This week at TechRaptor I had the chance to exchange some emails with Dean Razavi, the lead developer on Vidar. We recently covered him on To The Green and were talking on different subjects about his game, indie development and the industry in general. The only editing that has ocurred is ordering follow up questions after the original.
TR: Can you tell us about yourself, what got you into game development, and your experience?
Dean: I’ve had a bizarre journey here. I started out studying music, went to law school and have been a lawyer for about 4 years, and am now doing game development. Even though it’s so far removed from everything else I’ve done, it feels like a homecoming. I’ve grown up with games, I’ve celebrated games, my wedding was game themed. So developing has felt really…right. As a hobby for about 10 years now, I’ve staged what I call “Hunts:” large-format urban scavenger hunts that are focused more on completing challenges than finding something and taking a picture of it. At one point, my last one had my friends wearing mirrors on their backs and chests and trying to aim a laser pointer light to reflect on all of them – a real world Zelda mirror shield puzzle. All involve riddles of various forms, and a lot of times the players have to find unique ways to combine various tools they find.
From all of these, it seemed like making video games was a natural progression, particularly a puzzle game.
TR: How much do you think your legal training has helped you with doing indie development? What sort of skills overlaps are there?
Dean: Logic is logic, and I really think attorneys make great developers (and developers would make great attorneys). From a pure programming perspective, the same brain that can figure out how 10 different laws work together can also make that assessment in code. Beyond that, I think it takes a dramatically inflated sense of self-worth to have the skin thick enough to become an independent developer, and lawyers have those in spades 😉
TR: There’s been a lot of talk about things like emergent stories and living worlds but for the most part those tend to be more marketing speech than anything else. With Vidar, you seem to have taken the idea of stories emerging from events and what the player does to heart – what led you to do this, and what have you done to avoid the mistakes others have made?
Dean: Playing Skyrim and being told by an NPC that I should check out the mage’s college, when I was in fact the head of that college, was one of the biggest let downs of my gaming life. Which I guess means I’ve led an extremely charmed life, at least when it comes to video games.
Before I got to “emergence” or “procedural generation” and all those things floating out in the indie dev universe, I had an entirely different model for Vidar. I looked at games which use the conceit of a central town that gets built up as you progress, where the blacksmith arrives and now sells better weapons or the alchemist arrives and has with him some great potions. I wanted to turn that on its head and ask, what happens if those people leave? And what if it’s random? Can you still have progression, or really even a game at all? Does it feel right? It turns out, it does and it doesn’t, and that was what the extremely early prototypes of Vidar looked like. When I finally had the realization that the storyline of my blacksmith was way more interesting than whether or not she was there to sell you an upgraded sword, I took out combat entirely. Every NPC became a quest giver, and I never looked back.
Everyone tackles emergence and randomness in their own way. I’ve seen some games occasionally try to generate characters, and it’s never sat well with me. It’s like a computer has buckets of possible options for an NPC with no concern for what kind of character it creates. So you get Mary Goldberg, an 8 year old black man who has scoliosis and gives you a quest to rescue his daughter. That’s an extreme example and easy to address with procedural generation, but you get the point – these characters feel like 2d cardboard punch-outs that exist purely to give you a quest. Maybe because it’s what I started with, but I’m not doing that in Vidar, I’m using predefined characters so I can make them feel more real, more relateable. And in fact, the core living world aspect of the game is focused really on two things; what quests you’ve done and who is still alive. By constantly examining just those two factors, I’m able to have a great deal of control over how that world develops. The term “living world” may also be a misnomer – Vidar is anything but living.
It’s important to study other games, too, and the biggest lesson I’ve learned in the past 2 years is from Papers, Please. That entire game takes place in a passport check booth. The scope of the game is seriously a room smaller than my apartment. And yet it tells a grand story, because it’s so localized. So to avoid the mistakes others have made, I’ve kept Vidar local – there are only 24 villagers, and we start killing them off pretty early. The smaller the world, the bigger the story you can tell.
I’ll also note that when I was demoing the game at MAGFest, more than a few people asked “man so like, how many endings are there in the game?” And the first time I heard the question I was actually a little struck – there are really only 2, either you stop the Beast before everyone in town is dead, or you become it’s 25th victim. And it’s because the Beast is the MacGuffin of the game – it doesn’t matter anymore than the statues in Final Fantasy 6, the Crucible in Mass Effect 3, or whatever the heart/darkness/key thing is that Mickey’s trying to protect in Kingdom Hearts. Vidar is all about the journey the individual citizens take through the game, and the possible combinations of those is something astronomical. But game designers so far have done a good job of convincing people that emergent stories mean different endings, and have totally forgotten about the story part.
TR: What is your opinion of the state of storytelling in the games industry?
Dean: There are some amazing stories that continue to be told in games, and the fact that emerging narrative hasn’t reached its full potential doesn’t mean that other games aren’t doing something tremendous. Triple As sometimes get slammed for this, but it’s only partially deserved. I won’t spoil anything, but Guild Wars 2 has done some really amazing narrative twists this past month, and they’re an MMO. The genre notorious for some of the worst storytelling in the industry, and they’ve got me on the edge of my seat.
There’s a reason the novel has existed for thousands of years, but the choose-your-own-adventure book has fallen out of favor. A story “on rails” (as the games industry might put it) gives the author so many more tools to make it infinitely compelling. In 2013 we saw The Last Of Us and Bioshock: Infinite, and while last year AAAs didn’t wow us with their stories, it’s ok to have an off year.
And indeed, Telltale is doing some amazing things too, so maybe choose-your-own-adventure isn’t quite dead.
As we get to explore new forms of storytelling, there are inevitably going to be bumps in the road. But that’s what the indies are for. I’m working on building emergent stories through changed circumstances; Noir Syndrome by Glassknuckle Games tries to generate a new murder mystery every game; a former Blizzard developer is working on a system to handle emergent quest lines. AAAs can’t take the risk in investment for a system that might not work, but we certainly can!
TR: The demo has the ice caves, but what are the different levels you plan on including in the game and how does an individual area effect puzzle designs?
Dean: Yeah, so we’re mid-Kickstarter right now which is primarily to fund the art. And the art for the ice cave has been complete, so that’s why you’re seeing a lot of it in the demo. The rest of the zones look like some great MS Paint rectangles right now 😉
There are at least 3 more “biomes” included the game. The creatively named Dark Cave, Water Cave, and Boulder Cave will all be added in addition to the Ice Cave. Each has a core mechanic that’s central to every puzzle there. So in the ice cave you see frictionless ice puzzles all over. In the dark cave, you’ll have a lantern and the light around you will constantly be closing in – you’ll need to continually refuel in order to see the puzzles. But there are tons of variations on each theme. In the ice cave, there are arrows that bump you in different directions, and those can get rotated around; there are doors that can be passed or block you, depending on if they’re up or down; there are tiles which break once you pass over them once, so you can’t really map out the entire puzzle from the start easily; there are imps that can knock you off the ice and turn the puzzles into something more focused on timing. It’s the same with each area of the cave – a core mechanic that guides what kind of puzzle you’re dealing with, and then individualized mechanics within that.
A favorite example are the wolves in the dark cave. Generally the dark cave is a frantic mess, where you’re trying to light all the torches in a room before they burn out, keep your fuel light up, and race over bridges that flicker on and off. To add to it, there are wolves that will try to track you down and kick you back to the entrance, and they get faster the less light your lantern gives off. A side area of the cave uses the same mechanics for a completely different puzzle – you need to lure 10 of these wolves into areas where you can confine them by closing a door. Now you’re only kind of trying to avoid the wolves – you actually need them to chase you.
TR: For the various puzzles that are random, how do you plan on making sure the difficulty doesn’t swing too much and that the puzzles make some sense in relation to the quest at hand?
Dean: There are three layers of “randomness” to the puzzles. First is the actual rooms you get while exploring the cave. In the ice area, you’ll get about 4 or 5 rooms filled with ice puzzles, but the game has 10 to choose from when placing them. Second is the path through these rooms. Each room contains anywhere from 2 – 6 puzzles, but you won’t reach all of them. Blocks and paths will be created to force you through in a particular direction to see only some of the puzzles. If in the demo you played the nun’s ghost quest, you’ll see this pathing thing right off the bat. The arrow at the entrance to the cave is either pointing right or down, forcing you to one of two possible puzzles. Finally, the last layer is the content of the puzzles themselves. The placement of the rocks and arrows and doors consists of a single option pre-designed just for that puzzle space. There are about 10 options per space, meaning that assuming you get the same room and same path as last time, it’s still unlikely you’ll get the exact same puzzle option.
By using a puzzle bank like this, I can actually control flow while still keeping the game random. Sometimes I want to be able to teach you a specific skill (like how to jump on floating boxes in the Water Cave). I can have 30 different puzzles which use the mechanic in different ways but all designed to make sure you learn the skill.
Difficulty balance is tough. Like, really tough. Tweaking one little thing oftentimes results in a huge change in completion rates. What’s also interesting is that when I’ve demoed the game, the majority of players fall into two categories: they solve the puzzle in half the time allotted, or giving them 3 days wouldn’t be enough for them to solve it. It’s not a nice, neat bell curve where I can try to aim the peak at a comfortable completion time. At that point, it becomes a game design decision. I may ultimately add a “normal” and “hard” mode to Vidar to accommodate both groups, I’m not yet decided!
I’m using an analytics company called Indicative to keep track of pretty much everything in a playthrough. So, for example, most of the puzzles when you’re trying to save Sandor’s son Erik have between a 65% and 70% completion rate, and while that’s a little low, at least it’s consistent. One option has a completion rate of 27%. That particular option is way too hard, but I’ve left it in for the demo because it’s fun to watch people struggle sometimes. I’m not looking too hard at the numbers now, but you can bet that in beta I’m going to have those charts on every monitor in the apartment.
TR: A lot of game developers seem worried about people not experiencing their ‘full game,’ yet with Vidar you have made it impossible for that to happen in a single playthrough – was that something you were aiming for from the outset or more a consequence of the type of story Vidar became?
Dean: Making sure that you can’t see all of Vidar was an extremely deliberate choice. I have this fantasy of people playing the game 3 times and then finally logging on to the wiki to compare notes and realizing that they’ve only seen 20% of the game. So much of Vidar has been inspired by events in my life, and this choice is one of them. The Germans call it Torschlusspanik, or Gate Closing Panic. It’s that tummy-feeling or that ennui you might get from knowing that every day, there are fewer options open to you. Maybe the fact that I’ve bounced from composer to lawyer to game developer gives you a hint that I’m certainly plagued by it 😉
Games tend to avoid Torschlusspanik. There are occasionally the super rare treasures or secrets that you have to get now or you’ll never be able to get it again, but it’s rarely anything critical for the player to have. In fact, in most RPGs, the game gets broader as you play. You start in a small village, you only have one party member and one skill, and by the end there are a dozen people to choose from, 30 different mechanics, 80 cities, 200 minigames, and an airship. You’re just before the climax of the game, and the world is your oyster. That’s so not how the world works. For every person that turns 30 (which is pretty much retirement age for most RPG heroes) and says “now I really get to do what I’ve always wanted” there are a thousand who say “I’m 30 and have achieved nothing that I wanted to.” We’re preoccupied with what we haven’t done except for in games.
So I absolutely wanted to convey that to anyone who plays Vidar. Every expected nicety in the game is a quest reward that you can’t get if that villager dies. That timer in the demo which tells you how long you have left? That’s a quest reward. So are sprint shoes, maps, quest journals, save points, all of the things a player wants to have in their game. And the chances to get those rewards go by quickly – the gates are really closing in Vidar, and it’s time to panic.
TR: You mention putting a lot of the niceties that people expect as quest rewards – these days we see a lot more handholding for games at the start with stuff like quest markers and such. What in particular makes you want to invert that trend and make easing things or ‘quality of life’ stuff rewards?
Dean: I think that handholding has just always been in games, but doing it in a creative gameplay way is really tough. The famous example is Mario 1-1, where the first goomba and first question block are placed in just such a way that most players will instinctively jump to avoid the goomba and end up seeing that the block will give them an item. That’s handholding, but it’s not a giant quest marker or block of text explaining what to do. It’s handholding by conditioning the player, and if you do that right, you don’t need a quest marker.
Ultimately, my incentive for changing the trend in Vidar was to make the game itself play different every time. I’m someone who was obsessed with Chucklefish’s “Starbound” for like easily a year before it came out. I made my preorder the first day, I panicked when the beta launch date was smack in the middle of my honeymoon. But over a year of playing it now brings up the familiar refrain for most sandbox games. While sandbox promises replayability, it’s only the map that changes. The way you play the game is still the same, just a different planet with different trees. “Replayability” actually means more than what we as an industry have described before – it means that the next time the player hits “New Game,” they have to engage the game differently than they did before.
I don’t pretend that Vidar is the absolute answer to that, but it’s a step in that direction. A player who doesn’t have a timer explores less, because they don’t know how much time they have. A player who doesn’t have a quest journal is writing down a lot of notes in the real world (or maybe forgetting to do quests the whole time). A player who has the ability to teleport through the dungeon has a different game than the player without. So I’m hoping that giving the player only some of these things will make them play Vidar differently.
Make sure to check out page 2 of the second half!