The hidden logic of PTE Academic scoring: why most candidates plateau at 65

Thousands of PTE Academic candidates each year plateau somewhere between 58 and 65, often after multiple attempts, and cannot understand why. Their English feels strong. They study hard. Their writing and reading seem accurate. And yet the score report keeps coming back below the target they need for university admission or skilled migration. The reason is almost always the same: they are preparing for PTE as if it were a human-marked exam, when in fact it is scored by an algorithm that rewards very different features.

At Shoreline Languages, PTE preparation is one of our most in-demand courses, and the gap between what candidates think the algorithm is scoring and what it is actually scoring is the single biggest source of plateau scores we encounter. This post walks through the logic of how PTE Academic is actually scored, with concrete task walk-throughs, the specific features the algorithm rewards in Speaking and Writing, and the strategic mistakes that cost candidates the ten or fifteen points they need to reach 65 or 79.

The most important fact about PTE: a machine reads your script

PTE Academic is scored by Pearson's automated scoring engine. Your Speaking, Writing, Reading, and Listening responses are all converted into data streams that the algorithm evaluates against features it has been trained to recognise. There is no human examiner reading your essay or listening to your pronunciation. Your Speaking responses are analysed by an automated speech-recognition model trained on millions of samples from second-language speakers worldwide. Your Writing is analysed for measurable features (grammar patterns, spelling accuracy, content-word coverage, appropriate length, written discourse markers). Your Reading and Listening responses are scored by exact matching against algorithmic keys.

This has profound implications. Features that impress a human marker, creative phrasing, emotional emphasis, natural hesitation, sophisticated rhetorical structure, may be invisible or even penalised by the algorithm. Features that the algorithm rewards, steady pacing, accurate content-word reproduction, consistent oral fluency, can feel unnatural to a learner trained on human-graded speaking practice.

Candidates who do not understand this usually come from an IELTS background where their teachers have taught them to speak naturally, pause for emphasis, and vary their tone. These are exactly the behaviours that can hurt your PTE Speaking score. Unlearning them is often the single largest gain available in PTE preparation, and it is what our teachers spend the first two weeks of every PTE programme addressing.

How scoring actually works: communicative skills and enabling skills

Your PTE Academic score report shows four communicative skills (Speaking, Writing, Reading, Listening), scored from 10 to 90, and six enabling skills (Grammar, Oral Fluency, Pronunciation, Spelling, Vocabulary, Written Discourse), scored on the same scale. You need to think of these as two different layers.

The communicative skill scores are what most candidates focus on, and what most universities and visa categories specify as requirements. Skilled migration to Australia typically requires 65 in each of the four communicative skills. Postgraduate admission to Australian universities typically requires 50 to 65 depending on the programme. But these communicative scores are derived from the integrated performance across task types, many of which contribute to more than one skill simultaneously.

The enabling skill scores tell you why your communicative scores sit where they do. A Speaking score of 58 with an Oral Fluency score of 45 tells you that your fluency is the specific feature dragging Speaking down. A Writing score of 62 with a Written Discourse score of 50 tells you that your essay structure, not your grammar or vocabulary, is the bottleneck. Candidates who ignore the enabling skills are essentially flying blind; candidates who read them carefully know exactly what to fix.

The other strategic insight most candidates miss is cross-skill contamination. Your Listening score is not just affected by the dedicated Listening tasks. It is affected by any task that requires you to process audio, including Repeat Sentence, Re-tell Lecture, and Write From Dictation. A weak Listening score is often pulling down your Speaking score through integrated items, and the fix is not more dedicated Listening practice but better performance on the integrated tasks, particularly Write From Dictation and Repeat Sentence.

The task types that dominate your score

Not all PTE task types are created equal. Some contribute heavily to your final score; others contribute modestly. A generic preparation plan that treats all task types equally is a preparation plan that wastes most of its time.

Write From Dictation: the single highest-leverage task

Write From Dictation contributes to both your Listening and Writing scores. It is straightforward in principle: you hear a sentence of between 8 and 12 words, and you type it word for word. But the scoring is unforgiving: each correct word scores 1 point, each missed, misspelled, or incorrect word scores 0. There are partial credits; a sentence of 10 words where you type 8 correctly gives you 8 out of 10, not zero.

Because the same sentence contributes points to both Listening and Writing, a candidate who reliably scores 90%+ on Write From Dictation can lift both their Listening and Writing communicative scores by 5 to 10 points. This is why we describe it as the highest-leverage task in the exam.

A worked example: you hear the sentence "Most universities now require first-year students to complete a compulsory academic writing module."

That is 12 scored words: most, universities, now, require, first-year (or first, year if hyphenated differently), students, to, complete, a, compulsory, academic, writing, module. If you type that perfectly, you score 12/12. If you miss "compulsory" (a common spelling trap: one "m," one "l"), you score 11/12. If you type "competition" instead of "compulsory," you score 11/12 as well (the incorrect word just doesn't count; it doesn't penalise further, but you have also lost the point for the missing correct word).

The technique that most reliably improves Write From Dictation scores is daily drilling with accuracy tracking. Twenty sentences per day, typed immediately after a single listening, with a running accuracy percentage. Most of our candidates start around 60% accuracy and reach 90%+ within three weeks.

Repeat Sentence: the second highest-leverage task

Repeat Sentence plays a sentence aloud (typically 7–10 seconds long) and asks you to repeat it word for word into the microphone, once. It contributes to both Speaking and Listening scores, and the algorithm scores you on three features: content (did you reproduce the content words), oral fluency (was your delivery smooth), and pronunciation (were the words intelligible).

The critical strategic insight is that the algorithm rewards content-word reproduction. If the sentence is "The professor suggested that students submit their assignments by Thursday," the content words are professor, suggested, students, submit, assignments, Thursday. Reproducing these accurately, even if your function words (the, that, their, by) drift, earns strong marks. Candidates who try to paraphrase the sentence, rather than reproduce it, often score badly because the algorithm is specifically looking for reproduction.

The technique that builds Repeat Sentence scores fastest is shadowing native audio. Shadowing means listening to a short sentence and speaking along with it, not after it, attempting to match the rhythm, stress pattern, and pacing of the original. Thirty minutes of shadowing per day, over two to three weeks, produces measurable gains in both Repeat Sentence scores and in Oral Fluency more broadly.

Read Aloud: the task candidates over-prepare in the wrong way

Read Aloud presents you with a 60- to 90-word text on screen and asks you to read it aloud into the microphone. You have 30 to 40 seconds to prepare. It contributes to Speaking and Reading scores.

The algorithm scores you on oral fluency (steady pacing), pronunciation (intelligible word production), and content (did you read the text accurately). It does not reward dramatic reading, carefully articulated diction, or slow emphatic delivery. It rewards smooth, natural pace with accurate word production.

Most candidates make one of two mistakes. Either they read too slowly, trying to articulate every syllable clearly, which the algorithm scores as low oral fluency; or they read too fast and skip or slur words, which the algorithm scores as low content. The ideal pace is a natural conversational pace, around 160 to 180 words per minute, with linking between words (for instance, "bit of" pronounced as "bi-tov", not as three separate careful syllables).

A Read Aloud at 150 wpm with perfect articulation of every syllable often scores worse than a Read Aloud at 170 wpm with natural linking and slightly imperfect pronunciation of two or three words. This is not what most candidates expect.

Summarise Written Text: the rule-driven task

Summarise Written Text presents a 200- to 300-word passage and asks you to produce a single-sentence summary of between 5 and 75 words. It is scored on content, form, grammar, and vocabulary.

The form rule is unforgiving: your response must be a single sentence. Two sentences score zero for form regardless of content quality. Over 75 words scores zero. Under 5 words scores zero. Candidates who write two well-constructed sentences of 60 words each receive zero for form and therefore have their entire response heavily penalised.

A worked example: given a 250-word passage about the impact of remote working on urban centres, a Band 79+ response might look like:

Although remote working has reduced demand for traditional office space in major cities and contributed to declining foot traffic in central business districts, it has simultaneously enabled workers to relocate to smaller cities and regional areas, driving growth in previously under-developed regions and reshaping the geographic distribution of economic activity.

That is one sentence, 54 words, using compound subordination ("although...it has"), captures the two main ideas of the passage (impact on cities and on regional areas), and uses appropriate Written Discourse markers. It will score full marks on form, high marks on content, and strong marks on grammar and vocabulary.

The Oral Fluency problem

Oral Fluency is one of the six enabling skills, and one of the most misunderstood. It is not a measure of how fast you speak or how fluent you sound to a human listener. It is a measure of how consistently you speak, without long pauses, false starts, or hesitations, at a pace close to natural native speed.

Candidates who pause to think, restart sentences, or speak very slowly will score low on Oral Fluency even if their English is otherwise strong. The fix is counterintuitive: keep speaking, even if what you say is imperfect. A smooth, steady response with small grammatical errors will score higher than a precisely correct response with long pauses.

In concrete terms: in Describe Image, a 30-second response delivered smoothly with one or two small errors will score higher on Oral Fluency than a 20-second response with perfect grammar but two 2-second pauses. The algorithm is explicitly measuring pause frequency and pause length. Two pauses of 2 seconds each are a catastrophic Oral Fluency signal.

The shadowing technique described above is the most reliable way to build Oral Fluency. A second technique, which we use with every PTE candidate at Shoreline Languages, is topic sentence starters. Candidates memorise a small library of high-frequency opening sentences for Describe Image and Retell Lecture ("This graph shows the changes in...", "The lecture focuses on..."), which means they are never hesitating at the start of a response. Combining this with shadowing-built Oral Fluency across the body of the response typically lifts Speaking by 8 to 12 points within four weeks.

The Pronunciation problem

Pronunciation is another algorithmic scoring feature that behaves differently from human assessment. The algorithm is trained to recognise the phonemes, word stress patterns, and sentence-level rhythms of a wide range of English accents, including those of second-language speakers. It is not penalising you for having an accent. It is penalising you for specific features that make your speech harder for its speech-recognition model to parse.

The highest-value pronunciation work for PTE focuses on three features in this order:

Word stress. English words have a specific stressed syllable, and misplacing that stress can render a word unrecognisable to the algorithm. "Photograph" is stressed PHO-to-graph; "photography" is stressed pho-TO-gra-phy; "photographic" is stressed pho-to-GRA-phic. A candidate who says pho-TO-graph is very likely not being recognised correctly, and the algorithm scores that as pronunciation error.

Sentence stress. In English, content words (nouns, main verbs, adjectives, adverbs) receive stress, and function words (articles, prepositions, auxiliary verbs) are reduced. Candidates whose first languages are syllable-timed (Mandarin, Japanese, Spanish, Italian, Hindi, Tagalog among others) often produce every English syllable with equal weight, which disrupts the rhythm and makes their speech harder for the algorithm to parse.

Linking. Fluent English links words together: "get up" becomes "ge-tup," "what are you doing" becomes "wha-are-ya-doin." Candidates who produce each word as a discrete unit sound choppy and, more importantly, score lower on Oral Fluency because the algorithm measures the consistency of connected speech.

Candidates who drill individual vowels and consonants often see limited improvement because those features are not the primary driver of the Pronunciation score. Candidates who drill word stress, sentence stress, and linking typically see rapid gains, often 10 points in Pronunciation within two to three weeks.

Why IELTS-style strategies often hurt your PTE score

Candidates who have previously prepared for IELTS often bring habits that actively hurt their PTE performance. Because we see many candidates who have taken both exams, we know these patterns well.

In IELTS Speaking, pausing to think is a mark of a thoughtful answer and is often rewarded. In PTE, pausing kills your Oral Fluency score. The habit of gathering your thoughts before responding must be replaced with the habit of speaking continuously, even if the first few words are semi-formulaic.

In IELTS Writing, sophisticated vocabulary and varied sentence structure are rewarded. In PTE Writing, the algorithm primarily rewards accurate grammar, appropriate vocabulary, and structure that follows the task brief. Over-sophisticated vocabulary that produces collocation errors is actively penalised because the algorithm flags the errors, not the ambition.

In IELTS Reading, nuanced interpretation is often required. In PTE Reading, the algorithm rewards candidates who can quickly extract specific information and reorder content based on textual evidence. Deep interpretation is not the goal; rapid, accurate extraction is.

The fix is not to be less skilled. It is to align your output with what the algorithm actually rewards.

Reading a score report properly

Every score report you receive has diagnostic information that most candidates never use. Here is how to read one.

Imagine your score report shows: Listening 58, Reading 63, Speaking 55, Writing 61. Enabling skills: Grammar 70, Oral Fluency 48, Pronunciation 65, Spelling 72, Vocabulary 68, Written Discourse 54.

A naive reading is "I need to improve Speaking." A sophisticated reading is: "Oral Fluency (48) is dragging Speaking down, probably by 10–12 points. Listening is lower than Reading, likely because the integrated tasks (Repeat Sentence, Write From Dictation) are pulling Listening down. Written Discourse (54) is the single weakest enabling skill, which means my essay structure needs work."

The priority plan for this candidate is clear: daily Oral Fluency drilling via shadowing and extended Describe Image practice; daily Write From Dictation and Repeat Sentence drilling to lift the integrated Listening contribution; targeted Summarise Written Text and Write Essay practice focused on Written Discourse markers (however, moreover, in contrast, consequently) and clear paragraph structure. Within six weeks, this candidate can reasonably expect to move from 55–63 across the board to 70+ across the board.

At Shoreline Languages, our teachers do exactly this kind of report reading in the first lesson, with every PTE candidate, and we build a targeted eight-week plan from it.

What a genuinely PTE-aware preparation plan looks like

If you want to move your PTE score from 65 to 79+, the most productive plan usually has four features.

First: diagnose from your score report, not your instincts. Your official score report tells you exactly which enabling skills are pulling down your communicative skill scores. A candidate with weak Pronunciation pulling down Speaking is a different candidate from one with weak Written Discourse pulling down Writing, and they need different preparation.

Second: prioritise the high-leverage tasks. Write From Dictation, Repeat Sentence, Read Aloud, and Summarise Written Text should absorb the majority of your practice time. Describe Image, Retell Lecture, and the Listening multiple-choice tasks are secondary.

Third: drill with authentic timing and interface. Practice under the exact timing constraints and interface conditions of the real exam. Candidates who drill on paper, or without strict timing, perform worse on test day regardless of their language ability. The authentic PTE interface has specific timer behaviours, microphone prep periods, and transition speeds that affect performance.

Fourth: focus on algorithmic features, not natural speech. Smooth pacing, accurate content-word reproduction, consistent oral fluency, and word-level spelling accuracy are what the algorithm rewards. Optimise for these, not for what sounds good to a human.

How Shoreline Languages prepares you for PTE

At Shoreline Languages, our PTE preparation is built around one principle: we use your score report to target the specific task types where you are losing the most marks, and we drill the algorithmic features that actually move the needle. We do not guess. We do not use generic IELTS materials rebadged for PTE. We work with your actual score history to build a plan that addresses your specific weaknesses.

Every programme starts with a full mock test under authentic interface conditions, scored across all ten skill areas. From there, we design a six-to-twelve-week programme that focuses on the three or four highest-leverage tasks for your specific profile. We drill Write From Dictation daily with accuracy tracking, shadowing practice twice a week, and integrated task rehearsals under exam timing. Our teachers are trained in the specifics of the scoring algorithm and can tell you exactly why a response scored the way it did, not just that it was "good" or "needs work."

Our candidates typically move 10 to 15 points per skill within six to eight weeks, enough to cross the 65 to 79 threshold that is the single most common target for skilled migration. We are not the only way to prepare for PTE, but we are one of the few who prepare for it on its own terms rather than borrowing techniques from other exams. For candidates who need a specific score by a specific date, that precision is exactly what matters.

If you are stuck on PTE, here is what to do

If you have taken PTE once or more and are stuck below your target, the most valuable thing you can do is get an honest, detailed diagnosis of your score report from someone who understands how the algorithm actually works. We offer a free trial lesson that includes exactly this: a walk-through of your last score report with a clear explanation of which enabling skills are pulling down your communicative skill scores, and what a targeted plan to address them would look like. Bring your most recent score report, and you will leave with a concrete roadmap to your target.

The candidates who break through the PTE plateau are the ones who understand what they are really being scored on. If you would like that understanding, we would love to help you get there.