If a typing test counted actual words, results would be almost meaningless for comparison. "I typed at 60 WPM" on a passage of short, simple words (average 3-4 characters) represents genuinely less keystroke output than "60 WPM" on a passage of long, technical words (average 8+ characters) - despite the identical number. Anyone could inflate their score simply by choosing or writing text full of short words, and no two test results would be comparable unless the underlying text had matching average word length, which almost never happens naturally.
The standard: one "word" equals five characters
To fix this, typing tests (TypeClash included) use a standardized word unit of five characters, including spaces, rather than counting whatever the actual words in the passage happen to be. Total characters typed (correctly) get divided by 5 to produce the "word count" used in the WPM calculation, regardless of what the real words in the passage were.
This means "WPM" as reported by a typing test is really a proxy for characters per minute, scaled to a familiar unit - it's not literally telling you how many dictionary words you produced. A 60 WPM reading always represents roughly the same keystroke throughput (about 300 characters per minute), no matter what specific text was typed.
Why this is the fairer standard, not a workaround
This convention dates back to typing speed measurement long before computers - it was already standard practice in manual typing certification tests, specifically to make speeds comparable across different test materials. Modern typing platforms inherited it for the same reason: it's the only way a WPM number from one test is meaningfully comparable to a WPM number from a completely different passage.
Without it, comparing your score on a TypeClash Classic run to a friend's score on a different passage - or even to your own score from a different day - would be comparing two different things wearing the same label.
What this means practically for you
- Your WPM reading is consistent across different passages, difficulties, and languages (see our piece on how multilingual typing tests handle this), because it's always measuring the same underlying unit.
- A passage full of short, simple words won't artificially inflate your score, and one full of long technical words won't artificially deflate it - both get normalized to the same five-character unit.
- When you compare your WPM to a published "average typing speed" statistic, checking whether that source used the same five-character standard (most reputable ones do) matters more than people realize - it's the single biggest reason "average WPM" numbers you find online sometimes disagree with each other. The number on your screen isn't literally "words," but it's a much fairer and more comparable measurement than literally counting words ever could be.
A quick way to sanity-check any typing test you use
If you're ever unsure whether a platform is using the five-character standard, there's a simple check: type a passage of very short, simple words at your normal pace, note the WPM, then type a passage of noticeably longer words at the same physical typing effort and compare. If the two readings land close together, the platform is almost certainly normalizing by character count, not literal word count - which is what you want for a score you can actually trust across different tests. If the numbers swing wildly based purely on average word length in the passage, that's a sign the underlying scoring hasn't accounted for this at all, and any comparison you make using it should be treated with real skepticism.