Composition rules are universal. Signaled traits are not. Window light above eye level, Duchenne smile, jaw-preserving angles, body shot at slot 3. The audit names the photo that plays best as your lead.
Built around Penton-Voak 2001 (facial dimorphism), Ekman FACS (Duchenne smile), Willis-Todorov 2006 (100ms first-impression).
17-metric written report with gender-weighted scoring
The mechanical composition rules transfer directly: face dominance at 25 to 40 percent of frame, eye-level camera, eye contact, Duchenne smile, shoulders 15 to 30 degrees off-axis, focal distance around one meter. These are universal portrait conventions.
The traits the photo signals differ. Penton-Voak (2001) and subsequent sexual-dimorphism work show male faces benefit from photos that preserve jaw angularity, brow projection, and the shadow structure that gives the face its three-dimensional read. Female faces benefit from photos that preserve skin smoothness and warmth markers. Lighting choice is the main lever; window light from above eye level preserves the male-coded structural signal, where soft front-diffuse preserves the female-coded warmth and smoothness signal.
The other large difference for men is the social proof and lifestyle slots carry more weight on the message-first format. On Bumble specifically, women decide whether to write to a match, and the lifestyle and social slots are the second-stage filter after the lead. See Bumble photo strategy for the message-first treatment, and the women's guide for the parallel.
Window light from above eye level is the workhorse setup for male portraits. A north-facing window with the sun blocked by clouds, the subject standing 45 degrees off-axis, with the window roughly at the height of the top of the subject's head. The slight elevation creates a shadow under the jaw and along the cheekbone, which preserves the structural definition that the male-coded photo signal benefits from.
Flat front light is the most common mistake. Smartphone flash, an overhead fluorescent, or noon overhead sun all flatten the face and erase the jaw shadow. The same face shot in flat front light versus slightly elevated window light reads meaningfully different in side-by-side comparison; the elevated-window version reads as more masculine without changing the subject.
Direct hard side light (a bright window directly at 90 degrees on a sunny day) splits the face into hot and shadow halves. For men this can work as slot 4 or 5 mood-coded content (a candid working photo, an outdoor portrait) but it is too intense for slot 1. The lead photo should be soft-window-from-above. For the broader lighting fundamentals, see the lighting guide.
The most common male dating-photo mistake is the stoic-moody portrait at slot 1. The reasoning behind the stoic photo (looking mysterious or intense) does not survive contact with how viewers actually rate photos. Photo-rating studies consistently show Duchenne smiles outscore non-Duchenne expressions across rater genders. For men specifically, a soft closed-mouth half-smile with engaged eye crinkles tends to score as high or higher than a wide open-mouth smile, while preserving more jaw definition.
The Ekman-Friesen Facial Action Coding System defines the Duchenne smile as AU12 (lip corner pull) plus AU6 (cheek raise, eye crinkles). The AU6 signal is the warmth driver; AU12 without AU6 reads as performative. Three production tactics: think of a real recent moment that made you laugh, have a friend off-camera prompt expression between shots, or shoot in bursts and select the frame just after a laugh where the eyes still crinkle.
The audit at /audit flags Duchenne-true and Duchenne-false explicitly per photo, and the report includes the per-photo warmth read.
Camera at eye level. The duck-face selfie angle (phone held above) over-collapses the jaw for men and reads as performative. The chin-up angle (camera below eye level) exaggerates nostrils and the chin and reads weak. Eye level reads as peer-to-peer and preserves the jaw line cleanly.
Shoulder rotation 15 to 30 degrees off-axis. The three-quarter angle introduces depth and gives the camera something other than a flat outline to render. For men with strong jaw definition, the three-quarter angle is the lever that converts the jaw into the dominant visual feature. Combined with the slightly elevated window light, this is the canonical male portrait setup.
Posture: shoulders pulled back without becoming military-stiff, chest open, head level (not pushed forward into the camera, which is the "Reddit selfie" pose and reads insecure). Slight head tilt off vertical (5 to 10 degrees) reads casual; dead-vertical reads stiff. See the best face pose guide for the underlying conventions.
Lead wardrobe: well-fitted casual that you would actually wear on a first date. A solid-color t-shirt or henley, a button-down rolled at the sleeves, a casual sweater in cooler months. Avoid graphic tees, sports jerseys, and athletic wear at the lead because they read as either lazy or single-interest. Fit matters more than brand; a $25 well-fitted t-shirt outperforms a $200 poorly-fitted shirt in viewer ratings.
Variety across the 6 photos: lead in one wardrobe, slot 2 in different context, body shot in real-world clothing that signals lifestyle, activity in context-appropriate dress (sports, hobby, outdoor), social in casual context, warm close in different season or location. The variety signals that the photos are recent.
Two things to avoid: sunglasses on a shirt collar (reads dated), and clothes that are too formal (a suit-tie photo at slot 1 reads transactional or LinkedIn-imported). One suit photo at slot 4 or 5 in clear professional context is fine if professional credibility is part of the signal you want to communicate.
One full-body or three-quarter body shot at slot 3 is the rule. Real-world clothing, natural posture, in a location with visual interest, photographed by another person. The body shot pre-empts the body curiosity that would otherwise hijack attention in later slots. Mirror selfies are not a substitute; arms-length distortion plus the mirror context lower the read.
Shirtless is a separate question. Shirtless lead is a documented mistake (transactional read, lower message-first conversion on Bumble). One physique-in-context shot at slot 4 or 5 is acceptable if the physique is genuinely strong and the context is real (a beach with friends, a pool, a hiking shot where the shirt is off because of heat). The flexing gym mirror selfie is not in this category and underperforms.
For the deep 6-slot composition see dating app photos, and for the catastrophic mistakes see dating app photo mistakes.
Stoic-moody portraits at the lead. The intent (mysterious, intense) reads cold to most raters, and on the message-first format on Bumble specifically the cold read lowers the woman-writes-the-opener probability. Save mood photos for slot 4 or 5 if they are genuinely strong portraits with clear context.
Group photos at slot 1. The viewer cannot identify which person is you in the 100 millisecond first-impression window. Move groups to slot 5 with 2 to 3 friends and you visually obvious as the subject.
Fish or trophy photos. These cliched dating-app patterns are widely mocked and lower the read even from people who fish or hunt. If outdoor lifestyle is real, use a candid trail, lake, or campsite photo instead.
Gym mirror selfies. The mirror context and arms-length distortion combine to lower the read regardless of physique. Physique signals are better conveyed through fitted casual clothing, lifestyle context (hiking, sports), or one beach-or-pool shot at slot 4 or 5.
The complete catalog with research-backed fixes is at dating app photo mistakes.
The audit at /audit scores each photo on 17 published facial-perception metrics: symmetry, canthal tilt, fWHR, facial thirds, expression (Duchenne true or false), focal distance, lighting direction, crop, contrast, jaw definition, brow projection, eye area, lip area, nose projection, hair framing, background visual weight, and outfit fit. For male subjects, the report weights jaw definition, brow projection, and structural metrics slightly higher than for female subjects, reflecting the Penton-Voak dimorphism findings.
The output is a 5-page written report with per-photo scores, a recommended slot order, and a list of specific lighting or crop edits per photo. The lighting recommendations are tuned for the male-coded structural signal (window light above eye level versus soft front-diffuse). See the citations at the research base.
For a quick scored read without the written report, the free photo ranker covers the basics.
The $29 dating audit returns a 5-page written report with per-photo scoring, slot order, and lighting fixes tuned for the male-coded structural signal. The free photo ranker handles the basics.
5-page written report scoring 17 metrics with jaw definition and structural signals weighted for male subjects.
All free. All private. All instant.
Is your smile genuine or forced?
How close are your proportions to ฯ?
AI attractiveness analysis
Rate my face 1โ10
How attractive am I?
How symmetrical is your face?
Which photo gets more matches?
Best photo for LinkedIn
Your glow-up score
Hand-picked from 90+ tests, guides, and audits.
Quiz-format attractiveness scoring
LooksmaxHonest AI verdict in 30 seconds
LooksmaxGet your decile rank in 30 seconds
LooksmaxHonest AI verdict in 30 seconds
LooksmaxNo-fluff score with grounding science
LooksmaxPeer-reviewed scoring methodology