Form check: improve your lifts with better technique

    I hate seeing people chase heavier weights while their squat turns into a good morning. A formcheck isn't about vanity: it's the fastest way to tell if your nervous system is learning the right pattern or compensating into injury territory. I'll look at one angle at a time: bar path, knee tracking, spinal position. Fix the biggest leak first, then add load.

    Proper form is the cornerstone of effective strength training, yet many lifters struggle to assess their own technique. In other fields, researchers have long relied on rigorous evaluation frameworks [1] and advanced imaging to reveal hidden details [2], while psychologists note that unwanted thoughts about harm can affect performance [3]. Similarly, automated theorem provers use logical reasoning to verify correctness [4], and chemical studies examine reaction mechanisms at the atomic level [5]. These diverse approaches underscore the universal need for precise assessment, a principle that directly applies to form checking in the gym. Just as global software development teams adopt structured practices to ensure quality [6], and teachers play a strategic role in achieving educational goals [7], athletes must incorporate systematic form checks to prevent injury and maximize gains. Even socioeconomic analyses of land lease and migration [8] and pathway-targeted therapies for feline coronavirus [9] highlight the importance of identifying critical variables. By identifying relevant contextual factors [10] and addressing the gap between pain prevalence and treatment effectiveness in postoperative care [11], we can draw parallels to the need for accurate form feedback in strength training. With dorsi.ai, you get an AI coach that applies these principles to your lifts, ensuring every rep is safe and effective.

    Practical Playbook

    1. What does bad form actually look like?

      Bad form isn't always the dramatic collapse you see in fail compilations. It's the small stuff: knees caving inward on a squat, a deadlift bar drifting forward of midfoot. Most people don't catch it because it feels normal. That's the problem: your body will adapt to slop, and strength plateaus follow.

    2. Record yourself from two angles

      Set your phone at hip height. One angle from the front, one from the side. Film three working sets, not your warmup. Watch the video at half speed. Look for joint angles at the sticking point, does your back shift? Does the bar drift? You'll see in thirty seconds what you missed across thirty reps.

    3. Compare against a single reference point

      Don't watch five different coaches and get confused. Pick one standard: a textbook side view of a clean deadlift. Frame-match it. Does your hip rise first? Do your shoulders track over the bar? One reference eliminates the noise. If you're lifting alone, Dorsi's form feedback tool can spot mismatches in real time.

    4. Implement one fix per session

      You can't fix everything at once. Pick the single most obvious flaw, knee cave, rounded upper back, bar path, and drill it during warmups. One rep with perfect execution beats a set of five with four bad ones. Progress compounds faster than you think.

    Process at a glance1What does badform actuallylook like?2Record yourselffrom two angles3Compare againsta singlereference point4Implement onefix per session
    Process at a glance

    Common Mistakes

    • Mistake
      Fixing your form by watching your reflection in the gym mirror.
      Why
      Mirrors lie. The angle distorts depth perception and you're shifting your head to see, that changes the spine position you're trying to assess.
      Fix
      Record a set from the side at hip height. Review the video after, not during. You'll catch the real angles.
    • Mistake
      Rewatching the same lift until you convince yourself it's fine.
      Why
      Anchoring bias sets in. The first look is usually the most honest. After that your brain starts filling in 'good enough'.
      Fix
      Watch the recording twice max. If something looks off on first pass, it's off. Move on or mark it for an adjustment.
    • Mistake
      Trying to match the bar path of a powerlifter you follow online.
      Why
      Your femur length changes your hip hinge. Trying to replicate someone else's back angle often pushes your torso into compensation.
      Fix
      Learn the cues that apply to your proportions: hinge until you feel the hamstrings load, don't force the bar over midfoot.
    • Mistake
      Saving form check requests for when something hurts.
      Why
      Pain is the last signal. By the time it hurts you've reinforced a bad pattern for weeks. Small errors accumulate.
      Fix
      Log a form check every 4-6 weeks regardless of how you feel. Catch drift before it becomes injury.

    Sources we drew from

    1. 1
      Synthesising verified access control systems through model checkingPeer-reviewed

      Nan Zhang et al. · 2008 · Journal of Computer Security

      We present a framework for evaluating and generating access control policies.

    2. 2
      Expansion Microscopy for Beginners: Visualizing Microtubules in Expanded Cultured HeLa Cells.Peer-reviewed

      Zhang C et al. · 2020 · Current protocols in neuroscience

      Expansion microscopy (ExM) is a technique that physically expands preserved cells and tissues before microscope imaging, so that conventional diffraction-limited microscopes can perform nanoscale-resolution imaging.

    3. 3
      Maternal unwanted and intrusive thoughts of infant-related harm, obsessive-compulsive disorder and depression in the perinatal period: study protocol.Peer-reviewed

      Collardeau F et al. · 2019 · BMC psychiatry

      <h4>Background</h4>Unwanted, intrusive thoughts of harm-related to the infant are reported by the vast majority of new mothers, with half of all new mothers reporting unwanted, intrusive thoughts of harming their infant on purpose.

    4. 4
      A Prover for the muCRL toolset with applications -- version 0.1Peer-reviewed

      Jaco van de Pol · 2001 · Centrum Wiskunde & Informatica (CWI), the national research institute for mathematics and computer science in the Netherlands

      This document describes an automated theorem prover, based on an extension of binary decision diagrams.

    5. 5
      A computational study of the reactions of a β-diketiminatoaluminium( <scp>i</scp> ) complex with the hydrogen atom and the electronPeer-reviewed

      Iain McKenzie et al. · 2005 · Chemical Communications

      A computational study has been performed to examine the reactions of a model beta-diketiminatoaluminium (I) complex with the hydrogen atom and with the electron.

    6. 6
      AN ONTOLOGICAL FRAMEWORK FOR REQUIREMENT CHANGE MANAGEMENT IN DISTRIBUTED ENVIRONMENTPeer-reviewed

      Asma Khatoon et al. · 2014 · The Nucleus

      Global Software Development (GSD) is getting fame in the software industry gradually.

    7. 7
      Development of Knowledge Management System of Teachers’ CompetencyPeer-reviewed

      Annisa Saraswati et al. · 2019 · Jurnal Pendidikan Teknologi dan Kejuruan

      Teachers had a strategic role in helping achieving an educational purpose.

    8. 8
      Does Land Lease Affect the Multidimensional Poverty Alleviation? The Evidence from Jiangxi, ChinaPeer-reviewed

      Hui Xiao et al. · 2023 · Land

      This study uses field survey data from 382 families in the year 2020 in Jiangxi province, China, to explore the effects of land lease and labor migration, a well-known occurrence in China, on rural households’ multidimensional poverty stat…

    9. 9
      Identifying Natural Products as Feline Coronavirus M<sup>pro</sup> Inhibitors by Structural-Based Virtual Screening and Enzyme-Based Assays.Peer-reviewed

      Jiang Z et al. · 2025 · ACS omega

      The main protease (M<sup>pro</sup>) is a pivotal target in the life cycle of feline coronavirus (FCoV), which causes a high mortality feline disease, feline infectious peritonitis (FIP).

    10. 10
      The quest for organizational flexibility: Driving changes in business processes through the identification of relevant contextPeer-reviewed

      Monica Anastassiu et al. · 2016 · QUT ePrints (Queensland University of Technology)

      &lt;b&gt;Purpose&lt;/b&gt; - The purpose of this paper is to propose a method for identifying business process-relevant contextual information that is likely to impact on the process goal.

    11. 11
      The Effect of Benson Relaxation On Pain in Post Major Surgery PatientsPeer-reviewed

      Alivia Mona Kaparang et al. · 2022 · Malahayati Nursing Journal

      ABSTRACT In the United States &gt;80% of postoperative patients experience pain but only 30-50% receive effective treatment.

    Just show up. Dorsi handles the rest.

    • HRV-driven readiness — today's plan adapts to how recovered you actually are.
    • Adapts every session — no decision fatigue, no second-guessing your numbers.
    • Apple Watch native — log a set with your wrist, not your phone.
    Download on the App Store

    Related topics

    dumbbell bench press

    barbell rows

    bent over rows alternative