Prediction history
The fairest read is topic-level accuracy: whether Gormster pushed students toward the same examinable skills and contexts that appeared on the real written paper, not whether it cloned the exact wording.
Accuracy (Accounting for student choice)
90.4%
All-options Accuracy
82.5%
Direct hits
8/15
The headline score uses the real student-choice structure: best nine Section A short questions plus best two Section B long questions. This reflects the paper students actually sat, where they could avoid weaker-fit options.
Scored by appeared question/topic, weighted by the marks printed on the paper: direct hit = 100%, strong adjacent = 75%, partial = 50%, weak adjacent = 25%, miss = 0%. The all-options audit scores every question that appeared, including options a student did not need to answer.
Sources checked against SEC 2026 LC Computer Science Higher Level Sections A & B.
Gormster's written mock included a smart greenhouse context with a microcontroller, sensors, environmental data and automatic control. The real paper then used a greenhouse controller as a full long question, including embedded systems, sensor types, rules, stored readings, testing and design process.
The real long questions rewarded applied problem solving rather than isolated definitions. Gormster's priorities lined up strongly with information systems, algorithms, abstraction, embedded control, validation, UI design, security and social impact.
Computer Science 2026
Algorithms, trace tables and loop walkthroughs were a headline prediction.
Counted in student-choice scoreMarks
6
Prediction
Direct hit
Score
100%
CPU, memory and storage comparisons were explicitly predicted.
Counted in student-choice scoreMarks
6
Prediction
Direct hit
Score
100%
Boolean logic, gates and compound conditions were explicitly predicted.
Counted in student-choice scoreMarks
6
Prediction
Direct hit
Score
100%
Binary, hexadecimal and data representation were explicitly predicted.
Counted in student-choice scoreMarks
6
Prediction
Direct hit
Score
100%
Systems, networks, privacy and security were predicted, though cloud storage was not named as the exact frame.
Counted in student-choice scoreMarks
6
Prediction
Strong adjacent hit
Score
75%
Pseudocode, algorithms and loop construction were explicitly predicted.
Counted in student-choice scoreMarks
6
Prediction
Direct hit
Score
100%
Sorting and algorithm walkthroughs were explicitly predicted.
Counted in student-choice scoreMarks
6
Prediction
Direct hit
Score
100%
Ethics, evaluation and system limitations were predicted, but AI tools were not named as the exact topic.
Marks
6
Prediction
Partial hit
Score
50%
Software/system design was adjacent, but modular design was not a headline prediction.
Marks
6
Prediction
Partial hit
Score
50%
Usability, accessibility and universal design were explicitly predicted.
Counted in student-choice scoreMarks
6
Prediction
Direct hit
Score
100%
Computational thinking and decision logic were predicted; heuristics was a strong adjacent hit.
Counted in student-choice scoreMarks
6
Prediction
Strong adjacent hit
Score
75%
Only broadly adjacent through computational-thinking theory; Turing machines were not specifically predicted.
Marks
6
Prediction
Weak adjacent hit
Score
25%
Information systems, data types, UI/accessibility and validation were all in the predicted lane.
Counted in student-choice scoreMarks
38
Prediction
Strong adjacent hit
Score
75%
Algorithms, abstraction, computational thinking and impact analysis were top predictions, but Euler paths/Konigsberg bridges were not named as an exact prediction.
Marks
38
Prediction
Strong adjacent hit
Score
75%
The Gormster mock used a smart greenhouse with a microcontroller, sensors, environmental data and automatic control.
Counted in student-choice scoreMarks
38
Prediction
Direct hit
Score
100%
Using the same choices students had in the real exam, Gormster's 2026 Computer Science predictions matched at 90.4%. That is the most student-friendly score because nobody had to answer every option on the paper.
The stricter all-options score is 82.5%. That counts every question on the paper, including options students could skip, so it is a tougher audit of the predictions.
Section A's best nine scored 94.4%, and Section B's best two scored 87.5%. The greenhouse long question was the standout exact-context hit.
Bottom line: if you sat the Gormster mock and then revised the course topics you struggled with, you would have been very well set for the real paper.
