Skip to main content
Custom Solutions · Gaming & Esports

CS2 Team Balancer with FACEIT API Integration

Automated CS2 team balancer that pulls live FACEIT stats and creates fair teams using Bayesian scoring and a multi-objective balancing optimizer.

Cube · Bali, Indonesia
Read the build
Screens

See it on screen

Cube screenshot 1
Cube screenshot 2
Cube screenshot 3
Cube screenshot 4
The change

From manual to automatic

One change at a time — swipe through the four.

  1. 01 / 04
    By hand

    React 19 + Vite web

    Handled manually, step by step.

    Automatically

    Now automated

    Built React 19 + Vite web application with multi-step wizard flow: landing → team count selection → player nickname input → loading → results

  2. 02 / 04
    By hand

    FACEIT Data API v4

    Handled manually, step by step.

    Automatically

    Now automated

    Integrated FACEIT Data API v4 for real-time player data: profile lookup, Elo rating, skill level, lifetime stats, and last 30 match history

  3. 03 / 04
    By hand

    Bayesian weighted scoring

    Handled manually, step by step.

    Automatically

    Now automated

    Developed Bayesian-weighted scoring algorithm combining Elo with adjusted K/D, recent sample size, ADR/headshot context, and player archetypes

  4. 04 / 04
    By hand

    Multi objective team

    Handled manually, step by step.

    Automatically

    Now automated

    Implemented multi-objective team optimizer with locked groups, team locks, slot mirror balance, captain parity, star/weak distribution, and swap-based improvement passes

What we shipped

Inside the build

Built React 19 + Vite web application with multi-step wizard flow: landing → team count selection → player nickname input → loading → results

React 19ViteJavaScriptVercel Serverless FunctionsFACEIT Data API v4

Integrated FACEIT Data API v4 for real-time player data: profile lookup, Elo rating, skill level, lifetime stats, and last 30 match history

React 19ViteJavaScriptVercel Serverless FunctionsFACEIT Data API v4

Developed Bayesian-weighted scoring algorithm combining Elo with adjusted K/D, recent sample size, ADR/headshot context, and player archetypes

React 19ViteJavaScriptVercel Serverless FunctionsFACEIT Data API v4

Implemented multi-objective team optimizer with locked groups, team locks, slot mirror balance, captain parity, star/weak distribution, and swap-based improvement passes

React 19ViteJavaScriptVercel Serverless FunctionsFACEIT Data API v4

Built serverless API endpoint on Vercel with batch processing (5 players per batch) and rate limiting (100ms delay) to respect FACEIT API constraints

React 19ViteJavaScriptVercel Serverless FunctionsFACEIT Data API v4

Created bilingual interface (English/Russian) with React Context-based i18n system and language toggle

React 19ViteJavaScriptVercel Serverless FunctionsFACEIT Data API v4

Developed results display with color-coded K/D ratios, skill level indicators, team average stats, and balance spread percentage

React 19ViteJavaScriptVercel Serverless FunctionsFACEIT Data API v4

Implemented responsive design optimized for both desktop and mobile use during LAN events

React 19ViteJavaScriptVercel Serverless FunctionsFACEIT Data API v4

Added support for 2-10 teams with dynamic player count validation (5 players per team)

React 19ViteJavaScriptVercel Serverless FunctionsFACEIT Data API v4
Results

Measurable results

Swipe the metrics — each one with what it changed in the business, and why it mattered.

  • Process Elimination

    Instant balanced team generation from player nicknames — zero manual stat lookup or spreadsheet work required

    Instant team generation removes the organizational friction that discourages community match organizers, increasing event frequency and player engagement.

  • Statistical Rigor

    Bayesian-weighted composite scoring combines Elo and recent K/D for more accurate skill assessment than Elo alone

    More accurate player assessment leads to consistently competitive matches, improving player satisfaction and community retention.

  • Competitive Integrity

    Multi-objective optimizer targets low power spread, slot balance, captain parity, and balanced star/weak distribution

    Fair matches keep players engaged and returning — lopsided games are the primary reason players leave custom match communities.

  • Scalable Architecture

    Real-time FACEIT API integration processes up to 50 players with rate-limited batch fetching in under 30 seconds

    Reliable performance at any group size ensures the tool works for small friend groups and large LAN tournaments alike without degradation.

Stack

Built on what they already ran

No rip-and-replace, no new SaaS to buy — we connected the tools already in the building.

Runs on
React 19ViteJavaScriptVercel Serverless FunctionsFACEIT Data API v4React 19ViteJavaScriptVercel Serverless FunctionsFACEIT Data API v4React 19ViteJavaScriptVercel Serverless FunctionsFACEIT Data API v4React 19ViteJavaScriptVercel Serverless FunctionsFACEIT Data API v4
Reviews

In the client's own words

I'm genuinely speechless — this thing is insane. The tool works exactly as needed and completely eliminates the headache we had organizing local and online tournaments. I'm honestly blown away.

Slava — Founder at Cube

Related work

More from the same playbook

Let's talk

Got a process like this —
still running by hand?

A 30-minute discovery call. We listen, sketch the schema live, and tell you whether we're the right fit — straight.

— or —[email protected] ↗