Pavel_Volkov_Presentation

1.

Physics modeling: using physics and
simulation systems to provide realistic
physics in games
Author: Volkov Pavel Maksimovich

2.

Motivation
Core of game engines
Used far beyond games
A genuinely hard skill

3.

The Problem
No way to check yourself
Reasoning stays hidden
Knowledge stuck to one
case

4.

Algorithm Selection
0
Breadth of use
Conceptual relatedness
Pedagogical tractability
THE FOUR SPAN DIFFERENT SIMULATION TYPES
SPH
·
Verlet
·
XPBD
·
MPM

5.

Goal & Research Questions
GOAL Build an environment that teaches game-physics algorithms by having learners implement them — and test whether
that understanding transfers to an unseen problem.
OBJECTIVES
1
RQ1
Choose algorithm
2
Split each into small, testable functions
3
Build the levels, live canvas and checker
4
Design a capstone that demands transfer
5
Run a study to measure whether it works
Does building the algorithms yourself, with instant feedback,
create real conceptual understanding?
RQ2
Does that understanding transfer to a new problem the
learner was never taught?

6.

How It Works
Interactive client
Godot 4
→
Validation server
FastAPI / Python
→
Study harness
Browser

7.

The Experiment
THE Flow
Pre-questionnaire
Velrlet integration level
Exam
Post-questionnaire

8.

Results
20 / 31
3.9–4.5
passed the unseen task
average survey score
65% completion
out of 5
ROPE CAPSTONE OUTCOME
POST-SESSION SURVEY (N = 31, 1 –5)
Quiz gating
Passed
Timed out
Skipped
3,9
Theory
4,1
Understanding
4,2
Code tasks
4,4
Visualisations
4,5
0
0,5
1
1,5
2
2,5
3
3,5
4
4,5
5

9.

Future Work
A controlled comparison
A wider audience
Adaptive quizzes
More physics methods

10.

Thank you