A Chinese company approached me with a specific ask: build a casual, hero-based PvP game in the spirit of League of Legends and Dota 2 — but designed for everyday players, not hardcore fans.
I had never played or designed in this genre. Rather than treating that as a disadvantage, I treated it as a design constraint: I had to understand the genre from first principles, not habit. I found Auto Gladiators — a popular Dota 2 mode — and studied it carefully. That became the reference point.
The core tension of the entire project was established immediately: take a deep, complex genre and make it feel effortless — without hollowing it out.
A casual hero-based PvP game for WeChat — a platform dominated by lightweight, accessible experiences with massive reach in the Chinese market.
Auto Gladiators (Dota 2). Rich, deep, heavily systemic. The opposite of casual — but the closest thing to what the client wanted, at its core.
Zero experience with auto-battlers. I approached the design through research, play, deconstruction — and a commitment to understanding what makes the genre fun before simplifying it.
Designing a casual auto-battler from scratch — in a genre I'd never worked in — meant facing three distinct design problems simultaneously. Each one could break the game if handled poorly.
The reference (Auto Gladiators) was built for hardcore players. Stripping it down for casual audiences risked destroying what made it compelling. The core loop had to survive the simplification intact.
Six heroes, each needing a distinct role and feel — without overlapping so much they became reskins. Every hero had to offer a genuinely different playstyle, not just a different coat of paint.
37 cards across three tiers. Every card needed to be viable — the system had to reward strategic thinking, not just push players toward obvious meta choices.
The card system was the strategic backbone of Battle of Legends. Players build their approach by selecting and upgrading cards mid-match — responding to both their own hero's strengths and what they observe in their opponent.
The challenge: with 37 cards, it's easy to end up with a handful of dominant picks and a graveyard of unused options. I spent months iterating on cost, upgrade curves, and effect values to prevent any single strategy from becoming the obvious answer.
Every card needed a reason to exist. Not just in theory — in practice, against real opponents, with different heroes. The tiered structure helped: Normal cards are reliable workhorses, Epic cards offer specialization, and Legend cards are high-commitment power spikes.
The real work was in the numbers — setting upgrade costs and effect values so that no tier felt mandatory, and no single card felt like the only choice.
The hardest design constraint: six heroes, each with a distinct identity and playstyle — with enough overlap in the underlying systems to be balanced, but enough difference to feel unique. I ran dozens of test iterations to catch the moments where two heroes started feeling interchangeable.
Some heroes were sharing too many traits — which was turning them into reskins with different names. I caught this through repeated playtesting: when two heroes started solving the same problems in the same way, I knew something had to change. The fix was always in the details — a stat shift, a modified ability interaction, a different upgrade curve. It took many iterations, but the goal was clear: every hero had to offer something no other hero could. — Reza Hassani, Principal Game Designer