wordle/_bmad-output/brainstorming/brainstorm-wordle-clone-2026-07-07/brainstorm-intent.md

Wordle Clone -- Brainstorm Intent

Core Premise

A 5-letter, 6-attempt word-guessing game with green/yellow/gray positional feedback -- familiar Wordle mechanics reimagined around fairness, learning, and adaptive challenge rather than raw competition.

Key Differentiators from Original Wordle

Adaptive Difficulty

• Track per-word global average attempts. Words self-sort into difficulty tiers organically -- no manual labeling.
• Hidden MMR/skill rating drifts with player performance vs. global averages. Surfaceable as badges, not a raw number.
• System starts players at mid-complexity words. Difficulty adjusts over time based on their performance against global averages for each word.

Computer Baseline Solver

• Pre-compute solves for every word using known algorithms. This produces a difficulty reference independent of player volume.
• The same solver engine powers: word difficulty ranking, difficulty-tier assignment, and solvability validation.
• Post-solve: show optimal solver replay as a teaching moment alongside the player's own path through the word.

Percentile Ranking (Not Absolute)

• Result display: percentile among all players who guessed the same word (e.g. "top 15%" not "#14,327").
• More meaningful and less demoralizing than absolute rank. Removes toxicity from leaderboards.

Streak Quality Weighting

• Streaks on harder words count more than streaks on easier words.
• Percentile ranking plus quality-weighted streaks together create a "leaderboard without toxicity."

Standardized Hints (Attempts 5-6)

• Weak hint on attempt 5, stronger hint on attempt 6. Same hints for every player on a given word.
• Keeps global stats comparable. Fair: strong players solve in <=4 attempts anyway. Serves as a retention mechanic for struggling players.

Post-Solve Learning Layer

• Word definition, etymology, example sentence, and picture/illustration shown after solving.
• Transforms the game into a vocabulary-learning experience.

Word Bank

• Curated by recent internet usage frequency (last ~5 years). No obscure or archaic words.
• Words are validated by the computer solver baseline before entering the pool.

Daily Word --> Infinite After-Play

• One daily word shared by all players (the "event").
• After completing or failing the daily, players may continue with more words from the pool as long as they want.

V1 Scope

In:

• 5-letter word length, 6 attempts, green/yellow/gray feedback with highlighted keyboard
• Adaptive difficulty with computer baseline solver and organic word-tier emergence
• Percentile ranking on results screen
• Quality-weighted streak tracking
• Standardized hints on attempts 5 and 6
• Post-solve learning: definition, etymology, example sentence, picture
• Daily word with infinite after-play flow
• Word bank curated by recent internet usage frequency
• Lifetime statistics with attempt-count histogram
• Computer-guess subgame (player picks word, computer solves -- teaches strategy, measures word complexity)

Out (post-V1):

• Real-time 2-player mode (both guess same word simultaneously)
• Variable word length (3-7 letters)
• Position-blind hard mode (yellow/gray only, no green)
• Rotating-target-word chaotic variant
• User-submitted words (rejected permanently)
• Time-penalty mechanics (rejected permanently)

Core System Insight

The computer baseline solver, adaptive difficulty engine, and word-bank curation form a single integrated system. One solver drives ranking, tiering, and validation. Percentile ranking and quality-weighted streaks together make leaderboards meaningful without breeding toxicity. Standardized late-game hints are a retention feature, not a compromise.