Chapter 26 Quiz — Combat Design
Fourteen questions. A mix of multiple-choice, short-answer, and applied-design. Aim for 11/14 to call it passed. An answer key with explanations follows the questions — read the explanations even on items you got right. The "why" matters more than the correct letter.
Questions
Q1 — Multiple Choice
Frame data describes, in order, the three primary phases of an attack as:
A. Start, middle, end
B. Startup, active, recovery
C. Buffer, commit, cleanup
D. Input, processing, output
Q2 — Multiple Choice
The primary purpose of hit pause (hit stop) is to:
A. Give the server time to confirm the hit in multiplayer games
B. Communicate the impact moment by briefly freezing both attacker and target
C. Hide animation blend imperfections during the transition from attack to idle
D. Prevent rapid-fire inputs that could break the combat system
Q3 — Short Answer (1-3 sentences)
What is the difference between a hitbox and a hurtbox? Why should they be abstracted shapes (capsules, boxes) rather than derived from the rendered character mesh?
Q4 — Multiple Choice
A dodge roll's i-frames (invincibility frames) are important because:
A. They allow the roll animation to blend smoothly with other animations
B. They skip collision detection entirely, saving CPU time
C. They create a narrow window during which the player survives attacks that would otherwise hit, producing the "just barely dodged" feel
D. They prevent the player from being hit by their own hitboxes
Q5 — Short Answer (1-3 sentences)
The chapter argues that fewer combat verbs produce deeper combat, not shallower. Name a specific game cited as evidence, state how many verbs it has, and briefly explain how the constraint produces depth.
Q6 — Multiple Choice
The "tell-to-active ratio" rule of thumb is:
A. Tells should be at least half the length of the active window
B. Tells should equal the active window exactly for perceived fairness
C. Tells should be at least 2:1 longer than the active window so the player has time to read and respond
D. Tells are not necessary for small enemies; only bosses need them
Q7 — Multiple Choice
In a hybrid combat system like Dark Souls, what does FromSoftware's choice to allow almost no cancels on attack recovery achieve?
A. It simplifies the animation pipeline
B. It reduces the number of network packets required per attack in online play
C. It forces players to commit to attacks, making the timing of each swing a meaningful decision
D. It limits the maximum DPS a player can achieve, balancing the game
Q8 — Short Answer (2-4 sentences)
Describe one concrete difference between pure number-tuning difficulty (HP and damage scaling) and mechanics-changing difficulty. Which approach does the chapter favor, and why?
Q9 — Multiple Choice
The typical structure of a multi-phase boss fight presented in the chapter is:
A. Phase 1 → Phase 2 → Phase 3 with identical movesets that scale in damage
B. Intro → Phase 1 → Transition → Phase 2 → (optional Phase 3) → Kill Moment
C. Tutorial → Boss Intro → Combat → Cutscene resolution
D. Easy phase, hard phase, and a gimmick phase
Q10 — Multiple Choice
The chapter's design rule for Phase 3 of a boss fight is that it should:
A. Introduce the most difficult and unique attack in the fight
B. Reuse phase 1 mechanics but at phase 2 speeds
C. Almost never introduce a mechanic the player hasn't already seen — the player's job in phase 3 is execution, not learning
D. Be at least as long as phase 1 and phase 2 combined
Q11 — Applied Design
You are designing a light attack for the player character in a 2D action-platformer. The game runs at 60 FPS. Propose concrete starting values (in frames) for:
- Startup
- Active
- Recovery
- Hit-pause on connection
Justify each value briefly.
Q12 — Short Answer (2-4 sentences)
What is the rock-paper-scissors trap in encounter design? Describe one of the three fixes the chapter recommends.
Q13 — Multiple Choice
A lock-on camera in 3D combat is most appropriate for:
A. Large multi-enemy encounters in wide arenas
B. Stealth gameplay where the camera should be player-controlled
C. Duel-focused combat where a single enemy is the primary threat
D. Games with extensive vertical gameplay, like platformers
Q14 — Short Answer (2-4 sentences)
The chapter claims that a good boss fight has a "rhythm." Describe a specific technique for playtesting whether your boss has one, and what an arrhythmic boss feels like to play.
Answer Key and Explanations
A1: B — Startup, active, recovery.
These terms are imported directly from the fighting-game community, which formalized them decades ago, and they have become the standard vocabulary for any real-time combat system. "Start, middle, end" is descriptive but imprecise; "buffer, commit, cleanup" is not in common use. If you mixed these up, go back and re-read the "Hitboxes, Hurtboxes, and Frame Data" section — this vocabulary will come up in every combat conversation you have as a designer.
A2: B — Communicate the impact moment by briefly freezing both attacker and target.
Hit pause is the single most important game-feel technique in combat. Without it, attacks look like they pass through enemies; with it, each hit reads as a connection. Answer A confuses hit pause with netcode features; C is a minor side-effect but not the purpose; D is unrelated. Implementing hit pause in your prototype is, in my experience, the single highest-impact combat-feel change you can make.
A3: Sample acceptable answer:
A hitbox is attached to the attacker and is active during the attack's active frames; it defines where the attack can deal damage. A hurtbox is attached to any character and defines where that character can be damaged. They should be abstracted shapes, not derived from the rendered mesh, because the designer needs direct control over the "feel" of attacks — a hitbox slightly larger than the visual weapon reads as generous, a hurtbox smaller than the crouching silhouette rewards crouching defensively. Using the mesh as the collision source couples gameplay to visuals in a way that makes tuning impossible.
A4: C — They create a narrow window during which the player survives attacks that would otherwise hit.
I-frames are pure design intention. They are what let a dodge roll dodge rather than merely reposition. The "I barely dodged that" feeling is an i-frame-timing feeling. Answer A conflates animation blending with invincibility; B misunderstands the implementation (collision is still running, but hurtbox is disabled); D is nonsense.
A5: Sample acceptable answer:
Hyper Light Drifter has three verbs (sword, gun, dash), and the combat is rich because the three verbs have deep timing relationships with each other. Dash provides i-frame dodging; sword hits refill gun ammo, so weaving melee and ranged is a resource loop; enemies are designed to pressure all three verbs. The depth emerges from interaction, not from list length. (Hollow Knight with four verbs and Sekiro with a short verb set routed through the central deflect mechanic are also valid citations.)
A6: C — Tells should be at least 2:1 longer than the active window so the player has time to read and respond.
Human reaction time on an expected cue is ~180ms (11 frames at 60 FPS); on an unexpected cue, ~250ms (15 frames). An attack with only 10 frames of tell is genuinely unreactable. 2:1 is a starting rule, not a hard law, but it is where experienced combat designers calibrate to.
A7: C — It forces players to commit to attacks, making the timing of each swing a meaningful decision.
This is a deliberate design choice, not a technical limitation. By removing cancel safety-nets, FromSoftware makes every attack a real risk, which in turn makes the player read enemies carefully before swinging. The feel of the game — weighty, committed, tense — is produced by the cancel policy. Elden Ring relaxed this slightly with jump attack cancels, a small concession to accessibility.
A8: Sample acceptable answer:
Pure number-tuning difficulty inflates enemy HP and damage, making fights longer rather than more challenging. The player's actions, reads, and strategies are unchanged — only the time-on-task grows. Mechanics-changing difficulty modifies enemy behavior (new attacks, AI coordination, changed affordances like fewer heals), requiring the player to actually play differently. The chapter favors mechanics-changing because it produces a genuinely different experience rather than a time-tax. The Bloodborne BL4 community is cited as evidence — the game's difficulty survives even with minimum stats, showing that the real difficulty is mechanical, not numerical.
A9: B — Intro → Phase 1 → Transition → Phase 2 → (optional Phase 3) → Kill Moment.
This skeleton is the baseline, not a law — but every shipped modern boss fight uses some subset of these beats. Answer A describes a lazy boss; C confuses boss structure with narrative structure; D is not a structural rule.
A10: C — Almost never introduce a mechanic the player hasn't already seen.
Phase 3 is an execution test, not a learning test. Players who die in phase 3 to a brand-new mechanic blame the game for changing the rules. Players who die in phase 3 to a tightened combination of learned mechanics blame themselves. The second is the experience you want. Hollow Knight's Nightmare King Grimm is the textbook example — phase 3 is pure combination of phase 1 and phase 2 attacks.
A11: Sample acceptable answer:
Startup: 6 frames (100ms at 60 FPS). Short enough that the attack feels responsive, long enough for a brief wind-up animation.
Active: 3 frames. Short active windows prevent "hit through" bugs and force precise spacing. A frame count of 3 is standard for light attacks.
Recovery: 12 frames (200ms). Long enough that a whiffed attack has a real cost; short enough that combat tempo stays brisk.
Hit-pause on connection: 5-6 frames. Short enough not to interrupt rhythm; long enough to read as an impact.
These values place the total animation at 21 frames (350ms), right in the pocket of light-attack conventions from Street Fighter to Dead Cells. Further tuning happens in playtest.
A12: Sample acceptable answer:
The rock-paper-scissors trap is designing combat around rigid type matchups where each enemy has a specific weakness that the player must match with a specific tool. When the matchup is too strict, combat devolves into menu-swapping — the player stops fighting and starts sorting. Three fixes are suggested: (1) make matchups multipliers rather than absolutes, so every weapon works but some work better; (2) make switching itself a fun combat verb (as in Devil May Cry), so switching IS combat rather than interrupting it; (3) limit matchups to specific high-salience enemies like bosses, so trash mobs don't demand menu management.
A13: C — Duel-focused combat where a single enemy is the primary threat.
Lock-on excels in 1v1 because it frames both combatants and angles the player's movement vector usefully around the target. It struggles with multiple enemies (can only focus one), large enemies (camera tilts into the sky), and vertical gameplay (locks the pitch). Answer A describes a free-camera use case; B is stealth, which usually uses player-controlled cameras; D is unrelated.
A14: Sample acceptable answer:
One technique: record 60 seconds of your boss fight and close your eyes, listening only to the audio. Can you hear a rhythm in the attack cycles and the player's responses? If yes, your boss has a musical pulse — which is what you want. An arrhythmic boss feels either metronomic and boring (attacks on a perfect clock, no variation) or chaotic and unfair (attacks at random intervals, no predictable pulse the player can learn). Good boss rhythm is jazz-like: a common pulse, with tasteful variation around it, letting the player learn the spine while staying awake to the variations.