As coaches we all understand the importance of evidence based, research guided practice. If the preponderance of research indicates a particular training methodology, recovery method, nutrition strategy etc. is effective then it makes sense for us coaches to adopt these elements into our practice where appropriate.
But what do we do when we have questions that aren’t currently being answered by the available literature? And how transferable is much of the evidence in research to the everyday training and performance of our athletes?
While I’m not suggesting we abandon our evidence-based practice I do feel that we too often take conclusions from research and apply them wholesale to our practice without questioning what other conclusions could be derived from study results, and what way any conclusions we make integrate with other elements of our programming and into our wider philosophy as coaches.
As an example, I’ve long had a preference for cold water immersion (CWI) as a means of recovery from training for no reason other than I enjoy it and found it beneficial myself. So, I’ve always advocated it’s use with my own athletes. Then, several years ago research began to emerge showing that CWI improved recovery by blunting the adaptive process. In other words, we recover quicker but we make less gains from the training session we are trying to recover from.
As a young coach my inclination was to abandon CWI and stop recommending it to the people I work with and I’m pretty sure that’s exactly what I did. Then I heard another coach mention how he advocated the use of CWI in-season when maximising performance was the priority and encouraged limited to no use of CWI in the pre-season period when strength and hypertrophy adaptations where a bigger priority. This made total sense to me, and I began to give the same advice to the players and athletes I worked with.
In the preceding years I’ve continued to read studies and perform further research and I’ve worked with many more teams and athletes. What I’ve concluded for myself is that there is no one way to do anything. Brian Cody and Jim Gavin have vastly different approaches to preparing teams, yet both have had unimaginable success. The keys to success in sport are not black and white, increasing human performance is achieved through discerning between the many shades of grey. For instance, is it actually best practice to avoid CWI during the pre-season? If a player really believes in its efficacy and has a positive history of using it around some exceptional performances, is it good to create a possibly negative association with CWI in her mind? What about with highly competitive squads? I’ve been involved with several inter-county squads over the last 7 years and competition among places within a squad is probably the standout measure for later success of the team. Every pitch session is a chance to perform and stake your place. Obviously, we want to overload players during the pre-season and our job as S&C coaches is to prepare them so they’re in the best physical condition for the business end of the season. But are we then doing a disservice to the players as individuals and the team as a whole if we don’t give them the opportunity to recover as quickly as possible from all of that overload and allow them to compete well at pitch sessions? Also, is there a compounding interest achieved from completing more sessions at a greater intensity and higher quality that the research can’t demonstrate? For the record, I don’t have answers to these questions. These are just some of the things I think about and to my knowledge there is no long-term study looking at CWI in field sport players and how it affects their physical and technical & tactical development.
Being guided by research is important. Being curious, questioning your own convictions and going through periods of inquisition is equally important. These great articles from John Kiely ((PDF) Periodization, Planning, Prediction: And why the future ain't what it used to be! (researchgate.net), (PDF) Periodization Paradigms in the 21st Century: Evidence-Led or Tradition-Driven? (researchgate.net) were hugely influential on my own philosophy and developing a more critical and sceptical mind. Again, I’m not suggesting that we ignore current research and literature, rather that we constantly question our own conclusions and thought processes and that we try to find answers for ourselves rather than making assumptions. To that end I want to advocate for the process of self-experimentation and performing case studies on ourselves as coaches, and, where appropriate, with the people we work with.
One of the key issues with academic research is that negative result studies are rarely if ever published (Why it's time to publish research “failures” (elsevier.com)). I’m sure many of us coaches often have the same questions as each other. Yet we can’t rule out most of the possible answers we have to our questions because the research isn’t there to support that. For instance, one of the questions I’ve investigated for myself is “Does a trap bar deadlift 1RM influence the load at which a person’s peak power will occur on trap bar non-countermovement jump”. If the answer is “Yes” then that would make prescribing loads for Trap Bar jumps much easier. I investigated this about two years ago. We tested more than 15 male club GAA players on consecutively heavier trap bar NCMJs using a push device, we estimated their 1RM based off a max 3RM effort and we looked for trends. We found none! Maybe some of the players had a low training age, maybe the push band isn’t accurate enough, maybe our testing methods weren’t specific enough and we allowed for too much noise. Another possibility is that we didn’t P Hack the results (What is p-Hacking and how you should avoid it? (edupristine.com)) or extrapolate real world meaning from results that were merely statistically significant (Statistical significance—meaningful or not | Law, Probability and Risk | Oxford Academic (oup.com)). What I do know is that there certainly are other coaches out there who tried to do a similar experiment. Maybe some of them did it with a view of publishing it only to find there were no trends or positive results. If we’d have known of other experiments like that, we could have perhaps performed ours with a different set of methods or maybe even not bothered performing it at all, satisfied with the available info that it wasn’t worth investigating.
To that end I want to share a small case study I performed on myself recently. A question I’ve routinely asked myself is “Is there a particular explosive exercise I use with my players that is better for power development than the other explosive exercises I use?”. More specifically, I wanted to know if any one of the power clean, power snatch, barbell squat jump and trap bar jump is better than the others for training maximal power development. I based my little case study off the assumption (not very clinical, I know) that training at or near maximal power output is the best way to develop maximal power. Who knows if that’s even the case, again there’s no research that I can find clearly demonstrating this, but we’ll roll with it anyway!
So before we start I should do a quick round up on some of the research that is available but doesn’t directly answer my question, there’s obviously a lot more studies that I could mention here but these few are just a sample of some of the issues we can have when trying to apply research to practice. This study (https://pubmed.ncbi.nlm.nih.gov/8289617/) looked at heavy squat vs depth jumps vs loaded jumps at the load that produced peak power and how each exercise effected performance on a variety of tests including 30m sprint, CMJ and SJ. It showed very favourably for the loaded jumps. Good info to start with but it didn’t look at differences in loaded explosive exercises (which is what I want) and the study notes the participants could “perform a half squat exercise with a load greater than bodyweight”. So possibly not a very strong cohort which could explain why loaded jumps improved performance in the various tests to a greater degree than plyometrics. The study was published in 1993 so it was undoubtedly very valuable info back then, but not enough to guide us further today.
This excellent paper (https://www.researchgate.net/publication/223128818_Developing_Maximal_Neuromuscular_Power_Part_2_Training_Considerations_for_Improving_Maximal_Power_Production) answers lots of questions. I’d highly recommend reading this one and reading part one of the same paper. It discusses the various methods we can use to improve muscular power as well as how optimal load (i.e., the load that peak power occurs) is very effective for improving power output. But it also mentions that the available literature is based on short term studies (8-12 weeks) on under trained subjects. Also, as I mentioned earlier, increased power production or rate of force development (RFD) on a gym-based exercise is no guarantee of increased power or RFD in pitch/sport based movements. This study (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6724584/) compared barbell squat jumps, trap bar jumps, and barbell shrug jumps and their associated peak power outputs at various loads relative to the bodyweight of the lifter. Getting much closer here! The study demonstrated well that the trap bar jump and the barbell jump shrug elicited higher peak power numbers than the barbell squat jump. One major issue I have here is that I never programme jump shrugs. The technique for the jump shrug as described in the study is a long way away from how we’d coach a shrug pull or a power clean. Additionally, each lift was prescribed at 0, 20, 40, 60 or 80 percent of bodyweight. Very useful and interesting information but not what I’m currently grappling with. A person’s ability on each of those exercises would differ based on their physical architecture and their own individual training experience. So, it would be more interesting and useful for me to see the output of each exercise at loads based on either a 1RM for the given movement or perhaps even on a rate of perceived exertion for each exercise, i.e., the participant would work up in 10kg increments until they felt they were at an RPE of 9 over x reps at y weight. Lastly, while I like the fact that they used a countermovement on the exercises as that’s a little more sport specific in my opinion, I’m sure it adds more noise. Did they descend deeper or give themselves a longer eccentric action across the sets? Did they perform the countermovement at the same depth and speed on the barbell squat jump as they did on the trap bar jump? Lastly, the hip dominant, posteriorly shifting type movement used on the jump shrug would surely be much different to a knee dominant, quad heavy movement they’d have presumably used on the barbell squat jump and trap bar jump? The hips muscles are powerful extensors and would definitely contribute to a high-powered jump. But on the field of play most jumps are ankle and knee dominant and I’d have preferred to see the three movements performed as closely to that type of pattern as possible.
As for my n=1 self-experiment; I’ll discuss it in the following format:
· I’ll describe the methods/procedures I used for the various tests.
· I’ll provide the results.
· I’ll discuss why I think the tests resulted in the way they did.
· I’ll go through all the things I would change if/when I do the testing again.
· In true sport science research style, I’ll wrap up by making some very inconclusive conclusions.
Methods/Procedures:
· As mentioned, I chose to measure my outputs on 4 different exercises – a Barbell Squat Jump (SJ), a Trap Bar Jump (TJ), a Block Power Clean (PC), and a Block Power Snatch (PS).
· I performed sets of 3 reps at each load with 30s between each rep and 120s between each new set & load.
· I moved up in load in 10kg increments each new set from 30kg up to whatever load I considered to be a 9RPE. Through my own training, I had an idea what my top set would be on some of the exercises already.
· I measured each rep using the new Flex Stronger device from Gym Aware. Not research grade but this isn’t a peer reviewed article! However, having used the band device from Push extensively for a few years I feel that the Flex Stronger device is much more consistent. It uses a laser field instead of an accelerometer to measure bar velocity. Additionally, the number of exercises you can measure using the Flex Stronger is much more limited when compared to Push, which I consider a good thing for reliability and validity. You can check this study out if you’re interested in ensuring the validity of the Flex Stronger device (https://www.researchgate.net/publication/340529748_Criterion_Validity_and_Interunit_and_Between Day_Reliability_of_the_FLEX_for_Measuring_Barbell_Velocity_During_Commonly_Used_Resistance_Training_Exercises)
· I took 2-3 days between each session, and I did no other training during this period other than a very small amount of accessory work after some of the jumps.
· I performed all the exercises with no countermovement from a controlled start. The TJ was initiated from the ground, PC & PS were completed from blocks and the SJ was initiated from a seated position on a box. You can see videos of some reps from each of the exercises further below.
The results:
· First a quick discussion on this – I didn’t know whether it was best to take the best rep of each set or the average of the three singles/rest pause reps of each set. This study (https://www.jsams.org/article/S1440-2440(16)30154-2/fulltext) showed that the average of three jumps was more valid than the highest jump when looking at a simple CMJ. I’m not sure if that would transfer across various types of loaded jumps/weightlifting type movements but my guess is it would. In any case, I tend to programme these movements for multiple reps per session whether that be time-based singles and doubles, standard sets of 2-5 reps or rest pause reps like I did in this case study. So, the average of each set worked better for my purposes.
Below you can see the average of the peak velocity and peak power for all 3 reps of the 4 tests with the best set in bold:
As you can see, the three reps I completed at 100kg on the PC had the highest average peak power at 2687w. Next was the PS with 2445w at 80kg followed by the 90kg TJ at 2248w. The SJ came in last with 2014w at 70kg.
Some discussion of the results:
· I trained in Capital Strength under the tutelage of Harry Leech, Weightlifting Ireland’s most esteemed coach, for several years. I’ve consistently included variations of snatches and cleans in most of my own programme for the last 6+ years. So, whilst I wouldn’t consider myself a good lifter by the standards of the Weightlifting Ireland, I would say I’m better at those two particular lifts (relative to my other more traditional lifts like squats, deadlifts etc.) than 80%-90% of non-weightlifting, non-crossfit athletes. In other words, my results might not hold true for most athletes that solely use a power clean or power snatch to improve their physical output in another sport.
· I rarely use the trap-bar in my own programming. But I use it all the time in programmes for my athletes. They may perform much better than I did on the TJ relative to their other lifts.
· I’m much more comfortable back squatting than I am trap bar deadlifting. That said, I’m not surprised that overall, the TJ outperformed the SJ. Jumping and landing with the bar on my back felt much more difficult than jumping and landing inside the trap-bar with the weight in my hands. Not very surprising given it’s a more mechanically advantageous position and if in danger I could always drop the trap-bar mid rep, not so with a heavy bar across my back!
· I was surprised the PS performed so well. I’ve never had an issue with power cleans, they feel very similar to a full clean for me and I’ll often use them if I want to limit my overall squatting volume. However, power snatches have never been an exercise I particularly enjoyed. The timing and execution of a power snatch feels totally different to me in comparison to a full snatch and I love full snatches, so I nearly always stick with them. But it was nice to see an 80kg PS had almost as high an output as the 100kg PC. The PS also had by far the fastest velocities of any of the exercises, which makes perfect sense because the bar has travel significantly further in the PS than the other three exercises. Something for me to consider if I want to limit total volume-load or focus more on what you might call a speed-strength emphasis.
· While the TJ performed quite well, it felt very, very heavy. Uncomfortably heavy to jump with. I’d certainly try to drop it upon landing if I was using it at such heavy loads for training purposes but, to be honest, I’m not sure I’d ever prescribe it at such heavy loads. Not very scientific or based on fact, but it felt immanently more dangerous doing a 100kg trap-bar jump compared to power cleaning 100kg. Maybe that’s just me.
https://youtu.be/W72FzOvhvE8 - Examples of the different exercises used
There are some things I would change if/when I redo the test:
Firstly, my set up for each lift would change.
· I’d definitely perform the SJs from a higher box next time before I totally rule them out. I feel like the depth of the box made the movement more hip-dominant than would be natural for me in a regular jump. You can see a clip of the SJ in the video below to see what I mean.
· Likewise, I would perform the TJ from blocks. While I felt more knee dominant in this position the amount of knee bend I had felt very unnatural to me. If I was performing an unloaded counter movement jump (which I think is the joint angles I should be replicating) I wouldn’t descend so low and enter such high knee flexion so higher blocks would be better.
· I was pretty happy with the start position of PC & PS. If I was being very nit-picky, I’d start the PC from a slightly higher position, maybe 3-5cm higher. That way I’d be well clear of my knees and starting from a position that’s truly in the second pull portion of the lift. But most people wouldn’t even have an issue with that.
After that if I was to test it all a third time, I’d be interested to see what difference a countermovement makes. While there’s huge benefits to training starting strength or the ability to overcome inertia, I’m curious to see what making the two jumps more natural and relevant to the movement patterns we see in sport would do to the results. Whether I’d compare them to power cleans and power snatches from the blocks, the hang or the hang with a countermovement; I’m not sure yet.
Other than that, I think I’d keep the sets, reps, rest, build up etc. the same. That all seemed to work well.
So, what are my conclusions?!
· Well, as is typical, I’m left with more questions than answers. I’m pretty sure that in comparison to the TJ, the SJ would underperform for almost everyone, not just me. So going forward I may look to only using the trap-bar for heavy loaded jumps. It’s safer anyway, there’s no need for a power rack to do it and there’s no issue with an athlete using a trap bar for their jumps and still using a back squat as their main lift if that’s their preference. So, whereas I’ve always previously matched the loaded jump to the main lift of the athlete (i.e trap bar jumps with trap bar deadlifts and barbell jumps with barbell squats) I may only programme trap bar jumps if the results from my self-experiment continue to hold true on others.
· If I’m honest, I wouldn’t have been surprised if the TJ outperformed the PC. The fact the PC outperformed the TJ across all sets surprised me. As I mentioned, the TJ didn’t feel very good at higher loads. Additionally, for all the criticism they get regarding their perceived complexity and high learning curve, power cleans aren’t actually that complicated or hard to teach to a moderately trained athlete. As long as the coach is proficient at the exercise themselves and, probably more importantly, is good at coaching in general, there should be no issues.
· The PS was probably the most surprising result for me. It performed well against the TJ and outperformed it at heavier relative loads/higher RPEs. So, it could be a good option when looking to emphasise speed whilst keeping the total power output high.
· Aside from the argument that the mobility, kinetic linking, and coordination & timing needed to successfully perform the olympic lifts and their derivatives make them a worthwhile endeavour in the first place, one of the main advantages they have over loaded jumps is their higher rate of force development (RFD). Because the lifts consist of a controlled accelerated pull up to the power position followed by a very quick and explosive extension to propel the bar upwards, the time the athlete has to impart full force into the bar is very limited. Therefore, more power is created in less time when compared to the loaded jump variations which allow for much more time to accelerate the load as much as possible. This has very important implications for sport performance. Even if the TJ outperformed the PC in a marginal way. One could still make the argument that the PC is more beneficial to sport performance as the RFD is much higher.
Beyond the straightforward conclusions I’ve made above there are bigger questions that we should be asking about this experiment and any other research we come across whether peer reviewed or not. How does any of this crossover to the sports we’re training our athletes for? I’m a big believer in the position that a more generally athletic player makes a better player in any sport. But how general is too general and how specific is too specific when it comes to training? We have a very finite amount of time with our athletes. They have a finite amount of adaptive resources and energy. What are the things we should focus on and what are the things we might be wasting time on? If a power clean is better than a trap bar jump that’s cool but is either as good as an extra speed session or various types of unloaded jumps i.e., the actions they do in the game? Can I cover the entire force velocity curve by ticking off both ends through lifting heavy and jumping and running fast? Are Olympic lifts and loaded jumps more of a fast strength exercise that are best used in place of traditional strength exercises when reducing fatigue is paramount? These are just some of the thoughts I constantly rack my brain with.
The meat and potatoes of the self-experiment are of definite importance. I’d love to hear from any coaches who have done something similar or who are now planning to do something similar after this article.
However, what’s most important for me is that we as coaches adopt an attitude of curiosity. We’re open to learning and we maintain a white-belt mentality. We’re sceptical. Sceptical about what others put forward as facts or knowledge but even more importantly, sceptical about what we consider our truth. We look for answers and when we don’t find them, we try to create them for ourselves. It’s a process of self-doubt, searching, discovery and then even more self-doubt! But it’s fun and it makes us better!
Thank you to Conor for this really interesting blog which really delves into the issues surrounding exercise selection for optimal power development and the challenges that exist when interpreting the research.
Conor Clifford is a Strength & Conditioning coach based in Dublin. He has his own coaching company called Armoured performance which caters for athletes of all levels. You can find him Instagram as @armouredperfromance. Conor is an experienced coach who has worked with a wide range of individuals and teams including Shamrock Rovers, Dublin U21 Footballers and most recently the Dublin Senior hurling team.