Things have gotten crazy in the gaming sphere over the last couple weeks man. Polygon has the best recap I think. Aside from that, I’ve seen thousands upon thousands of words written on recent subjects like feminism and the gaming community from a lot of smart people. I think right now I’m going to mainly focus on Anita Sarkeesian’s most recent Tropes vs Women in Video Games episodes — “Women as Background Decoration.” Not only is it a good critique of a lot of video games, it also brings up interesting questions about game design and technology that the video doesn’t really cover.
Basically all of Sarkeesian’s videos have been elaborate highlights of how generally bad video game writing and storytelling still is. Mainstream news articles have pointed out how the series has essentially applied standard film or book criticism to video games, under which a great many hold up terribly. This criticism is just centered on the portrayal of female characters.
The “Background Decoration” segments make a lot of good points about how what appear to be human beings are often relegated to simple things to be used. Often this means what appear to be female human beings being relegated to sex objects for (usually) male consumers to use with impunity. Often this is depicted quite violently. When so many examples are shown together it reveals how insane the trends can be. I think this happens not only because of writing failures, but also the challenge of reconciling character interaction with technology. Basically, the way character interaction is handled mechanically in a lot of games robs a great swath of characters of agency, which looks terrible under a microscope.
Let’s take prostitutes in Grand Theft Auto as an example. One of the most well-known points of controversy in the game is how players can pick up hookers, then kill them to get their money back while rarely facing consequences. Sarkeesian points out that this is pretty much the full extent of interaction you get with these people. Even if one or two prostitutes might become actual characters in a game, the vast majority are still sort of just “things” to be used and thrown away. The main reason this happens is because Rockstar simply didn’t design the game’s AI or scripting to go any further.
The same goes for games like Assassin’s Creed II where courtesans are kind of just a weapon. The mission where you can cause a guy to serial murder a bunch of courtesans — noted in the video linked above, is particularly creepy. It’s a failure of Ubisoft’s writing, but also a limitation of the game’s code. The same goes for the random crimes Sarkeesian talks about in Watch_Dogs. I haven’t played Watch_Dogs, but John Davison’s critique of the game on the John & Garnett podcast went over the same problem with the game’s AI for random pedestrians. He pointed out how when you analyze a non-player character’s personal information, leave, then come back to that same character, the game randomly creates a whole new set of information for them. He also expressed annoyance at how after a while you end up hearing hookers spit out the same proposition lines on the street over and over. Basically, it seems to be really easy to run into the limits of the AI in Watch_Dogs which snaps the game’s suspension of disbelief in half. Being able to sit there and repeatedly watch a guy gut a woman on the street in Red Dead Redemption, as Sarkeesian points out, makes those characters look like little more than a theme park attraction. Maybe that’s because a lot of games are designed like theme parks as a way to reconcile tech limitations, with these displays of characters being an unfortunate result. This is all really an expansion on the problem of having NPCs repeat lines of dialogue.
This brings up the question for the aforementioned games: How complex should developers make AI and character scripts in order to avoid the problem of making characters look like disposable objects? Should every prostitute you see in GTA have a daily schedule, a pimp she owes money to (or who may come after you for killing her), and a family to feed? Should every character you see on the street have a wide range of possible responses to the player’s actions? Maybe just recording a few more lines of dialogue for each character might help.
One set of open-world games my mind keeps coming back to when thinking about this subject is Bethesda’s Elder Scrolls games along with Fallout 3 and Fallout New Vegas. To me those games put players in surprisingly gender-balanced settings, but the more important thing is how complex their NPCs are mechanically. Most characters in games built on that framework, even random bystanders and merchants, are made to have a basic set of wants and needs as well as daily schedules. They sleep, wake up, eat, and do their jobs at certain hours of each day. They have friends and family they interact with (who react to how you treat them). Essentially, they commit actions and reactions independent of the player. These games also have rudimentary law and order systems that respond to your actions. This AI setup definitely has limitations, and background decoration characters do exist in games like New Vegas, but the boundary of interaction for players to run into is further out than in most open-world games I see.
Reasons like this are actually why I’ve started to like games of the Computer RPG tradition a lot. Games like Fallout and Mass Effect are built on a foundation that offers a lot more and much deeper character interaction than most game worlds. Of course this is probably because character interaction and malleable stories are a pillar of role-playing games all the way from their tabletop roots. GTA on the other hand was originally a game about having fun with cars in urban environments. That type of gameplay is supposed to be pure fun, like a toy, but now technology has expanded it. That originally simple concept has been put into increasingly “real” worlds, and the more detailed those worlds get, the more graphic and ultimately unsettling some of their factors and consequences become. That goes not only for graphics, but also AI and how the characters work. Watch_Dogs is supposed to be a game about people’s electronic information — an inherently complex concept. Perhaps Ubisoft didn’t fully succeed at presenting this in a subtle and tactful manner.
The reason character interaction in games so often falls straight into an AI uncanny valley of sorts is probably because writing characters for audiences to actually interact with is uncharted territory. Game designers have been doing it for a few decades but how often have people really thought critically about how NPCs should be written? It’s certainly a problem no other entertainment medium has had to solve. Sarkeesian is just one of the first people to take a deep critical eye to the problem through a highly-accessible channel.
It’s also probably possible for developers to write things around their current technology limitations. If giving every NPC complex enough AI to make them seem like a “full” character puts to much strain on CPUs or the developer’s budget, perhaps further exposition can be written in elsewhere. Maybe some games are reaching too far with how large and interactive they try to make their worlds, and not enough critics are pointing this out in regards to characters and narrative. If Watch_Dogs was going to be about people’s digital information, did it have to take place in a huge metropolitan area with players snooping in on thousands of characters? Would Watch_Dogs have made its point more effectively if it had a more small-scale setting where you hacked and interacted with a relative handful of characters with deeper, more defined writing? Maybe that point can be made for a lot of open-world games.
All this doesn’t really forgive the writing issues we see in other parts of games like cut scenes or scripted events that play out in front of players. When those events start a game basically turns into a movie for a couple minutes, and at that point you can reasonably critique it next to the rest of film. I imagine most cut scenes in gaming don’t hold up well under that kind of scrutiny. Back in the 90’s we didn’t scrutinize non-interactive events that much because we were still amazed at the novelty of their existence. Cut scenes blew us away on the PS1 simply for being there. We’ve probably reached a point where they need to be put under heavier writing scrutiny, but we aren’t. Most game reviewers aren’t film critics after all.
That brings us all the way back to the problem of this industry still trying to figure out how to make good interactive storytelling which is a whole other discussion. Maybe Sarkeesian’s series is bringing us to the question: “How do we write socially responsible and provocative interactive storytelling?”
- Castle Vidcons is back. http://t.co/pauhWIIyKM
- Nice Silent Hill remix. http://t.co/Y7LsuR3LQy
- Apparently a while ago that Ninja Turtles spoof series switched to World Star: http://t.co/GH55JIWgto
- The latest soft beverage! http://t.co/pwLYatTRYj
- Nice interpretation of Miyamoto’s recent comments on hardcore vs casual gamers: http://t.co/rZ0VQtXMQc
- http://t.co/Jx6Sajy0rx A game based around the idea of media bias sounds interesting not only for its subject matter, but also as a game mechanic.