GamesBeat: Do you foresee much difference between early access and full launch?
Hecker: I have a ton of stuff I want to do. I don’t know how long it’s going to be in early access. Probably at least two years. I say two years for everything. Four years ago I said two years. Subnautica was in early access for four years. Knock on wood, the game is coming out.
I’m proud of how the game is coming along. The interaction of the community, the ideas — at elite levels people are playing the game I wanted to make. It’s different, because you have to listen to the game and go with it. It’s evolved. But from the very beginning, I wanted to make a competitive multiplayer game about subtle human behavior, and that’s the game they’re playing. It looks amazingly better than I thought it would ever look. I’m proud of it. I just have much more I want to do.
I want each character to act differently. Some people like drinks and some people don’t. As you ease that stuff in — the first idea everybody has about is that all the characters should have different personalities. I thought that was way too complicated — not to program, but to learn as a player. But then DOTA came out, and I realized there was no such thing as too complicated. People will build this encyclopedic knowledge.
At elite levels you’d shoot the guy, or at least be really suspicious. But the first time you, as a newbie sniper, learn about these traits and start getting suspicious of someone who doesn’t take the drink as this guy—the spy says, “Why did you shoot me?” “Oh, you didn’t take a drink back there. Your character drinks too much.” That educates both players. The spy realizes they have to start learning these traits. It eases you into this. Both players are clueless at the beginning, while at elite levels it’s just part of the game. There’s this ramp. It rewards study. The replay system gets to that as well. I have ideas for about 20 traits like that.
Here’s what the old UI looks like. There are still pockets of old UI in here. The new UI fades and does all the pretty video game stuff. If you’re about to click a button that takes you to old UI, I’ll put up a skull and crossbones. “You’re about to see something really ugly!” It’s mostly only elite-level sections where you’ll find that. If you’re just going to play the game, do the tutorial, head to matchmaking, you won’t see it. But once you get deeper into it, you’ll find places where the old UI is still there and I haven’t had time to replace it. Replay is one of them.
This replay system is off the hook for an indie game. People tell me it’s better than most triple-A game replay systems. I just loaded this game that was done earlier this year. It loads them all the way back to 2014. You can load really old games, because it’s backwards compatible. I’m playing this guy WarningTrack who’s way better than me. You can see, when you go to the results screen, he shot me, which is not surprising, because he’s way better and it’s a skill-based game.
I haven’t even counted them recently, but I have more than 300,000 replays on the server, all the games recorded since I got spectation in. I want to make a replay database that’s publicly accessible. “Give me all of Dean’s games on Pub where he was playing these missions and he won as spy.” I’m about to play you in a tournament and I want to see how you play. You can review these things, and even play sniper against them. You can play sniper against my past spy. If there’s no one in matchmaking, you can play games the server feeds you. This all works for spectating live or archived replays.
GamesBeat: What gave you away in this game?
Hecker: He had me highlighted from the statue visit. Then you can see he shoots me right after I bug the ambassador. You can see the missions I did. I inspected statues here, but I got the highlight for it. You can fingerprint at statues as well. One of the tells is an audio tell, contacting the double agent. Look at how many lowlights he got from this. It means no one in a conversation could be it. A couple of his highlights were at statues, but as soon as the banana bread goes off, he was able to lowlight that person and this person at the window. Now he’s narrowing down his suspects.
What happens is, the bug is a hard tell. My arm has to come out. You can see it here. It’s about to bug. I’m the woman with the cool hair. See that arm go out? He sees me standing next to the ambassador with 23 seconds left. Then I’m stupid. I need one more mission, and it’s really late in the game. See this white quad? That’s the guest list. One of the missions to purloin the guest list off the waiter’s tray when you take a drink. I do that.
Let’s see where he’s looking. Here’s the split screen. He’s basically looking right at me, because I’m standing right next to where the ambassador was. He can see it’s gone now. There’s no quad there. He takes the shot because I just took a drink. One of the flaws in my play style is I push too many missions late. I have to rush at the end like that, and they know this about me. They just wait and see me panic at the very end, and then they shoot me. Those play styles come out and you can see that in the replays.
Going on, I don’t believe you can make things esports. There’s all these companies that are like, we made this game and now we’ll make it an esport by spending a ton of money. Unnamed companies that I respect highly, but I don’t think that’s how esports happen. That’s not how football worked. It’s organic. I can’t pay millions of dollars to make this an esport anyway. I’m just trying to put in all the stuff that makes a highly competitive multiplayer game even more compelling and castable on Twitch and whatnot. We’ll see what happens.
The community is small. They do ladders, where I’m way lower-ranked than I should be, but it’s good when people beat you at your own game. The first time that happens, it’s like seeing your kid walk. They organize all that. I haven’t had to organize anything yet. It’s all community-based.
I want to do another feature after early access. I want to do recommendations with the spy. You saw that guy highlighted people who go to statues. I want the spy to be able to tell people in a conversation, “The statues at this party are really awesome. Check them out.” That increases the likelihood that everyone in the conversation goes over to the statues. Now all of a sudden this dude has way more highlights than he knows what to do with. Once you have too many highlights, it’s worse than not having any. Who are you going to watch?
Having the spy be able to puppeteer the party is another design thing I want to do that’ll make the game even deeper. It’s a 1,000-hour game right now, but I want it to be a 5,000-hour game. That gets you in Counter-Strike/DOTA territory.
One thing that led me to make the game this way was on Spore. Rob Pardo, during Spore, gave this talk. He was at Blizzard at the time. At GDC Austin he talked about Blizzard’s design methodologies. One of the points in this talk that’s stuck with me since then — this is 2004, 2005 — was that Blizzard does this depth first, accessibility later development style. Make the 300-hour core game loop, then make it accessible. It’s easier to make something accessible.
I sent that out to the Spore team. “We should do this. We’re doing the opposite. We’re making it accessible but I’m not sure we have the depth yet.” It’s hard to go the other direction. When you have an accessible thing, making it deeper is the hard part of game design. Making things accessible is just more tutorials, ease people in, beginner modes, things like that.
I vowed, after Spore, when I started working on SpyParty full time, that I would make the exactly opposite mistake. I’d go deep first and accessible last. You had to read a four-page manual to play the game before the tutorial came in. At PAX we taught people how to play. But now there’s a full tutorial. Zach Gage, who’s an iPhone game developer, convinced me to do a multi-modal tutorial where it’s text and voice. Zach had just finished the Tharsis tutorial. I tried to encode how we teach people to play in person at PAX. That’s me talking over it.
This thing about having you talk and be able to read it at the same time makes retention way better. But because you’re doing a tutorial for multiple audiences at the same time — there are people who’ve never played a video game before. You have to teach them what WASD is. There are people who’ve played games, but never played a game like SpyParty. There are people who’ve played SpyParty, but want a refresher. And then there are elite players who just want to see what’s going on. It’s skippable in multiple stages. That finishes the text. That goes to the next thing entirely.
Let’s be clear. Everybody hates tutorials. And the only good way to do a tutorial, the only good way to teach people to play a game, if you’re not there in person — it works at PAX, but it doesn’t scale very well – is the StarCraft way, which is you make two video games. You make the single-player campaign mode and you make the multiplayer mode. I can barely make one video game, let alone two. I knew I was going to have to do a normal tutorial. My motto on Spore was “overdeliver late.” The replay system is totally nuts compared to what most games do. I try to make this stuff really good. We tested it at PAX and it did teach people to play. Once it went in, I had players record their sessions on YouTube or Twitch so I could do even more user testing.
Doing depth-based testing for a new feature — like the traits feature, adding different behaviors to the characters – that I can test myself. I’m pretty good at SpyParty. I can play with other pretty good players. The really good players will tell me how it is. But getting newbie feedback on how the tutorial went is hard. There’s a message of the day on the main menu, and this would print out a thing that said, “Hey! Tutorial is brand new! Record it on YouTube and tweet it at me for testing!” That’s one of the cool things about having a good community.
Matchmaking is still coming in a little hot here, but I’m pretty sure it’s going to work. The little tests we’ve done so far have gone well. It’ll be in there for Steam. I have to. You can’t have a non-matchmaking — the lobby right now, here’s what it looks like. You cannot expect someone to come into this and invite someone to play. That’ll still be there for the old-timers. They like to hang out in the lobby. We call them lobby furniture. But the way people play games now is matchmaking.
I think it’ll help me play more. The little test we did, it’s really nice to just drop in, play some games, and get out. In the lobby, right now—I made the game, so people want to talk to me. “How’s that feature coming?” If I get into a match with someone everybody watches and broadcasts it.
GamesBeat: Do you just play continuously with your group?
Hecker: In the matchmaking it’s different. That’s one thing that makes it lighter. There’s a social contract when you go to a lobby and invite somebody by name and you’re chatting. You’re going to play 10 games, 20 games? How long? It makes for a heavy weight. You’re going to be going for an hour or whatever. It’s great with friends, but what matchmaking does—I decided to make matchmaking as light as possible. It’s only one game.
Players were originally suggesting — this is a super asymmetric game. People talk about StarCraft as asymmetric because the Zerg are different from the Protoss. This is two completely different games in the same space. The only way to make sure that’s totally balanced is to alternate. I was worried that it wouldn’t be fair to only play one game, but it actually works pretty well. The server keeps you balanced in terms of how many spy games you play versus how many sniper games. It tries to alternate you. But it alternates you across players.
Now all of a sudden you’re not even socially contracted to play two games with someone. You go in, play a game, and then you’re in a new game. It worked out well. It felt good as an alternative to a heavier match.
GamesBeat: What about people who just want to keep going?
Hecker: I don’t know if this is a terrible idea for Steam, but there’s text chat available. Hopefully people don’t discover that, maybe? I finally put a kick command in the lobby. I didn’t have banning, a way to ba people off the server, for five years. I didn’t need it, because the community was so cool and our social norms were so good. But I finally chickened out. I figured I’d better have this in here before Steam goes live, the ability to ban somebody. Hopefully I won’t have to use it. I can switch the chat to emoji chat if I need to. Just, “Good game! Wanna play some more?”
This might not make early access, but it’ll be weeks after. If you’re playing a matchmade one-game match, I want you to be able to drop in. But friends will be able to just meet each other in the lobby. Friends won’t have to go through that.