Kenneth Chen

Kenneth Chen

Digital Media PhD Student

Kenneth is a doctoral student in the Digital Media PhD program. In his series “Kenneth on Games” he writes about his passion for games and game design.

Nier: Automata has the unenviable task of merging open-world RPG design with hack-and-slash combat. These two directions don’t have to be mutually exclusive, but sometimes the designers need to take a side. Platinum Games is practically the face of the spectacle fighter genre, so it’s no surprise that they lean towards their comfort zone when the option is available. This divide can be noticed in one of the most mundane details: how the characters jump.

WARNING: Mild spoilers for Nier: Automata, footage from a boss battle is used to illustrate a point.

Aerial Physics

Surprisingly, designers spend a lot of time perfecting the physics behind a good, solid jump. There are all sorts of tips and tricks that can make a character’s jump feel snappy and responsive, or smooth and ethereal, or just outright clunky. A lot of games will let you jump higher if you hold down the button longer, for instance.

But one of the most basic rules for good platforming controls is to have the jump happen instantaneously. As soon as you press the button, the character should jump. This is how Mario jumps, and how Meat Boy jumps, and how Ori jumps, and how Faith jumps. It’s a standard design pattern that very few games will violate if they have sections that require precise platforming.

View post on

Nier: Automata is one of those very few games. When you jump, your character takes a brief moment to crouch down and prepare for the jump. Even though it’s very fast (maybe around 1/10th of a second), it’s long enough to be noticeable, and it makes the jumping feel just a tiny bit sluggish, a little bit laggy. Not enough to ruin the experience, but sometimes you’ll get annoyed about missing what should have been an easy jump.

This is because Platinum Games has historically not been in the business of making platformers. Raiden in Revengeance had a delayed jump too (while ninja running). In these games, jumping onto platforms was a way for the level designers to make scenarios that weren’t just flat ground surfaces. Often in Revengeance or any other spectacle fighter, the battles will happen on flat ground surfaces anyway, with height elevations used as occasional level design gimmicks.

Rather than precise platforming, these games needed to fit jumping into the actual game loop: combat. It might be a way to launch an aerial combo, or to dodge an incoming attack, or to enable some kind of ground slam attack. With all of these considerations, platforming responsiveness falls down the list of priorities. If jumping needs to be balanced by adding a windup animation, then so be it, at the cost of feeling slightly more clunky.

In a traditional linear action game, the designers have a lot of control over when the player needs to jump and how these situations are arranged. But in an open world game where players spend a lot of time traversing through the environment, there’s more to worry about than just combat. Platforming is suddenly important again rather than just being an add-on. Eliminating the windup animation starts to sound like a good idea.

However, better responsiveness also turns jumping into a viable defensive option in a system that’s balanced around having only one. The dodge mechanic is very strong and very satisfying, carrying enough depth to fulfill all of the player’s defensive needs. It displaces them, it turns them invincible, it enables a counterattack if timed perfectly, it cancels out of nearly every other basic action.

Dodging is a core part of Nier: Automata, more so than blocking was in Revengeance. A second defensive option could throw this carefully designed system completely off its axis. By making jumping more unresponsive, players will unconsciously gravitate towards the much more fluid dodge mechanic instead. But this is an awkward and unintuitive way to handle the problem.

Is it possible to make a jump that feels good for precise platforming, knowing that players will be spending a lot of time exploring and navigating through the environment, while also reducing the jump’s effectiveness as an evasive maneuver in combat? Can we get the best of both worlds? Here’s my attempt.

The Best of Both Worlds

Make jumping instant, but when you jump you leave behind a “ghost hitbox” that lingers at the ground position you were at when you started the jump. This ghost hitbox will last a very short time (as long as the windup animation takes). If it is hit, you will take the hit. None of this is explicitly conveyed to the player in any way.

This is a very weird solution and it cheats against the player. But it also ensures that damage will go through in any situation where it would have gone through with the windup animation, while also being instant. If the player takes damage this way, they can rationalize it as a strange bit of lag rather than an actual game mechanic, despite the fact that it is an actual game mechanic.

But game design is all about cheating the player. It’s just that most of the time, we don’t notice that it’s happening. Aim assist. Ledge detection. Input buffering. All of these mechanics exist under the hood to affect the player in ways that they shouldn’t be affected in order to nudge them in the designer’s intended direction. They do what the player thinks would have happened, even though the buttons that the player pressed would not have caused it to happen.


Back to the topic of jumping, developers will often implement a small buffer time to count a jump if you have recently run off a ledge. This is one of those very specific mechanics that is completely invisible, but it just makes jumping feel more responsive and fluid. In this illustration from Toto Temple Deluxe’s dev blog, they will count the jump as the first one for 3-4 frames. On the other hand, Dustforce will not do this at all and they explicitly tell you this through menu hints. It may make Dustforce’s jump feel a little more difficult from a usability standpoint, but it’s necessary because it’s a precise speedrunning platformer game.

In Nier: Automata, it feels like this buffer time is comparatively long. This is an effective way to increase jumping’s reliability for precise platforming without also increasing its reliability in combat, but it seems to last much longer than 3-4 frames. If you’re aware of it, you can abuse it for platforming sections. The reason why they need such an exaggerated buffer time is because the windup animation before jumping makes it too difficult for platforming, so they needed a band-aid mechanic. If the problem can be solved at the root, there wouldn’t be a need for external solutions.

However, if jumping were to feel more responsive (even though they are not actually more responsive from a hitbox/combat point of view), wouldn’t the shift in perspective already change the way players act? There have been cases where this has happened. In League of Legends, sometimes the patch notes listed changes for a champion that didn’t actually exist because a designer forgot to commit a change, and yet their win rate changed. What if making jumping more responsive would make players use it more often as a placebo effect?

View post on

This leads us to the hard question: what would change if verticality was more deeply integrated into Nier: Automata? As it stands, verticality is mostly a gimmick that is used occasionally, but the player will be spending 90% of their time on the ground. 3D hack-and-slashes that have truly embraced verticality are rare. The Otogi series pulled it off by enabling players to dodge vertically as well as horizontally. A similar comparison of verticality integration can be made between the Smash Bros series and a more traditional fighting game like Street Fighter V. In the latter, the air game is one tactical component in the core game loop, whereas in Smash Bros the air game is the core game loop itself.

Imagine if ghost hitboxes and windup animations were just removed, and jumping was embraced as a viable evasive maneuver just like horizontal dodging. Make it so that jumping turns you invincible and enables you to perform a counterattack if timed perfectly. Certainly, this would help the platforming, but what effects would this have on the combat? Would it feel faster-paced, or more methodical? Would players feel more cunning, or more reflexive? Would players use jumping and dodging equally, or would they favor one over the other? What if jumping enabled a different set of counterattacks (for example, jump counterattacks launch enemies whereas dodge counterattacks knock them away) so that the choice becomes a tactical decision?

These are the kinds of questions that are better answered through playtesting rather than armchair theory. Game design may not have even been the primary goal. Maybe there’s some kind of deep lore significance behind the way that androids jump. From what I’ve heard about Yoko Taro, anything is possible. But Platinum Games did a good job with the limitations they had, even if it meant that they had to make some this-or-that decisions.