多人竞技游戏实力对比平台:从 Valorant Demo 开始Competitive Skill Battle Platform: Starting with a Valorant Demo
一次产品脑暴:用房间、游戏适配器、评分算法和排名动画,做一个可以扩展到 Apex、CS2 等游戏的多人实力对比平台。A product brainstorm for a multiplayer skill comparison platform built around rooms, game adapters, scoring algorithms, and ranking reveal animations.
This is not a normal stat lookup site. It is a multiplayer comparison experience. Players join a room, choose a game, upload or sync recent data, and the system generates overall scores, ability dimensions, titles, and rankings with game-specific logic.
First MVP
The first stage is a Valorant mock-data version: 5 players, 20 matches each, a single-player rating algorithm, a multiplayer ranking algorithm, a FastAPI backend, a Next.js frontend, ranking reveal animation, and a comparison dashboard. The goal is to validate the core experience before real APIs.
Scoring Model
Valorant starts with six dimensions: Aim, Impact, Survival, Team, Consistency, and Recent Form. The initial overall score uses Aim 25%, Impact 25%, Survival 15%, Team 15%, Consistency 10%, and Recent Form 10%. Later versions can adjust weights by role.
Architecture
The first version should not hard-code Valorant everywhere. The backend keeps a GameAdapter layer: ValorantAdapter, ApexAdapter, CS2Adapter, and more can parse raw data and return one unified Player Rating Result. The frontend only needs overall_score, sub_scores, ranking, awards, and title.
Out of Scope for Now
The first version skips Riot login, real private player data, databases, user accounts, payments, machine learning, and screenshot OCR. It focuses on making the flow from multiplayer data to scoring, ranking animation, and comparison report work.
Next
The priority is to set up the monorepo, define GameAdapter, prepare Valorant mock data, finish the rating algorithm, expose demo APIs, then build the room, ranking, and comparison pages. Once the Valorant demo is stable, Apex can test whether the multi-game architecture holds.