Features

• 🔍 Hybrid Multi-Query Search - Combines strict phrase matching with broad keyword search
• ⭐ Smart Scoring Algorithm - Ranks results by relevance instead of just filtering
• 💾 Save Favorites - Bookmark games for quick access later
• 🔖 Recent Searches - Resume previous searches instantly
• 🎲 Random Search - Discover games with randomized filters
• 🌐 Multi-language - UI translations with Groq-powered game summaries
• 📱 Responsive Design - Works on desktop and mobile

How the Search Works

When you search, FindThaGame runs two parallel queries:

1. Strict Query (limit 100): Searches for the full phrase in name and alternative_names.
2. Broad Query (limit 300): Searches individual keywords across name, alternative_names, summary, and keywords.

Results are merged, deduplicated, scored, and the top 200 are returned.

How the Scoring Works

The system calculates a Total Score for each game to determine its ranking. Instead of filtering games out, it adds or subtracts points based on how well they match the input.

The calculation happens in 4 steps:

1. Text Score ($S_{text}$)

Checks if the search query appears in the game's text fields.

$$S_{text} = \text{BaseMatch} + (\text{KeywordHits} \cdot 0.4)$$

Match Type	Field	Points
Exact Match	Title == Query	+3.0
Partial Match	Title contains Query	+1.5
Keyword Field	Matched in keywords	+0.8
Alt Name	Matched in alternative_names	+0.5
Context	Matched in summary/storyline	+0.3

Note: Determining the "BaseMatch" takes the highest scoring single match found.

2. Metadata Score ($S_{meta}$)

Adds points if the game matches the selected filters (Genre, Theme, Mode, Perspective).

$$S_{meta} = \sum \frac{\text{Matching IDs}}{\text{Requested IDs}}$$

Example: If 2 genres are selected and the game has both, it gets +1.0 point.

3. Multipliers ($K_{total}$)

Boosts or reduces the score based on critical constraints.

$$K_{total} = K_{plat} \cdot K_{cat} \cdot K_{status} \cdot K_{company}$$

Filter	Logic	Multiplier	Result
Platform	Match	x1.0	Score stays same
	No Match	x0.3	Score drops 70%
Category	Main Game	x1.0	Score stays same
	DLC (Unwanted)	x0.5	Score drops 50%
Status	Cancelled	x1.5	Boost (only if selected)
Company	Developer	x2.2	Huge boost
	Publisher	x1.8	High boost
	Porting	x1.3	Medium boost
	Supporting	x1.2	Small boost

4. Final Calculation

Combines the base score with multipliers and applies final bonuses/penalties.

$$FinalScore = ((S_{text} + S_{meta}) \cdot K_{total}) + A_{bonus}$$

Where $A_{bonus}$ contains:

• Date: $-0.1$ points for every year outside the range.
• Age Rating: $+0.2$ points if it matches.

Getting Started

Prerequisites

• Node.js 18+
• A Twitch Developer application (for IGDB API access)

Installation

1. Clone the repository

   git clone https://github.com/kbtale/findthagame.git
   cd findthagame
   

2. Install dependencies

   npm install
   

3. Configure Environment

   Create a .env.local file with your Twitch credentials:

   TWITCH_CLIENT_ID=your_client_id
   TWITCH_CLIENT_SECRET=your_client_secret
   AI_API_KEY=your_ai_api_key
   

4. Run Development Server

   npm run dev
   

Credits

• Game data provided by IGDB API
• Translations powered by Groq AI
• Created by Carlos Bolivar

Contributing

Contributions are always welcome! Please check out the Contributing Guide and Code of Conduct before submitting a Pull Request.

If you encounter any security issues, please refer to our Security Policy.

Support

If this tool helped you find a lost memory, consider buying me a coffee ☕

License

This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details.