There are a lot of ephemeral factors that impact a player’s ability to enjoy a game, especially on the PC, where the huge number of potential system configurations mean that it’s impossible to playtest every possible arrangement. And I’m not talking about things like writing, or level design, or even bug-testing. I’m talking about the little things, the ones that can ruin a game, despite never being the features that go on a box in big bright marketing-speak.
Things like: How fast does the cursor move in the menu? Is my character’s walking speed fast enough to get around without being annoying? Are there things that take two button clicks, when, with slightly better design, they could take one? Game designers have learned, over the years, how to make games feel more convenient, without sacrificing difficulty or quality (and the fact that they’ve had to learn over time is what can make the experience of going back and playing older games such a daunting one). We play games to enjoy ourselves, whether that means being challenged, or amused, or scared. We don’t play them to have our time wasted.
The single biggest offender (for me, at least) are load times. When a game can’t get a handle on these, it turns every choice to load a save, or enter a new area, into a game of time management. Do I have the patience to wait for this to load? Or do I just skip the potential content in that area? How many times, when you were playing Mass Effect 1, did you think about checking if your squadmates down in the Normandy had something fresh to say, but said ‘Screw it’ when you remembered the 20 second elevator ride between you and them? Those moments of frustration damage player mood, and they discourage exploration.
The Thrill of Experimentation
All of this is to say that, if more games used the saving/loading system in place in Tom Francis’s excellent stealth/hacking/spring-loaded-pants game Gunpoint, gaming would be a much better place. The system is very, very simple: The game seamlessly autosaves every few seconds. When you die, you’re offered several choices. You can reload the most immediate save, a save from about 7 seconds ago, or one from 20 or so seconds ago. You can also restart the mission.
The genius of these saves is that they’re perfectly timed to the game’s pace. The first option is usually good enough to reverse whatever dumb thing you did to get yourself killed. The one before it undoes the decision that put you place to make that dumb decision, and the oldest save is almost always good enough to undo everything you’ve done wrong in the level, period (and if not, the levels are bite-sized enough that a restart isn’t that onerous).
This does two things. One, it makes it clear how well Gunpoint’s individual moments are designed, with clear, interesting choices offered to the player every five seconds or so. “Do I leap on a guard, making him a potential witness?” “Do I trigger a lightswitch that’s been wired to that door and hope I can dodge the guard?” “Do I smash through the window in a Hail Mary desperation move?” The save mechanic makes it clear just how fast-paced and exciting Gunpoint can be.
More importantly, though, it makes choices like those in the above paragraph possible, because it completely removes the idea of ‘failure’ from the game. There are only ideas that work, and ideas that don’t. Every move forward in the game is an experiment, one that outputs information about how the world works, and without the player being punished or penalized by having their progress or their time stolen from them when the result is negative. In games with limited saves and long load times (and ‘long,’ in this case, means more than two seconds or so, long enough to feel like the game’s limitations are holding you back), trying things that might be damaging or ultimately pointless is discouraged, because the player is punished for putting the game into a failure state. In Gunpoint (where experimentation is key, because the game’s central ‘Crosslink’ mechanic encourages players to rewire levels extensively, often with unforeseen consequences), experimentation is actively rewarded with new information to tackle the game’s levels.
Risk and Reward
I’m not arguing that every game should adopt this system. I wrote recently about how failure is used in the stealth game Monaco, where being caught by guards is the start of a crescendo of violence and escape. Monaco is a game that’s fundamentally about risk, about wagering your characters’ lives to accomplish your goals, and to remove risk from that scenario would be folly. The ability to undo a mistake would rob the game of all of its tension. The thrill of Monaco is in escaping by the skin of your teeth against tough odds. It can be emotionally devastating to fail a mission at the last moment, but deeply thrilling to succeed.
Gunpoint lacks those same highs (and lows) – it rarely engages the emotions (through gameplay, at least – the writing can be quite evocative). Instead, it offers a more sedate experience, presenting the world as a laboratory for the player to explore. Its most thrilling moments are the joy of discovery, when an elaborate series of traps you’ve Crosslinked together go off perfectly. And the save system enables those mechanisms to be explored and found, with no threat of pain or death.
Load times are a sign of technical limitations, but they are also a question from the game designers of how much the player will put up with. Will we accept 2-second load times, in exchange for a larger level or better graphics? A 10-second one, to get nicer-looking textures? I think it’s significant that Tom Francis, who for many years reviewed games for PC Gamer before trying his hand at game production, answered that question definitively with Gunpoint: No. Stealing time from the players is always, for any reason, wrong.
I tend to agree.