The central claim is not merely that safer exercises feel better or look cleaner. The stronger claim is that a safer, better-organized lower-body pattern may let the body produce meaningful force with less wasted motion, less harmful compensation, and more repeatable quality under load and fatigue.
In that context, “safety” does not just mean lower injury risk in a vague sense. It means preserving joint relationships, reducing unnecessary stabilization noise, limiting excessive side dominance, and allowing more correct repetitions to accumulate. That is important because the body does not only need to produce force. It needs to direct force, transfer force, and repeat force without the kinetic chain beginning to leak.
To understand that, the reader first needs to know what researchers are actually measuring. When a paper reports peak force, it is measuring the highest amount of force produced during the movement, usually in Newtons. That tells us how much raw output was generated. When a paper reports average velocity, it is describing how fast the bar or body moved across the repetition. That matters because force and velocity together help determine power expression. If force stays similar but velocity drops, that tells us the movement may still be strong, but the expression of that strength has changed.
When a paper reports electromyography, or electromyographic activity, it is measuring the electrical activity associated with muscle activation. More electromyographic activity does not always mean “better,” but it often tells us the body is doing more muscular work to stabilize or execute the task. When a paper reports joint force at the tibiofemoral joint or patellofemoral joint, it is estimating the internal load experienced by the knee structures. And when a paper reports a limb symmetry index, it is comparing one limb to the other, with 100% meaning perfect symmetry and lower values indicating greater asymmetry.
The Safety Squat Bar data are useful because they show that a more organized squat pattern can preserve meaningful force output. In one 2024 study, peak force during an acute high-intensity session was 2443.0 ± 46.6 N with the traditional Olympic bar and 2622.9 ± 65.8 N with the Safety Squat Bar, and that difference was NOT statistically significant.
What that means in plain language is that the Safety Squat Bar did NOT meaningfully reduce the lifter’s ability to produce force in that test. The same study found average velocity was 0.42 ± 0.04 m/s with the Olympic bar and 0.38 ± 0.05 m/s with the Safety Squat Bar. That means the bar moved a little slower with the Safety Squat Bar, even though peak force stayed similar. The practical meaning is not that the Safety Squat Bar is “weaker,” but that it changes the movement solution. It can keep force production in the same neighborhood while altering how the body organizes the lift. This suggests a more upright, more manageable pattern can preserve substantial output rather than sacrificing it entirely.
The 2024 biomechanical comparison between the Safety Squat Bar and the traditional barbell adds another layer. That study focused on kinematics and kinetics. Kinematics means how the body moves, such as joint angles and torso position. Kinetics means the forces and moments acting through the joints. The authors found that the Safety Squat Bar changed the movement mechanics, including posture, while preserving similar knee-joint demands under the tested conditions. It is not simply about “lifting the most weight.” It is about whether the body can hold a stronger pattern while still producing training-relevant force. If a bar variation gives you similar joint demand where you want it, while improving the movement organization, that supports the idea that safety and performance do not always oppose one another. Sometimes a safer pattern is simply a more efficient pattern.
The hand-supported squat data are especially important because they show what happens when you reduce stabilization demand directly. In the 2025 study by Takeshita et al, participants used 12 ± 5% of body weight through the hands. That means the hands were providing a modest but meaningful amount of assistance. With that support, tibiofemoral joint force dropped from 321 ± 38% body weight to 278 ± 42% body weight, and patellofemoral joint force dropped from 427 ± 46% body weight to 356 ± 61% body weight. In plain language, the knee joint experienced less internal loading when some of the stabilizing and support burden was shared by the hands. That does not prove that support “improves stability” as a motor skill, but it does prove that support can reduce internal cost while keeping the squat pattern intact. This suggests that a more supported setup may allow more tolerable reps, more volume, or better recovery when the goal is to train a pattern without burying the athlete in unnecessary joint stress.
The Smith-machine data help explain the stabilization side of the equation. In the classic comparison, free-weight squats showed 43% higher average electromyographic activity across the measured muscles than Smith-machine squats. Specific muscles were also more active in the free-weight squat: gastrocnemius by 34%, biceps femoris by 26%, and vastus medialis by 49%. What does that mean? It means the free-weight squat required the lifter’s body to generate more muscular activity to control and stabilize the movement. That is a sign that the free squat asks more from the self-organizing system. But that is not automatically a good thing. It means the free squat contains more coordination demand and more stabilization demand. If the athlete is well prepared, that can be valuable. If the athlete is leaking, that same demand can become noise. So the data do not say “Smith is better.” They say the Smith machine removes some of the stabilization problem. We are distinguishing productive force expression from force wasted on managing instability.
This is where side dominance enters the picture. The asymmetry study after anterior cruciate ligament reconstruction is powerful because it compares a simpler task, the squat, to a more demanding task, the drop jump. Popovich and his colleagues measured vertical ground reaction force limb symmetry index and knee net joint moment limb symmetry index. Vertical ground reaction force is the force the ground applies back into the body. The limb symmetry index tells us how evenly each leg shares that force. Knee net joint moment is a way of estimating the turning force around the knee joint; again, symmetry tells us how evenly the legs are handling the work. In that study, vertical ground reaction force symmetry dropped from 90.0 ± 11.0% during squatting to 79.7 ± 14.9% during drop jumping. Knee net joint moment symmetry dropped from 66.4 ± 25.6% during squatting to 55.8 ± 17.6% during drop jumping. In plain English, the movement became more asymmetric when the task became more chaotic and demanding. That is a major defense of the Strong Angle system. Leakage in the simpler pattern does not disappear under higher demand; it often gets worse.
The correlations in that same study matter too. The researchers found moderate correlations between squat asymmetry and drop-jump asymmetry, with correlation coefficients around 0.58 and 0.61. A correlation does not prove causation, but it does tell us that the people who were more asymmetric in the squat also tended to be more asymmetric in the drop jump. That supports the logic behind our contact ladder. If the lower-level pattern is already leaking, the higher-level pattern is likely to reveal even more leakage. So the contact ladder is not just a convenient progression. It is a screening and development strategy based on the idea that unresolved problems at lower demand are often amplified at higher demand.
That is also why the plyometric progression literature matters. The 2021 ACL reconstruction commentary by Buckthorpe proposes a four-stage criterion-based plyometric progression. “Criterion-based” means progression is not driven only by time or grit. It is driven by whether the athlete demonstrates the qualities needed to move to the next level. That aligns very well with your model. The contact ladder is essentially saying that people should earn higher-demand contacts by demonstrating enough alignment, symmetry, control, and tolerance at lower-demand levels first. The research language is clinical, but the meaning is the same: do not jump into chaos until the body has shown it can organize itself in simpler tasks.
Training the pattern rather than hoping for indirect transfer is also well defended by the specificity literature. The 2025 systematic review by Saaterbaken et al found that dynamic resistance training produced about a two-fold larger effect size in trained dynamic strength than in non-trained isometric strength. In simple terms, the body gets better at what it practices. If you train a dynamic pattern, the biggest gains show up in that dynamic pattern, not equally everywhere else. Organized, pattern-specific conditioning of the kinetic chain is superior to assuming that generic loading will automatically transfer into power, power endurance, landing, deceleration, or real-world execution. The specificity review does not prove every Strong Angle claim, but it strongly supports the foundational idea that direct pattern practice beats indirect hope.
The hex-bar deadlift data extend that same logic to another pattern. In the 2011 study, subjects lifted a heavier one-repetition maximum with the hex bar than the straight bar, 265 ± 41 kg versus 245 ± 39 kg, and also produced higher peak power, 4872 ± 636 W versus 4388 ± 713 W. One-repetition maximum is simply the heaviest load a person can lift one time. Peak power combines force and velocity, so it reflects explosive output more directly than force alone. The important point is that the hex bar changed the geometry of the lift in a way that let people lift more and express more power. That supports the broader inference that setup matters. A better-organized pattern can improve usable output, not just comfort. It also reinforces that training is not just about how much force exists in the system; it is about how well the system is positioned to express it.
The rear-foot-elevated split squat data from Helme et al help on the asymmetry side. That exercise was found to be a valid and reliable measure of unilateral leg strength symmetry. “Valid” means it measures what it is supposed to measure. “Reliable” means it gives consistent results across testing. The key variable in that study was again vertical ground reaction force symmetry, meaning how evenly the lead leg expressed force. This matters because unilateral exercises often expose differences that bilateral exercises can hide. They make side dominance easier to see. That does not mean bilateral training is bad. It means bilateral training alone may not tell the whole story. So step-ups, split squats, and lunges plausibly fit the Strong Angle model because they offer clearer opportunities to see and correct asymmetry before it gets amplified in sprinting, jumping, or chaotic movement.
Correlation tells us whether leakage in one task tends to travel into leakage in another task. And criterion-based progression tells us that the best next step should be earned, not assumed. Once you understand those variables, your understanding becomes much clearer: the best exercise or machine is not simply the one that looks hardest or loads the most weight. It is the one that gives the body the best chance to produce meaningful force, with manageable cost, with lower leakage, and with better transfer into the next level of demand.
This is exactly where the UpSquat sits. The current research does not directly prove that the UpSquat is the safest machine or the best machine. That claim still needs direct validation. But the data do support the underlying logic to justify it. If a machine can help the athlete maintain better alignment, reduce excessive frontal-plane drift, reduce side-dominance leakage, and accumulate more correct reps under speed and load, then the surrounding literature suggests that machine should improve the quality of force expression and the quality of progression. That is still an inference, but it is a very reasonable inference.
The current evidence does not yet let you say, “The UpSquat is proven superior.” It does let you say, “The design hypothesis behind the UpSquat is highly consistent with what current biomechanics and progression research suggest should matter.”
The main challenge to the hypothesis is that more support is not always better. The Smith machine example shows that reducing stabilization demand can also reduce the need for the body to self-organize. That means support can be a powerful tool, but it cannot be the whole story. At some point, the athlete still has to transfer that organization into freer, faster, less predictable tasks. That is why the contact ladder matters so much. The answer is not “always use the most supported version.” The answer is “use enough support to train the correct pattern, then progressively remove support as the athlete earns more chaos.”
Conclusion: The data suggest that better-organized leg exercises can preserve meaningful output, reduce some mechanical cost, expose or reduce asymmetry, and prepare athletes more intelligently for higher-demand movement. The research does not yet prove every Strong Angle claim, and it does not yet validate the UpSquat directly. But it does support the core idea that the missing filter in lower-body training is often not more effort. It is better organization. If that is true, then the real bottleneck in conditioning is not simply how much force a person can produce. It is how well they can preserve force integrity as demand rises. That is the heart of the argument, and the data we reviewed does give it a serious foundation.
Bibliography
Buckthorpe, M., Della Villa, F., Della Villa, S., & Roi, G. S. (2021). Recommendations for plyometric training after ACL reconstruction: A clinical commentary. International Journal of Sports Physical Therapy, 16(3), 879–895.
Helme, M., Bishop, C., Emmonds, S., & Low, C. (2019). Validity and reliability of the rear foot elevated split squat 5 repetition maximum to determine unilateral leg strength symmetry. Journal of Strength and Conditioning Research.
Johansson, D. G., Marchetti, P. H., Stecyk, S. D., & Flanagan, S. P. (2024). A biomechanical comparison between the Safety Squat Bar and traditional barbell back squat. Journal of Strength and Conditioning Research, 38(5), 825–834.
Saeterbakken, A. H., et al. (2025). Task specificity of dynamic resistance training and its transfer to non trained strength tasks: A systematic review and meta analysis. Sports Medicine.
Popovich JM Jr, Tsai LC, Brito M, Xerogeanes JW, Lyle MA. Association of Loading Asymmetry During Squatting With Loading Asymmetry During Drop Jump After ACL Reconstruction: Implications for Rehabilitation Progression. Sports Health. 2026 Jan-Feb;18(1):145-153.
Schwanbeck, S., Chilibeck, P. D., & Binsted, G. (2009). A comparison of free weight squat to Smith machine squat using electromyography. Journal of Strength and Conditioning Research, 23(9), 2588–2591.
Swinton, P. A., Stewart, A. D., Agouris, I., Keogh, J. W. L., & Lloyd, R. (2011). A biomechanical analysis of straight and hexagonal barbell deadlifts using submaximal loads. Journal of Strength and Conditioning Research, 25(7), 2000–2009.
Takeshita, Y., Kiyama, R., Kawada, M., & Nakai, Y. (2025). Effect of hand support during squats on tibiofemoral and patellofemoral joint forces. Journal of Biomechanics, 189, 112814.
Staheli N, Cowley JC, Lawrence MM. Comparison of Olympic and Safety Squat Bar Barbells on Force, Velocity, and Rating of Perceived Exertion During Acute High-Intensity Back Squats in Recreationally Trained Men. Int J Exerc Sci. 2024 Aug 1;17(7):1120-1133
Prepared by Strong Angle Technology, LLC. The studies referenced are summarised in plain language; a copy of the full document is available on the UpSquat page.