PlainSpell Rankings
Most Confusable English Word Pairs
Word pairs most likely to be confused, ranked by a confusion-risk score built from word frequency and visual similarity.
- 50
- ranked entries
- 25
- #1 score
The verdict
“that vs this” leads all 50 ranked entries with 25 score, ahead of “that vs they” (45).
- 25
- #1 - that vs this
- 50
- ranked entries
- 25→176
- score range
According to Wiktionary data (CC BY-SA, analyzed May 2026). Rankings are computed directly from the corpus by the stated metric.
What This Ranking Tells Us
Confusable word pairs are words that look alike, sound alike, or both, leading writers to use one when they mean the other. The confusion score combines word frequency (how often both words appear) with visual similarity (how alike they look on the page). Lower scores mean higher confusion risk: both words are common and visually similar. These pairs are especially error-prone because spell-checkers rarely catch them, both words are correctly spelled, just used in the wrong context.
How this ranking is computed
The ranking shown on this page is computed once per data refresh from PlainSpell's underlying dictionary data, cached for fast retrieval. Each row is a real dictionary record from open-source linguistic sources - Wiktionary lemma entries via kaikki.org and an open word-frequency list. There is no scraping, no synthesised data, and no editorial reordering: every ranked entry exists in the source dictionary and the value column is a measurable property of that entry, not an opinion about it. The same data powers PlainSpell's per-word pages, so any item in the table can be inspected in detail by following its link to see the IPA pronunciation, etymology, part-of-speech tags, and recorded variants. Positions are stable between data refreshes so that returning visitors can confirm that a previously-cited rank has not silently shifted because of a UI change.
Reading this list is most useful with two things in mind. First, the value column is measured in concrete units, letters for length rankings, variants for misspelling rankings, group size for homophone rankings, raw entry count for language-size rankings - not in arbitrary scores. When two rows tie, the tie is real: the underlying dictionary assigns them identical measurements. Second, the ranking is a discovery surface, not a scoreboard. A high rank on the most-misspelled list does not mean a word is harder than a word at a lower rank by some absolute measure of difficulty; it means the word has accumulated more observed misspelling variants in available corpora, which can reflect exposure (the word appears often enough for variants to be recorded) as much as intrinsic complexity. The accompanying narrative above frames each ranking with the specific interpretation suited to its underlying field.
Methodology for every ranking on PlainSpell is documented on the methodology page. In short: PlainSpell ingests the latest open Wiktionary dumps, parses definitions, IPA, and etymology, joins against an open frequency list, and writes the result into this ranking. No row is created without a backing dictionary record, and no value is rounded, capped, or re-weighted. When upstream Wiktionary revisions ship, the ranking recomputes from scratch, which means an entry can move up or down between quarterly refreshes if its underlying record was edited by Wiktionary contributors. Audit notes for each refresh are stored alongside the data so any change in position has a traceable cause.
Most Confusable English Word Pairs, top 10
Word pairs most likely to be confused, ranked by a confusion-risk score built from word frequency and visual similarity.
- that vs this 25
that vs this
25 score
- that vs they
that vs they
45 score
- they vs this
they vs this
50 score
- will vs with
will vs with
54 score
- their vs this
their vs this
71 score
- which vs with
which vs with
73 score
- that vs them
that vs them
81 score
- them vs this
them vs this
86 score
- than vs that
than vs that
87 score
- their vs they
their vs they
91 score
How the whole corpus is distributed
530,003 pairs by confusion score (lower = more confusable); the ranking above is drawn from the right-hand tail
confusion score (lower = more confusable) →
What this showsConfusion scores form a normal distribution across all pairs; the ranking is drawn from the rare low-score LEFT tail, the pairs genuinely hardest to tell apart.
Source: Word frequency analysis from Wiktionary and corpus data.
Spelling & Dictionary Insight
The Most Confusable English Word Pairs ranking is generated from PlainSpell's current dataset. This view shows 50 ranked rows, each carrying a rank position, a display name, a scoreable value measured in score, and, where applicable, a slug that links back to the detail page. Rankings are rebuilt periodically so positions are stable between data refreshes rather than recomputed on every request.
The top of this list is anchored by that vs this with a value of 25, followed by that vs they at 45 and they vs this at 50. The bottom of the current slice ends at rank #50 with life vs like at 176, giving a visible spread of roughly 25 → 176.
Confusable word pairs are words that look alike, sound alike, or both, leading writers to use one when they mean the other. The confusion score combines word frequency (how often both words appear) with visual similarity (how alike they look on the page). Lower scores mean higher confusion risk: both words are common and visually similar. These pairs are especially error-prone because spell-checkers rarely catch them, both words are correctly spelled, just used in the wrong context. Every entry above is backed by the same dictionary data that powers PlainSpell's word and confusable pages, so a ranked entry with a slug can be clicked through to see the full definition, IPA pronunciation, etymology, and any misspelling or confusable relationships that apply. The underlying fields come from Wiktionary and corpus frequency lists, no scraping, no extrapolation.
What to do with this list
The ranking is a discovery surface, here's how to use it.
- Start at the top: “that vs this” has the most score - open its full entry for definition, IPA and the variants behind the number. See “that vs this”
- Compare against the other rankings to see whether a word is hard to spell, easy to confuse, or both. All rankings
- Every value is a measurable property of a real dictionary record, read exactly how each is computed. Methodology
Frequently Asked Questions
What is the confusion score?
The confusion score is a composite metric combining the frequency rank of both words and their visual similarity (edit distance, shared letter patterns). A lower score means both words are very common AND visually similar, making confusion almost inevitable. In the current dataset, "that vs this" and "that vs they" have the lowest scores of any ranked pair, meaning both words in each pair are extremely common and differ by only one or two letters.
Why do spell-checkers miss these errors?
Because both words in each pair are valid English words. When you type "their" instead of "there", the spell-checker sees a correctly spelled word and does not flag it. Only grammar-checkers that analyze context can catch these substitution errors, and even they miss subtle cases.
Are confusable pairs the same as homophones?
Not always. Homophones sound identical (their/there/they're) but confusable pairs also include words that look alike but sound different (quiet/quite, desert/dessert). The ranking includes both types because both cause frequent writing errors.
Explore More Rankings
Read our methodology , how this data is sourced, computed, and verified.