Hangman as Letter-Frequency Practice

2026-05-17 · Hangman

Hangman is the only word puzzle that taught me what English actually looks like.

I always knew abstractly that some letters are commoner than others. Scrabble puts E, A, I, O, N, R, T, L, S and U at one point for a reason.

But knowing a frequency table is not the same as having an instinct — the thing that lets you look at five blank squares and feel which letter to spend first.

Hangman built that instinct in me without me noticing. After a few weeks of daily hangman, I stopped consciously thinking about which letter to guess. My hand would just move to E, then T or A, then I or O. The order was not memorized; it had become reflex.

Where this pays off for me is the other formats on this site — the tile game especially, since it is the same question asked a different way: which letter is worth spending next.

Whether it makes you a faster reader, I genuinely do not know. That is the kind of claim word games make about themselves rather freely, and I have not seen it hold up. What I can say is that the sense of which letters carry information is shared across every puzzle here.

Every guess is an information bet

The one thing worth internalizing: every guess is an information bet.

You are not asking "is this letter in the word?" so much as buying an answer to "where is it?" — a correct E tells you which positions it fills, which is far more than a yes.

So the best guess is not your favourite letter. It is the letter that removes the most uncertainty, and for short English words that is E.

Across the 884 five-letter answers this game uses, E is 12% of all letters and appears in 462 of the words — a bit over half. Opening with E is right almost every time.

After E, the useful order stops matching the raw frequency table, because it now depends on what you already know. Two E positions tell you the shape is _E_E_ or E___E, and the best next guess is whatever is likeliest given that.

I used to think T and N were the strong follow-ups here. Counting the pool says otherwise: among words containing E, it goes R (185), A (175), T (127), S and L (121 each), and N is well back at 88. R and A are the pair worth reaching for.

The misses are the data

I am not going to hand you a decision tree. Building it by playing is the entire point.

What I will say is that the words you fail are the valuable ones. They tell you which letters your intuition has mis-weighted — keep losing to words with K or W in odd positions and you are under-weighting K and W.

One specific letter most players underweight at five letters: Y. Many five-letter words end in Y (HAPPY, SHINY, FANCY, BOSSY, FUZZY) and players forget to check this. If you have four positions filled and the last is open, Y is worth a guess before R or D.

Beyond frequency: partial-word constraints

Underneath the frequency game is a better one: using the letters you have as constraints.

Three revealed letters in a five-letter word narrows things brutally. Across this game's answer list, 72% of three-letter patterns leave exactly one possible word — at that point you are not guessing, you are reading it off.

_O_ED is the extreme case. A full dictionary has dozens of matches, but only one of them is in the list this game picks answers from, and it is NOTED.

For me, the daily practice has paid off in places I did not expect — typing speed, spelling under fatigue, even reading speed on dense academic text. The skills hangman trains are general. The puzzle is just where they are easiest to practice.

Six lives is plenty. Most of my recent hangman games end with one life remaining or two. The lives are the safety net for the misses you learn from. Use them.