A brainstorm session exploring what a Wordle clone becomes when you question every assumption — then rebuild from first principles.
July 7, 2026 · SCAMPER · What If · Worst Possible Idea · Kill the Crown Jewel
Scroll to explore
Technique One
The SCAMPER Method
Seven lenses applied to the core Wordle formula. Each letter opens a door to a different kind of reinvention — substitute the familiar, combine the unexpected, adapt to the player, modify what matters, put the mechanic to new use, eliminate the unnecessary, reverse the perspective.
S
Substitute
Variable word-length mode (3–7 letters) for fresh challenge dimensions. The 5-letter anchor stays for daily play, but substitution opens the game to broader word discovery.
C
Combine
Post-solve learning layer: definition, etymology, example sentence, and illustration. The game becomes a vocabulary builder — each solve is a micro-lesson you actually remember.
A
Adapt
Adaptive difficulty driven by global averages. Words self-sort into tiers — no pre-labeling. A hidden MMR/skill rating drifts with each performance, surfaced only as earned badges.
M
Modify
Percentile ranking instead of absolute position. "Top 15%" means something — "#14,327" means nothing. Streaks weighted by word difficulty: five hard-word wins beats ten easy ones.
P
Put to Use
Computer-guess subgame: you pick the word, the solver plays it back. Teaches strategy through observation. Optimal solver replays become a teaching moment post-solve.
E
Eliminate
No penalty for fast guessing — speed doesn't prove skill. No user-submitted words — curation matters. No obscure/archaic words — only vocabulary with recent internet-frequency usage.
R
Reverse
Reverse the roles. A computer-guess subgame where the player sets the target and watches the machine deduce. It measures true word complexity and teaches optimal strategy by demonstration.
Technique Two
What If Scenarios
Diverge without restraint. Each "what if" opens a branch — some lead to features, some to dead ends, and some reshape the entire product architecture. The branches that survive are the ones worth building.
What if the daily word was just the beginning?
accepted Daily word first, then prompt to continue — infinite play after the daily event. Two modes, one session.
What if we couldn't trust community-submitted words?
rejected No user submissions. Word bank curated from recent internet-frequency usage only — nothing obscure, nothing archaic, everything solvable.
What if word length was variable?
post-v1 3-to-7-letter mode unlocks entirely new strategy spaces. Daily word stays at 5 — the variable mode is an optional skill expander.
What if position didn't matter?
post-v1 Position-blind mode: yellow and gray only, no greens. A hard-mode variant that forces pure letter-set reasoning without positional crutches.
What if the target word changed each guess?
post-v1 Rotating target — chaotic, disorienting, possibly brilliant. A variant for players who've mastered the standard constraints and crave chaos.
What if two players raced on the same word?
post-v1 Real-time 2-player mode. Same word, simultaneous guesses. Social Wordle without the toxicity of direct competition — you share the word, not the scoreboard.
Technique Three
Worst Possible Idea
Sometimes you find the boundary by crossing it. The worst ideas are mirrors — they show you, by inversion, what you actually value. Here, the worst idea revealed the irreducible core of the game.
WORST IDEA
Replace green / yellow / gray position feedback with a single numeric distance score.
A single number — say, a Levenshtein distance or similarity percentage — replaces the colored tiles. No positional hints. No letter-level feedback. Just one opaque score per guess.
Position feedback isn't decoration — it's the compression algorithm that makes 5-letter words solvable in 6 attempts. Without it, you'd need 12+ guesses. The green/yellow/gray system is a brilliantly efficient information channel. It stays.
Technique Four
Kill the Crown Jewel
The crown jewel of Wordle is its tri-color position feedback. What if we killed it? Below, the tiles shatter — and what emerges from the wreckage is the realization of why they're indispensable.
G
Y
W
G
Y
Green / yellow / gray feedback is the irreducible minimum. It's the game's entire information architecture compressed into three colors. Remove it and you don't have a harder Wordle — you don't have Wordle at all. The constraint of 5 letters and 6 attempts only works because each guess delivers 5 data points of positional truth. Kill the crown jewel and the whole system collapses. So we keep it — and build everything else around it.
Synthesis
The Unified System
Every surviving idea from every technique converges into a single coherent architecture. The solver engine is the gravitational center — it drives difficulty tiering, validates word solvability, generates baseline percentiles, and powers the teaching layer. Everything connects.
Takeaways
Four Insights That Define the Architecture
01
One Engine, Many Purposes
The computer solver baseline isn't a feature — it's infrastructure. The same engine drives word difficulty tiering, percentile ranking baselines, solvability validation for curation, and the teaching replays. Build it once; it feeds everything.
02
Leaderboard Without Toxicity
Percentile ranking and quality-weighted streaks together create competitive motivation without the zero-sum anxiety of absolute ranks. You're not #14,327 — you're in the top 12% on a hard-word streak. That feels earned, not compared.
03
The Solve Is Where Learning Begins
Post-solve is the moment of maximum curiosity. Show the definition, etymology, and example sentence right then — plus an optimal solver replay. The player just invested mental effort in the word; now they'll actually remember what it means.
04
Standardized Hints Are a Retention Mechanic
A weak hint on the 5th attempt and a stronger one on the 6th — identical for every player — catches struggling players before they bounce. Good players solve in 4 or fewer, so the hint doesn't distort stats. It just keeps people in the game.