Today is the 10th anniversary of what is often called one of the best stealth games ever and one of my personal favorite games of all time. I spent all weekend re-examining Tom Clancy’s Splinter Cell: Chaos Theory partly because it’s still an excellent game and partly to re-approach it.
I stopped playing it for a while because after doing a perfect stealth run (no kills, no knock-outs, leave all enemies undisturbed, complete all side objectives) I became such a perfectionist I actually couldn’t enjoy the game anymore. This past weekend I decided to start my first expert mode run and generally approach things as if I were playing for the first time again. That’s obviously impossible, but I was able to flip the switch back from the speed runner mentality to that of someone who earnestly examines each situation in the game. This works extremely well because unlike the original Splinter Cell which is sort of a series of puzzles to be solved, Chaos Theory is a fluid simulation that can give you different results every time you play it.
While in the first level this weekend I placed a sticky camera on a wall by a passing guard and fired gas from it to knock him out. Another nearby guard saw this and came rushing over — straight into the gas while it was still firing. I laughed out loud and after trying several times, couldn’t reproduce that exact sequence of events. The simulation just popped out an emergent fluke event like a snowflake.
Clint Hocking — the game’s director, appeared on an episode of the Idle Thumbs podcast a while back and spent a lot of time talking about Splinter Cell. Hocking repeatedly described all his games as “simulations,” — that’s how he approaches game design and that’s pretty much what makes Chaos Theory what it is among the other SC games and among other stealth games. I’ve used the term “immersive simulation” on this site a lot, usually when describing games like Ultima Underworld, Thief, the original Crysis, or Deus Ex. As a stealth game, Chaos Theory is philosophically very similar to Thief and the original Deux Ex. Playing them feels the same. What’s fun about these kinds of games is they feel like toying with the elements of each situation.
The term “simulation” here to me means these games fill their environments with a bunch of factors, give you some goals and a bunch of tools, and simply let you use your tools to plug at those factors, getting various results and achieving objectives in various ways. Some might call it a cornerstone of PC game design — as most PC games are traditionally designed like computer programs for users to play with as they see fit. The “simulation” of Skyrim is so thorough (or perhaps not thorough) that you can render a character blind by placing a bucket on their head, and then steal from them with impunity. Chaos Theory brings together things like enemy AI, dynamic lighting, shadows, player sound versus ambient sound, Sam Fisher’s sticky devices, locks, computers, and electrical systems and simply asks players to solve problems with all these factors in play. I remember a promotional video for the game back in 2005 where someone at Ubisoft basically said “It’s your game. We won’t tell you how to play it.”
One thing some people might take issue with in this game is its quicksave-anywhere policy which is a change from its predecessors’ checkpoints. As I recall even the console versions of Chaos Theory have quicksave. Technically it might break the difficulty, but I think it fits with how it’s designed. Basically, Chaos Theory wants you to experiment, and always being able to reload let’s you explore the breath of what you can do in the game. This is why even as I die repeatedly on expert mode I’m not frustrated because I’m never failing the exact same rote task each time. Trying a new method upon every attempt and getting a different result keeps the game fresh, even if you die a lot.
In another level this weekend I snuck through a vent and opened a hatch into a conference room through the roof, but four guys in the room immediately saw the hatch opening. When I saw all of them bunched together looking up, I launched a gas grenade which quickly knocked all of them out. A supporting character over the communicator responded with the line “that’s one way to do it,” and Fisher responded, “Finesse is for the young and cocky.” This is perhaps the other half of why Chaos Theory works — it actually anticipates some solutions you might use and writes around them.
Other “simulation” games do this, but mainly it’s the ones that are still built around linear stories an campaigns. Gamse like this, Thief, Deus Ex, or Crysis aren’t Minecraft or Elite — they are actually quite linear. What impresses me the most about them is how they reconcile tight level design with a heavy emphasis on player choice. It’s a fine balancing act that’s pretty rare. Most of the games that achieve it stand out as some of the very best.
In fact, it’s simulation games that usually make the choice in level design and setting interesting because it becomes a series of functional possibilities in addition to the aesthetic choices. Chaos Theory’s first level is a set of colonial-era ruins which forces the security to be basic and low-tech — a good tutorial area basically. You start out there by manipulating a few security emplacements like generators, tents, and the odd laptop, but it’s mostly just terrain and enemies. The middle of the game immerses you in a penthouse and office building as you learn to use their high-tech security against your enemies. Then Chaos Theory hits you with a low-tech retreat in the mountains of Japan. Aesthetically it’s the perfect place for an NSA Ninja, but functionally it forces you to deal almost solely with tight corridors and enemy AI.
Hitman: Blood Money is actually one of my favorite examples of this, and another excellent stealth simulation game. I kept playing through the game just to see what kind of environment it would let me subvert next. Its Playboy mansion analogue in particular sticks out to be because of some of the tools it gave you. For instance, at the beginning of that mission the guy you’re supposed to kill is sitting in a glass hot tub suspended over a 70-foot drop which you can shatter if you’re quick.
A lot of people like to point out how how the latest SC game — Blacklist, has a similar focus on player choice and technically they’re right. Blacklist impressed me with how much latitude it gave me in every situation. The main difference is the factors each game presents. Blacklist is very much about traversal, crawling over and under what feel like video game levels at near Assassin’s Creed speeds. Chaos Theory is more about slowly subverting realistic environments. Also, while Blacklist’s missions are broken into little arenas (sort of like a stealth Halo really), every Chaos Theory level is one giant map that works as one “system.” In this new run instead of following waypoints like in modern games I constantly referred back to the 3D map screen to thoroughly analyze each place. Actually, one of Chaos Theory’s main flaws is that its map and information screens could have been designed better. Its map get’s confusing because it doesn’t indicate your exact position.
At the moment I’m also playing through Assassin’s Creed IV: Black Flag, and technically its individual missions take place within sandboxes. Creed does let you play with factors like bushes, alarm systems, pick-pocketing, climbing, and different ways of stabbing people. It’s probably the basic focus Blacklist came from. The difference from Chaos Theory and the other aforementioned games is the pacing as well as the focus on being a superhuman killing machine. Blacklist and Creed are about utilizing essentially superhuman powers in broad environments while the older simulation games are more about thinking your way through intimate environments with frail characters and advanced tools. Crysis might be the exception and a great compromise actually. The first Crysis gives you super powers enabling you to overcome challenges in both subversive and explosive ways, but you see it all from a very human first person angle, and the levels are intricate facilities rather than broad obstacle courses.
The Ubisoft of today has certainly gotten the idea that players like to “create their own experiences” as producer Jade Raymond noted. I’m just not sure it can nail this in the design since the same way it did with Chaos Theory 10 years ago. I haven’t played Far Cry 4 or Assassin’s Creed Unity yet but I hear those games push towards the same goal.
Fortunately, a few games today feel like they’re retaining the same philosophy that gave rise to Chaos Theory. While developing Dishonored Arkane was all about letting players use its tools however they wanted within dynamic environments. I’m also seeing a lot of the same kind of design in Metal Gear Solid V. Ground Zeroes simply puts you in a zone with some objectives and more abilities than every before, and fluidly reacts to however you choose to do those objectives. It’s almost the only stealth game where I don’t quit after tripping an alarm because I want to see how the game reacts to what I do next. There already exist videos of people exploiting the game in crazy ways, sort of like the “Nanosuit Ninja” videos for Crysis. Hitman developer IO Interactive seems to have gotten the message after the lukewarm reception for Hitman Absolution. An open letter claims the next game will “create living, breathing and believable levels which will allow gamers to play around with the AI to create those unique moments every fan of the Hitman franchise loves.”
One of my best definitions for a “simulation” game would be a game that makes me shout “Wait, I can do that?!” more than once.
There are a lot of other great things about Chaos Theory I don’t have the time to go into right now. Actually, I think it’s one of the best examples of a nearly “perfect” game. It has the singleplayer design I just described, a co-op campaign living by the same rules, fantastic multiplayer that was simultaneously unique and skill-based, a great soundtrack I play to this day, and looked amazing in 2005. It’s a very rare example of a game I really felt was worth $50. It’s the ideal so many AAA games try to achieve today when they cram in so much content. In my mind MGSV is shaping up similarly.
- Recent blog post from Hocking on how developing Chaos Theory affected his life. http://t.co/SSc9Hccqn0
- I haven’t played Splinter Cell Double Agent yet but I hear the version for the original Xbox is the best one and basically “Chaos Theory 2.”
- Impressive story from the early days of Demon’s Souls. http://t.co/jeNRQ4VFna
- One-Punch Man anime promo. https://t.co/LajuiXQTRV I think this could be big.
- Gangsta looks pretty cool too. https://t.co/fux17lu5PW
- How Bethesda (Elder Scrolls, Fallout 3) actually created Madden. http://t.co/sus2V6mvRW
- Demo links for cloned Steam Greenlight games leading to malware. http://t.co/SuRgJqukPy
- The creator of Senran Kagura is trying to out-do himself. http://t.co/SdNXaUVGe1
- A developer answers why so many of them are dodgy around supporting Nintendo platforms. http://t.co/Jhdrzq58Sy
- Beware the typo. http://t.co/AdeouvTlX3
- Weird story on the creator of Wonder Woman. http://t.co/HEbCrR2zC0
- Remember that Halo mod for ArmA 3? Now there’s a dynamic universal warfare scenario based on it. https://t.co/EgQRanklJL
- Stunning aerial photography comes to San Francisco. http://t.co/oTvaZo8pqk