Wordle Statistics from Our Verified Archive
Explore the letters in past answers. These Wordle statistics describe the verified puzzles available on Word Puzzle World, with the sample size and date range shown beside the results.
The archive behind these figures
1,978 verified past puzzles · 2021-06-19–2026-11-17
- Puzzles with a repeated letter
- 615 of 1978 · 31%
- Average standard vowel positions
- 1.79
- Verified past puzzles included
- 1,978
The current puzzle is excluded using our New York reference date, 2026-11-18. The available date range may contain gaps. These are archive word statistics, not player results or a complete history of every puzzle.
Browse the underlying past answersHow many vowels appear in an answer?
A, E, I, O and U count by position. Repeated vowels count each time; Y is excluded.
| Group | Puzzles | Sample share |
|---|---|---|
| 0 | 8 | 0% |
| 1 | 613 | 31% |
| 2 | 1151 | 58% |
| 3 | 203 | 10% |
| 4 | 3 | 0% |
| 5 | 0 | 0% |
Each puzzle belongs to exactly one row. Rounded percentages may not add up to 100%.
Which first letters appear most often?
| Group | Puzzles | Sample share |
|---|---|---|
| S | 295 | 15% |
| C | 170 | 9% |
| B | 149 | 8% |
| T | 131 | 7% |
| A | 125 | 6% |
| P | 122 | 6% |
| F | 108 | 5% |
| G | 105 | 5% |
View all first-letter counts
| Group | Puzzles | Sample share |
|---|---|---|
| A | 125 | 6% |
| B | 149 | 8% |
| C | 170 | 9% |
| D | 93 | 5% |
| E | 63 | 3% |
| F | 108 | 5% |
| G | 105 | 5% |
| H | 57 | 3% |
| I | 31 | 2% |
| J | 15 | 1% |
| K | 20 | 1% |
| L | 81 | 4% |
| M | 98 | 5% |
| N | 35 | 2% |
| O | 31 | 2% |
| P | 122 | 6% |
| Q | 21 | 1% |
| R | 84 | 4% |
| S | 295 | 15% |
| T | 131 | 7% |
| U | 32 | 2% |
| V | 34 | 2% |
| W | 68 | 3% |
| X | 0 | 0% |
| Y | 8 | 0% |
| Z | 2 | 0% |
One first letter per puzzle. These frequencies are not scores for choosing a starting word.
What goes into these Wordle archive statistics?
We include stored puzzle records whose date, number and five-letter answer have passed validation. The date must be earlier than the reference date used by this site. Today’s puzzle and future dates are left out, even when an adjacent-date page is available for a different time zone.
“Verified” describes those checks; it does not mean that this is an official New York Times statistics service. Definitions are a separate layer. A puzzle can be included in the letter counts without a dictionary definition, because its spelling is enough to calculate the measures here.
The earliest and latest dates describe coverage, not completeness. If some dates are unavailable, the page should show the smaller actual sample. Adding missing verified records can change the distribution without anything about the old answers changing.
Repeated-letter frequency counts puzzles, not extra copies. APPLE has a repeated letter because P occurs twice. A word with two different repeated letters still contributes one puzzle to the repeated-letter total. The measure answers “How many puzzles contain a repeat?” rather than “How many duplicate tiles appear?”
A small example of the arithmetic
Teaching example only: the following 100 records are invented. They are not this site’s current figures or real Wordle history.
Imagine a sample with this distribution:
| Standard vowel positions | Invented puzzle count |
|---|---|
| 0 | 5 |
| 1 | 30 |
| 2 | 45 |
| 3 | 17 |
| 4 | 3 |
| 5 | 0 |
There are 100 records and 183 vowel positions in total. Dividing 183 by 100 gives an average of 1.83. The two-vowel group contains 45 records, so it represents 45% of this invented sample.
If 28 of those records contain a repeated letter, the repeated-letter figure would be 28%. That count is calculated separately from the vowel groups. Repeated letters can be vowels or consonants, and a word can belong to any vowel group while also containing a repeat.
The live Wordle archive statistics use the same arithmetic with the actual verified records. The teaching numbers stay in this example and never populate the dashboard above.
Read a pattern without turning it into a forecast
A frequent first letter describes what occurred in the available sample. It does not establish what is due next. A letter that has appeared rarely is not owed a turn, and a familiar pattern does not make another pattern impossible.
The available archive can also be a selective window into the game’s history. A short period, missing dates or a later backfill can affect the figures. Keep the sample size and coverage beside any number you quote or compare.
For an active game, your own green, yellow and gray feedback is more immediate evidence. The five-letter word finder can organize those constraints. The starting-word guide discusses letter coverage and second guesses without treating this archive as a prediction engine.
Looking for a different kind of Wordle statistic?
Your personal win percentage, current streak and guess distribution belong to your own game record. This page does not read that record, combine player submissions or calculate a worldwide average number of guesses.
For the structure of the current answer, use Wordle difficulty today. That page explains an independent rule-based rating and lets you decide when to reveal structural clues. For help with the current board, open today’s Wordle hints.
These Wordle statistics are useful for exploring the available past answers and checking how the counts are made. Keep that scope attached to the numbers when you share them.
Word Puzzle World provides independent puzzle help. Wordle belongs to The New York Times. Play the official Wordle.