[dwaltrip]

Machines keep the weights on a track. With free weights, your muscles have to maintain the "track" all by themselves. My understanding is that doing so gives a much more comprehensive and well-rounded work out, as all of the stabilizing muscles must be utilized. It also provides more useful strength, as they aren't any "tracks" in regular life when we want to use our muscles for some activity.

The point about injuries, especially for squatting, is definitely a good one though. Although I would want to ask an expert about it.

[yomly]

Your point on stabilising muscles is on point.

This is why pistol squats are worth more than just a body weight * x two-legged squat. This is why yoga is valuable for more than just flexibility.

Isolation of a muscle is all great and can make you look strong, but your body is actually a non-trivial network of muscles distributed across your entire body.

Unless your aim is to lift a big number or look big in a specific way, a gymnasts are a better proxy for balanced strength and they rarely do more than BWEs.

Personally, I do love a good old fashioned barbell squat though...

[VLM]

I do not think the trigonometry works that way if you resolve the vector forces. Surely if I chest press away from my chest, I'm not working the muscles that lift upward in a shoulder press I agree with you on that. That would be awful for the development of my muscles that push upward, but they are not sentient about not getting exercise and they get an excellent workout when I do the shoulder press perhaps ten minutes later, although the shoulder press doesn't work my chest press muscles at all, on a long term scale it doesn't matter. To get a "push upward" stabilizing muscle workout at the same time as I do a chest press, I'd have to lift sqrt(2) times the weight upward at a 45 degree angle, and this is getting kinda ridiculous compared to just hit the shoulder press and the chest press. And of course there are incline settings on the chest press machine that I use...

I'm trying to think of a machine I used today with a "track" where the track would be unavoidable via other exercise or bad. For the leg press I don't want my knee flexing sideways if at all possible but my hips and ankles are movable so kinetically in an engineering force diagram how would my knees "know" that my back was pushing against essentially a chair as opposed to pushing to stand up? If for the sake of argument I accepted that my knees have stabilizing muscles etc, you and I know via cognition and vision that my back was pushing a chair up a track, but from an engineering stress and strain diagram my knees aren't sentient and can't know the difference. The bicep machine and pec fly machine and leg extension machine have a ridiculous array of ball bearings and axles such that I'm only constrained in the circular force direction and the other axes are free over at least a small range, think of like how a car suspension has the wheels continue to turn regardless of the angle the suspension components are hanging at; I could design an inferior machine to do pec fly but my gym thankfully doesn't own one like that...

The row machine is interesting in that via the magic of cables its even less constrained than the oars on my little rowboat where I actually use rowing muscles IRL... The lat pulldown is similar in that if I were to climb a tree I'd be more constrained than I am in the machine...

The leg press is another example of the trig not "working" in that if your lifting form is any good there won't be more than a couple pounds of sideways vector force on your knees if lifting free weights, so putting low/zero side load on the knee in a leg press machine is the same force on the knee. If you run the math on the forces the stabilizing forces are very small with proper free weight form... stand on one leg and hold the other leg out 20 degrees the force on my knee should be mathematically larger than a dead lift freeweight or a leg press.

Furthermore there is infinite ink spilled on the topic of proper form for freeweight lifting, but real world muscle use rarely permits perfect freeweight form, simplifying it to the same problem.

Surely, if machines provided an inferior more dangerous higher risk of injury workout, then people would use free weights to recover from machine injuries, yet that never happens, so inductively I find it highly unlikely.

In regular life I don't want to lift heavy objects. I like the general health benefits and stress reduction and bone density of lifting, I'm not training specifically to deadlift car engine blocks instead of using an engine hoist like a normal person. There are probably people training specifically for vocational purposes where that wouldn't apply, soldiers perhaps.