KDU CGT study materials
KDU CGT study resources

KDU CGT flashcards

KDU CGT ยท Build recall with a focused review session.

32 cards. Read the term and the definition here. The deck at the top of this page is optional.

Verification vs validation
Verification checks that the product is built according to specification (reviews, static analysis). Validation checks that the product meets user needs and intended use (playtesting, acceptance testing).
Entry and exit criteria
Entry criteria are conditions required before testing starts, such as a compilable build and passing smoke test. Exit criteria are conditions to stop testing, such as all planned cases run and no open blockers.
Vertical slice
A small end-to-end playable segment that includes all major systems and production quality. It validates fun, performance, and pipeline before full production and should be tested across all included systems.
Regression testing trigger
Any code or content change can introduce regressions. Changes to shared systems such as input, save format, or network protocol require wider regression scope than isolated UI changes.
Boundary value analysis
Test values at and just outside the edges of input ranges: minimum-1, minimum, minimum+1, maximum-1, maximum, maximum+1. For an inventory cap of 99, test 98, 99, and 100.
Equivalence partitioning
Divide inputs into classes expected to behave identically and test one representative value per class. For health, classes might be 0, 1โ€“99, and 100, with boundaries tested separately.
State transition coverage
Model game states and events, then design tests to visit every state and traverse every valid transition. Also test invalid transitions to confirm they are rejected or handled safely.
Pairwise testing
A combinatorial technique that generates a test set covering every pair of parameter values at least once. It greatly reduces the Cartesian product while catching most interaction defects.
Test case anatomy
A complete test case includes a unique ID, title, preconditions, steps, test data, expected result, actual result, priority, and traceability link to a requirement or user story.
Exploratory testing charter
A time-boxed, goal-directed testing session with a written charter and session notes. It complements scripted cases and is effective for emergent gameplay and usability defects.
Hitbox and hurtbox
A hitbox is the region that deals damage; a hurtbox is the region that receives damage. Combat testing verifies their alignment with animation and that hits register only when intended.
Damage formula verification
Confirm that final damage equals base plus scaling minus mitigation, with critical multipliers and damage-over-time ticks applied correctly. Test minimum, maximum, and immunity-frame cases.
Save data migration
When save format changes between versions, test that old saves load correctly, are upgraded without data loss, and that newer-version saves are rejected with a clear message.
Economy duplication exploit
A defect where a player gains currency or items without paying, often by reconnecting or interrupting a transaction. Test by disconnecting mid-purchase and verifying server-side reconciliation.
Frame time budget
At 60 frames per second each frame has 16.67 milliseconds; at 30 FPS, 33.33 milliseconds. Exceeding the budget on any thread causes visible hitching or stutter.
1% low FPS
The average frame rate of the slowest 1% of frames during a session. It reveals stutter that average FPS hides and is a key performance acceptance metric.
Soak testing
Running a long session, often 8 to 24 hours, to detect memory leaks, handle leaks, and thermal throttling. Compare memory use at start, midpoint, and end.
Compatibility matrix
A prioritized grid of hardware and software combinations to test, including OS versions, GPUs, CPUs, RAM, resolutions, and input devices. Prioritize by market share and support policy.
Text expansion in localization
Translated text can be 30% or more longer than English, causing overflow in UI. Test with German and Finnish, check right-to-left layouts, and verify font glyph coverage.
Accessibility testing checks
Verify remappable controls, subtitle size and background options, colorblind modes, screen reader support, motion reduction, and hold-to-toggle alternatives for all actions.
Certification gate
A pass/fail review against platform requirements before release. Failure can delay launch and incur resubmission fees, so test current published requirements for each platform.
Suspend and resume compliance
Platforms require games to handle suspend and resume without data loss or corruption. Test mid-cutscene, mid-transaction, and during network activity, then verify state restores correctly.
Storage full behavior
When storage is full or removed, the game must not lose save data silently. It should show a clear error and allow recovery. Test with corrupted and newer-version saves too.
Achievement unlock verification
Test that each achievement unlocks under its stated condition, in the correct order, and is not unobtainable. Verify on a clean profile and an existing profile with prior progress.
Server authority
The server validates and owns critical state such as position, ammo, and currency. Test by modifying client values and confirming the server rejects or corrects them to prevent cheating.
Latency and jitter simulation
Introduce artificial round-trip delays such as 50, 100, 200, and 400 milliseconds plus jitter. Verify client prediction, server reconciliation, and interpolation keep movement acceptable.
Host migration
When the host leaves a peer-to-peer session, another player must take over without losing game state. Test migration during combat, loading, and transactions, and verify no ghost sessions remain.
Matchmaking bucket test
Verify that players are grouped by skill, region, and latency within policy. Test party sizes, backfill, cancel during search, and timeout behavior to avoid cross-region matches beyond policy.
Severity versus priority
Severity measures impact on the game, such as a crash or data loss. Priority measures urgency to fix. They are independent: a cosmetic typo on the store page can be high priority.
Bug life cycle states
Typical flow is New, Assigned, In Progress, Fixed, Ready for Test, Verified, Closed. Alternate paths include Reopened, Deferred, Duplicate, Not a Bug, and Cannot Reproduce.
Defect density
The number of defects found per unit of work, such as per feature or per thousand lines of code. It helps compare quality across modules and prioritize additional testing.
Traceability matrix
A table linking requirements to test cases, defects, and results. Coverage equals covered requirements divided by total requirements, and uncovered high-risk requirements must be reported.