TurtleDock
SchoolDeck · Feature

Patterns in the marks, and the labels we refuse to attach

Performance analysis surfaces patterns across a student’s assessment history: a subject slipping against their own prior performance, a component they consistently lose marks on, a change that coincides with attendance. It describes what happened. It does not predict what a child is capable of.

This is descriptive. Any tool claiming to forecast a child’s potential is selling a labelling system, and labels applied to eleven-year-olds are very hard to remove.

Written by Checked by Latha KumaresanLast reviewed

A subject teacher sees their own subject. A class teacher sees forty students across all of them. The pattern that matters — a student who has slipped in three subjects since September, in the same weeks their attendance changed — is visible to neither without somebody sitting down with a spreadsheet.

Compares
A student against their own history
Surfaces
Change, not absolute level
Goes to
The class teacher, not the parent app
Never
A predicted grade, rank or potential
01

Against themselves, not against the class

Ranking a student against their cohort tells you something they and their family already know. Comparing them against their own prior performance tells you something new: this student is doing worse than they were, which is actionable regardless of where they sit in the class.

The same logic applies to the strong student who has quietly dropped from ninety to seventy-eight and is still comfortably top of the section. Nobody notices, because the ranking has not moved.

02

What we will not build

  • No predicted board grade. A number a family will treat as a verdict, generated from partial data on a fourteen-year-old.
  • No stream or career recommendation. That conversation belongs to a teacher and a family, with a person accountable for it.
  • No risk score on the student record. A label that follows a child between teachers and between years.
  • Nothing pushed to parents automatically. A pattern goes to the class teacher, who decides whether it is a conversation and with whom.

Each of these is technically straightforward and several are routinely sold. We think they cause more harm than they remove, and the harm falls on children who cannot contest it.

FAQ

Questions people ask

Phrased the way they arrive, answered so each one stands on its own.

Does it predict board exam results?

No, and it will not. Predicted grades on partial data become verdicts a family acts on, and the student carries the consequence.

Do parents see this analysis?

Not automatically. It goes to the class teacher, who decides whether and how to raise it. Automating that conversation does harm.

Is it comparing our students to other schools?

No. Comparison is against the student’s own history, within your school. There is no cross-school model.

Related

Read next

Provenance

Who checked this, and against what

Accountability

Checked by the Implementations desk on 6 September 2026. Corrections go in the page, dated, rather than quietly.

No call required

Tell us your enrolment, we will tell you the number.

Pricing is on the page, in the open, with the counting rule written next to it. Nobody has to book a call to find out what it costs.

See pricing

Bring your own structure. We configure it live on the call.

Forty minutes with an implementation engineer — your classes, your fee heads, your grading scheme, entered as we go. No slides, no sample school.