> ## Documentation Index
> Fetch the complete documentation index at: https://docs.draftliftai.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Readability Scoring

> Readability scoring is table stakes. What DraftLift adds is slop detection.

# Readability Scoring

Readability scoring is table stakes. Every writing tool does it. What DraftLift adds is slop detection — active quality enforcement that catches the patterns making your content sound machine-generated.

Metrics update as you type, so you see the impact of every edit in real time.

## Scoring methodology

DraftLift uses two established readability formulas:

### Flesch-Kincaid Grade Level

Estimates the U.S. school grade level required to understand the text. Based on average sentence length and syllables per word.

| Grade level | Audience                    | Example platforms             |
| ----------- | --------------------------- | ----------------------------- |
| 5-6         | General public              | Social media, X posts         |
| 7-8         | Broad professional audience | LinkedIn, email newsletters   |
| 9-10        | Educated readers            | Blog posts, articles          |
| 11+         | Specialist audience         | Technical or academic content |

### Flesch Reading Ease

A 0-100 score where higher means easier to read:

| Score  | Readability                                    |
| ------ | ---------------------------------------------- |
| 90-100 | Very easy — understood by an 11-year-old       |
| 60-70  | Standard — plain English                       |
| 30-60  | Difficult — college-level readers              |
| 0-30   | Very difficult — academic or technical writing |

## Additional metrics

Beyond the two core scores, the editor sidebar tracks:

* **Word count** and **character count** — Essential for platform-specific length requirements
* **Reading time** — Estimated time to read at average speed
* **Average words per sentence** — Shorter sentences improve readability; aim for 15-20 words on average
* **Paragraph count** — More paragraphs with fewer sentences generally improve scannability

## Anti-slop quality system

This is where DraftLift goes beyond scoring into active quality enforcement. The [Anti-Slop Quality System](/guides/anti-slop) analyzes content for patterns that mark generic AI writing:

### Critical patterns

Issues that make content sound obviously machine-generated:

* Robotic sentence structures ("It's not X, it's Y")
* AI vocabulary crutches (utilize, leverage, synergy, paradigm, myriad)
* Infomercial hooks ("The best part?", "Ready to level up?", "Here's the thing")
* Vague attributions and fabricated statistics

### Warning patterns

Issues that weaken writing quality:

* Em-dash overuse
* Arrow symbols and emoji clusters
* Formulaic triple-item lists
* Binary opposite structures ("It's not about X, it's about Y")

### Style signals

Subtler patterns worth reviewing:

* Filler word density
* Adverb overuse
* Copula avoidance (unnecessarily replacing "is/are" with action verbs)
* Synonym cycling (using different words for the same concept to sound varied)

Each flagged pattern includes a severity level and a specific suggestion. Patterns are highlighted inline in the [editor](/features/editor) so you can fix them in context.

## Platform-specific targets

Readability targets vary by platform. Match your score to your audience:

| Platform         | Target grade level | Why                                          |
| ---------------- | ------------------ | -------------------------------------------- |
| X / Twitter      | 5-7                | Must be instantly understood while scrolling |
| LinkedIn         | 7-9                | Professional but accessible                  |
| Short-form video | 5-7                | Scripts should sound natural when spoken     |
| Email newsletter | 7-8                | Scannable in a busy inbox                    |
| Blog post        | 8-10               | Readers expect more depth                    |
| YouTube script   | 6-8                | Spoken content should feel conversational    |

<Tip>
  Don't chase the lowest possible score. A blog post about database architecture should be at a higher grade level than a LinkedIn motivational post. Match readability to your audience and platform.
</Tip>