Free Candy For Everyone

The first talk I attended today was Cliff Bleszinski discussing Gears of War. This is a game that I’ve watched some people play and that there’s been a lot of discussion over by people in the industry. It apparently won Game of the Year at the Game Developer Choice awards last night. It’s a polished game that focuses on doing a few things, and doing them well. Oh, and it’s a third person shooter.

This guy Cliff B has gotten a lot of attention paid to him in the industry, but I’d known really very little about him. Until my current job, I paid little attention to first or 3rd person shooters. But it’s a bit more relevant now, so I’ve been trying to be a bit more interested.

His first game was actually something I did play, a long time ago, a PC platform game called Jazz Jackrabbit. That very well might be the last game of Epic’s that I played. Well, I need to pick up a copy of Gears of War, because it’s supposed to be rockin’ awesome.

On to the talk. Gears has sold approximately 3 million copies - and since it’s an Xbox 360 exclusive title, that makes that more impressive - and it has an average review score of 94%. I cannot even imagine that. Must be awesome.

He gave a little bit of history - it started out as Unreal Warfare. They were all excited about games like Tribes at the time and thought it would be large scale, vehicle based, territory control. That game was not what they wanted to make for this title, so it ended up becoming part of the next Unreal Tournament, the Onslaught game mode. They decided they really wanted a fast, tactical action game, focused on the character of the player.

He talked about a few of his game design ideas that didn’t make it in because they strayed from that core concept of a fast tactical action game. “Combat Cash” was going to be a way that you could loot the bodies of your enemies for profit, but that would slow down the pace - who has time to loot during frantic combat?

Also cut, “Morale Meter”, a morale system where your squadmates could react when things went poorly. But this led to a negative feedback loop where if you didn’t do things exactly great at first, your squadies ended up cowering in the corner - they preferred a positive feedback mechanism.

Cliff brought up and anecdote about programmers not always listening to designers that made me laugh - because it’s too true. He was explaining how Fraps is a great development tool - it’s a video capture tool that lets you take video of your game for later playback. It’s great for proving that something occurred. Typical designer-programmer exchange “The AI was in the corner twitching violently!” … “I don’t believe you, nah.” Enter Fraps, “I have a video to prove it!” Yeah, I’m just as bad as the next engineer, we never believe those silly designers when they allege these crazy bugs that cannot possibly happen given what we know is in the code. Yeah, I know, we can be difficult.

He made an interesting point about how he cannot really enjoy games anymore because he’s always a critic. He’s thinking about how he’d change things to make it better and cannot just focus on enjoying the experience. Interesting…

Back to some of the game design choices - they decided to make it 3rd person, not 1st, to differentiate it from Unreal, and because the use of cover is such a huge part of the game, they really wanted to see the player character as he went into cover. Once you go 3rd person, it makes camera a huge part of design. A bad camera system can destroy an otherwise good game.

Some things Cliff talked about in regards to their camera:
1. Offset the camera. They’re a shooter, and they don’t want the player smack in the middle of the view obscuring things. Putting him offset to the side made the center show you a clear view of the action, not your own backside.
2. No running towards the camera. Nope, you’ve got your gun out and it should always be pointed at your enemy, so you cannot run towards the camera.
3. Use FOV tricks where necessary - widen the camera view when in cover, for example.

He talked about the run of the player character. You cannot shoot while running, which was both a game design choice and something that saved them animation work since they then didn’t need to make run and gun animations. He called it the “Roadie Run”, after concert roadies who come scuttle out in a low crouch on stage when something goes wrong, as though being scrunched down will make them less visible. The player only moves 1.2 times faster when running, but through some fun with the camera, it feels much faster - they widen the FOV, lower the camera, and make it wobble. These things give the illusion of speed without having to have the player dart around the level.

They decided cover would be a fundamental aspect of the gameplay. They wanted it to be fundamental and prevent people from run-and-gun style play - they penalized players for not using cover by increasing the AIs accuracy when the player wasn’t in cover.

He made some more good comments on using test levels for development as “control” levels, as well as controller schemes, weapon design, and multiplayer maps. It was really fascinating stuff, and I haven’t really played the game!

I was very impressed by the talk. Cliff is clearly a very smart guy. There are a lot of people who get hyped up in the game industry. Frequently, they’re famous just because they know how to talk to game journalists and wow them. Cliff seems to be well-known for his merits, and that’s always something it’s good to see. He has firm ideas about what makes good gameplay, is willing to listen when people tell him something won’t work, and seems like a generally real person. Just my impressions from seeing the guy at a distance in a large conference room filled with a few hundred folks!