Summary
Devil May Cry 3: Dante’s Awakening (2005, PlayStation 2), developed by Capcom, is Bycer’s chosen benchmark for technical depth in action game combat. After the poorly received Devil May Cry 2, DMC3 rebuilt the series with a focus on expressive, player-controlled combat that rewarded improvisation, weapon-switching mid-combo, and aggressive play without repetition. Bycer considers it the top game in the action genre for technical expert play and the clearest demonstration of how far action game combat complexity can be taken (Bycer, 20 Essential Games to Study, see source-20-essential-games-to-study).
Design lessons
- Combat depth comes from combinatorial tools, not complex inputs. Dante’s moveset was built from his two equipped weapons, each with its own move list, plus a combat style that modified his base capabilities. Players crafted their own combos by switching between weapons mid-attack. The learning curve was high, but the ceiling was higher — and critically, the depth was accessible gradually rather than all at once.
- Style ranking drives improvisation. The game’s style meter graded players on not getting hit, maintaining pressure, and — most importantly — not repeating the same combo string. Players who defaulted to the same sequence of inputs were explicitly penalised by the ranking system. This mechanic forced players to explore the entire toolkit rather than settling on one safe strategy.
- 1v1 vs group combat are different design problems. Bycer distinguishes two approaches: games designed around 1v1 encounters (faster, more precise, ebb-and-flow positioning) versus group-focused games (wider, slower combos targeting multiple enemies simultaneously). DMC3 was fundamentally 1v1 — enemies in groups were collateral to the focused engagement with one target. This orientation enabled the precision and depth that group-focused combat cannot support.
- Random boss patterns create organic, skill-expressive fights. In contrast to fixed-pattern bosses (which become puzzles once memorised), random-pattern bosses pull from a set of attacks in unpredictable order. Expert players know what attacks are possible but never know what is coming — so every fight is reactive, not scripted. The final Vergil boss fight is Bycer’s primary example: Vergil’s devil trigger activation changed the fight’s conditions dynamically (Bycer, 20 Essential Games to Study, see source-20-essential-games-to-study).
- Styles as personalisation of playstyle. The four combat styles (Trickster for mobility, Swordmaster for melee depth, Gunslinger for ranged enhancement, Royal Guard for defensive skill) let players orient their approach to Dante without locking them out of the rest of the toolkit. Switching styles was possible between missions, allowing playstyle experimentation across runs.
Key mechanics
- Weapon switching mid-combo: each weapon has its own move list; switching during an attack string creates player-authored combo sequences.
- Style system: four styles modify Dante’s base capabilities; only one active at a time; encourages replay with different playstyle orientations.
- Style meter: grades combat performance on variety, aggression, and damage avoidance; SSS is the highest grade.
- Random boss attack patterns: bosses pull from a fixed set of attacks in unpredictable order; proximity to the player influences attack selection.
- Vergil as recurring boss: fought three times across the game, with the final encounter deploying the devil trigger mechanic as a dynamic escalation.
Historical context
DMC3 released in 2005 as a prequel and course correction after DMC2 was poorly received. It was directed by Hideaki Itsuno and is widely regarded as the creative high point of the original DMC series. Its combat system influenced a generation of action games, including Bayonetta (designed by a former Capcom team), and it remains a reference point in discussions of action game design (Bycer, 20 Essential Games to Study, see source-20-essential-games-to-study). See also god-of-war for a contrasting approach to action boss design.