There is some current discussion on foot position in the squat
following a post by Kelly Starrett, and I’ve been asked to comment. This article is not intended to criticize him or anyone else, nor is it intended to stand as irrefutable fact. Its purpose is to quickly organize my thoughts on the topic and answer the requests for my input; I'm sure I've left out a number of things I want to say. Use the available information to make your own decisions on training and coaching.
Arguments for Toes Forward
The following is a quick and very basic summary of the arguments for squatting with the toes forward as presented by Kelly Starrett (view the post and video here)
- Potential knee injury magnitude is reduced with a reduced valgus/rotational force.
- Squatting toes forward is motor learning to ensure this stance when jumping and landing; squatting with toes out teaches athletes to jump and land with the sub-optimal toes-out position.
- Landing with feet out means potential for valgus knee movement.
- Squatting with feet/knees out requires constant focus to maintain position—loss of focus means valgus knee movement.
- We need to prepare athletes in a way that limits the magnitude of potential injuries.
The first question to ask is why do we squat with the toes out? There are two basic possible answers. The first is that we rotate the feet outward to match the direction of the thighs to ensure the knee is hinging soundly rather than twisting. The second is that the individual has flexibility limitations (namely limited ankle dorsiflexion, and/or tight adductors and/or internal hip rotators) and is forced to excessively turn the toes out and roll onto the inside edges of the feet to circumvent normal ankle and hip movement when trying to achieve a deep squat position. Clearly the latter is not a legitimate reason to squat with the toes out, but this also has no bearing on the argument because it’s entirely unrelated—this is excessive and unintentional outward rotation.
While the hips have a large range of possible motion, there is a fairly narrow plane in which they can flex maximally. This plane varies slightly among individuals based on anatomical differences, but it’s always a position of some external rotation. This is easily demonstrated by squatting to full depth in a narrow stance with the feet straight forward—you’ll find it impossible to maintain an extended lumbar spine as the femurs hit the end range of motion against the pelvis. Now take the feet out a bit wider and turn them out a little and notice the feeling in your hips—the femurs will more freely move up in the hips. This can also be demonstrated in a seated position by pulling your knee straight up and back to your chest—you’ll feel the hip max out. Now from this position, bring the knee out to the side and you’ll find you’re able to pull it farther back without feeling the compression of the femur against the pelvis.
This is why we squat with the knees out to some degree (that degree, as mentioned previously, a bit different from athlete to athlete)—to allow the fullest hip range of motion possible, which means the greatest depth possible while maintaining proper spinal extension.
Which brings us back to the feet. As I stated above, we turn the feet out to match the direction of the thighs. That is, if the thigh is exiting the hip at about 25 degrees from the centerline, we turn the feet out about 25 degrees from the centerline. If the feet are also placed at the correct width, this means that the knee hinges as it’s intended rather than being twisted as it flexes. Below is a photo showing what happens to the knee when the feet are straight forward (or nearly so) and the stance is not narrow with the knees straight forward as well. Now we’re not talking about potential catastrophic valgus knee failure—we’re talking about repeated stretching of inelastic connective tissue surrounding the knee, leading to chronic knee pain, joint laxity and the resulting increased injury potential, as well as the potential with each squat of an acute injury. This photo is of a relatively wide stance, but you should be able to visualize a narrower stance.
Photo by Rob Macklem, best weightlifting photographer EVER
Clearly the above photo is a maximal-depth squat as performed by weightlifters. So what about non-weightlifters? The short answer is that the shallower the squat, the less relevant the toe position is. You can do a quarter squat with your toes turned in without hurting yourself, although this isn’t exactly a great idea. An athlete who is squatting only to horizontal or barely breaking horizontal, particularly with the hips-back, limited-knee flexion posture typically used with such squatting, will usually be able to do this with the toes forward because the flexion of the knees is minimal and consequently, the rotation never reaches the degree it would with complete knee flexion (as pictured above). If you insist on squatting with the toes forward, I would recommend keeping the depth limited.
Consider a powerlifting squat (or an archetypal one at least)—the feet are extremely wide, the toes possibly turned out less than many weightlifters’, and the depth to horizontal thighs at the greatest. The potential for a knee to drop inward abruptly and injuriously is a real possibility in such a stance because the feet are far outside the knees, meaning the joint is not directly supported by the lower leg. Further, with the toes and knees pointed straight forward, think of the lateral force on the knee (imagine what you would feel in your knees if doing the side splits with your feet elevated and toes straight forward). If you’re not very actively pushing out and engaging the lateral hip musculature, you’re in trouble. This wide stance and limited outward rotation of the feet also binds up the hip capsule and makes it easier to support the bottom position of the squat. Try a squat with an extremely wide stance and you’ll feel what I mean—you’ll find that you’re forced to stop around parallel because your hips bottom out.
I don’t teach a powerlifting squat for a number of reasons not requiring discussion here. My lifters squat with their feet under their thighs. Remember our discussion about aligning the feet and the thigh to ensure proper hinging of the knee—this only works if the feet are under the legs. If they’re outside or inside the leg significantly, even with the foot parallel with the thigh, the knee will be forced to rotate. I actually like my lifters’ stance to be very slightly outside directly under the legs—this allows the hips to sit slightly between the thighs to help absorb the force at the bottom of the squat somewhat rather than a more abrupt, jarring stop at the bottom as occurs if the thigh and lower leg are perfectly stacked. But this is an extremely minor deviation and creates insignificant rotation at the knee.
Without getting entangled in the interminable weightlifters vs. powerlifters argument, let’s consider the weights squatted by each, the positions of those squats and the gear involved. The world record powerlifting squat is over 1000 lbs (more or less depending on the federation). This is fully geared and to approximately parallel depth with a wide stance. Now consider some squats done by weightlifters. One that comes to mind was done by Pat Mendes—an 800-lb squat
to full depth with nothing other than neoprene knee sleeves. For those who will argue that it was just bouncing his ass off his calves, how about Mikael Koklyaev squatting 794 lbs
relatively slowly with a weightlifting singlet, belt and neoprene knee sleeves—and some impressive depth to boot. The point is simple: even though the absolute weights squatted by weightlifters are lower than the top weights lifted by fully geared powerlifters, weightlifters’ knees, hips and ankles are moving through much greater ranges of motion with nearly as much weight and with virtually no supportive gear, and as a consequence, it seems reasonable to state that no other athletes are putting those joints through as much work in the squat. Where are their feet? Nearly invariably pointed out.
In fact, let’s not even limit ourselves to weightlifters or deep squats. Take a look at Chuck Vogelphol squatting 1175
or Mike Miller squatting 1220
. Feet? Pointed out.
The knees going valgus during a squat has a couple possible causes. One is a squat stance wider than flexibility or anatomy can support—that is, the feet are outside the knees and the knees simply cannot be pushed out far enough to remain over the feet. Another is ankle inflexibility; this will also be accompanied by a wider than desirable stance along with excessive external rotation. In both of these cases, the valgus knee movement will occur during both the eccentric and concentric portions of the squat.
More common is valgus knee movement during only the recovery from the bottom of the squat. If the leg position is correct on the way down, flexibility and stance can’t be the problem; in this case, the issue is related to strength and activation of the lateral hip musculature. The athlete is either weak here or for some reason is not properly engaging the muscles necessary to maintain proper positioning and movement. If the athlete is strong and properly engages, there is no need for a particularly high degree of focus on preventing valgus knees; no more focus on this is required than on extending the knees to stand from a squat if the movement is learned and practiced correctly and weaknesses are addressed.
Finally, if the stance is of proper width, valgus knee movement, while not ideal and certainly offering a potential for strain in various locations, is by no means any particularly great opportunity for serious acute injury. The knees are still supported to a large degree by the lower leg vertically—they’re not unsupported and buckling in as they would with a squat stance too wide.
Spend some time on youtube watching the elite level weightlifters. You’ll notice the overwhelming majority squat with the toes out. Yet ACL tears and strains and other serious acute knee injuries are extremely uncommon among weightlifters, despite the ballistic nature of the majority of their squats.
In short, squatting with the toes out to a correct degree (i.e. not excessive rotation to accommodate ankle inflexibility) and with a proper stance width will not set an athlete up for sudden valgus knee failure, and will in fact help keep the joints moving through their intended planes of motion safely.
Jumping, Landing and Field Play
My knowledge and interest does not extend very far into athletic endeavors outside of competitive weightlifting; however, there are a few points I want to address with regard to this part of the argument.
If we agree that jumping and landing and similar sport-specific actions are best performed with the toes forward, this does not mean we have to agree that the toes-forward position is how we should be squatting. I won’t argue with this notion, because not only is this not an area in which I have extensive knowledge, I also have no objection to it.
But gym training does not and cannot perfectly mimic play on the field. If we had to perform exactly in the gym as we did on the field, we could do virtually nothing in the gym. Strength training and the like is intended to develop basic athletic qualities, not finer sport-specific motor patterns—the latter are learned, developed and practiced with sport-specific training on the field. If we’re going to say that because we want our football players defaulting to a toes-forward stance, we’re going to have them squat this way, that same rationale would need to be applied across the board, which would mean, for example, that we would need them to squat on the balls of their feet rather than with more pressure on their heels because no one sprints or cuts or jumps flat-footed.
To compare the squat to sprinting, jumping and landing is somewhat problematic because of both depth and stance. While landing from a jump may be done in a squat-width stance, both sprinting and jumping will nearly always be performed with a narrower stance. Further, these athletes are not absorbing the force of landing from a jump by squatting—knee and hip flexion is of a limited range of motion. Likewise, neither the support nor the drive of a sprint involves anywhere near the degree of knee and hip flexion seen in a squat. And finally, jumping is not performed from a deep squat position.
Additionally, most of these actions involve primarily single-leg support and drive—the positions of a unilateral leg movement are not identical to a bilateral one. For example, even though I teach my lifters to squat with the toes out, they don’t do single-leg work, such as lunging (or splitting in the jerk) with the toes out—but these are different activities that require and allow different positioning. The key to stable unilateral movement and support is strength and control at the hip and ankle—learning to engage the muscles that keep the leg where we want it and making them strong. Interestingly enough, the kind of lateral hip engagement and strength that prevents the knee from going valgus in a single leg support is very similar if not the same used to keep the knees out over externally rotated feet during a bilateral squat.
As far as motor learning goes, consider repetition volume. How many times in a given period is an athlete squatting with the toes out, and how many times in that same period (or over a career) is that athlete jumping, landing or planting a foot and driving in a sprint with the toes oriented as they should be? The volume of the latter eclipses the former by orders of magnitude, and as a consequence, there should be no concern about the motor patterns being altered undesirably.
Athletes need to be taught and practice proper jumping, landing, cutting and footwork mechanics in their sport-specific training, not in the weight room.
Training for Success
The primary goal for any coach or trainer is to prevent your athletes from getting injured. Injured athletes can’t play or train. This should always be kept in mind when deciding how to train an athlete. However, there’s a difference between training in a manner that prevents injury and one that plans for failure. I prefer the former. In other words, train wisely and safely for success rather than modifying your training to prepare you for impending doom.
Squat in the way you believe and demonstrate to be the safest and most effective for you, and teach your athletes to do the same.
This is strength and performance speak and how we train athletes. I am with your view and believe KStar's to be very clinical. It may even hold tru during unloaded squats but not during loaded squats. It is in my estimation that a volleyball player tearing the ACL from landing is either a fault of not teaching athletes proper landing or lack of strength and not because feet are pointed out. Good stuff Greg.
Your site is one of only a few that I check EVERY SINGLE DAY I log on the net.
Keep up the good work.
"This is also the reason why we teach the loaded squat the way great lifters squat. The test squat and the loaded squat are two different patterns needed for two different skills."
I would even underline the "for you" part.
I can honestly say I learnt to squat with toes out and i even had the correct range of motion until i stood up and then my knees would start tracking in a bit. In more recent times, that is 1.5 months ago, i tried a less of a toes out squat, similar to the Snatch above only with my feet under my thigh in correct alignment, and both my front squat position and Back squat position has bettered and the squat gone up 20kg and no knees tracking in. It works for me and it was something i tried, by chance, before these articles came out.
Don't dismiss what Kelly says nor what Greg says but listen to what greg said last and decide for yourself.
Notice I said MOST people, for as Mark H. pointed out, there are a few for whom a toes out position degrades the efficiency and safety to some extent due to anatomical reasons. So, in the end I would suggest only a small percentage of lifters to go toes straight, so long as the knee and hip didn't suffer at the expense of the position.
In the end I agree that sports specific applications and squatting are two separate things with their own inherent movement patterns. Trying to box the vast majority of people squatting in an unnatural position to compliment other sporting endeavors may not be the wisest thing to do, although the local PT's might like it!
Thanks again for your article.
I think what you said about taking anatomical differences into account is really really important. FWIW, my toes point out when I stand naturally.
In other words as you got wider you would have to rotate the feet out more... Any advice on what width is strongest given a good base of flexibility?
Try a variety of positions and I think this will make more sense. Yes you can change the width and angle of the feet to some degree and still have the legs stacked over the feet - but you have to factor in the rest of the posiiton/movement, not just the legs and feet in isolation. In terms of what is strongest, that will depend on the strengths of each athlete with regard to hip vs quad strength and the like.
The whole point to neutral foot and knees out A.K.A 'K-Starr's version', is that this creates larger external rotation force at the hip, creating a stable joint positon; resists valgus force at the knee and establishing a stable arch in the foot by locking up the ankle joint. The foot is designed to absorb force by collapsing, and transmit force by stiffening up, this is why the ankle joint locks in this way!
if you cannot squat this way, you are missing range somewhere and thus will need to address this IF you want to squat this way. If you don't, turn your feet out and do what is comfortable, just be aware this isn't a 'perfect-world' technical model, as described by Kelly, and your tendency to collapse into a valgus position may be slightly greater. You can still be a great athlete without adopting a 100% biomechanically optimal technique. Your willingness to adopt this 'optimal' position just depends on your capacity to adapt and grow as an athlete.
In regards to adopting these principles for Olympic lifting, it becomes more difficult to apply an optimal technical model to every athlete. Using one example, lower limb length is a huge deciding factor as to the foot placement within the set position, with many tall athletes adopting an excessively externally rotated hip position (duck feet) simply to clear the knees during first pull. This is far from 'optimal' but allows the athlete to complete the movement. Within that, the coach should seek to improve mobility and capacity to create torque where possible, but how far do you think a 6ft 5 basketball player will have to drive his knees out if using a neutral foot postion!? Probably not going to happen.
Just my thoughts..
Thank you for your follow up comment. I've never understood the reasoning behind KStar's recommendation until now. Now I understand because of your sound structural explanation. I will reassert my efforts to mobilize and achieve optimal positioning; given that I am quite short, I do not think it will be a problem for me in lifts, etc.
Best and thanks again,
So you're creating a "stable joint position" at the hip by sacrificing the mechanics of the knee? I don't understand why the concern for torque at the hip (which is arguably not that important if you consider the combined enormous squat weights and the common foot position of world class competitive weightlifters) eclipses the concern for functioning knees?
Outside of instances in geared competitive powerlifting, how many of these catastrophic valgus knee explosions are occuring? Even IN geared competitive powerlifting? That injury is a product of a wide stance and weights that are only possible with supportive gear (and if we're being honest, other "gear"), not of the angle of the feet. Using any problems of such a squat (wide-stance competitive PL) as supporting evidence regarding how weightlifters, but especially recreational lifters of fitness enthusiasts, should squat makes absolutely no sense.
There is an enormous difference between possessing the mobility to be able to sit into a squat with your feet forward (which I agree is something that athletes should be capable of doing) and actually training the squat with that stance. And there is an enormous difference between a real squat (i.e. a full-depth Olympic style squat) and a partial depth squat used by another athlete for the sake of developing basic strength (e.g. a parallel or even just below parallel squat) - the latter allows more lattitude with the angle of the feet because of the greatly limited degree of knee flexion occuring, particularly when you consider that those using such a squat are also typically keeping the shins and hips back significantly, further limiting knee flexion.
Calling a toes-forward squat position "100% biomechanically optimal" is, I would say, irresponsible at best without a large amount of qualification and education.
If it feels better, do it.
I wouldn't be worried about stability - if that position is strong and comfortable for you, it's the one to use. I would just be cautious of pushing your knees significantly outside your feet. If I'm understanding your description correctly, I would turn the toes out a bit more with the heels in the same position to keep that same width but get the knees into better alignment.
Thanks for not shying away from addressing these topics and debates. This was helpful and will help.