Whole Training

Dr. Steve Bain, a longtime friend of GMS and quite a smart guy in his own right, has written an article for publication in Volleyball Magazine. His summary of the article:

“On behalf of my Co-Author, Dr. Carl McGown, please find attached our manuscript titled, “Motor Learning Principles and the Superiority of Whole Training in Volleyball”. This manuscript is being submitted to Coaching Volleyball Magazine as a response to “Whole vs. Part Training”, which was published in the 2010 April/May issue. The primary aim of this submission is to offer the volleyball coaching community an educational and evidence-based rebuttal of the opinions presented in the Weitl article. Fulfilling this aim requires extensive documentation and substantial references to the motor learning literature and we are hopeful that you will find this effort an important and timely contribution to the journal. The manuscript submission includes an image of the, “Facts Not Opinions” inscription, which is etched in the stone facade over the doorway to Kirkaldy’s museum in London. Please consider using this figure as an introduction to our manuscript.”

The entire document can be downloaded as a PDF here.

11 comments on Whole Training

  1. Anonymous says:

    >Excellent article and much appreciated. My question is why Dr. Bain chose to only view the problem from the scope of motor learning. I believe that also considering the structural (and not just neurological/biomotor) adaptations involved in sport skill acquisition.

    To use the volleyball spike as an example, obviously the primary influences on an athlete's ability to perform the skill are spatial awareness, timing, when and how much force to apply, etc., which fall under the guidelines described by Dr. Bain. However, I do believe that a significant determinant in an athlete's spiking ability is how much force they can apply to the ball. Now, much of that is influenced by the same factors (firing muscles with proper timing will result in faster torso rotation, resulting in a faster armswing; contacting the ball at the proper point on the hand results in more force, etc.), however, an athlete's physical structure also plays a large role in force production.

    To use an example, all things being an equal a 200lb man will hit a ball harder than a 90lb girl.

    Additionally, I'm sure many high school coaches have had a tennis player come out to try volleyball and find that they can overhand serve the first time they try, and with more power than other girls who have played volleyball for a year or two in middle school.

    The primary influence in causing structural adaptation such as muscle growth is simply the training volume applied over time.

    So, again using the example of a volleyball spike, it is obviously easier to apply a higher training volume using part and blocked training. I think it is difficult to argue that, after a year of practice, the athlete who hits 1,500 balls per week using part, blocked methods, will have stronger, more developed musculature in the shoulder girdle than an athlete who hits 300 balls per week using whole, random methods. I think it is also difficult to argue that the second athlete will show an increased mastery of the skills in game situations.

    My question now relates to the principles of diminishing returns. I'll assume one can hit 5x as many balls in a part, block setting- not unreasonable I think. Obviously, there will be a huge difference in skill acquisition between somebody who hits 0 balls per week and 100 balls per week (using whole/random). There will be less of a difference between somebody who hits 100 and 200 and less still between 200 and 300 and so on and so forth. So my question then becomes, might there be some intersection of whole/random and part/block training such that skill acquisition is "sufficient" and training load is higher to elicit a stronger anatomical adaptation. Meaning, which athlete is likely to be the best:

    (1) 300 balls per week whole/random?
    (2) 250 whole/random and 250 part/blocked?
    (3) 200 whole/random and 500 part/blocked?
    (4) 100 and 1000?
    (5) 0 and 1500?

    Quite obviously it is not (5). And (1) is likely to be the "most skilled." Yet this raises another fundamental question:

    Is the most skilled athlete necessarily the most effective?

  2. Anonymous says:

    >Whether you watch Lebron James dominate the NBA at 20 years old or the big girl who swims and plays softball run off 15 serves in a 6th-grade game, it is obvious that any sport this side of chess involves some intersection of skill and strength.

    However, the (whole/random) methods that science (and experience) has revealed to be best for developing skill do not, in all cases, mesh with what science (and experience) has revealed the be best for developing strength. Generally speaking, skill is more important to sports performance than strength.

    I also would contend that there exists "tipping points" of sorts where a little increase in strength can go a long way. The weak girl with excellent technique whose serve struggles to reach the net will never be an effective server until she gets stronger. In this case, I would argue that part/blocked (or at least blocked) training would be superior until her serves are consistently clearing the net.

    I don't pretend to have any answers, but I hope that I'm asking some of the right questions and can continue to learn. I would love to see Doctors Bain or McGown address their considerations on structural adaptation/specific strength development and its interaction with motor learning.

    Joseph Trinsey

  3. CJM says:

    >Joseph, please email me at info AT goldmedalsquared DOT com and I'll get you Steve's and Carl's email addresses. You can continue the discusssion with them.


  4. Anonymous says:

    >Reading this article makes me think two things:

    1. The science behind whole vs part practice, and more specifically random vs blocked practice, lends itself very well to the questions of practice organization and drill design. Contextual interference is a key component to the discussion as it promotes longer, more permanent learning for athletes. However there is a specific application to serve receive training I think that gets easily overlooked.

    We need contextual interference to promote learning and both random and whole practice to promote learning that actually transfers to match day performance. My recommendation is this: For every serve receive drill you run, require that every 5 serves the passers switch between forearm and overhead passing, without exception and regardless of how much the initial performance visibly drops. And of course this is all off a live serve adjusted for ability at the discretion of the coach.

    2. Since random practice is necessary for transfer to actual match performance, athletes must also understand that initial practice performance will be degraded under random conditions. However, given that permanent enhancement of performance is a slow process and from the athlete's perspective it will appear nothing is being gained from random practice, goal setting for practice needs to be modified to match the realities of the situation. During the first couple practices of a new season, take a few simple stats to develop an average of what the actual performance of the athletes is, and then slowly increase requirements rather than setting some arbitrary number out there and making demands to meet it. The realities of practice will be better matched with the realities of the athletes perspectives so they will enjoy each training session more while also becoming better accustomed to the pressure of increasing demands.

    I'll stop for now as I could write posts twice as long as this for each individual bullet point.


  5. Tom says:

    >The neuronal explanation for these effects are perhaps best exemplified by our own observations
    (Bain and McGown), of inexperienced coaches training novice players where the instructor(s)
    become frustrated by the performance variability and lack of successful repetitions of new
    learners. As a consequence, these inexperienced coaches limit or abandon whole teaching
    methods for part, and random practice for blocked

    This is the deal. As long as coaches can "SEE" the armswing improve on the box, we'll keep having these debates.

  6. CJM says:

    >Dear Joe (Part I)

    Before I respond to your blog post I want to frame it in the same way that I provide feedback to our players when we are correcting them in practice; i.e. "This is information, not criticism". Hopefully, you will receive what I am about to write in that spirit. Knowing Carl as I do, I am sure that was his intent as well when he responded.

    With that brief preamble, your response to the "whole training" article that Carl and I wrote, and the comments/questions you posed, underscore the conundrum that widely held beliefs on motor skill development and strength training, which are cited as unequivocal "truths", have little or no factual basis or support in the scientific literature. Even so, such beliefs are held up as fact and are deeply embedded in coaching culture and training methods at all levels of the game (and unfortunately, in most other ball sports as well). That this is so is exemplified by the editors of "Coaching Volleyball", who were happy to publish our article as an important contribution to the "debate". As a scientist, I find it rather ironic that in the whole vs. part debate, opinions seem to carry as much weight as fact.

    In any event, there are many excellent articles that serve to address the issues you raise in your post and I have attached three of them to this email.

    With respect to your beliefs about strength training, please see the review article by Witten's group and that of Jensen. Witten's review clearly debunks what we "know" about strength training, substantiating that "less is more" (or is as effective as more) and the same training influences that optimize motor skill learning (random practice, contextual interference, etc.) have the same potency in strength training. This is an extremely important idea in physiology and has influenced my own work in musculoskeletal adaptation where we have shown that the skeleton essentially ignores stimuli that are not novel and responds more profoundly to exercise when rest periods are inserted at regular intervals. The molecular and cellular mechanisms for this have been revealed but go beyond the scope of this response.

    The paper by Jensen essentially explains why motor learning of a specific skill is inseparable from the force variables that control its execution. The Adkins' paper (Jeffrey Kleim's lab) also has much to say on this topic but the important point is, "corticospinal system is not only plastic but that the nature and locus of this plasticity is dictated by the specifics of the motor experience. Skill training induces synaptogenesis, synaptic potentiation, and reorganization of movement representations within motor cortex. Endurance training induces angiogenesis in motor cortex, but it does not alter motor map organization or synapse number. Strength training alters spinal motoneuron excitability and induces synaptogenesis within spinal cord, but it does not alter motor map organization. All three training experiences induce changes in spinal re?exes that are dependent on the speci?c behavioral demands of the task." (Please note how important specificity is to all parameters of the motor experience ~ this is crucial)

    It follows from the above that the structural adaptations are tightly coupled to the movement patterns and force variables engendered by neuromuscular activity. Again, the scientific literature is full of evidence demonstrating this relation but if we simply imagine how a child (or a young animal) learns and adapts to its surroundings it is immediately obvious why this must be so. Ecologically, we are reared in settings boundlessly rich in visuomotor content and environmental variability. Said another way, our neuromuscular systems are genetically tuned to adapt to random stimuli not "block training".

  7. CJM says:

    >Dear Joe (Part II),

    When these factors are considered In the context of the volleyball spike (which you posed as an example), the force variables associated with the dynamics of skill execution are inseparable. As the execution of this skill in competition is random (i.e. no two spikes are identical), the force variables required to execute the skill can ONLY be optimized by random practice, which, over time, will produce the maximal amount of optimal solutions for expert task execution. Again, look at the work of Jensen and Kleim (and others), and you will realize why there is essentially no transfer of block training to the acquisition of skilled performance. It simply doesn't train the right neuronal systems. This is also why that unless the 200 lb male and the 90 lb female are both expert volleyball players, there is no way to predict which one will have a more powerful spike. Similarly, the tennis serving example you pose is a "well known fact" but it is simply not true. ALL of the research has demonstrated that skills are specific to the task and while the tennis player might be a good server there is no scientific rationale to predict that this will be so. I have to admit at this juncture that I have erroneously applied some of these well known "facts" myself. Best example is that when I first started coaching volleyball I tried to recruit the tall basketball players to play middle because I just "knew" they would be awesome slide hitters. The only thing I learned was that they weren't and all it did was piss off the basketball coach. Give me a 5'7" kid that has been in our program since Jr. High and we'll run circles around 6 foot basketball players that don't have volleyball skill training.

    See Witten to clarify your statement about training volume and its influence on muscle growth. Current research is revealing that this is simply not the case.

    Yes, you can achieve a higher training volume with blocked training. Wonderful, but my question for you then is this; "If we know conclusively that block training doesn't create motor maps in the brain that are representative of movements we need in competition; and we know conclusively that force variables are tightly coupled to execution of the specific movements; and we know conclusively that block training does not lead to skill transfer (i.e. motor learning); why would we train that way?" Further, if you look at the references in our article about random training you will find that if you control for the amount of time practiced, your assumption that 1500 blocked spikes per week is superior to 300 random is false when it comes to measure of skills in game situations.

    To your question on the principles of diminishing returns ~ emphasizing again that the adaptation is specific to the task ~ there is for all intents and physiological purposes no "intersection" of whole/random and part/block. It does not exist. These activities stimulate different brain pathways, optimize different neuromuscular functions, and are, at least from a motor skill perspective, completely independent . Again, if you control for amount of time practiced, random will be far superior to block for all of the reasons previously cited.

    Finally, scientific research has revealed that skills AND strength are specific to the task. Serving is a great example of a closed skill that can (and is) trained in a block fashion to increase serving strength. However, in our gym, we know that eventually this skill will be executed in a random environment with environmental cues and stressors not normally present in practice. Thus, to optimize the acquisition of this skill we train it randomly and use contextual interference, which we know will optimize BOTH skill AND strength acquisition.

  8. CJM says:

    >Dear Joe (Part III),

    Joe, I hope that you will receive this information in the spirit it was intended and I encourage you to continue to ask probing questions that stimulate thought provoking discussions such as this. Carl has been instrumental in provoking my own personal pursuit of the answers to these fundamental questions and I hope that in some small way this response has been helpful in your own quest.

    Best wishes,


  9. Junien says:

    >Dear Steve,

    I am not sport science profession, just a coach who love volleyball and coaching and educating.

    The journal as well as the the discussion are inspiring. It is totally different from what Asian's sports culture: coaching from experience.

    However, as a coach, I still see lots of successful examples on whole/random training and failures on part/block training.

    I agree with the journal: whole training should come first to stimulate connector and part training could come later for expert performance.

    But there is a problem: what would be the standard of evaluation? In other words, how we evaluate player's performance? What would be the indicator or the performance?

    Best Regards,
    Alan Chew

  10. Steve and Betsy says:


    If I understand your question correctly, then the measurement standard to evaluate practice efficacy must be linked to competitive performance. As such, skill is essentially the successful replication of a functional movement to achieve desired outcome. Take a skill, for example passing, establish a standard (let's say 50% perfect,) and measure passing performance in competition over time. This can easily be applied to any skill to measure improvement over time.

    Best wishes,


  11. Steve and Betsy says:


    If I understand your question correctly, then the measurement standard to evaluate practice efficacy must be linked to competitive performance. As such, skill is essentially the successful replication of a functional movement to achieve desired outcome. Take a skill, for example passing, establish a standard (let's say 50% perfect,) and measure passing performance in competition over time. This can easily be applied to any skill to measure improvement over time.

    Best wishes,


Comments are closed.