Pages

Tuesday, October 8, 2013

gtav review: chaos, linearity, and the open world test

Sometimes while playing GTAV, I didn’t think the designers of the game knew what made it fun.

The important part, of course, is that the game is still fun. How can it not be? It’s Grand Theft Auto in a new city, with Rockstar’s ever-present attention to detail and quality, and the reviews have been universally lavish. Going around jacking cars, causing chaos, and escaping from cops is the same as it has been (with some improvements) since GTA3. [1]

But other times, like thirty seconds after I start doing yoga or in yet another shootout, I sigh and wonder if maybe someone has been whispering evil things into the designers’ ears. Things like, “the game needs more distractions that have nothing to do with the enormous world in the game,” or “players really just want to have gunfights with bad guys all the time.”

Of course, the game needs missions and things outside of the main campaign around the city. But these should always remind the player that they are playing Grand Theft Auto. Many of the game’s side-missions accomplish this beautifully; the ones that have Franklin working with a paparazzi are perfect at merging an unglamorous occupation with extant mechanics and a healthy dose of satire (this sort of mashup is a GTA series staple, and among the reasons I’ve grown to love it), but I have no idea what golf, tennis, yoga, etc etc etc are doing in the game. One of the properties players can buy unlocks the opportunity to slowly meander underwater in search of nuclear waste. I wish I was making that up. I cannot imagine the state of mind one would have to think this is what people are looking for in this series.

If it was limited to entirely skippable and irrelevant diversions, that wouldn’t affect my opinion of the game hardly at all. But it’s not. The main missions, the heart and soul of the game, fall into the same trap that GTA4 and San Andreas[2] did: here are some guys in a small area. Kill them all. GTA4, in a decision I would use a time machine to go back and change long before I considered using the device to assassinate Hitler,[3] introduced cover-based shooting to the series; GTAV, on more missions than not, will have you slowly advancing through conveniently-placed half-height blocks and other miscellaneous objects in order to kill all the enemies.

This is dumb. The strengths of the game are not in gunplay. Despite promises with each release that the shooting is way better this time really we mean it, the only way to get through these is still with habitual abuse of the autoaim system in a way that feels cheap and gamey every firefight. The best way to dispatch unseen enemies is releasing aim, then re-aiming to lock onto them and hit them with a shotgun blast. This is needed because the way players are intended to use the autoaim, by flicking the right stick from one side to the other to go to the next enemy, is a bit finicky and will sometimes unhelpfully aim slightly next to the now-dead body instead of finding the guy with a shotgun who is having no trouble blasting you.

This inspires me to come up with this simple test for all future developers of open-world games: when doing a mission, is the player doing anything that they could not have done in a specific genre game (eg racing, shooter, helicopter sim)? If they are not, it should probably not be in your open world game. The most notable breaking of this rule was most egregious in Far Cry 3, a critically-praised game that had a few interesting tasks for the player when roaming around (but not too many), and a main story that had the player repeatedly walking down extended linear hallways over and over to trigger the next half-cinematic. There are plenty of fine linear games that do their thing and they are more than welcome to do that, but let’s separate the two categories of linear and open-world games a bit more here. Players who pop in an open-world game should be able to get away from the rigid structure of linear games, and today’s open world games aren’t providing this.[4]

The open world with linear quests is a formula that feels built on budget-dictated compromise, whereas its inverse of a linear story with several ways to get past each obstacle feels much more rewarding. Deus Ex is the canonical example of this,[5] but games like MGS3 that people actually bought did it as well, and Hitman: Blood Money implemented it absolutely beautifully in a way that is unlikely to be replicated any time soon. The times when GTAV lets the player approach something however they want provide some of the best missions: in one, the player has to hijack a slow-moving truck in a heavily-defended caravan, and I wanted to play that mission over and over trying out different tactics. On the other hand, an open-ended assassination mission just led me to stand by the mission’s start point, wait an eternity for him to leave the hotel, and snipe him, then run away. Mission complete. Hooray?

The big introduction to the series is heists, and I was extremely impressed by how they were executed. The structure of a heist gives a great natural progression to the missions,[6] the heists themselves have multiple ways the player can choose to do them,[7] they’re affected in noticeable ways by the NPCs the player decides to hire for that mission (and while that’s all scripted, it still led to great moments of “thank god I hired THAT woman” and “oh no, why did I pick this dumbass?”), and the heists themselves are the most fun missions in the game.

I can’t discuss this without spoiling something about the heists: there are barely any. In fact, there are five, and one of those has no options for hiring people, while another one has no options to it (and only one person to hire) and is just another “kill all the enemies” mission. The game makes a huge deal out of showing you how to do heists and how the people will get better the more you use them, which makes the fact that you’ll only use your hacker once more in the entire game feel really shitty.

A radical idea, then: the game’s big new mechanic/theme is heists, and the heists are super fun, and the idea of them is that you use the same people from one to the other, so why weren’t there more? Like, a LOT more? Why aren’t 90% of the main story missions either heist or heist-setup-related? Instead, it’s more killing of guys.

For decades now, action games have half-assedly implemented stealth systems. The difference between a true stealth game and the accursed half-assed stealth system is pretty clear: in a stealth game, you can stealth around all the time, not just in the 10% of it when it explicitly tells you “HEY, BE SNEAKY NOW OR YOU’LL AUTO-LOSE.” Real stealth games have levels of enemy awareness between having no idea where the player is and actively shooting at them. Real stealth games don’t have cinematics that trigger enemies all automatically knowing the player’s location, with no chance to re-stealth. (Real stealth games have re-stealthing, by the way.) One of the game’s least excusable anti-fun interactions is that when a player misses a shot with a silenced gun on a foe. The foe will immediately know where the player is, start shooting, and all the other enemies start running toward you. For the sake of fun gameplay, missing a shot should do nothing, and alerting an enemy should leave a good half-second window for the player to recover by taking them out before they tell their buddies where you are.

GTAV apparently thought enough of its stealth mechanics to make it one of the stats that characters have. The stats in this game are a completely unnecessary addition to the game that serve no real gameplay purpose. The characters already had differentiation via their different abilities, and there was never a time playing on a character when I consciously thought “gee, I wish I was playing a character that was better at this thing I’m doing.” And if it doesn’t affect gameplay, then what’s the point?

The writing in the series has long been the shining example of the best the video game world has to offer, and that continues here… mostly. Michael and Trevor have a great dynamic,[8] but Franklin is just kind of… there. I assume he’s supposed to be some sort of audience stand-in, because he’s mostly reasonable but wants to generally earn money and maybe kill/rob people if necessary, or whatever. It’s like his entire storyline was cut at the last minute. The satire that runs through the game (as it has through all Grand Theft Auto titles), though, fell much flatter in this one than in previous games. It’s like they were going through the motions: gotta make fun of Big Tech Trends one and two and the latest best-selling game series, right? Part of what felt off about it, though, was that after a dozen years of us living in the modern era of Grand Theft Auto, hasn’t the series itself elevated to the point where, had it been any other game, it would be the one getting parodied? Why isn’t the son in the game playing a game based on GTAV, since it’s the best-selling game in the world?

Does this come across as nitpicking a game that’s really good? It should. I spent about 30 hours in GTAV over four days, and I worked a full shift on two of those days. I enjoyed the game. But when this much time and money (and money and money and money) is sunk into a AAA project that gets unanimously adoring reviews, we should examine each aspect of it a bit closer.


[1] This is slightly unfair to the newer iterations of the game. If one goes back to the GTA3-based games, it’s pretty obvious that the escaping-from-cop mechanics were, well, bad. You were basically not supposed to get away from them once you got to a certain level. The whole “find a pay-and-spray” thing was just never a good idea, and the game improved substantially when it went away from that model entirely.

[2] I did not enjoy San Andreas very much. The missions were tedious, and the open world combined with a lack of anything interesting in the empty space made the game, well, boring.

[3] Much less risky. Assassinating Hitler never works out well.

[4] Game developers today seem extremely hesitant to release a player in an open world with the intent of genuinely letting them wander. This used to be very common: even those series we think of as linear, like Final Fantasy, used to have extended portions of the game where the player was simply thrown into a place and expected to stumble across something resembling a plot five hours in. There’s a reason we’ve moved away from this, of course: it sucked. It was a confusing mess with no reasons given to the player as to why they’re doing what they’re doing. A better-executed example of a good open world would be the first Fallout, where the player has a set end goal that they need to accomplish, and the feeling of wandering with only a vague direction in mind perfectly mirrored the vast world with only scattered civilization. Sure, it could have used a bit more guidance at times. But these pseudo-open-world games with a rigid linear quest and irrelevant durdling available are just total bullshit.

[5] Only a few thousand people bought Deus Ex when it first came out, but everyone who did started an unreadable blog about game design.

[6] This reminded me of the wonderful set of Thieves’ Guild sidequests in Elder Scrolls: Oblivion, where the player undergoes a series of smaller heists that all yield one important item that is needed at a critical time in The Big Final Heist. Instead of seemingly-unrelated quests involving just stealing stuff, I felt like I was really doing something productive toward a set goal, and the payoff during and after TBFH made everything worth it.

[7] Okay, two ways. Still counts.

[8] This includes a car ride where Michael accuses Trevor of being a hipster, and it rivals Portal in how good it is as comedy writing.