Machines vs free weights for strength and power: Which is better?

If you spent any time on fitness training social media around July of 2023, you probably came across the growing chatter about a specific study that looked into the role of machines in athletic training.

Many coaches online proclaimed this study as a big win for machines, starting the move towards kicking the stigma surrounding machine-based training for athletic development. While I'm all for more inclusive training equipment utility — machines are still unfairly considered toxic by many in S&C — I just don't quite think the analysis on this specific study has been accurate.

Let’s approach this research from a practical perspective and dive into the nuance.

Before you go raging in the comments…

Any time I explore a potentially specific and potentially contentious topic only, I like to lay out my overall position:

• Machines and barbells are both fantastic tools. As a strength coach and in my own training, I utilise both extensively.
• I strongly agree with the idea that machines should be de-stigmatised in general training. However, they aren't the miraculous solution to all training woes, and may still be limited in athletic development contexts where sports performance and injury mitigation are both key outcomes.
• An effective training program shouldn't rely exclusively on one modality or equipment type. You would expect a carpenter renovating your kitchen to show up with a whole range of tools for the job, coaching is no different.
• Conducting research is complex, as is creating a well-structured training program.

Free-weight versus machine-based exercises: Challenging a traditional assumption

Methods and program design

The study in question (full paper linked here) took 34 fit male subjects and put them through an eight-week strength training program. All subjects used the same four exercises; squat, bench press, prone row, and shoulder press with half using barbells, and the other half using machines. The researchers assessed the subjects' athletic performance through various tests, including jumps, sprints, changes of direction and balance.

They trained with a periodised linear program guided by velocity targets and velocity loss thresholds for autoregulation.

Results

Both groups improved on virtually all tests, as is to be expected with 8-weeks of structured velocity-based training.

However, the surprising finding that has everyone interested is that there were no statistically significant differences in the performance gains made between the two groups. This led the authors to conclude that machines are just as effective as free weights for strength and conditioning (S&C).

Non-significant differences

The changes in performance between groups did not reach statical significance.

Put another way, both training groups improved, by amounts so similar that we can’t confidently determine which training method was better.

However statistical rigour is vital in science, so this study end up a draw. Machines -1, free weights -1.

This point invites caution. Unlike in football where a nil-all draw favours the underdog, in science, a nil-all draw is a call for a rematch. It urges us to delve deeper, reevaluate our methods, and try again.

If we were to ignore statistical significance for a moment and look at the average improvements across the tests.

While not statistically significant, the free weights group did show a greater degree of improvement than the machines group for every test. Not evidence by any means but for me as a coach working in the real-world with lots of decisions to make and lots of variables at play this finding and the conclusion that machines deliver results at parity with free-weights is met with a lot of skepticism.

Dissecting the study

This is just one one study, and it's not without its flaws. here are some of the limitations I can see from a coaches perspective that contribute to my hesitation in agreeing with the authors conclusions:

Short time frame

The study only lasted 8-weeks, a really short amount of time in training.

Would machines lead to an earlier plateau in strength or athletic gains?

Would the athletes get bored on the machines?

Would the coaching time used on free weight movements need to be factored into the study?

Would a third group that did a blend of equipment actually do the best?

Impacts on injury risk

Short studies are impossibly limited in the ability to determine the flow on effects for injury rates, both in training and competition. Even long studies are limited as injuries are such multi-factorial events.

What are the impacts on injury rates over 6, 12, 24 months?

Do athletes who lift only with machines have more or less joint wear and tear over years of training?

Does training with free weights lead to more or less gym injuries? Which modality puts more stress on the back?

These might be impossible to answer even with 1000s or athletes tracked for years and years.

Program was not reflective of real-world athletic training

The exercise & test selection was incoherent.

The athletes only used four exercises, three of which were upper body-focused: bench press, row, and shoulder press. Only the back squat/hack squat was included for lower body training. Despite this the speed and power testing was focused on lower body only.

Would the inclusion of unilateral exercises, hinge movements or an explosive movement have changed the outcomes direction?

Improving future studies

To better inform our understanding of machine versus free weights, we need more comprehensive and practical studies.

Here are some ideas of what factors could be included in a follow up study on this topic:

• Include upper body tests, such as a throw or punch test.
• Employ a full-body training program that includes a diverse range of movements like a hinge, split squat, lateral lunge, and loaded jump.
• Enhance real-world applicability by including a plyometric or movement skill element to the program for both groups.
• Include a blended-group that does some machine based and some free-weight training.

Sports science research is just part of the evidence base

The study highlighted here and on social media is a great starting point for more conversation about the role of varying equipment in training.

However, we still need more research to fully understand the different utilities and potential best practices for athletic transfer. Until then, let's continue to use all the tools at our disposal, including both machines and free weights, to help our athletes reach their highest potential.

Remember, the right tool is often determined by the job at hand.