Gamers love a challenge, but more than that, we love a steady challenge. We love a game that starts off by being challenging and remains so throughout its five to ten to twenty to hundred hours of play. So, naturally, we love a game with a consistent difficulty curve, one that ramps up at a stable rate, never spiking and never slouching. But the problem with this kind of difficulty curve is that it’s so predictable. A game usually introduces us to all of its mechanics by the halfway point, and then spends its latter half simply throwing tougher and tougher opponents at us. It’s true for any genre or game. Call of Duty, Bayonetta, Dead Space, Need for Speed, Sleeping Dogs—by the time I’m halfway through (sometimes sooner), I’ve seen everything the game has to offer.
Purely by coincidence, two recent games that I’ve played bucked this trend by introducing new mechanics — or at least new contexts for old mechanics — during their climatic final levels. Rather than petering out with an ending that I’m likely to forget in a few days, these games end with a bang earned through the introduction of something new. Not only did I remember these endings, I remember loving these endings. All games should strive to end on such high notes, to have our final memory together be a good memory.
Driver: San Francisco is the best example of how to make an ending stand out from the rest of the gameplay. Driver is a racing game with one clever hook: you can “shift” out of your car into another, like having an out of body experience. The narrative justification for this is that our hero, John Tanner, is in a coma, and the game is actually taking place in his mind.
The game makes good use of this trick in various ways. Sometimes we’re in a team race, shifting between teammates in order to ensure both cross the finish line; other times, we’re trying to take down street racers, but instead of chasing them in a cop car, we can just shift into oncoming traffic and take out the racers head on. Driver slowly introduces us to the many ways “shifting” can be used to augment typical racing game tropes, but then it throws all that out the window for the penultimate mission.
As we near the end of the story, we learn that the bad guy Jericho can also shift between cars. Then he does something that we’ve never seen before. He starts throwing cars at us. In the penultimate mission after Tanner realizes this is a dream, we get the ability to throw cars too.
The controls don’t actually change to incorporate this new mechanic. We still shift into cars as we’ve always done, but now instead of taking control of the vehicle, we automatically chuck it into the air towards Jericho. It’s an entirely new mechanic, but it uses controls we already understand, so there’s no need for an extended tutorial teaching us how to use this new system. Driver does an excellent job building up to this moment by introducing each concept one at a time: Tanner can shift, Jericho can shift, Jericho can throw cars, and now Tanner can throw cars. Each concept gets us accustomed to the overall concept, and then they all come together in s crazy ending that easily becomes the most memorable part of the game. This is a singular moment, and it’s immediately intuitive.
Transformers: Fall of Cybertron doesn’t introduce any new mechanics near its conclusion, but it mixes and matches its mechanics and context to create situations that feel new. It does this multiple times in such rapid succession that it creates a grand sense of scale.
First, the game has us flying between two huge ships trying to destroy the tow cables attaching one to the other. At first this seems no different than any of the previous flying vehicle levels, but then you realize that your machine gun isn’t effective against the cables. You have to switch into robot form to use your rocket launcher, and since you’re in space, your robot form isn’t limited to fighting on a horizontal plane like it was before. Suddenly, the predictable pattern of transforming and shooting has changed: You have to balance the momentum provided by the jet form with the firepower provided by the robot form, a balance of movement that you haven’t had to consider up until now.
Then, in a boss fight, you have to grapple the boss and call in a mortar strike while he’s on the ground. These are two mechanics we’ve used before but not at the same time. By combining them for this one battle at the very end, the game maintains a sense of escalation. It seems like my arsenal is still growing.
The final scene is a one-on-one fight with Megatron, and the game itself transforms from a shooter into a beat-em-up, but the game has prepared us for this. Previous chapters had us controlling big robots that just stomped through the level, and the one-on-one fight controls are no different than those. The left trigger raises a shield, and the right trigger swings a sword. However, the new context changes everything because now our opponent can fight back. Instead of using the shield to protect against petty gunfire, we use it to protect against heavy attacks. Now blocking matters, and we have to think about the timing of our attacks instead of just mashing “stomp.” The game uses old systems in new ways to create unique moments that stand out and make the ending more memorable.
These games understand good pacing. The excitement of a big battle doesn’t just come from its difficulty or spectacle, but also from our familiarity with the systems at play. Games are all about learning systems. This is what we do as gamers. This is what we love. By giving us something slightly new to learn at the last second, something we haven’t seen before but that controls in a way that we already understand, these climactic moments avoid feeling like the predictable peak of a steady difficulty curve. They feel special. Driver: San Francisco and Transformers: Fall of Cybertron have such great endings because their endings represent the narrative and mechanical culmination of everything that has come before — and more.