to talk about that without significant spoilers. If you don’t want to get spoiled on Mass Effect 2, Dragon Age: Origins and a bit of Oblivion/Morrowind, stop now!
---
Understatement of the week: creating an interactive story is tough. I don’t think any game has nailed it, or maybe it’s just that the medium is still so new that we haven’t developed a common language yet.
One of the biggest stumbling blocks is handing players the keys to the car, when the game’s path runs right along a cliff. Do you create a wall along the cliffside, so the player simply can’t drive off? Do you drop the wall at some point, when you think the player’s had enough practice?
What about a scenario where the player can choose not to change a tire at mile 5, and that’s going to determine whether he drives off the cliff at mile 28?
That’s what I’m talking about here: choices you make while playing a game that have a profound effect on what happens later in the game.
Elder Scrolls Style
In Elder Scrolls: Morrowind, killing an NPC required for the main story popped up this cryptic message:
With this character’s death, the thread of prophecy is severed. Restore a saved game to restore the weave of fate, or persist in the doomed world you have created.

This message was essential for Morrowind. In a huge, open world environment, it’s likely you’ll come across a character required for the similarly huge main story. Some of those characters have bad attitudes, and whether you’re playing a do-gooder or the type who kills those foolish enough to mouth off, you’ll probably end up attacking one of them. If not for the message, you’d continue playing for many hours and never realize you just created at best a changed ending, or at worst a broken script that completely blocks off the main quest.
Oblivion took a more heavy handed approach: NPCs required by the main story simply can’t be killed. You see a special cursor when you mouse over them. This includes characters joining to fight with you, like Martin.

You can take advantage of their “required” immortality by hiding when their health gets low. They fall in combat, you wait it out while in the shadows, and a minute later they rise and start fighting the enemies again. Clumsy, maybe... but less intrusive than a pop-up telling me to load a saved game.
Dragon Age Style
Dragon Age: Origins features a smaller world and fewer side quests than Morrowind or Oblivion, but its story threads are even more complex and intertwined. It’s great that there are multiple endings for my character and for her team, but sometimes the paths toward those endings are so convoluted that they feel unfair. I actually replayed the last 1/3 of the game just to get a different ending, and in the process I realized that one small (and at the time, insignificant) choice completely changed the entire outcome of all my hours of play.
Before I say anything else, BIG kudos to Bioware for making characters I care about so much that I’m willing to dig around on the Internet and replay the game just to have a better ending for them! 
I was playing with Abby, my dwarven rogue, and I invested in the romance path with Alistair. He’s an awesome character--very well written, funny, believable, and endearing.
Seriously, if you don’t want massive spoilers, stop here!
Alistair is also the bastard son of a dead king. One potential ending for the game puts him on the throne as a good king, still in a relationship with a female player character. Other potential endings have Alistair as a mediocre king, in or not in a relationship with a female player character, not king at all, or as king-for-about-30-seconds-but-now-dead.
My first play-through, Alistair ended up as a mediocre king who couldn’t be in a relationship with Abby, and who took one for the team and died at the end of the final battle. I understood the story power behind his sacrifice and completely understood the choice that caused that ending... but I felt unsatisfied with the romance ending and Alistair’s so-so performance as king-for-30-seconds. I started looking around the web to see what I may have missed.
Bioware folks, correct me if this is wrong, but it looks like one conversation choice in one dialogue is what created that ending. Alistair has a personal side-quest to reunite with his long lost half-sister. She’s a mercenary shrew and he’s shaken up afterward. There are a variety of things you can say to him at one point in the follow-up dialogue. If you don’t pick the exact right response (which is something about needing to look out for himself sometimes), Alistair becomes a mediocre king and won’t stay in the relationship with the female player character.

You have no idea when you’re in the dialogue that one choice can change the ending of the game (at least in terms of your character’s relationship with Alistair, and his performance as king). There are several choices in that conversation that seem just as kind, or strong, or wise, or guiding for your character to say. Why would one dialogue choice outweigh all of the other conversations my character had with Alistair over all those hours of play?
In contrast, I absolutely understood why Alistair died on that first play through. I was told that the Grey Warden who killed the archdemon would also die. Morrigan came to me with a proposed solution, and I said no. At that point, I was consciously choosing to have one of the Wardens die, and I knew it might be me or Alistair (or the third Warden).
The difference between the two is clear. In both cases, I was making a choice during a dialogue that would radically change the game ending for my character. In one case, I understood that decision and therefore also understood the results. In the other case, I had no idea I had even made a decision and therefore I didn’t understand the results and I felt cheated.
Mass Effect 2
You can see this same dichotomy in Mass Effect 2. There are two distinct points where decisions you make affect what happens in the game: how quickly you go to the Omega 4 relay, and who you pick for various stages of the final missions.

The first choice is the one that’s unclear. Throughout Mass Effect 2, you get missions that sound urgent but you actually have as much time as you want to complete them. You can explore the galaxy, collect resources, upgrade your ship and crew, and complete side-quests to your heart’s content even if one of your crew members just got an urgent message from his dying (and believed already dead) father.
Yet you cross an invisible line when some of your ship’s crew is kidnapped. You receive a mission that seems just like all the other missions in the game. If you take your time responding to that mission--which you probably also sense kicks off the end sequence for the game--you’ll discover that some (or all) of your crew is dead. You don’t make that discovery until later in the game, though, and you can only save your crew by responding to the mission within a certain, unspecified time frame.
How would a player ever understand that choice--or that he is even making a choice--under those circumstances? How can that feel anything but unfair in a game that doesn’t ever put a visible time limit on any mission (including this one)?

You can contrast this with great decision points during the final mission sequences. You have to pick which members of your crew lead other teams, with specific goals and with specific skills required. You know it’s a life-or-death situation. You know the team member is going to be hacking security systems. If you pick a team member with low Tech skills and you fail the mission, you absolutely know what you did wrong. The ending feels fair and when you try it again, you make a better choice.
Looking Ahead
Thinking about these situations made it more clear for me why it’s hard to create interactive stories, where players really have control over how the story progresses and how it ends. There’s a constant struggle between clear communication (the “threads of prophecy” telling me to reload my saved game) that breaks immersion, and the larger sense of storytelling that is at its best when today is made up of all the small choices you made yesterday.
When you look at the design choices that worked, though, you can start to see a clear path. Morrigan’s offer in Dragon Age is a good example. I didn’t know exactly who would die if I turned her down... but I knew it would probably be either my character or Alistair. I was accepting the consequences when I said, “No.”
The choice of team members in Mass Effect 2’s end sequence is another good example. It was entirely true to the moment that I, as team leader, had to assess the situation and pick the person best suited for the work. The NPCs helped make that choice clear by outlining the work involved in the task, without ever breaking immersion.
Here’s how you could handle the two situations that felt less fair.
For Alistair’s romance and his suitability as king, have that one dialogue choice set that path but also have an accumulation of other, similar dialogue choices also set that path. There’s no reason my character bolstering his confidence in every other conversation in the game has no effect on Alistair. You can see this in how you gradually build relationships with your team members as a whole--why not extend it to really important things... like the game’s ending?
For Mass Effect 2, it’s simpler: don’t do it. If you’re going to have something irrevocable happen if the player doesn’t do X in Y seconds, you have to make that super clear. You can make it clear by creating a certain mission type that’s timed, so when you get the “rescue your crew” mission you recognize that type. You can also make it clear and maintain immersion by having one crew member sending secret emails or video logs to you, making it clear that he is getting closer and closer to being killed. Either way, you have an obligation to communicate to players that action (or inaction) in this circumstance is being timed.
This ties into whether games should be “realistic,” which I’ll tackle in my next blog update. If you’re reading this and you haven’t played Dragon Age or Mass Effect 2 yet, what are you waiting for? They’re fantastic games with terrific characters, and it’s only through developers like Bioware pushing the envelope that we start to see what works and what doesn’t.

• Inside the belly of the beast

Sat, 16 Jan 2010 09:01:47 -0800

I’m almost done with Myst III: Exile, and I’ll post my thoughts about its puzzles and overall design after I finish it tonight. I want to talk for a minute about level design today.
Edanna is an age in Myst III. It’s a rocky island in the middle of a vast bright blue sea. The pale stone spire twists and turns, ranging from sheer rock at the peak to a shaded, lush jungle below. You wind your way from top to bottom before you return to Myst III’s home age, J’nanin.

I bet this age was the toughest to build, especially considering 3D technology at the time. Organic modeling is always challenging, and at many points in the level, you feel like you’re actually walking around inside one vast organism. Parts of the level are flat-out beautiful.

It’s also impossible to navigate. You can’t tell what’s a branch, a log, a stone ledge or a path. You walk right by essential turn-offs and never realize you could have gone that way. You move in circles, you hit dead ends, and even walkthroughs can’t help you. There are too many mushrooms for “There’s a left turn right by the mushrooms” to be much help.

This sort of overly organic, “inside the creature” level design is all too common. Some of my favorite games of all time have the organic stunt level, and in my opinion, they were always the weakest, most frustrating levels in the game. Just a few examples:
System Shock 2 had the level where you go inside the body of The Many.

Halo 3 sent you inside an infected ship to defeat the Gravemind.

You know the level design is confusing when walkthroughs have to diagram where you can walk.
Even the otherwise fantastic Dr. Ned DLC for Borderlands sent you into the guts of Dr. Ned’s fleshy laboratory. I was playing online with a friend, and both of us said, “Oh no!” as soon as we dropped down into that part of the level. Thankfully, it was short and completely linear, but we still felt lost at first.
I understand why these sorts of levels sound cool, but they go against how the human mind works. We seek order in our environment, and we look for distinctive landmarks when we’re navigating. When you put us in an environment without clearly distinguished areas and with no easily discernible pattern, we flounder. It’s the worst sort of maze.
It also goes against everything we’ve learned in playing other levels in the game. These levels are usually toward the end of the game, I would bet in an attempt to “give some variety.” You toss everything we’ve learned about how the world works in your game world right out the window, which feels like a rip-off.
It always comes at a point in the game where you feel like you’ve mastered the mechanics of the world and now you want to focus on strong execution against those mechanics. Then the designers essentially flip you off and send you somewhere that doesn’t obey any of the rules.
Good level design is instruction: you’re teaching players how your world works in the early levels, then testing them on their retention toward the end of the game. Tossing in the obligatory “inside the belly of the beast” level is just as absurd as giving the students of your English class a Calculus mid-term. You know... just for variety.

• Myst vs. Riven

Sat, 9 Jan 2010 11:25:14 -0800

I’m done with Riven and moving on to Myst III: Exile. It’s interesting to think about the changes between Myst and Riven. Myst’s puzzles were both more contained and more clear, and the change in tone for Riven’s puzzles makes the entire game both subtle and overly complex.
Containment
Containment is important for puzzles, especially if your audience is fairly casual (like Myst’s). You’re more likely to put the pieces of a puzzle together if they’re in reasonable proximity to each other, especially when you’re navigating a world (even in 2D).
In Myst, you enter a cabin where there’s a picture of a giant tree, an unpowered furnace, and a safe. The actual giant tree is located just outside, clearly visible, and about 2 clicks/screens away. Yes, the code for the safe is somewhere else, but it’s easy to understand that you need to open the safe to power the furnace and do something to that tree.

In Riven, you come upon a strange device on an island that causes water domes to rise on parts of the island when you push various buttons on the device.

In a location somewhat close by (probably 12-15 clicks/screens, involving an elevator), there is another device that raises metal bars as you push buttons to form what looks like topography.

Finally, in an entirely different section of the world, more clicks/screens away than I can even estimate and at least one vehicle ride and a couple elevators away, there is a third device that lets you drag and drop a couple marbles onto a grid. The colors of the marbles are significant and you can only understand them by finding specific bits of color at various locations throughout the world.

(Marbles edited out so it’s not a spoiler)
Yep, that’s one puzzle. I won’t reveal how it ties together, but it’s an essential puzzle for finishing Riven. It doesn’t just break the law of containment--it shatters it and buries the pieces all over the map. It’s not just a tricky (impossible?) puzzle to solve, but it sets an expectation for the player: anything you see may be a puzzle, and when your world becomes filled with potential clues, there’s no way to sort them out. It’s easier to spot a zebra when it’s with a herd of horses than when it’s hanging out in a herd of its zebra buddies.
Clarity
It’s also easier to spot a zebra when you know you’re looking for one. Solving a puzzle is very difficult when you don’t even know what the puzzle is.
In Myst, you enter the spaceship and find an organ and a control panel with levers that play a musical scale as you move them up and down. There’s a window above the levers, currently blank/black. It’s pretty easy to understand that this is a puzzle involving musical notes from the organ and the arrangement of the levers.

In Riven, as you explore the world, you find lots of fun interactions that cause animations. There are interesting D’ni gadgets, buttons, levers, viewers, creatures, books and other interactive objects that help make it a more interesting and “alive” game than Myst.

The vast majority of these interactive objects are just for entertainment. A couple are clues for puzzles. One particular set is an essential element for solving one of the central puzzles, a show-stopper that prevents progress in the game.
You may not have even noticed the items in this set--they’re usually tucked away on side-screens. One of the items is even sitting on a desk, disconnected from where they are usually found. You may or may not have clicked on them to discover that they play a sound. You certainly didn’t connect those simple interactive objects with an essential puzzle... that you hadn’t even seen yet.
Even after you see the puzzle, it’s challenging to connect it to the items in the set because they’re actually just one of the steps in the solution. I can’t be specific about what they are without spoiling the puzzle (although realistically I would bet 98% of all Riven players look it up in a walkthrough anyway), but imagine seeing a car on the street one day and the next day trying to understand that you needed to know what song was on the radio when that car passed. You had no way to know that you needed to remember either of those things, or that they could possibly be connected.
Why This Is A Problem
I’m sure there are people who figured out those Riven puzzles without help. There are probably enough of them to fill a small stadium. I wouldn’t be in that stadium (and not just because I hate sports), but that’s OK.
What’s less OK is that you want to sell your game--and, more important, the next game in the franchise--to more people than the 1000 who fit in that stadium. And those Riven puzzles are simply not a great fit for the audience Cyan wanted... and the audience they already had after Myst.
The key here is to remember containment and clarity. Make it easy to figure out that you want to find a zebra, and even make it easy to find the zebra. The challenge should come from getting that zebra into the corral.

• Thinking about time

Sun, 3 Jan 2010 12:00:42 -0800

I finished Myst the other night, and started on Riven. I’ve been thinking about a lot of things, including the relative complexity of Riven as compared to Myst, and especially as compared to newer games like Dream Chronicles and Enlightenus. I’ve also been thinking about the passage of time.
As I walked around the islands of Riven last night, I really enjoyed the distant cries of birds, the trickle of water, the buzz of insects. The scenery was stunning, even at 640X480 resolution, even after all these years.
I also found myself rushing along the path, moving quickly from one screen to the next to reach a puzzle or interaction. At some point, I realized that I’ve changed and games have changed. The journey has become less important than the end result. I wasn’t enjoying the scenery of Riven the way I used to, because I have become unaccustomed to lingering.
It’s ironic because my personal project is called The Journey. I’m sure it will be renamed at some point, or at least subtitled, but that’s what I’ve called it for years. It’s a story and world that I can’t forget, a project that has been started and restarted more times than I can count.
While I work on the first 3D scene for The Journey, I’ve started watching Michael Palin’s Around The World In 80 Days. He talks several times in the episodes about how plane travel changed the world, that travelers no longer enjoy the journey or understand the distance.
I think the same thing is true for games, and for my life. I’ve become so focused on the end result that I don’t think about the experiences along the way. It’s one of the primary goals in creating a new world: The Journey is a journey for me as well.

• Ringing in the new year...

Wed, 30 Dec 2009 09:06:38 -0800

I’ve started working on a personal project that, for me, is a return to where I started. I design games because of Myst. I had always played games, and always wanted to make games, but Myst gave me incentive and inspiration. It gave me the drive to spend all of my free time learning to create 3D images and script in Director, all while working 50+ hours a week as a tech writer. I learned, I worked, and in the end I created my first game, sold it to Microprose, and the rest of my life followed.
I’m replaying all of the Myst games, starting with the original, and since I’m not ready to talk about my personal project quite yet, I’ll talk about Myst instead. I’ll update with my thoughts as I progress, again, through the game that got me started and all of its sequels.
Right now, I’m about half-way done with the first world I chose to explore, Selenitic.
As I completed the organ/musical note puzzle, it struck me that we’re inconvenienced now by games that make us take notes, especially if that note-taking involves real pen and paper. I used ATL-TAB to take notes in Notepad, and yes, I felt inconvenienced. Later Myst games added the ability to take screen shots and save them for later review, but I’m at least three games away from that innovation. I’m not quite dedicated enough to do what one fan did, and create my own handwritten, Myst-style journal of my notes.

Apparently I’m not the only one who was artistically inspired by Myst! Creating a game that inspires others to create is, to me, a marvelous achievement and just as the look of Myst inspires my art, its essence inspires my goals.