Answer the chaotic 15-dimension SBTI Test, unlock the hidden drink branch if fate allows it, and get one of 27 cursed personality endings instantly.
This adapted SBTI test keeps the original weird energy: 30 core questions, one random gate question, a hidden drinking branch, and 27 result posters waiting to roast you.
core questions
dimensions
possible endings
A gate question is always inserted at random, and a hidden drink question appears only if your answers make the algorithm nervous.
SBTI is a deliberately unserious personality quiz built around 15 dimensions instead of the usual four-letter format. It mixes sincere self-report prompts with cursed internet humor, then maps your answers onto a library of absurd archetypes.
The main model uses 30 regular prompts. A random gate question always sneaks in, and one extra drink-related question appears only for certain answer paths.
Self-esteem, attachment, worldview, execution, social style, and more all feed the final pattern match behind the result card.
Most outcomes come from the standard type library, but the original test also includes a hidden DRUNK ending and a HHHH fallback for especially unclassifiable brains.
The quiz runs directly inside the `/free-tools/*` experience with integrated layout, progress feedback, and result cards.
The appeal is not clinical accuracy. It is the mix of self-reflection, internet slang, and being personally dragged by a suspiciously specific result card.
Everything on this page is rebuilt to fit the existing free-tools pattern while keeping the original scoring behavior intact.
The page plugs into the same free-tools shell as other interactive tools, including metadata, breadcrumbs, FAQ markup, and long-form support sections.
Regular questions still map to the same 15 dimensions, with the same pattern matching, similarity scoring, and fallback thresholds throughout the test flow.
Questions, result names, descriptions, and dimension labels are fully adapted for English readers instead of appearing as partial or untranslated copy.
All 27 result posters are stored in the local public assets folder so the page renders reliably without depending on external image hosts.
The rebuilt version keeps every question visible and easy to tap on small screens, while preserving progress feedback and result readability.
Chinese keeps the original voice, while English readers get a localized version that preserves the same joke structure and tone where possible.
A few answers before you hand your self-esteem to a cursed personality algorithm.
No. It is an entertainment-first personality quiz. Some prompts are sincere, some are jokes, and several result descriptions are intentionally exaggerated for comedic effect.
This implementation includes the same 27 endings bundled in the source: 25 standard results plus the hidden DRUNK ending and the HHHH fallback result.
One gate question is always inserted at random. If you answer that you enjoy drinking, the source logic unlocks a second hidden alcohol question that can trigger the DRUNK ending.
The scoring comes from your selected option values, not from the display order. Question order is shuffled for variety, but the 15-dimension totals are calculated the same way each run.
It is a similarity score from the original matching logic. The algorithm compares your 15-dimension pattern against the standard type library and ranks the closest match.
If your best normal match scores below 60%, the quiz intentionally falls back to HHHH as a joke ending for people whose brainwaves do not sit cleanly inside the standard library.