Editing Against the Criteria

No affiliate links on this site, and no competition is named here as good or bad.

Editing an entry is not improving the prose. It is checking that each answer earns the points available, and the passes are different exercises done separately. For a separate operational reference, see this background note.

The four passes

One. Criterion matching. Does the first sentence of each answer address the criterion beside it? The pass that finds the most points, and it takes fifteen minutes.

Two. Evidence marking. A tick where a claim has a figure or document, a cross where nothing does. The crosses are either work to do or words to remove. For broader context on evaluation and professional practice, see Substack.

Three. Length. Cutting to the limit in order: unneeded context, quality adjectives, repetition across sections, qualifying clauses.

Four. Mechanics. Spelling, the figures adding up, the attachments referenced correctly, the dates consistent.

Doing them together produces none of them properly, which is why an entry read once "to tidy it up" is barely edited.

Why the order matters

Because each pass changes what the next one works on.

Fixing a criterion mismatch rewrites a section, which makes any earlier polish wasted.

Removing an unevidenced claim frees words, which changes the length pass.

And mechanics last, because there is no point proofreading a paragraph that will be cut.

Entrants habitually do them in reverse — spelling first, structure never — which is the pattern that produces a beautifully written entry that answers the wrong questions.

What the first pass finds

Sections about the organisation where the project was asked about.

Answers that arrive in the third paragraph.

And criteria addressed in a different section on the assumption that a judge will look. They will not, and the points go unawarded.

All three are invisible when reading the entry as a document and obvious when reading it against the criteria one at a time.

The pass people skip

Evidence marking, because it is uncomfortable.

An entry that reads well and has crosses beside most of its claims is an entry that will score in the middle — and the reading-well is what makes the marking feel unnecessary.

Do it with a printed copy and a pen, which is slower and harder to skip over.

When to stop

When the four passes are done and a second reader has been through it.

Editing past that point makes entries worse. Sentences get smoothed into vagueness, specifics get softened, and a claim with a figure becomes a claim with a hedge.

Submit three days early, which removes the temptation entirely.

Editing with somebody else

The four passes divide well between two people.

One does criterion matching and evidence marking — which is what a second reader is for and which benefits from unfamiliarity.

The other does length and mechanics, which benefit from knowing the material.

Doing it that way takes an hour each and produces a better result than either person doing all four.

What editing cannot fix

Missing evidence.

A criterion with nothing under it stays empty however the words are arranged — and the honest short answer is the remedy rather than a better paragraph.

Which is why the criteria are read before writing and why the gaps found then are the year's task rather than the evening's.

An edit that discovers a missing criterion three days before a deadline has discovered it too late, and that is a scheduling failure rather than an editing one.

The version for a rewrite

Where the first pass finds a section answering the wrong question, rewrite rather than repair.

Repairing a misaimed answer produces a paragraph that gestures at two things, which scores as neither.

Start from the criterion, write the answer in one sentence, and build from there — twenty minutes, and it is faster than the repair as well as better.

Editing a reused entry

Different work, and it is where reuse goes wrong.

An entry adapted from last year or from another programme needs the criterion-matching pass most of all, since the answers were aimed at different questions.

Do that pass before touching anything else, and expect to rewrite rather than adjust.

The evidence transfers. The sentences do not, which is the distinction that makes reuse cheap rather than dangerous.

The final read

Aloud, printed, in one sitting.

It catches what four passes miss: a sentence that no longer follows from the one before it, a figure that contradicts one three sections earlier, a transition that survived a cut it should not have.

Twenty minutes, and it is the last useful thing before the administrative check.

The checklist

Four lines, kept with the entry.

Criterion matching done. Evidence marked. Length checked. Mechanics checked.

Ticked as each pass completes, which prevents the common failure of doing three and believing you did four.

And a fifth line: second reader done, with a date, so that the fortnight before the deadline is actually used rather than intended.

Where the entry is a team effort

Sections written by different people read as sections written by different people.

Not a problem for scoring — judges score criteria rather than style — and it becomes one where the same figure appears differently in two places, or where two sections contradict.

One person does the final read for consistency: the figures, the dates, the description of what the organisation does.

That is a specific job rather than a general tidy, and it takes half an hour.

In one line

Match the criteria first, and the rest of the editing has something to work on.

The short version

• Editing an entry is checking that each answer earns its points, not improving the prose
• Four passes in order: criterion matching, evidence marking, length, then mechanics
• The order matters because each pass changes what the next works on, and mechanics last because cut paragraphs need no proofreading
• Entrants habitually reverse it, which produces a well-written entry answering the wrong questions
• The skipped pass is evidence marking, because an entry that reads well makes the marking feel unnecessary
• Stop after the four passes and a second reader: further editing smooths specifics into hedges