PelletPursuit

CS105 Pellet Pursuit — Grading Rubric

Overview

100-point rubric for the Pellet Pursuit final project. Students implement core engine methods, three ghost personalities, difficulty tuning, a custom bonus item, and file I/O, and deliver a short in-class demo. Extension tasks are a bonus.

---

1. Engine Implementations — 50 points

Short but critical methods spread across the phase guides. Each one has a visible, testable consequence when missing.

Method	File	Points	Broken state without it
handleKey() + isWall() + canMove	Player.java, GameMap.java	10	Player sits still, ignores input, phases through walls once moving
draw() tile loop + eatDot()	GameMap.java	5	Maze invisible; dots never disappear and level never ends
collidesWith()	Ghost.java	5	Ghosts never kill or get eaten
updateBonusItems()	GameApp.java	5	Bonus items never collected or removed
saveToFile() + loadFromFile()	ScoreTree.java	10	Scores reset every run
insert() + collectDescending()	ScoreTree.java	15	Leaderboard never populates

• 45–50: All six groups fully correct and visibly working.
• 30–40: Four or five groups correct; one or two missing or broken.
• 15–25: Two or three groups correct; others still use placeholders.
• 0–10: One or fewer groups correct; game is largely non-functional.

---

2. Ghost Implementations — 30 points

All three ghosts (Patrol, Shy, Ambush) must behave distinctly from each other and from the chase-only placeholder. collidesWith() (category 1) must be working for ghost interactions to be gradeable.

• 25–30: All three ghosts behave as specified. Frightened mode works correctly for each. No crashes or infinite loops.
• 15–20: Two ghosts fully correct; one is partially implemented or has a minor behavior bug.
• 5–10: One ghost fully correct; others still use the placeholder or have major bugs.
• 0: No ghost implementations beyond the placeholder, or code crashes.

Per-ghost breakdown (use to allocate partial credit):

Ghost	Full	Partial	None
Patrol (pink) — patrol corner + chase when close	10	5	0
Shy (cyan) — flee until cornered, then attack	10	5	0
Ambush (orange) — target ahead of player’s heading	10	5	0

---

3. Difficulty & Game Feel — 10 points

Item	Points
Maze is visibly different from the default layout	5
Custom bonus item works, differs from Cherry (points, color, label), and has a custom draw() with a time-based visual feature	5

LevelConfig tuning is not separately scored but is part of the In-Class Demo — students should be prepared to explain what values they changed, why, and how the game feel differs between level 1 and later levels.

---

4. In-Class Demo — 5 points

Two minutes at the projector. Run your game and explain one design decision (ghost personality, maze layout, bonus item, or difficulty tuning).

Score	Criteria
5	Game runs, student explains their choice clearly
3	Game runs but explanation is vague or incomplete
0	No-show, game crashes, or explanation is absent

---

5. Documentation — 5 points

Students are expected to comment their own implementations, not the pre-existing engine code.

• 5: Each student-written method has a short comment explaining the approach (not just restating the code).
• 3: Some methods commented but not all, or comments only restate what the code does.
• 0–2: Very few or no student-written comments.

---

6. Extension Work — 5 points bonus

Base tasks are worth 100 points. Extension points are a bonus (max 5).

Students choose one of the three extension paths:

Option 1 — Pellet Pursuit add-ons

Task	Points
Per-level layouts (getLayout() returns different maze per level)	+1
Player movement re-implemented (all directions + tunnel wrapping)	+2
BFS replaced with a different pathfinding strategy (must behave differently)	+2
Custom sprite images (PNG in resources/game/images/, loaded via loadImage())	+1

Option 2 — Free-form project (early finishers)

Student completed all four phases and built an additional project of their own choosing with remaining class time.

• +5 Project runs and student explains it clearly at the demo
• +3 Project runs but explanation is vague or incomplete
• +0 Nothing additional to show

No data-structure requirements — base tasks already cover them.

Option 3 — Custom project track (approved proposals only)

Student completed Phase 1, submitted an approved proposal, and built a different project for Phases 2–4. Graded against the same four rubric categories as Pellet Pursuit — see mapping below.

Apply the standard rubric categories (sections 1–5) to the custom project, with these translations:

Standard category	Custom project equivalent
Engine implementations	The method groups named in the approved proposal
Ghost personalities	Three concrete subclasses with visibly distinct behavior
Difficulty & game feel	Something tuned/designed that affects the experience
Demo	Run and explain one decision

The custom project earns the same base 100 points as Pellet Pursuit; there are no additional extension points for doing Option 3.

---

Project Checklist

Use this before submitting. Each item should be verifiable by running the game.

Engine implementations

• isWall() returns true for wall tiles
• canMove in Player.update() uses isWall()
• handleKey() responds to all four arrow keys and WASD
• Walls stop the player and ghosts from passing through
• Maze is visible — walls, dots, and power pellets all render correctly
• Dots and power pellets disappear when the player walks over them; level ends when all are eaten
• Ghost kills player on contact; frightened ghosts can be eaten
• Bonus item appears, can be collected, and disappears when expired
• Scores persist between sessions (quit and relaunch to verify)
• Leaderboard on the Game Over screen shows top scores in descending order

Ghost personalities

• Patrol heads toward a corner when the player is far away
• Patrol switches to chasing when the player gets close
• Shy moves away from the player in open corridors
• Shy turns and chases when trapped in a dead end
• Ambush approaches from in front of the player, not from behind
• All three ghosts turn blue when a power pellet is eaten
• Frightened behavior is distinct from normal behavior for each ghost

Difficulty, maze & bonus item

• Maze is visibly different from the default layout
• Custom bonus item has unique points, color, and label (not Cherry)

Personalization

• GAME_TITLE changed from “Pellet Pursuit”
• Each ghost has a custom name in getName()

Code & documentation

• No compilation errors or runtime crashes
• Each student-written method has at least one explanatory comment
• No placeholder return statement left in a finished method

---

Notes for instructors

• Grade engine implementations first — if isWall() or collidesWith() are missing, ghost behavior cannot be fairly assessed.
• Patrol’s frightened mode (flee away) and Shy’s frightened mode (chase instead of flee) are intentionally opposite — both are correct.
• A ghost that always chases (placeholder) scores 0 for that ghost even if the code compiles and runs.
• Extension points are additive and do not replace base task points.

Custom project track:

• Require Phase 1 to be demonstrably working before accepting a proposal — the gate exists to filter out students who want to escape the work, not to reward them for getting stuck early.
• Reject proposals that don’t concretely name where each required structure appears. “I’ll use a list somewhere” is not enough.
• Limit approvals to however many students you can realistically support during lab time without neglecting Pellet Pursuit students.
• Grade using the standard rubric; the category translations in section 6 map each Pellet Pursuit concept to its custom-project equivalent.

This site is open source. Improve this page.