Word Games

Learning New Words Through Puzzles

📅 May 13, 2025⏱ word-games min read
Run a rack through a solver and you will get valid words you have never met. Most people either ignore all of them or try to learn all of them, and both are wrong. The useful move is triage, because unfamiliar words are not equally worth your time.

The problem with learning everything

A seven-letter rack run through a solver can return sixty or more valid words. On a typical rack, perhaps fifteen will be unfamiliar. At three racks a game, that is forty-five unknown words per session.

Nobody learns forty-five words a session. Trying produces a list you abandon within a week, which is worse than learning three, because three learned is three more than none.

So the question is not how to learn unfamiliar words. It is which ones.

Three buckets

Learn. Short, likely to recur, or built on a pattern you keep meeting. These are worth real effort.

Note. Interesting but not urgent. Write them down, do nothing else, and see whether they appear again. Recurrence is a reliable signal of usefulness.

Ignore. Long, obscure, or dependent on a letter combination you will rarely hold. These consume study time and return nothing.

Most unfamiliar words belong in the third bucket, and being willing to discard them is what makes the method work.

What goes in the Learn bucket

  • Two and three-letter words. Always. They recur constantly and there are few of them. If a solver shows you a valid two-letter word you did not know, learn it that day.
  • Words using awkward tiles. Anything with Q, Z, X or J. These are the words that rescue otherwise stuck racks, and there are not many of them.
  • Words built on a pattern you already know. If you know RETAIN and the solver shows RETINA and RETSINA, that is a family worth taking whole.
  • Anything you have now met twice. Second appearance is the cutoff. A word that shows up twice will show up again.

What to ignore without guilt

  • Eight and nine-letter obscurities. They almost never get played, and when they do you rarely have the tiles.
  • Words needing a specific rare combination. A valid word requiring both a Z and a W is not a practical investment.
  • Inflections of words you already know. If you know the base, the plural and past tense usually follow without study.
  • Anything that only exists in one word list. If it is not valid under the list you actually play, it is trivia.

The last one matters more than people expect. Word lists differ, and a word that works on a website may be rejected in your game, so check which list a tool uses before investing in its output.

The three-word rule

Cap it. Three new words per session, chosen from the Learn bucket, and nothing more.

Three per session at four sessions a week is over six hundred words a year, all of them drawn from real situations and filtered for relevance. That is more than almost anyone achieves by trying to learn everything.

If there are more than three candidates, choose the shortest. Short words recur more often, which means more retrieval practice for free.

What to do with each one

Four steps, about two minutes total per word:

  • Look up what it means. Not optional. A word learned as form only sits at a level you cannot use.
  • Find its relatives. A contains finder will show other words sharing its unusual feature, and learning the group is barely harder than learning the one.
  • Note where you met it. The rack, the game, the situation. Context gives you a retrieval cue that a bare list does not.
  • Scramble it for later. Write it as an anagram and solve it in two days, which converts the note into a retrieval test.

The Note bucket is doing real work

The middle bucket looks passive and is not. Writing a word down without studying it does two things.

It creates a weak trace, so the next encounter feels like recognition rather than a first meeting — and recognition on second contact is a much better starting point for learning.

And it produces the recurrence signal. When you find you have written the same word twice in a month, that word has demonstrated its own usefulness and can be promoted to Learn with confidence, rather than on a guess.

Keep the Note list somewhere low-effort. A single page, one word per line, no definitions. If it takes more than three seconds to add a word, the list will not survive.

Checking before you commit

Two checks worth running before putting a word in Learn:

Is it valid in your list? Different tournament lists accept different words. Verify against a tool that states which list it uses.

Is it actually rare? Some words feel obscure and are common in play because their letters are common. An anagram dictionary shows you what else those letters make, which is a decent proxy for how often the word will come up.

A word made of A, E, I, N, R, S and T will appear far more often than one requiring a V and a K, regardless of how familiar either sounds.

What this looks like after six months

A Learn list of maybe three hundred words, all of them met in real play. A Note list that turns over constantly, with most entries never promoted. And a large number of words you saw once, ignored, and never needed — which is the correct outcome, not a failure.

Try These Tools Free

270,000+ words, under 200ms, no signup needed.

Open Anagram Solver →

Frequently Asked Questions

Should I learn every unfamiliar word a solver shows me?
No. A single rack can return fifteen unknown words, which is unmanageable. Sort them into learn, note and ignore — and most belong in ignore.
Which unfamiliar words are worth learning?
Two and three-letter words, anything using Q, Z, X or J, words built on a pattern you already know, and anything you have now encountered twice.
Which should I ignore?
Long obscurities, words needing a rare letter combination, plain inflections of words you know, and anything not valid in the word list you actually play under.
How many new words per session is realistic?
Three. At four sessions a week that is over six hundred a year, all drawn from real situations — more than almost anyone manages by trying to learn everything.
What is the point of writing down a word without studying it?
It creates a weak trace so the next encounter feels like recognition, and it produces a recurrence signal. A word you write down twice in a month has proved it is worth learning.
How do I tell whether a word will actually come up?
Look at its letters. A word made from A, E, I, N, R, S and T will appear far more often than one needing a V and a K, regardless of how familiar either sounds.
M
MyAnagramSolver Editorial Team
Word game strategists and puzzle enthusiasts. We test every technique on real games.
Published: May 13, 2025 • Updated: September 22, 2026