BaaS

BaaS is a simulation framework for testing neuroadaptive game difficulty in Tetris. The system models how EEG-like emotion classifications can drive adaptive controllers that change game speed during play.

The current project focuses on three simulated users:

User	Short label	Skill profile
Frustrated Freddy	FF	Beginner
Average Amy	AA	Intermediate
Hairy Henry	HH	Advanced

The current experiment setup uses:

Component	Current options
Controllers	PID, Feedforward
Classifiers	SpeedOnly, Noisy
Target emotion	HAHV

---

Run the main experiment

The main experiment searches over PID and Feedforward controller parameters, then keeps only the best configuration for each user/controller/classifier condition.

python experiment.py

Main output will go to the experiment_results folders.

Emotion distribution from the best controller settings

Short labels:

Label	Meaning
FF	Frustrated Freddy
AA	Average Amy
HH	Hairy Henry
PID	PID controller
ff	Feedforward controller
SO	SpeedOnly classifier
N	Noisy classifier

Example: AA-ff-SO means Average Amy + Feedforward + SpeedOnly.

---

Run the delayed experiment

The delayed experiment tests robustness when the control loop is delayed.

It varies two delay values:

Delay	Meaning
Controller delay	The controller receives an older classifier probability vector
User-change delay	The user's emotional state changes after a delayed number of pieces

Run:

python experiment_delayed.py

Main output will go to experiment_delayed_results.

Enjoyment percentage across controller/user delays

In this plot:

Axis	Meaning
X	Controller delay
Y	User-change delay
Z	Mean enjoyment percentage

Lower delay generally produces higher enjoyment.

---

Run a visual simulation

Use the visual simulation to watch one selected user play under the selected controller/classifier setup.

python simulation.py --user "Frustrated Freddy"
python simulation.py --user "Average Amy"
python simulation.py --user "Hairy Henry"

The visual simulation opens a Pygame window and waits for ENTER before starting, so the display can be arranged before the run begins.

---

Notes

• MAX_MOVES means the maximum number of Tetris pieces the autoplay system attempts to place.
• A trial can end before MAX_MOVES if the game reaches a terminal state.
• The main loop performs one classifier prediction, one controller update, and one Tetris piece placement attempt per move.