· 5 min read

Defense should buy something

A successful block, parry, dodge, or dash should change the next moment by buying time, space, position, posture damage, or initiative instead of only postponing the same pressure.

Jacques Callot's small etching shows two duelists meeting crossed blades, each holding a sword and dagger as they lean into the exchange.
Jacques Callot, CC0 1.0

Block the whole string. Six attacks land on the guard, the health bar stays put, and the enemy returns to neutral. Then it starts the string again. The defense worked in the narrowest possible sense: nothing hurt you. After a few repetitions, though, the question becomes what the successful read was supposed to change.

The attack starts before the hit argued that a threat has to reveal its decision before the player needs to answer it. That is only half of the exchange. Once the player gives the right answer, the game has to decide what that answer buys.

It does not always need to buy a free attack. It can buy time, distance, a better angle, posture damage, a resource swing, or simply the ability to keep doing what the player was doing. The useful test is whether the state after good defense differs from the state before it.

Not taking damage is the floor#

Fighting games make this question easy to measure because both characters recover on a clock. Capcom's official Street Fighter V seminar on advantage and disadvantage defines advantage as the gap where one character can move while the other still cannot. A punish is an attack that lands while the original attacker is still stuck in that disadvantage.

The numbers are genre-specific, but the idea travels well. Defense is partly about deciding who owns the next useful moment. If blocking a dangerous move leaves the attacker recovering for longer, the defender can move, reposition, or attack. If the move is safe, the reward may only be that the pressure ended without damage. Both are valid, but they produce very different rhythms.

This is why a defensive mechanic cannot be judged only by its invulnerability window or damage reduction. The aftermath matters just as much. A generous dodge that always returns the player to a losing position can feel more exhausting than a stricter parry that clearly hands them the next beat.

GuardParry or deflectDodge or avoidDash-through
What it costsLow timing demand; may cost structure, stamina, chip damage, or positionPrecise timing and a read on the incoming attackTiming plus a direction or body-position choiceMovement timing while tracking several threats at once
What success can buySafety and information while the enemy finishes its sequenceInitiative, posture damage, stun, or a direct counter windowSpace, angle, balance recovery, or access to the enemy's recoveryContinued motion and offense through a projectile pattern
Failure when the payoff is missingThe player survives but never gets closer to actingThe risky input feels like a stricter version of blockThe player exits danger and immediately has to flee againEvasion interrupts the game instead of preserving its flow
Different defensive actions can pay out in different currencies. They do not all need a punish window, but success should usually alter the next decision.

Sifu makes the payment visible#

Sloclap describes Sifu's defensive options in terms of what they create rather than only what they stop. In its official combat-system overview, normal blocking fills the player's Structure gauge. Dodges create space. A correctly timed parry deflects the blow, unbalances the opponent, and opens a counterattack or throw. Avoiding a strong hit can recover balance and lead directly back into offense.

Those defenses therefore sit on different points of a risk-and-payoff curve. Guarding is dependable but cannot be the permanent answer because Structure eventually breaks. Parrying asks for more timing precision and pays by disturbing the opponent. Avoiding changes position and can recover the defender's own balance. The buttons are not interchangeable shields with different animations.

Sloclap's own boss advice makes the cadence explicit: observe patterns, look for the opportunity near the end of an attack cycle, then counter immediately after a successful avoid or parry. The defense is satisfying because the enemy's sequence has an exit. Learning the pattern eventually changes the player from someone weathering it into someone waiting for the handoff.

Sekiro lets defense advance the win condition#

Sekiro: Shadows Die Twice goes further. Activision's official mechanics guide explains that a deflect both prevents the incoming hit and damages the enemy's Posture. A successful deflect can also create a Counter-slash opportunity, and breaking Posture exposes the enemy to a Deathblow.

That changes the meaning of defense. Waiting safely is not separate from winning; good defense is one of the ways the player moves the fight toward its finish. The result is the familiar sword-clash rhythm where attacking and deflecting can alternate without feeling like offense has stopped. Both sides are still contesting the same resource.

The important part is not that every action game should copy Posture. It is that the game has made the reward for reading an attack concrete. A deflect is not merely a smaller health loss. It leaves the opponent in a measurably worse state and may open the next action immediately.

Some defenses buy motion instead of a turn#

A bullet-hell shooter would become strange if every successful dodge stunned the room and presented a polite counter window. Housemarque's official Returnal survival guide tells players to keep moving, use hip fire for awareness, and dash through most enemy projectiles without taking damage. The dash pays out by preserving movement through the pattern.

That is still a state change. The player has crossed the dangerous line, kept a route open, and often maintained fire while the projectile field continues behind them. The reward is continuity rather than turn ownership. This fits a game where several threats remain active at once and attention, not a one-on-one exchange, is the scarce resource.

The counterexample is useful because “defense should buy something” can otherwise turn into another rule that flattens genres. Sometimes survival is the reward. Sometimes the whole point is to stay inside pressure while finding a path through it. The question is whether the defensive action supports that intended rhythm or just inserts a pause before the same problem resumes unchanged.

Check the frame after success#

When a defense feels dull, I would stop tuning the input window for a minute and inspect the state immediately afterward. Is the enemy still committed? Did spacing change? Did either side spend or gain a resource? Can the player act sooner, reach a weak point, recover balance, escape the corner, or continue attacking? If every answer is no, the mechanic may be asking for skill without giving that skill a consequence.

The opposite failure exists too. If every dodge guarantees a huge punish, attacks stop carrying much threat once their tells are learned. Some moves should be safe. Some blocks should leave the defender under pressure. A long boss string can ask for several correct responses before the opening appears. The payoff only needs to be clear enough that the player can understand what another successful defense is moving them toward.

The Callot duel in the header freezes two blades at contact. The useful part of a parry is not that steel touched steel. It is where the blades, bodies, and initiative can go next. A defensive mechanic deserves the same review: after the hit did not happen, what changed?