How to Play Poker Face Test: The Complete Guide to Getting the Best Score

Your score on Poker Face Test isn't just about willpower. It's about setup. The AI camera reads your face 10 times per second, measuring micro-expressions, eye movements, and facial geometry. A poor setup means the camera misses real reactions or penalizes you unfairly. This guide covers everything that affects your score.

Camera Setup

Desktop users:

Position your webcam at eye level. A camera looking up at you from below or down at you from above distorts facial geometry and produces less accurate readings. If your laptop camera is below eye level, raise the laptop on a stand or books until the camera is level with your eyes.

Distance matters. Sit 50-70cm from the camera. Too close and the camera clips the edges of your face. Too far and expression detection becomes less accurate because facial features are too small in the frame.

Center your face in the preview window before starting. The face mesh overlay shows you exactly what the AI is reading. Make sure your entire face is visible including your forehead and chin.

Mobile users:

Hold your phone vertically at eye level. Do not tilt the phone sideways. The game is designed for vertical orientation.

The biggest mistake mobile users make is holding the phone too low, looking down at it. This creates an unnatural face angle that the AI reads differently from a neutral resting face. Hold the phone up so the camera is directly in front of your face.

If you're playing on a table or surface, prop the phone against something stable so it doesn't move. Any camera shake gets picked up by the face tracking and can affect your score.

Distance on mobile: 40-60cm from your face. Closer than desktop because mobile cameras have wider lenses.

Lighting

Lighting is the single biggest factor most players ignore.

What works:

• Natural daylight from a window facing you (not behind you)
• A lamp positioned in front of you at face level
• Even, diffused light that illuminates your whole face without shadows

What kills your score:

• Backlight — sitting with a window behind you puts your face in shadow. The AI struggles to read expressions accurately on a dark face.
• Side lighting — strong light from one side creates shadows that make one half of your face harder to read
• Overhead lighting — harsh shadows under your eyes and nose affect landmark detection
• Flickering light — TV screens or certain LED lights that flicker can confuse the detection

Simple test: Look at your face in the camera preview before starting. If you can clearly see both eyes, both eyebrows, your nose and mouth without harsh shadows, your lighting is good.

Posture and Head Position

Keep your head still and level. The AI establishes a baseline of your head position before the game starts. Any movement from that baseline gets detected and penalized.

Desktop: Sit with your back straight, head level. Don't lean forward toward the screen during shocking images — that head movement registers as a reaction.

Mobile: Keep the phone stable. If you're holding it, grip it firmly with both hands so hand tremors don't translate into camera shake.

Head tilt: Even a slight habitual head tilt affects scoring. During the calibration countdown, hold your head in the most natural, comfortable neutral position you'll maintain throughout the game. The AI calibrates to that position, so consistency matters more than perfection.

Eyes

Eye movements are heavily weighted in the scoring algorithm.

Blinks: Every blink is detected and penalized. You cannot avoid blinking entirely, but reducing blink frequency during the 3-second image windows helps your score. Practice slow, controlled blinking rather than rapid blinking.

Winks: A wink is detected as a major reaction. Avoid any asymmetric eye movement.

Eye widening: Surprise causes involuntary eye widening. The AI detects this through changes in eye aspect ratio. When you see something shocking, the instinct is to open your eyes wide. Practice keeping your eyes at a consistent level of openness.

Looking away: Resist the urge to look away from disturbing or funny images. Gaze shift is detected through changes in facial geometry. Keep your eyes on the screen.

Mouth and Jaw

The mouth region contains 20 landmark points. Any movement is detected.

Smirks: A slight lip corner pull registers before it becomes a visible smile. The algorithm detects early smile indicators. If you feel a smirk coming, the best technique is to slightly press your lips together — not tight enough to show tension, but enough to resist the lip corner movement.

Jaw: Surprise causes an involuntary jaw drop. Even a millimeter of jaw movement registers. Keep your jaw relaxed but controlled — not clenched tight, which creates visible muscle tension, but not loose enough to drop on impact.

Breathing: Breathe through your nose during the challenge. Mouth breathing causes constant small mouth movements that affect scoring.

Eyebrows

Eyebrows are the most expressive and least controllable part of the face.

Surprise raises both brows involuntarily. This is one of the hardest reactions to suppress because it happens before conscious awareness.

Disgust furrows the brows. Cringe content triggers this.

Practice: In a mirror, practice keeping your brows at a neutral, level position while imagining surprising or disturbing content. The muscle control needed to suppress brow movement is trainable.

Before You Start

The calibration countdown matters. During the 3-2-1 countdown before the game starts, the AI reads your resting face and sets your personal baseline. How you hold your face during those 3 seconds becomes the reference point for everything that follows.

During calibration:

• Hold your face in the most neutral, relaxed position you can maintain
• Don't smile, raise your brows, or tense any muscles
• Keep your eyes at their natural resting openness
• Breathe normally through your nose

A good calibration gives you the best possible starting point.

Mental Techniques

Don't anticipate. The worst thing you can do is try to prepare for what the image might show. Anticipation creates micro-expressions before the image even appears.

Focus on a fixed point. Pick a spot on the screen and keep your gaze there. This reduces involuntary eye movement when images appear.

Slow breathing. Slow, controlled breathing activates the parasympathetic nervous system, which reduces involuntary facial reactions. Breathe in for 4 counts, out for 4 counts.

Detach. The players who score highest describe a mental state of detachment — observing the image rather than experiencing it. This is trainable with practice.

Common Mistakes

• Starting before the camera has fully loaded and calibrated
• Poor lighting that forces the AI to work with a shadowed face
• Camera below eye level creating an unnatural face angle
• Tensing face muscles to suppress reactions — tension itself registers
• Looking away from disturbing content
• Holding breath and then exhaling visibly when an image changes
• Rapid blinking from nervousness before images appear

The Video Challenge

The video challenge scores differently from the image challenge. Instead of 10 separate 3-second rounds, you're scored continuously for the full video duration.

This means:

• One big reaction in the middle of a 60-second video costs you more than the same reaction during a 3-second image
• Recovery matters — if you crack, regaining composure quickly limits the damage
• The longer the video, the more opportunities for involuntary reactions

Additional tips for video:

The audio is part of the challenge. Funny sounds trigger facial reactions even when the visual doesn't. Keep your face neutral even when you hear something amusing.

Videos often build to a climax. Knowing this, avoid anticipating the peak moment — that anticipation shows on your face before the moment arrives.

Practice Makes Perfect

Your score improves with practice. The AI is consistent — it measures the same things every time. The variable is you.

Play the 10 image challenge first to understand what triggers your reactions. Your score breakdown shows which images caused the most damage. Use that data to identify your weak points.

Then play the video challenge. Then challenge a friend.

The more you play, the better your score. That's the whole point.

Start now at http://pokerfacetest.com/