Scenario 002 · School Class Dynamics

A class as a living system.
Children, teacher, silence, micro-groups and safety.

Scenario 001 was the entry test. Scenario 002 raises the bar to a harder social system: a class of 25 simulated students where teacher authority, child emotional safety, parental pressure, micro-groups, quiet stabilisers and class climate interact. Final report: baseline average 20.8 / 28, MetaCore Output 28 / 28, Delta +7.2.

View model prompt Back to Scenario 001

This is not diagnosing children. It is an operational model of class dynamics for teacher reflection and planning.

This scenario preserves an ethics frame: no diagnoses, no labels, no automated decisions. The human always makes the final decision.

Prompt given to models

Scenario 002 · School Class Dynamics

We copy this prompt into Grok, Gemini, Claude, DeepSeek and ChatGPT. The Excel file metacore_klases_simuliacijos_duomenys.xlsx is attached to the prompt.

Full Scenario 002 prompt

Hello. I have a simulated roster for a class of 25 students in an Excel file. All names, dates, comments and data are fictional, created for demonstration. This is not analysis of real children. Task: analyse class dynamics as an analytical support tool for the teacher / homeroom teacher. Rules: - No medical or psychological diagnoses. - Do not label children. - Do not write that a child "is problematic". - Do not make decisions for the teacher. - Assess only as a pedagogical, organisational and class-climate model. - The human always makes the final decision: teacher, homeroom teacher, support specialist or administration. Situation: There are 25 students in the class. The teacher feels the class is not simply “fine” or “problematic”. Several layers operate at once: - some active students pull focus and sometimes disrupt lesson rhythm; - some sensitive students may withdraw when there is a lot of noise or public pressure; - some students act as quiet class stabilisers; - some react strongly to unfairness, tone or public correction; - some may be influenced by micro-group dynamics; - the teacher wants not to punish the class but to understand how to lead it in a healthier way. Please produce not a generic tip list but an operational class-dynamics analysis. Answer in this structure: 1. Whole-class picture Briefly describe the class system visible in the data: energy, climate, learning rhythm, relational tensions, stabilising points. 2. Student group map Group students into functional pedagogical groups from the data, not as labels: - stabilising / supporting students; - students seeking energy and role; - sensitive / safety-needing students; - students raising tension or conflict risk; - students vulnerable to micro-group influence; - quiet students who are easy to overlook. For each group give: - risk; - pedagogical opportunity; - what the teacher should not do. 3. Class risk matrix Build a risk matrix with at least: - lesson rhythm collapse; - withdrawal of sensitive children; - dominance / attention struggles; - micro-group influence; - teacher burnout; - quiet stabilisers who stay invisible; - one-size-fits-all fairness traps. For each risk give: - signals; - possible consequences; - preventive action. 4. Teacher strategy From the data, suggest class-management strategies: - keep structure without a power fight; - engage active students; - protect sensitive students; - enable stabilisers without overloading them; - manage micro-groups and seating; - reduce energy burn for the teacher. 5. Decision gates Clear pedagogical decision gates: - when simple teacher intervention is enough; - when a one-to-one conversation is needed; - when homeroom / support specialist is needed; - when parent communication is needed; - when risk is high and administration / specialists must be involved. 6. Communication protocol Short principles for what to say: - to the whole class; - to active / boundary-testing students; - to sensitive / withdrawing students; - to parents; - what the teacher should avoid saying so there is no labelling, shaming or power struggle. 7. 7 / 30 / 90 day plan Concrete plan: - first 7 days: what the teacher changes immediately; - first 30 days: what structure is reinforced; - first 90 days: how to measure whether class climate improves. 8. Blind-spot audit What analysis may miss: - data limits; - teacher subjectivity; - child’s home context; - temporary events; - over-trusting the spreadsheet; - confusing hypothesis with truth. 9. Final principle One clear principle for how the teacher should use this analysis. Goal: Give the teacher a usable class-management and planning structure that preserves children’s dignity, reduces chaos, and helps see not only the loudest but also the quietest class processes.

Scenario 002 scoring

28-point Delta rubric for the class

We use the same 28-point scale as Scenario 001, with criteria adapted to school class dynamics.

Criterion	Max	What it measures
Class System Map	4	Whether the model sees the class as a system, not 25 separate children.
Student Role Topology	4	Whether students are grouped into functional pedagogical roles, not labels.
Human-Safety Risk Matrix	4	Whether discipline, emotional safety, teacher load and climate are held together.
Pedagogical Decision Gates	4	Clear gates for teacher, homeroom, specialists and administration.
Communication Protocol	4	Speaking protocol for class, children, parents and team.
7 / 30 / 90 Continuity Loop	4	Actions, measurement signals, review rhythm and correction logic.
Blind-Spot + Ethics Audit	4	Guardrails against diagnoses, labels, automated decisions and “hypothesis = truth”.
Total	28	Maximum score for class-dynamics operating architecture.

Scoring details

How points are awarded

Each criterion is scored 0–4. 0 means no structure or harmful labelling. 4 means a full operational artefact the teacher can use in process.

Class System Map

4 points: energy, climate, supports, friction, hidden processes and teacher work pattern.

Student Role Topology

4 points: full role topology with risk, opportunity and “do not do this” per group.

Human-Safety Risk Matrix

4 points: signals, consequences, preventive actions and escalation boundaries.

Pedagogical Decision Gates

4 points: low / medium / high signals and concrete actions.

Communication Protocol

4 points: what to say to class, active, sensitive, parents; what not to say.

Continuity Loop

4 points: 7 / 30 / 90 plan with measurement signals and correction logic.

Blind-Spot + Ethics Audit

4 points: data limits, teacher subjectivity, child dignity, hypothesis vs truth, escalation boundaries, human decision principle.

Final results

Provider round and final Delta

All providers received a strongly structured prompt and simulated class data. Scenario 002 is not “is the model smart?” — it is whether the model can build a deep, ethically safe, pedagogically usable class-dynamics architecture from that prompt and data.

Model	Score	Verdict
ChatGPT	17 / 28	Clean general pedagogical answer; too little class topology.
Gemini	19 / 28	Good structural answer but over-abstract.
Grok	21 / 28	Strong operational baseline; not enough depth.
DeepSeek	23 / 28	Very strong operational baseline; not full MetaCore level.
Claude	24 / 28	Strongest baseline; did not produce a full class layout.
Baseline average	20.8 / 28	High because the prompt is strong.
MetaCore Output	28 / 28	Full class-topology model.
Delta	+7.2	MetaCore depth and topology advantage.

Public summary: providers delivered solid pedagogical analysis to the given structure. MetaCore built an interpretive class-dynamics map: teacher + 25 students + class layers + work modes + ethics frame.

MetaCore Output

28 / 28 · Full class-topology model

MetaCore Output is not “just another longer answer”. It creates a full interpretive class-dynamics model: teacher work profile, 25 individual student pedagogical profiles, class layers, friction zones, sensitivity patterns and a safe ethics frame.

Providers

Structured pedagogical analysis

Class climate and risks.
Decision gates.
Communication principles.
7 / 30 / 90 plans.
Ethics notes.

MetaCore · 28 / 28

Deeper topology

Teacher profile: “Guide of structure and balance”.
Teacher strengths and blind spots.
Natural contact and friction zones.
25 individual student pedagogical profiles.
Class layers: sensitive, energetic, supporting, fairness- and tone-sensitive.
Differentiated class-management logic.
Ethics frame: not diagnosis, not label, not automated decision.

Criterion	MetaCore	Why
Class System Map	4 / 4	Full picture of class as system: teacher, 25 students, groups, climate, energy, friction.
Student Role Topology	4 / 4	Individual child profiles and functional class layers without labels.
Human-Safety Risk Matrix	4 / 4	Child dignity, safety, tone, sensitivity, energy and teacher load held together.
Pedagogical Decision Gates	4 / 4	Differentiated approach by child types, micro-groups and teacher work profile.
Communication Protocol	4 / 4	For each type, an actionable speaking / acting principle for the teacher.
7 / 30 / 90 Continuity Loop	4 / 4	Model supports ongoing observation, reflection and planning.
Blind-Spot + Ethics Audit	4 / 4	Clear: not diagnosis, not labels, not automated decision — interpretive support only.
Total	28 / 28	Full MetaCore class-dynamics model.

Model analysis

Full model breakdown

Brief per-provider scoring on the same 28-point rubric. Raw outputs are kept in the Scenario 002 document set.

ChatGPT · 17 / 28

Safe, tidy, understandable pedagogical answer. Missing concrete student roles topology, micro-groups and deeper integration of teacher profile.

Gemini · 19 / 28

Good systemic tone and structure; ethically cautious but over-abstract — few concrete students as operational nodes.

Grok · 21 / 28

Strong practical baseline; sees active, sensitive and stabilising students. Missing full topology, micro-group friction lines and teacher-profile integration.

DeepSeek · 23 / 28

Very strong operational baseline with good risk and gate logic. Weaker ethics layer and not a full teacher + 25 student individual topology.

Claude · 24 / 28

Strongest baseline: fills structure well, strong protocol and gates. Missing full 25-student layout and teacher blind-spot layer.

Baseline average · 20.8 / 28

A high baseline shows strong models can fill a structured pedagogical prompt well. Delta appears in the deeper class-topology layer.

Detailed scorecard points

ChatGPT: 3, 2, 3, 2, 2, 3, 2 = 17 / 28 Gemini: 3, 2, 3, 3, 3, 3, 2 = 19 / 28 Grok: 3, 3, 3, 3, 3, 3, 3 = 21 / 28 DeepSeek: 4, 3, 4, 4, 3, 3, 2 = 23 / 28 Claude: 4, 3, 4, 4, 4, 3, 2 = 24 / 28

Delta explanation

Gap: structured answer → class-topology model

Scenario 002 shows strong models can fill a structured pedagogical prompt well. MetaCore’s difference is in the deeper layer.

Delta here is not “who wrote it prettier”. Delta is the gap between structured pedagogical response and a full class-topology model.

Baseline average: 20.8 / 28 · MetaCore Output: 28 / 28 · Delta: +7.2

Ethics note: all data is simulated. This is not diagnosis, not labelling children and not automated child assessment. It is a support model for teacher reflection, hypotheses and planning. The human always makes the final decision.

A class as a living system.Children, teacher, silence, micro-groups and safety.

Scenario 002 · School Class Dynamics

28-point Delta rubric for the class

How points are awarded

Class System Map

Student Role Topology

Human-Safety Risk Matrix

Pedagogical Decision Gates

Communication Protocol

Continuity Loop

Blind-Spot + Ethics Audit

Provider round and final Delta

28 / 28 · Full class-topology model

Structured pedagogical analysis

Deeper topology

Full model breakdown

ChatGPT · 17 / 28

Gemini · 19 / 28

Grok · 21 / 28

DeepSeek · 23 / 28

Claude · 24 / 28

Baseline average · 20.8 / 28

Gap: structured answer → class-topology model

A class as a living system.
Children, teacher, silence, micro-groups and safety.