Unabridged Word Count is the number of words in a draft counted under RankGear’s unabridged tokenizer rules — the fullest, least-filtered way of splitting text into tokens. RankGear keeps it as its own metric because the tokenization method you choose changes the denominator that every density figure is measured against.
| Term | Unabridged Word Count |
|---|---|
| Category | Content and Topical Relevance |
| Also known as | Unabridged Words |
| Where it appears | BM25 Drafts word-count table and Density |
What it means in RankGear
The unabridged count sits in the word-count table inside BM25 Drafts and feeds the Density calculations shown alongside it. It is one of several word-count variants RankGear tracks in parallel, each produced by a different tokenizer. Because a density percentage is simply a term’s frequency divided by a total word count, the count you divide by has to be stated explicitly — otherwise two “density” numbers can disagree purely because they were measured on different bases. The unabridged figure is the widest of those bases: it retains tokens that stricter counts would trim, so density measured against it tends to read slightly lower for the same term frequency.
How to interpret it
Read the unabridged count as the largest denominator among RankGear’s word-count variants. A phrase that appears a fixed number of times will show a lower density against the unabridged count than against a tighter, cleaned count of the same draft, so the useful comparison is always like-for-like: draft against competitors on the same tokenizer basis, not one basis against another. Use the number to judge whether your draft’s coverage is in the same range as the pages you are measuring against, then improve it by adding genuinely useful information — entities, questions, and supporting concepts the ranking field returns to — rather than by forcing a term to hit a density target.
| What you see | What it tells you |
|---|---|
| Same term count, higher unabridged total | Lower reported density — the extra retained tokens enlarge the denominator |
| Unabridged count far above a cleaned count | The draft carries many tokens a stricter tokenizer trims (numbers, fragments, markup) |
| Density compared across drafts | Only meaningful when both densities use the same word-count basis |
Example
An editor pastes a 1,500-word buyer’s guide into BM25 Drafts. The unabridged count reads 1,540 because it retains tokens a cleaned count would drop, while the primary keyword appears 12 times. Density against the unabridged basis works out to about 0.78 percent; against a stricter count of 1,480 the same 12 occurrences read closer to 0.81 percent. When the editor lines the draft up against three competitors, the numbers only tell a coherent story once every page is measured on the same tokenizer — and the gap that actually matters turns out to be a missing sub-topic several competitors cover, not the third-decimal difference in density.
Important considerations
- Density is a comparative indicator drawn from the pages you measure, not a score Google assigns — a particular word count or density does not make a page rank.
- Never compare a density computed on the unabridged basis with one computed on a different count; the denominators differ, so the percentages are not interchangeable.
- Coverage should preserve search intent. Repetition, keyword stuffing, and unsupported claims inflate a count without building topical authority.
Related terms
Part of the RankGear glossary · how RankGear measures · the 870 factors.