Everyone has a physical type — almost nobody can put it into words. Tap the AI-generated faces that catch your eye and the MyType FYI model maps your exact bone-structure taste, your hidden gems and your personal top 2% in real time.
Tap every face you find attractive.
Your unconscious taps lock onto classical bone structure, Greek nose profiles, expressive dark eyes, and high cheekbones, while passing on generic high-consensus faces.
Faces everyone rates highly that also match your personal MyType FYI vector.
Overlooked by the crowd (35th–58th percentile), yet 92%+ in your model.
Faces 80%+ of raters chase that your taps consistently pass over.
Cinematic archetypes whose proportions cluster closest to your top taps.
Drop a photo of your partner, crush or favorite celebrity to see how closely it aligns with your calibrated MyType FYI vector. Processed entirely in your browser — the photo is never uploaded.
Why did the MyType FYI quiz explode across X (Twitter), TikTok, and Reddit? Because seeing someone’s Hotties You Like next to their Hidden Gems (Mids You Love) and Popular Faces Not Your Type exposes the invisible architecture of human attraction. Cycle through real anonymized rating sessions below to see how dramatically individual taste diverges.
Lion · Tortured artist · Long hair
High cheekbones · Warm smile
Buzzcut · Finance bro · Older man
If you ask someone to describe their ideal partner on a traditional my type quiz, they will recite broad demographic clichés: “tall, dark hair, kind smile” or “blonde girl next door.” Yet when that same person swipes through faces on a dating app or walks into a crowded room, their eyes lock onto a hyper-specific combination of craniofacial geometry in less than a tenth of a second. Here is how the MyType FYI methodology decodes that hidden signal.
In behavioral economics and evolutionary psychology, researchers draw a sharp boundary between stated preferences (what you consciously believe and tell others you want) and revealed preferences (what your micro-decisions actually select when presented with real visual stimuli). Most online my type test questionnaires fail because human language was never designed to parameterize a human face.
You cannot easily write down the exact millimeter ratio between a cheekbone prominence and a mandibular angle, the subtle slant of a medial canthus, or the way a straight Greek nose bridge meets a heavy brow ridge. However, your visual cortex processes those exact structural vectors effortlessly. By replacing verbal questions with rapid 12-face contact sheets, the MyType FYI test bypasses conscious rationalization and records your raw perceptual attraction.
Why can you tap through a 12-face crowd on MyType FYI in just a few seconds and still produce a highly stable statistical model? Because human facial evaluation is astonishingly fast and consistent over time. In a landmark study published in Psychological Science (Willis & Todorov, 2006), researchers demonstrated that a 100-millisecond exposure (one-tenth of a second) to a stranger’s face yields nearly the exact same attractiveness and trait judgments as unlimited viewing time.
Follow-up studies in Emotion (Bar, Neta & Linz, 2006), PNAS (Ballew & Todorov, 2007), and Social Psychological and Personality Science (Gunaydin, Selcuk & Zayas, 2017) confirmed that split-second evaluations made from a still photograph still predicted how much participants liked that person after meeting them face-to-face a full month later. Because rapid visual attraction is both immediate and durable, the MyType FYI tap-and-replace interface captures high-signal psychometric data without fatigue.
Primary literature: Willis & Todorov (Psychological Science, 2006) · Bar, Neta & Linz (Emotion, 2006) · Gunaydin, Selcuk & Zayas (SPPS, 2017) · Ballew & Todorov (PNAS, 2007).
When you take the MyType FYI quiz, the 12-face grids are not random static pages. Behind the scenes, every synthetic portrait in the catalog is embedded across dozens of continuous latent dimensions—spanning craniofacial bone structure (jaw width, chin projection, nasal bridge contour, brow thickness), eye geometry (hunter eyes vs. doe eyes, scleral show), skin tone, hair texture, and styling vibe.
During your first four crowds on MyType FYI, the sampler maximizes diversity to establish your baseline across the entire facial manifold. From Crowd 5 onward, the algorithm shifts into active learning: each new 12-face crowd mixes high-probability faces that your current vector predicts you will love alongside “boundary-testing” faces that probe the exact edges of your type. Before you tap a single portrait in a new crowd, the model ranks all twelve faces from 1st to 12th. Afterward, it measures how often a face you tapped was ranked higher than a face you skipped—reporting your real-time out-of-sample ROC-AUC accuracy (where 50% is a random coin toss and 85%+ indicates a tightly converged personal aesthetic model).
One of the most common myths on the internet is that 90% of people are competing for the exact same top 5% of “objectively attractive” faces. Large-scale data collected through the MyType FYI attraction test proves the exact opposite: human visual taste—especially when rating women, and to a surprising degree when rating men—is vastly more idiosyncratic than anyone realizes.
Average top-2% overlap: take any two people who rate the same pool of faces on MyType FYI and compare the portraits in each person’s personal top 2%. On average, two raters share only 3% of their top picks. Even the single most similar pair of raters in the dataset shared just 28%.
If attractiveness were a single universal ladder, the highest-rated face in the MyType FYI catalog would land in almost everyone’s top tier. Instead, empirical distributions show that even the #1 highest-consensus face in the entire dataset lands in the top 10% for only 19% of raters, while falling into the bottom 40% for 42% of raters. Meanwhile, a face with a dead-average 50th-percentile market score still lands in the top 10% for roughly one in five people.
There is an interesting gender asymmetry in the MyType FYI data as well: raters evaluating men’s faces exhibit higher consensus (sharing roughly 43% overlap in broad attractiveness tiers), whereas raters evaluating women’s faces diverge dramatically across facial archetypes, bone structures, and styling vibes.
When AI researcher Matthew Fisher built the original MyType FYI model and tested where his own wife, Julie, landed on his preference chart, he uncovered a striking pattern (Fig. 5). On his personal MyType FYI axis (left to right), Julie sat at the 98th percentile—right inside his personal top 2%. Yet on the general crowd-consensus axis (bottom to top), other raters placed her face near the 55th percentile, right in the middle of the market.
In the book Moneyball, a small-market baseball team beat the richest franchises in sports by identifying undervalued players whose on-base statistics translated to wins, even though traditional scouts overlooked them. Dating suffers from the exact same market inefficiency. On conventional swipe apps, millions of users exhaust themselves competing for a tiny handful of 95th-percentile consensus “hotties,” while completely overlooking the 55th-percentile hidden gems (“mids you love”) who are a 98th-percentile match for their personal MyType FYI vector.
When you add the requirement that attraction must be mutual, the math becomes even clearer: if two people are each inside the other’s personal top 1%, that is a 1-in-10,000 coincidence by random street encounters—requiring roughly 24 years of five-minute speed dates to stumble upon by chance. Understanding your type via a visual preference test turns dating from a status competition into a targeted search for mutual hidden gems.
When the MyType FYI test went viral, one unexpected search term surged alongside mytype fyi, my type fyi, mytype.fyi, mytypefyi, mytype test, mytype quiz, and my type quiz: “Greek nose.” Here is why specific bone-structure labels on MyType FYI result cards resonated with millions of users—and how to explore all 77 documented facial types in the MyType FYI directory.
In classical anthropometry and aesthetic portraiture, a Greek nose (also referred to as a straight classical nasal profile) is defined by a remarkably straight nasal dorsum descending from the nasion (the bridge between the eyebrows) to the nasal tip without a pronounced dorsal hump or deep nasal scoop. Unlike a button or upturned celestial nose, a Greek nose creates strong vertical midface architecture and a sculptural side profile that stands out immediately in a MyType FYI 12-face crowd.
Why did so many people discover a preference for the Greek nose only after taking the MyType FYI quiz? Because modern social media filters artificially shrink and flatten noses toward a single homogenized neotenous template. When users tapped through unfiltered, diverse AI-generated bone structures on MyType FYI, thousands realized that their highest-rated faces—especially within the Greek Nose & Classical Profile, Mediterranean Look, Big or Strong Nose (Men), and Prominent Character Nose (Women) clusters—shared strong, character-rich nasal bridges paired with expressive dark brows.
A straight dorsal bridge, minimal nasofrontal depression, defined cheekbones and intense brow framing. Frequently co-occurs with Mediterranean, Levantine and classical European bone structures in MyType FYI models.
Whether you want an instant 90-second visual readout with zero email gates or a deep 20-crowd statistical exploration, here is how our browser-native MyType FYI test compares to account-locked platforms and older text-based personality quizzes.
| This siteMyType FYI instant web app | Account-gated face tests | Text “my type” quizzes | |
|---|---|---|---|
| Sign-up / email login | None — first tap starts it, 100% anonymous | Google OAuth, X login, or email magic link | Often gates results behind email capture |
| Time to result card | ~90 seconds (6–10 crowds) + one-click unlock | 10 to 15 minutes (20+ crowds minimum) | 3 to 5 minutes of multiple-choice reading |
| Measurement method | Revealed preference via 12-face AI contact sheets | Revealed preference via 12-face AI crowds | Stated preference (self-reported clichés) |
| Moneyball matrix & scatter plot | Full 4-row matrix + live 2D map | Included after account verification | Not available (generic paragraph) |
| Crush / partner photo check | Free, 100% on-device, zero upload | Paid add-on ($10–$25) or geo-restricted | Not available |
| 9:16 story card export | Instant high-DPI PNG with your picked faces | Requires account session link | Text screenshot only |
Detailed answers on how the MyType FYI test calculates your physical type, why specific craniofacial traits like the Greek nose appear on result cards, and how your privacy is protected.
The MyType FYI test (often searched as mytype.fyi, my type fyi, mytype test, or mytype quiz) is a viral visual attraction assessment that maps your subconscious physical preferences without asking a single text question. Instead of checking boxes for height or hair color, you tap the faces that catch your eye in 12-portrait AI crowds. Within a few rounds, the MyType FYI statistical model identifies the exact bone structures, eye shapes, nasal profiles (such as the viral Greek nose), and aesthetic archetypes you naturally gravitate toward—along with your 'Hidden Gems' (faces overlooked by the crowd that sit in your personal Top 2%).
No. While the original account-gated version of mytype.fyi requires an email magic link or Google/X OAuth login before saving your crowds, this MyType FYI workspace runs 100% in your browser with zero sign-up required. You can tap faces immediately on the first screen, view your live You-vs-Everyone Moneyball scatter plot after every tap, unlock your full diagnostic card in 6 to 10 rounds, and download a high-resolution 9:16 PNG story card completely anonymously.
Thousands of users sharing their MyType FYI result cards on X (Twitter), TikTok, and Reddit discovered that their top positive trait pill was 'Greek nose', 'Roman nose', or 'Big strong nose'. Because most people cannot articulate craniofacial anatomy in everyday conversation, seeing 'Greek nose' on their MyType FYI readout triggered a massive wave of searches from people looking up what a classical Greek nasal profile looks like (a straight, unbroken bridge descending from the nasion with minimal dorsal hump).
Every AI-generated portrait in the MyType FYI pool has a baseline 'Crowd Consensus Percentile' (how often everyone else taps that face) and a 'Personal Type Score' (how strongly your individual model predicts you will tap it). 'Hotties You Like' are high-consensus faces (top 30% of the market) that you also love. 'Hotties You Skip' are universally popular faces that leave you completely cold—revealing where your taste diverges from the crowd. 'Mids You Love' (or Hidden Gems) are faces rated around the 35th to 60th percentile by the general public that score in the 95th to 99th percentile for your specific MyType FYI vector.
Empirical analysis of MyType FYI rating sessions shows that when two people evaluate the exact same pool of faces, their personal Top 2% favorites overlap by only 3% on average (and at most 28% for the most similar pair). Even the single highest-rated face in the entire dataset lands in the Top 10% for only 19% of raters and falls into the bottom 40% for 42% of raters. Just like Moneyball in baseball, modern dating is inefficient when everyone competes for the same consensus faces while ignoring the 'undervalued' faces that sit right inside their personal Top 2%.
Psychologists distinguish between 'stated preferences' (what you claim in words when someone asks 'what is my type?') and 'revealed preferences' (what your visual cortex responds to in 100 milliseconds). Landmark studies by Willis & Todorov (2006) and Gunaydin et al. (2017) proved that human observers form stable attraction judgments within a tenth of a second based on subtle geometric ratios—jawline gonial angle, brow ridge prominence, interpupillary distance, and smile warmth. Because the MyType FYI test (often searched as my type fyi, mytypefyi, or mytype test) learns strictly from your rapid taps rather than verbal self-reports, the MyType FYI model uncovers the hidden patterns behind attractions that surprise you.
No. Every portrait in the 12-face MyType FYI crowds is a 100% synthetic, AI-generated adult face created specifically for visual preference modeling. No real person is ever rated or ranked in the MyType FYI crowds, and you never need to upload a photo of yourself to take the mytype quiz and discover your type.
Once your MyType FYI preference vector is calibrated, you can optionally test any portrait using the browser-only Compatibility Checker inside your MyType FYI report card. The image is analyzed entirely inside your device's HTML5 Canvas memory (zero bytes are ever transmitted to a server) to estimate how closely its visual contrast and proportion profile aligns with your MyType FYI Top 2% Moneyball zone.
The viral mytype.fyi experiment (also searched as my type fyi, mytypefyi, mytype test, or my type quiz) introduced the 12-face AI contact sheet method for measuring revealed visual attraction. While a traditional my type test asks multiple-choice text questions, our instant MyType FYI web app lets you jump straight into rating 12-face crowds with zero email login, live Moneyball scatter plots, and instant 9:16 result cards.
Two minutes, twelve faces at a time. No account, no upload, no catch.
Start the MyType FYI test
