Original tool
How much does your opening word reveal?
Enter any five-letter word. The analyser plays it against every possible EvoWordo answer and measures how much of the dictionary should remain after the colours come back.
AROSE
Accepted guess and possible answer.
Information score 97.3/100
- Expected left
- 60.9
- Information
- 5.790 bits
- Unique letters
- 5/5
- Position fit
- 8.5%
Average candidates remaining from 2,254 answers. Lower is better.
How evenly the possible tile patterns split the answer pool. Higher is better.
A repeated letter usually tests less on the first turn.
Average frequency of these letters in these exact five positions.
What works
- EvoWordo accepts this as a guess.
- It is also a possible answer, so a lucky first guess can win outright.
- Five different letters, so no slot is wasted on a repeat.
- It hits 5 of the 12 most common answer letters (E, S, A, R, O).
Trade-offs
No obvious first-turn penalty under these measures.
| Word | Expected left | Information | Score |
|---|---|---|---|
| AROSE | 60.9 | 5.790 bits | 97.3 |
| SLATE | 55.8 | 6.073 bits | 97.5 |
| CRANE | 83.0 | 5.647 bits | 96.3 |
| CRATE | 78.6 | 5.745 bits | 96.5 |
| RAISE | 52.6 | 6.009 bits | 97.7 |
Expected remaining answers and entropy are measured against the current EvoWordo answer pool. Different metrics rank words differently, so this is not a claim that any opener is the best. The Daily Challenge schedule is never used.
What the numbers mean
A useful opener separates the answer list into many small groups. The returned tile pattern tells you which group contains the hidden word.
Expected candidates
The average number of answers still compatible with the result. This accounts for how likely each colour pattern is. Lower is better.
Information
Shannon entropy measures how evenly the opener divides the answer pool. Higher values mean the result is expected to teach you more.
Position fit
Common letters are more useful when they are also tested in places where answers frequently use them. This is descriptive, not a win probability.
Method and data
This is calculated from EvoWordo's own production dictionaries rather than a generic online word-frequency list.
The answer pool contains 2,254 curated answers. Every submitted word is scored against all of them with the same green-first duplicate-letter algorithm used by the game server.
Answers are grouped by their five-tile pattern. Expected candidates is the probability-weighted group size; information is the Shannon entropy of those groups. The reference table is recalculated from the same source data, so it changes when the dictionary changes.
The analyser never reads today’s puzzle or its schedule. It receives only the general answer pool stored with the site, so it cannot reveal the current Daily Challenge. Data reviewed 3 November 2026.
For practical advice on turning the result into a second guess, see the strategy guide.
Try it on the board
The Daily Challenge uses the same scoring rules and answer pool as this analysis.