Hitbox vs Hurtbox Explained: The Invisible Boxes That Decide Every Hit in Games

The difference between a hitbox and a hurtbox is the difference between the box that does damage and the box that takes damage, and once you understand that single sentence you understand why your favorite fighting game character has a “good poke”, why Dark Souls feels stricter than Elden Ring, and why a Smash Bros. player can stand on top of a punch without flinching. Every game with collision is built on these invisible rectangles, capsules, and polygons. Players rarely see them, but designers spend months tuning them, frame by frame, pixel by pixel. This guide breaks down what hitboxes and hurtboxes are, how they differ, why they matter for game feel, and how knowing the difference will change how you watch a fighting game forever.

Table of Contents

• What Is a Hitbox?
• What Is a Hurtbox?
• Hitbox vs Hurtbox: The Core Difference
• Other Boxes You Need to Know: Hurtbox, Pushbox, Throwbox, Blockbox
• Why Fighting Games Live and Die by Hitboxes
• Hitboxes in 3D Games: From Souls to Shooters
• Disjointed Hitboxes and Why They Feel Unfair
• How Designers Tune Hitboxes Frame by Frame
• FAQ

What Is a Hitbox?

A hitbox is the invisible region attached to an attack that, when it touches a vulnerable target, registers a hit. It is the offensive box. The sword’s edge, the bullet’s path, the fist at the end of a punch. When a hitbox overlaps with something the game considers vulnerable, damage happens.

Hitboxes are usually simple geometric shapes. A rectangle in 2D, a capsule or sphere in 3D. They are simple on purpose. Collision detection runs many times per second, sometimes 60 or 120 times, and the engine has to test thousands of these intersections against every other relevant box in the scene. The simpler the shape, the cheaper the math, the more boxes you can have on screen without dropping frames.

A hitbox is also temporary. It only exists during the active frames of an attack. A punch in Street Fighter might have 12 frames of startup (the windup), 3 active frames (where the hitbox is alive), and 14 recovery frames (where the character is locked in animation). For 3 frames out of 29, that hitbox can hit something. Miss it, and the punch was just air.

What Is a Hurtbox?

A hurtbox is the invisible region attached to a character that, when touched by an enemy hitbox, takes damage. It is the defensive box. The body that can be hit. The head, the torso, the limbs that count as “you” for the purpose of receiving pain.

Hurtboxes usually wrap around the character’s vulnerable silhouette. A standing fighting game character has a tall hurtbox covering the body. A crouching character has a shorter, wider one. When that character ducks under a high punch, the hurtbox shrinks below the line of the incoming hitbox, and the punch sails over their head. No collision, no hit.

This is why every “low profile” move in fighting games works the way it does. The animation is flashy, but the real mechanic is simple. For a few frames, the character’s hurtbox is missing some vertical real estate. Anything that needs that real estate to connect, misses.

Hitbox vs Hurtbox: The Core Difference

Here is the cleanest way to remember it. The hitbox is what the attack does. The hurtbox is what the character is. A hit happens when an active hitbox overlaps with a hurtbox that does not belong to the attacker. That is the entire rule.

Three consequences fall out of this:

• You can have a hitbox without a hurtbox. Projectiles often do this. A bullet has a hitbox but no hurtbox of its own, because the bullet cannot be damaged. (Some games, like Cuphead, give projectiles hurtboxes specifically so the parry mechanic has something to grab.)
• You can have a hurtbox without an active hitbox. A character standing still has a hurtbox but nothing offensive going on. They can be hit but cannot hit.
• Hitboxes and hurtboxes on the same character can be in totally different places. This is where things get interesting, and where the next sections live.

Other Boxes You Need to Know

Hitbox and hurtbox are the famous pair, but most games have a small zoo of invisible boxes doing different jobs.

Pushbox

The pushbox is what keeps two characters from occupying the same space. It is collision against other characters, not against attacks. When you walk into your opponent in a fighting game and they slide back instead of phasing through you, that is two pushboxes refusing to overlap.

Throwbox or Grabbox

Throws often use a special hitbox that ignores normal blocking and looks for a specific kind of hurtbox, usually called a throwbox. This is why you cannot block a throw the way you block a punch. Different category of collision, different rules.

Blockbox

Some games separate the “can be hit” region from the “can block” region. The blockbox decides whether your raised guard intercepts the incoming hitbox before it reaches your vulnerable hurtbox. This is how high and low blocking work in many fighting games.

Fighting games are the genre where players openly discuss hitbox vs hurtbox interactions in casual conversation. Watch any top player talk about Street Fighter, Tekken, Guilty Gear, or Mortal Kombat for five minutes and you will hear the words “disjoint”, “active frames”, “shrinks the hurtbox”, “the hitbox is huge”. They are reading the invisible geometry in real time.

The reason is that fighting games run on tight, predictable math. Every move has documented frame data. Startup, active, recovery, on-hit advantage, on-block advantage. Each of those numbers describes how the hitbox and hurtbox behave during the move. A “fast” jab is fast because its hitbox appears in 3 frames. A “safe” move is safe because, even on block, the player recovers before the opponent can punish.

If you have ever wondered why a competitive scene argues for months about whether a single move should lose 2 active frames in a balance patch, this is why. Two frames at 60fps is 33 milliseconds. That is roughly half the human reaction time floor for a trained player. Removing those two frames can flip a move from “free win” to “guess and pray”.

Hitboxes in 3D Games: From Souls to Shooters

In 3D games, hitboxes still exist, but they live in three dimensions and they often look like capsules wrapped around bones in the character skeleton. A first person shooter character might have separate hurtbox capsules for head, chest, arms, legs. Each one assigns a damage multiplier. That is how headshots work. The “head” hurtbox is small, the modifier is large, and a bullet hitbox that intersects it deals extra damage.

In Soulslike games, hitboxes are notoriously generous on enemies and stingy on the player. A boss’s giant sword often has a hitbox that lingers a frame or two longer than the animation suggests. That delayed swing is a hitbox active frame extending past the visual peak of the attack. It is not a bug, it is the developers giving the move more presence than the animation alone provides.

Roll invincibility, or “i-frames”, is the inverse. During a dodge animation, the player’s hurtbox is temporarily disabled or moved out of the world entirely. Any hitbox that passes through that empty space connects with nothing and the attack whiffs.

Disjointed Hitboxes and Why They Feel Unfair

A disjointed hitbox is one that exists in a place where the character’s hurtbox does not. The classic example is a sword swing. The blade has a hitbox at the tip. The character holding the sword has a hurtbox around the body. If the opponent attacks the blade itself, there is no hurtbox there to hit. The sword swing wins, the opposing attack hits nothing, and the player on the receiving end yells at their screen.

Super Smash Bros. is the textbook case. Characters like Marth and Lucina have disjointed sword hitboxes that extend well beyond their bodies. Characters with their fists, like Little Mac, have hitboxes that overlap their hurtboxes almost perfectly. This is why “swordies” are considered safer pokers and pure brawlers have to commit harder to every offensive option.

Disjoints are not bugs. They are deliberate design tools. A designer who wants a character to feel ranged, careful, and zone-control oriented will give them disjointed hitboxes. A designer who wants a character to feel risky, in-your-face, and high-reward will keep the hitbox tied to the body.

How Designers Tune Hitboxes Frame by Frame

Hitbox and hurtbox tuning is one of the most thankless jobs in game development. The work is invisible by definition. When it is right, players say the game “feels good”. When it is wrong, they say the game “feels off” and cannot explain why.

The basic workflow looks like this. The animator delivers an attack animation. The combat designer opens an editor that shows the model with collision boxes drawn on top. They scrub through the animation frame by frame, drawing or adjusting hitboxes to follow the weapon, the fist, the foot. They mark the active frames. They tweak the hurtbox to follow the body’s actual silhouette during the move, shrinking it during dodges, extending it during awkward poses where the character leans out.

Then comes the part nobody enjoys. Playtesting. The designer plays the move thousands of times, fights it from every angle, and decides whether the hitbox feels too big, too small, too early, too late. A move that lands “too easily” gets a smaller or shorter hitbox. A move that whiffs visually but still hits gets the hitbox shrunk, or the active window cut. The tuning is iterative, painful, and almost entirely about feel.

This is also why patch notes for fighting games are so granular. “Move X, frame 6 hitbox reduced by 5 pixels on the vertical axis.” That single line represents a designer staring at a frame for an afternoon and deciding it had to go.

Related reading on the topology of hidden game systems: how procedural generation builds worlds you never see twice, how cartridge saves stored your progress before cloud existed, and how speedrunners abuse the same hitbox geometry to clip through walls.

FAQ

Is a hitbox always a rectangle?

No. 2D games often use rectangles because they are computationally cheap and easy to author. 3D games use spheres, capsules, or compound shapes built from multiple primitives. Some modern engines support arbitrary convex polygons. The principle is the same regardless of shape: it is a region the engine tests for overlap.

Can you see hitboxes in normal gameplay?

Usually no. They are debug tools, not gameplay elements. Some fighting games include a “training mode” toggle to show hitboxes for learning purposes. Modders and tournament broadcasters sometimes overlay them too. In normal play, players read hitboxes through animation, hit sparks, and frame data, not by seeing the boxes directly.

What is the difference between a hitbox and a hurtbox in one sentence?

The hitbox is the part of your attack that can damage someone. The hurtbox is the part of your body that can be damaged.

Why do some games have “bad hitboxes”?

When players complain about “bad hitboxes”, they usually mean one of three things: the hitbox extends further than the animation suggests, the hurtbox is bigger than the model implies, or active frames last longer than they should. All three are common in early development and get tuned during balance passes, sometimes for years after launch.

Do hitboxes exist in non-combat games?

Yes. Anything that involves picking up an item, triggering a door, walking on a pressure plate, or interacting with an object uses some form of collision volume. They might not be called hitboxes, but they are the same idea: an invisible region the engine tests for overlap.

Conclusion

Hitboxes and hurtboxes are the closest thing video games have to a hidden language. Every hit you land, every dodge that saves your life, every “how did that miss” moment in your favorite fighter, all of it traces back to two invisible boxes deciding whether geometry overlapped on a single frame. Once you see the boxes, you cannot unsee them. The game gets deeper, the design choices get sharper, and that one cheap move you always complained about suddenly has a name and a reason.

---