You did the right thing. You couldn't tell if your squat was any good from the inside — nobody can — so you filmed it and asked for help. Maybe you posted it to r/formcheck. Maybe you dropped it in the gym group chat, or sent it to the strongest person you know.
Then the answers came back, and they didn't agree on anything.
Go deeper, you're cutting it high. You're already past parallel, stop before you round. Chest up. You're leaning too far forward. Widen your stance. Narrow your stance. Looks fine to me, add weight. Five people, five verdicts, and at least four of them can't be right at the same time. So you're back where you started, except now you're also worried, because a stranger with a lot of upvotes told you your back is going to explode.
The contradictions aren't because form checks are useless. They're because of who you asked, and what you gave them to work with.
The crowd can't see your anatomy either
Start with the deepest reason, because it's the one nobody says out loud. Even a well-meaning, knowledgeable person looking at your clip can't see the thing that actually determines what your squat should look like: your skeleton. How long your femurs are relative to your torso, how deep and which way your hip sockets are angled, how much your ankles bend. Good form is a range set by those proportions, and they're invisible from the outside. A ten-second video from one angle doesn't reveal them.
So the person answering does the only thing they can: they compare you to the shape they think a squat is supposed to make — which is usually the shape their squat makes, or the one from the coaching video they learned from. When your forward lean is actually correct geometry for a long-femured body, they read it as an error and tell you to fix a thing that isn't broken. They're not careless. They're guessing at a body they can't measure, using themselves as the ruler.
Everyone answers from the lift that fixed their lift
There's a specific trap that makes online form advice louder and more confident than it should be. The person most eager to help you is often the person who recently fixed the same-looking problem on their own body. It worked, it felt like a revelation, and now they see that fix everywhere. Bad back day taught them to brace harder, so every rounded-looking rep gets "brace harder." A cue that unlocked their depth becomes the cue they hand to everyone, whether or not it addresses what's actually happening in your rep.
This is why you'll get advice that's genuinely good — for someone — and completely wrong for you. Each answer is a true story about the body that gave it. Stacked together in a comment thread, a dozen true stories about a dozen different bodies read as pure noise. The upvotes don't sort for correct; they sort for confident and relatable, which is not the same thing.
A single clip is a thin slice of the truth
Even setting bodies aside, you handed the crowd almost nothing to work with. One set. Often one rep. One angle, frequently the worst one — straight-on, which hides depth and spinal position. No idea what your third set looks like versus your first, when fatigue starts bending the movement. No baseline, so nobody can tell whether what they're seeing is your normal or an off day.
A form problem is rarely a single frame. It's a pattern — the knee that caves a little more each rep as the set wears on, the depth that quietly creeps up as the weight climbs. You can't see a pattern in a highlight. Neither can the stranger you sent the highlight to. The mistakes that matter happen in about 200 milliseconds and never reach your conscious brain; a thumbnail-sized clip in a busy feed doesn't catch them any better than you did in the moment.
The instinct is right. The jury is the problem.
None of this means "stop asking." Filming your lifts and getting an outside read on them is exactly the correct instinct — it's the whole fix for a sense of your own body that's confident and unreliable. The problem isn't looking outside yourself. It's outsourcing the read to a rotating crowd of strangers, each seeing a sliver, each measuring you against themselves.
What actually moves a lift forward has three properties a comment section structurally can't:
- It knows your proportions, so it can tell the difference between a compensation and a body doing what its geometry requires. A forward lean isn't a verdict until you know whose femurs are attached to it.
- It watches more than the highlight — every rep, not the one you were proud enough to post — so it catches the pattern instead of a lucky or unlucky frame.
- It's consistent over time. The same evaluator, watching you across sessions, can say "this is better than last week" or "you only round on the last two reps." A new set of strangers every time you post can never tell you whether you're improving, because none of them saw the last one.
You can approximate this yourself. Pick one angle — roughly 45 degrees to the side, hip height — and film from the same spot every time so your videos are actually comparable. Record your top sets, where form breaks first, not your easy ones. Compare yourself to yourself over weeks, not yourself to a stranger in a thread. And if you want a human read, get it from one knowledgeable coach who watches you repeatedly, not a jury that changes every post.
That consistent, body-aware read is the thing we're building Flexion to give you: it watches your reps — all of them, from the angle that shows the most — and calls out the specific thing to fix and why, the same way every session, so you can actually see the line move. It's the opposite of a comment thread. One read, tuned to your body, instead of ten guesses tuned to everyone else's.
But the first step costs nothing and doesn't need us: stop crowdsourcing your form to people who can't see your bones. Film consistently, compare yourself to your past self, and be very skeptical of any stranger who's certain about a body they've watched for ten seconds.
