Sunday, March 31, 2013

Does Speed Work work? My response to Mike Tuchscherer’s article. Part 2

Does Speed Work work? My response to Mike Tuchscherer’s article
Part 2

Click here for the part 1 of this article.


One thing I forgot to emphasize and explain in Load-Velocity profile (from now on L-V profile or L-V trade-off) was that all reps are done at MAXIMUM INTENT to lift as fast as possible (within technical limit). In another words EFFORT of each rep, at each LOAD was MAXIMAL.  In some strength circuits this is called C.A.T (Compensatory Acceleration Training).

I have created a graph to explain that submax weights could be done with less than maximum effort.

Load-Velocity profile done with different effort levels

Using the above example I could lift 120kg at 0.55 m/s with maximal effort. With submax effort I will lift it slower (I have used percentage of the maximal velocity for a given load as percent of maximal effort, although this relationship might not be linear like this). Also, it might not be possible to slow down certain loads too much (especially not 1RM) due sticking points and TUT (endurance).

We use the word intensity to refer to a given percentage of 1RM. Recent paper by James Steele (thanks to Bret Contreras for pointing it out) argues against it and you can quickly see why in the above graph. I can lift submax load with maximal or submaximal effort. Steele recommends using intensity of load, intensity of effort to make things clearer:

In RT, we could talk of the intensity OF load as being the percentage of 1 RM or maximum voluntary contraction that is being used. Or we might talk of the intensity OF effort involved during an exercise with the caveat that we can only gain subjective measurement of the sense of effort through RPE, and measurement of motor unit (MU) recruitment in RT provides a physiological variable correlating with effort, with max MU recruitment representing max effort independent of load --- Steele J. Br J Sports Med Published Online First

Up to this point we have talked about only single rep velocity at certain load. Using less than maximal load one could perform couple of reps until exhaustion (technical or complete failure). On the following picture is the table by Dan Baker for experienced trainers.

Repetition maximum continuum
 This is another fundamental concept in strength training and it is called repetition maximum continuum or Load-Reps trade-off. 

Please note that there is no special reference to the effort of those reps (or speed) so they are usually done at self-selected pace. Also, there is an important difference in L-R relationship between exercises, level of lifters, etc. One could use Reps-till-fatigue to estimate 1RMs, although as said there is a big difference between exercises, lifters, etc thus most coaches use different tables.

In R-L continuum we are again talking about MAXIMUM reps done at certain load (%1RM). Using submax effort one could perform lower number of reps.  If we say that lifting for maximum reps athlete reaches maximum exertion (or exhaustion), then lifting for submax number of reps one reaches submax exertion.

This is one of the main principles of Mike Tuchscherer’s Reactive Training System. Mike uses RPE scale (Rate of perceived exertion) to quantify exertion for number of reps performed. You can read more about it HERE.

Another table is relative intensity which I’ve picked up from Donnell Boucher (click HERE). I am presenting here one example of such a table:

Relative intensity table

What is common to both of these tables is the concept of quantifying how hard a given set was (in terms of exertion at the end of it). This is useful for programming since all-out-sets demands a lot from the body and demands more time for recovery.

I will visualize it by using Dan Baker table

Performing submax number of reps at certain %1RM

 Please note that this is not universal percent table – it is only an example that you can use. For more info please refer to work by Mike Tuchscherer. I have provided outline of his system HERE and HERE.

As you can see there are two inter-related fundamental concepts L-V and L-R. It is crucial to understand them, but it is also crucial to understand the terminology issues. Thus we have

Intensity of load (%1RM)

Intensity of effort (lifting speed, intent to lift fast – concentric only or concentric-eccentric; tempo)

Intensity of exertion (expressed sometimes as 5[10] which means 5 reps of possible 10 reps; one could use RPE system as well or relative intensity table)

As you can see this is all related to one set. Then we have more programing variables relating to volume aspects, like number of sets, total number of reps, tonnage, rest between sets, etc.

The authors and coaches often confuse effort and exhaustion (I have wrote about it HERE). The example that James Steel gave depict the situation perfectly:

Along this vein of thought, however, we might also consider the differentiation between sense of effort and the physical sensations associated with exercise, which, considering the above definition of intensity, would also be inappropriately labelled as such. Indeed, these phenomena have also been recently called upon to be differentiated appropriately within the articles published in BJSM.16 17 Smirmaul16 offers a practical example appropriate to RT in this regard suggesting, “A short maximal voluntary contraction for leg extension, for example, will induce a maximal sense of effort while, initially, other unpleasant sensations will probably be modest. Repeating this maximal contraction several times, however, will increase these unpleasant sensations continuously, whereas sense of effort will always be the same (ie, maximal).” Indeed, I and my colleagues have questioned the use of RPE in RT to determine relative effort.6 Shimano et al9 showed that, when repetitions were continued to MMF, RPE was similar at 60%, 80% and 90% of 1 RM for most exercises;
* MMFMomentary Muscular Failure

Why is all of this important? It is important to differentiate between different methods and understand that a lot of authors forget to describe all related parameters, so we end up not knowing what the exact training was.  


One analogy I like to make is the example of DC motor with attached battery.  

Battery = Nervous System, motor = Musculo-skeletal system; Both = motor system

In this simple example motor is the muscle and battery in nervous system. For a same level of voltage motor will behave differently (in terms of angle velocity, torque, power) based on the load at his axis (inertial, friction, air, etc).

This is pretty much along the lines with Load-Velocity profile – velocity of the motor will depend on the external load, but also on the voltage (and in this case voltage is effort).

Effort = voltage

Thus lifting with maximal effort is like lifting with maximal voltage not matter what load is on the bar.

Another example might be lifting submax weights to exhaustion/failure (RE method). Suppose that the motor can fatigue and decrease his force/velocity output for the same voltage over time.  You attach certain load to the motor which is 50% of his maximal torque and use 50% of voltage. He is going to rotate slowly but as the motor get “tired” one needs to increase voltage to maintain rotation at certain velocity. After you eventually reach the highest voltage and are unable to increase more, the motor will stop.

This is along the lines of lifting submax weight until failure. At the end it will result with you recruiting you maximal voltage.

To make things more confusing sometimes the battery might get tired or reduce its voltage based on some feedback from the motor (e.g. temperature) to avoid complete motor breakdown (for this one needs more complex circuitry). Now we are talking about central fatigue as opposed to peripheral fatigue (at the level of the motor). Usually they are both involved in larger or smaller degree depending on what time frame we are talking about (single set, workout, week, etc), exercise, intensity, etc. You can read more about it in Cesey Butt article HERE.

Why am I saying all of this? Because according to Ralph Carpinelli it is not only the intensity of the load (%1RM) that determines motor recruitment (i.e. battery voltage):

…That is, the size principle does not support the popular resistance training recommendation to use a maximal or near maximal resistance. The size principle and interpolated twitch studies support the contention that if maximal
motor unit activation is desired, a maximal or near maximal effort at the end of a set of repetitions— regardless of the amount of external resistance—will elicit maximal motor unit activity. Effective resistance training does not require the use of a maximal or near maximal force to stimulate the available motor units and produce significant increases in muscular strength.

… Despite the plethora of opinions in the resistance training literature, the specific mechanisms of fatigue and exactly what constitutes an optimal stimulus for strength gains are unknown. If a maximal— or near maximal—effort is applied at the end of a set of repetitions, the evidence strongly suggests that the different external forces produced with different amounts of resistance elicit similar outcomes

…Despite the plethora of opinions in the resistance training literature, the specific mechanisms of fatigue and exactly what constitutes an optimal stimulus for strength gains are unknown. If a maximal— or near maximal—effort is applied at the end of a set of repetitions, the evidence strongly suggests that the different external forces produced with different amounts of resistance elicit similar outcomes

So the things are not so simple. I am not sure I completely agree with Ralph Carpinelli regarding the training recommendations (read the whole article), but at least this shed up some light on the issue that it is not only intensity of load (%1RM) that is important in increasing strength and maximizing muscle recruitment. Lifting submax weights to exhaustion will result in maximal recruitment as well.

I tend to believe that this is dynamic animal – there is certainly a dynamic threshold in intensity of load (%1RM) that a given trainee have to use and this might change over time. Also, there might be fatigue limit to performing submaximal weights to exhaustion and this might limit total work overall and thus limit strength gains.

I don’t want to turn this into discussion on lifting to failure or not or single set vs. multiple set, since I believe both are tools in your toolbox and needs to be done at certain periods for maximum results. What is important to get from this is that for maximal motor unit recruitment (according to Carpinelli) it might not be necessary to use maximal loads or near maximal loads.

And in the end this is the static picture I alluded to in the first part. We know that the goal is to increase 1RM, to increase battery voltage, to increase velocity around circa maximal weights, to improve technique (efficiency). But we don’t know what is the best method to come there. We know the destination, but we are not sure what the best journey is.

Will maximum recruitment yield most improvements in 1RM over time or is it something else? Force output? Number of grinding reps? Volume of lifts over 80%, over 90%? Total number of reps? At the end of the day we still don’t know.  That’s why we have programs that yield same or very similar outcomes by using totally different approaches – one might use low frequency of grinding reps (maximal recruitment, maximal exertion) usually known as HIT; one might use high volume and frequency of lower intensity of load (%1RM) like Sheiko. One  might combine different methods like Westside guys do to reach maximum recruitment in most things they do (ME for maximum recruitment with highest intensity of load; DE for maximum effort with lower intensity of loads for maybe maximum recruitment but with lower load; RE for maximum recruitment as the fibers get tired). Key message is that even if we know the static picture we still don’t know what change that static picture over time. And there are a lot of ways to skin a cat.


What Mike Tuchscherer believes [as far as I can tell from his writings] that will bring these improvements is the following – (1) force output during an exercises is very important stimulus and since force output changes with increasing loads [I will cover this soon] then higher loads are more important than lower, and (2) higher volume of work with this higher force output and (3) technique similarity. This is his rationale behind ditching lower intensity (50-70%) Dynamic Effort methods and putting more focus into higher intensity of load (sets @8RPE).

Basically, even if there is maximal muscle recruitment during Dynamic Effort method or Repetition Effort the force output is lower than with higher loads (%1RM) and hence will bring less improvements over time.  

I am not sure how correct this is, but  I wanted to explore my Force output during different loads. I am going to use my bench press (with pause) as an example. Here is my force output during set with 60kg:

Gym Aware output with 60kg bench press

Please note that the force is estimated using reverse dynamics using position change of the barbell. Refer to this paper for more info. 

To understand the forces involved I will create a simple mechanical model. 

Simple mechanical model...
What you can see on the above graph is estimated Muscular Force. When the muscular force is over gravity force (in this case 60kg x 9,81 = 588N) then barbell will start accelerating. If the muscular force is below gravity force the barbell will start decelerating, as you can see happening during the last 1/3 of the concentric phase. If the muscular force is below zero, that means I am actually pulling the barbell down (rowing) and as you can see that happens during the last period of the lift. This happens during the lighter weights (in terms of %1RM) and I believe in greater amount during the bench press than squat (your feet are not bolted to the ground so you can pull down actually).

Here is the table from Sanchez-Medina et al. Importance of the Propulsive Phase in Strength Assessment. Int J Sports Med 2010; 31: 123 – 129 where they showed relative contribution of the breaking phase in the bench press.
Breaking phase over  20-100 %1RM

According to this table - only loads over 80% 1RM produce no breaking phase. This is one of the reasons for using chains and bands when doing Dynamic Effort method – to extend the propulsive phase and create more force.

Here is my force output with 100kg.  

Gym Aware output with 100kg bench press

This time I need to overcome 100kg x 9,81 = 981N to get the bar moving. And there is not below zero force (no pulling the bar).

On the next table I have compared Peak Force output during concentric phase against different loads when reps are done with maximum effort (C.A.T.).

Peak force output over %1RM continuum in bench press

Peak force is the highest force achieved during the concentric phase. Using mean force output will yield weight of the barbell – so I have used peak force.

As you can see there is a curvilinear relationship (I have used polynomial equation on the graph and linear on the table, as you can see R2 is higher on graph). Bret Contreras posted similar picture from a study by Swinton et al. that is pretty much in line with this one.

According to my bench press data, to reach over 90% of Peak Force output one needs to use loads over 82% 1RM. To reach over 80% of Peak Force output one need to use more than 68% 1RM.

To make sure that this is also the case with squat I compared my squat numbers with even jump squat. 

Gym Aware output during squat jump with 20kg 
When I put everything in table this is what I get

Peak force output over %1RM continuum in squat

According to this sample, loads needed to reach over 90% of Peak Force are over 84% 1RM and to reach over 80% of Peak Force one need to use over 73% 1RM.  Here is the table

% Peak Force
>82% 1RM
>84% 1RM
>68% 1RM
>73% 1RM

It seems that Mike Tuchscherer is right in this regard – One cannot reach peak barbell force outputs with submaximal loads even if he utilizes compensatory acceleration.

[Disclaimer: The forces explained here are the forces acting on the barbell and not Ground Reaction Forces. To estimate those I have added 100% of my BW to the load . This is what I got:
Estimated GRF forces using 100% BW along with barbell weight

As you can see, using 100% BW into calculation I managed to produce the same peak force during the 20kg countermovement jump and 160kg squat. Does this mean I need to work more on my strength levels since I’ve utilized full its potential in the jump? I have no clue – I would need more data and need to think more about this since I just discovered it for the purpose of this article. Having a force plate might help as well.

This is also an example of result dependency on measurement method as I have alluded in this post on power measurement. Anyway, according to this study barbell kinematics should not be used to estimate barbell and body system center of mass in the back squat. So using barbell velocity as a representative overestimated barbell-body center of mass velocity and hence acceleration, power and force.

Anyway, if I stick to squat performance (without jump squat) to reach over 90% of peak GRF one need to use more than 74% 1RM and to reach over 80% of peak GRF one need to use more than 54% 1RM – a lot less than when we consider only barbell force. The question now is which one is more important – GRF or barbell force?

End of Disclaimer]

I believe that you are confused at this moment. Welcome to the club. To summarize – Mike Tuchscherer is right when he says that one can’t achieve peak forces with loads less than 80% 1RM when one take forces applied to the barbell only. In the case with peak GRF (in squat; estimated using LPT, but I would need force plate to be more certain) one needs loads higher than 70%.  Please note that Dynamic Effort uses 50-60% 1RM raw loads.

One thing to consider is that my C.A.T. with submax loads is NOT the same as Dynamic Effort. In Dynamic Effort one is performing both eccentric and concentric parts explosively for 2-3 reps, plus add chains or bands to the equation and the forces might be much higher than with my pause technique. But you also have sitting on the box that might actually decrease it. Without measuring it I can only speculate.

 So even if Mike is right when it comes to force outputs, does it means that force outputs are key stimuli for strength improvement, or something else? Only the experimental study could tell us this. And not done on college kids or weak coaches (point taken), but on real powerlifters. Changes should be estimated using smallest worthwhile changes. Anyone up for an experiment give me a call.


Conclusion is that the things are complex and there are a lot of ways to skin a cat. I cannot say with any confidence that DE works or not. Plus, one needs to take into account the whole training system and not only one method the system uses It is only tip of the iceberg. This brings me to my opinions that you should take with grain of salt.

I do think that Dynamic Effort is a little bit overrated. I am leaning more toward Mike Tuchscherer’s side, but I am maintaining my skeptical attitude. Mike is probably right when he says that force outputs are lower in DE then when working with higher loads (not necessary to exhaustion; in his system around 8RPE), along with different technique of execution which might not bring transfer. But again, are these the sole things behind strength increase stimuli? We don’t know.

One thing to consider is that DE might work for the other reasons instead of developing explosive force for blasting through the sticking point (this might be an interesting study to find out). When one works with ME twice a week and really pushing it, then he is not left with much than to do easier weights (50-60%) the next two day. This is similar to the polarization of the training for endurance runners.

This is why it is important to take a look at the big picture. Numerous champs are using this Westside, but numerous tall humans are playing basketball which doesn’t prove basketball make humans tall. One could use different methods during different phases.

This study (although sketchy) showed higher improvements in bench press in group performing it with C.A.T and before velocity drops below 80% of the  best rep, even when doing less volume compared to a group who did self-selected lifting speed to failure. I am very interested in this velocity based strength training at the moment – and I believe there is a huge potential in this approach.

One idea might be to utilize accumulation phase(s) using higher number of sets, lower number of reps with relatively higher %1RM (75-80%) with NO grinding at all and with C.A.T. (that is around 8RPE according to Mike) before switching to more Westside program for couple of weeks/months, where DE might be utilized as a complement to ME session.  

I hope you have enjoyed this article without any confident conclusion.

Friday, March 29, 2013

Does Speed Work work? My response to Mike Tuchscherer’s article. Part 1

Does Speed Work work? My response to Mike Tuchscherer’s article. 
Part 1


Mike Tuchscherer’s article Why Speed Work Doesn’t Work lifted a lot of dust lately. I have read Bret Contreras’, JL Holdsworth’s and Chad Smith’s responses and I have wrote some of mine ‘comment’ on Jim Wendler’s Facebook Wall

Today Mike contacted me and invited me to expand further and continue the discussion on his forum since he found my comment interesting. Thus I decided to write one blog entry as a starting point, along with explaining my rationale/viewpoint. 

I have huge respect towards Mike, what he is doing, both as a lifter and as a coach. I have learned a ton from him and he was really helpful with his responses to my emails and questions.   

Before I type anything I need to make couple of things straight. First of all I am not powerlifter nor I coach powerlifters. I do not coach strength athletes either, nor bodybuilders. I work with team sports athletes that demand a mixture of physical qualities ranging from power, speed, endurance, strength, but most importantly technical and tactical skill. I have some data on my own lifts by using Gym Aware LPT system which I will present shortly, but since I am weak as kitten (compared to elite powerlifters) take them with grain of salt (in more scientific term it is hard to make any inferences to elite powerlifting population from my own lifting samples). I will mainly reference some studies and use rational thought (that needs to be confirmed with scientific/empirical method). 


Load-velocity profile [trade-off] is one of the most fundamental concepts in kinesiology, emerging in characteristics of single muscle fibers, single joints and finally of multi-joint complex movements. I am not sure we completely understand why it happens and what is the exact mechanism - even at the level of basic muscle fiber – and not even at the level of complex movements governed by even more complex nervous system.   

Anyway, load-velocity trade-off is there even if we don’t understand it completely. To make things simpler I will refer to common strength training movements like bench press and squat when I discuss load-velocity trade-off. I have written couple of articles regarding velocity-based strength training and using load-velocity profile of the lifter for estimating his 1RM and prescribing training HERE (make sure to follow links in the text). 

One can look at load-velocity profile as a formula, where velocity of a movement [V] is predetermined by external load [L]. 

V (L) = a x L + b

Constants a represent slope and constant b represent the intercept.  In short: velocity of a movement depends on the external load.  Here is my load-velocity profile for the squat (pause around 1sec at the hole). 

Load-Velocity profile for squat

As you can see correlation is nearly perfect between the two. I took 160kg as my 1RM and associated speed was 0.3 m/s (mean concentric velocity). If I use regression formula (one could use =TREND function in Excel, or manually by using slope and intercept), velocity at 160kg is estimate to 0.301 which is practically exact (as can be seen by low SEE). Thus one could use this regression to estimate 1RM.

Note that 1RM doesn’t always happen at 0.3 m/s. My bench press is somewhere around 0.15 m/s and this might depend on the level of the lifter (beginner, intermediate, advance), movement (small vs. large muscle mass involved), type of lifter (ST or FT, grinder or explosive), etc. So it is individual.


What is the goal of powerlifting? The goal is to lift as much weight as possible without time reference. So one could lift 200kg in 4 sec and another might grind it for 20sec. Same results – so the movement velocity of 1RM doesn’t matter in powerlifting. 

Looking at the curve above the goal is to move 1RM to the right (i.e. from 160kg to 180kg). What happens with the slope of the curve doesn’t really matter – powerlifters are not competing who can generate more velocity with submax weights. This can’t be said for other sports!

Sometimes even if the 1RM doesn’t improve, if the velocity (and thus power output) at certain submax load (usually representing external resistance common to sport competitions [and NOT working around some magical load that produces peak power output]) improves that represent positive improvement. This is of no importance in powerlifting though.

Here is the possible scenario of me improving velocity without improving my 1RM which is still important in more [real] power dominant sports. 

Improvements in Velocity at without improvement in 1RM
And here is the possible opposite scenario of me improving my 1RM without much improvement velocities at submax load.

Improvements in 1RM without improvements in low resistance velocities

Third possible solution might be shifting the whole graph to the right by improving both 1RM and submax velocities. 

Improvements in both submax load velocities and 1RM

From a powerlifting standpoint the only important thing is improving your 1RM. Speaking of this I can improve my powerlifting result by not even affecting the load-velocity curve. You wonder how? By simply learning to grind more weight and in becoming more confident in it. Check the graph – I am learning to grind and I am able to lift 180 kg at lower speed. 

Improving grinding?

I am not sure if this is an improvement (development) or rather learning to express what one already have – and this goes pretty much in line with my develop~express concept. More experienced powerlifters and/or scientist might chine in on this topic. Also – can you change your grinding speed? Please chime in in the comments or on Mike’s forum

Going back to situation where the load-velocity curve improved so that 1RM improved, while velocity at submax weight didn’t (I am re-posting it again)

As you can see from the graph, as I have improved my 1RM I have also improved lifting velocities in the zone around 1RM. I am not sure if this is circular causation so the opposite is also true (i.e. improving velocities in the circa-maximal zone will improve 1RM). And I believe this is the CORE ISSUE here so I will bold it:

If improving 1RM also improves submax velocities, will improving submax velocities also improve 1RM? And who is first – chicken or the egg and what is the best way to improve each?

Since this is the chicken or the egg problem (as long as I don’t see experimental study with groups) automatically assuming process by the outcomes might be misleading. What do I mean by this? We have a tendency to believe/assume that using loads/velocities associated with one part of the curve will improve that same part of the curve. We can see this same rational flaw in distance running – and I HIGHLY suggest checking this great article by Steve Magness on Physiologial Model of Training (Note to Steve: if you are reading this I am still awaiting for the part two) and Attacking Adaptation from Multiple Directions.

 In plain English – will using circa maximal loads (90+%) improve 1RM better than submax weights over time? Will peak power be mostly improved by utilizing loads associated with it? Will VO2max be mostly improved by using VO2max workouts? Will lactate threshold be mostly improved by utilizing lactate threshold workouts? Is the best way to improve soccer skill by playing 10v10 all the time? And many more examples of develop~express confusion, misunderstanding of the specificity principle and training transfer along with highly mechanical/linear thinking.

I have written about this during 2009 in my Planning the Strength Training article, but I was mainly referring to the repetition continuum that you can see in most strength and conditioning textbooks: 

Repetition Continuum

It can be said that reaching of the different strength training goals (and thus motor qualities) is based on utilization of different loading protocols (weight, reps, sets, tempo, rest, etc.) or methods. So, each of the methods aimed at reaching different strength training goal utilize different loading protocols. This is based on the repetition continuum, or the ’idea’ that different goals can be achieved utilizing different reps per set. There is a dynamic interaction between the variables of reps, sets and loads. The load used (% of 1RM) ultimately determines how many reps per set are done. Reps per set (or set time) ultimately determines how many total sets must be done. The interaction between the three will affect what adaptation is seen. Although not all authorities agree, there is thought to be a continuum of adaptations which may occur with different repetition sets. This continuum is called repetition continuum. --- From Planning the Strength Training, 2009.

We have tendency to think this way and rationalize training methods, but it might be flawed. It is same as thinking that playing basketball will make you taller because basketball players are tall. Yes this is a bit extreme, but it is same flawed reasoning. 

Maybe I am nitpicking, but this might be reason of our frustration to understand same results with totally different programs and appreciate that what brought someone from A to B, might not bring him from B to C. 

Here is an example in running – the HIIT is quite popular and research is showing that runners who start doing more HIIT improve their performance and aerobic capacities more (there is the difference between improving VO2max and performance, see article by Steve Magness). And yet we see that most elite runners do mainly great volume of low intensity work and the ones who improved more did that by increasing low intensity volume instead of high intensity volume (see article by Seiler). 

When it comes to training we believe that the best way to improve 1RM (or strength) is to do loads very close to 1RM. And yet we have extremely strong lifters using Sheiko routines that are mostly 75-80%. It might be also interesting to read interview with Carlo Buzzichelli by Bret Contreras regarding this issue.

Enough for today – I have covered some basic theoretical/philosophical and maybe even practical aspects of this issue. I will leave you with the resources linked for now and continue soon with more practical words.

Stay tuned till part two…

Wednesday, March 27, 2013

Is power/speed reading in clean and snatch counterproductive (and other rant)?

Just had a great talk with Travis Fisher regarding velocity-based strength training (he is only one I know using it and trying to figure out a solid system) and we touched a little bit on using velocity/power readings during Olympic lifts. Could one predict readiness (day to day variability) or 1RM by using velocity as with bench, squat, squat jump? 

As I have wrote in couple of articles using velocity of a lift one could create load-velocity profile and estimate 1RMs pretty reliably (see the research papers by Juan J. González-Badillo, Mário C. Marques and Luis Sánchez-Medina) along with prescribing training volume (using velocity drop offs) and intensity (by using velocity instead of % of 1RMs) and taking into account day-to-day variability in readiness. I believe there is a huge potential in this approach and I plan experimenting more with it and maybe writing a starting workout guide with Travis. 

This works perfectly fine with bench press, bench throws, squat, squat jumps and other exercises. It works even better if you provide immediate feedback (average velocity works better with non-ballistic movements, while peak velocity might work better with ballistic movements like bench throw or squat jump) during the lifts (as show in research). 

I was wondering about could this be used in clean and snatch? Could you (1) estimate 1RMs in clean and snatch from submax velocities (peak or average), (2) should you provide instant feedback in terms of power or velocity of the lift and (3) can you judge the daily readiness from variability in power/speed from the same submax weight? I believe it is simply more complex than the thing with squats, squat jumps and other exercises. 

I am not aware of any studies in this regard and if someone knows them please give me a heads up in the comments. Or if someone wants to give me a funding or PhD scholarship to do those (and a lot more) let me know too :)

So until I get more empirical data I will based my opinion on rational thought.

In short I believe answer to those questions is… no. Let me expand. 

In balistic movements the goal is pretty straight forward – lower the weight (quickly or not depends if one wants to utilize elastic energy and reactive strength) and explode up as much as possible. Power, force and velocity are estimated/measured during the concentric phase of the lift that starts from the bottom position (lowest reach/distance) and ends up at the highest point in the movement. This might be tricky – it is questionable should one use the total distance of the barbell as concentric range (this involves flight phase) or only while the barbell is in the arms (in the case of bench throw) or until the full triple extension (in the case of squat jump). This again depends of the protocol used (protocol dependency) as I have pointed out in this article

As Dan Baker commented on it: it doesn’t matter as long as it is improving over time. But it makes comparison between exercises a lot harder. And this is especially important since we like to compare exercises and choose camps along with writing nice headlines: dynamic effort squats create more power than Olympic lifts; max power output happens around 50% of 1RM, the highest power output in jump squats is with bodyweight only, etc. But I digress. 

How the things are measured matters. 

My jump squat with 20kg using GymAware system. I have added phases with arrows. Notice that phase used for calculus includes both concentric phase and part of the flight phase. 

When it comes of Olympic lifts the things are not that straight forward as with other ballistic movements (even they have certain nuances).  They are a lot more technical and they have phases that are interdependent. And besides this not all phases are done with highest effort. For example, if you rush in the first pull you might end up screwing the second pull and totally losing the lift. 

I am not expert on Olympic lifts, but I think those who are will agree with me. 

If we measure power, force or velocity of a lift it is questionable where one should start and stop the measurement (as with bench throw and squat jump example). Should the measurement start from the ground or during the second pull? Should it end on the highest point of the lift, or at the highest position of the triple extension? 

Even if we get this measuring issue straight it is still questionable if one could use it to estimate 1RM or provide daily readiness estimates.  

Here is an example. Suppose my 1RM clean is 150kg. I do sets with 100kg. I attach the LPT to the bar to get power output or velocity. I do this submax clean without looking at the numbers. Then someone say try to generate more power or more speed by providing me instant feedback during the lift. How will this goal/constraint affect my technique? 

One might rush the first pull to get higher average power or velocity. One might increase the power of the second pull to get the higher bar height which might end up in doing power clean or even muscle clean. 

Please note that this might happen with the squat as well – one might end up being on toes or even leaving the ground. Thus measurement MUST be within the technical model of the exercises and sometimes providing a non-specific feedback might end up screwing up the technique itself.
At the end – what is the goal of the clean? To lift as much weight as possible, or to produce the highest power output (especially with submax weights)? 

I am not aware of anyone getting a gold medal in being able to produce most power output during the clean & jerk (who knows, maybe with Crossfit we might end up having those medals as well). Power output is the side-effect. It happens when you optimize the movement. I am not sure that the power output or velocity improvement during the submax lifts should be the goal and/or feedback with Olympic Lifts. 

Yes – you maximum power will be increased as your 1RM increase with the clean and snatch, but should the goal of training be improving power output with submaximal lifts per se? Or something else? 

My question would be what would be specific feedback during the Olympic lifts then? Could it be single blind (where the athlete is not aware of being tested/measure) power/velocity of the second pull (could this be used to create load-velocity profile and estimate of 1RM)? Or could it be something more elegant and complex than pure linear power/speed? Like rhythm, timing, depth? Olympic lifting is a skill.

This might be the same thing that happened to sprint analysis – vertical or horizontal forces, stride rate or stride frequency, flight time or contact time? Then we have coaches focusing on maximizing/minimizing those and forgetting about the bigger context, about the skill, about total movement, relaxation, rhythm, elegancy.

“Make things as simple as possible, but not simpler” as said by Albert Einstein.

 I have spoken with couple of athletes and coaches that said that the highest power and highest velocity happens when they relax and let it happen. My boxing coach said he was never able to knockout anyone when he really pursued it by aim to hit them hard – but instead it happened when he relaxed and let it happen. Sprinters reach highest speed and jumpers highest jump when they are most ‘relaxed’ and effortless. 

So the whole point of this rant if mind your feedback. Not all feedback is created equal. Check the bigger context. 

Another message is the importance of understanding how the things are being measured and estimated. So it is not so simple to state what movement created more power output – clean or squat jump, because they don’t have exactly the same movement goals and constraints.  

Safety (along with skill learning) might also affect the selection of power development exercises. For example in clean you accelerate the barbell up and catch it couple of inches below the top height. With squat jump you accelerate the barbell up in the air and then you drop for a lot more inches from the peak height down to the ground. This creates a lot of impact due the difference between peak height and start of the catch/amortization. One also need to take into account the eccentric part of the squat jump when you quickly lower the barbell and quickly reaccelerate it (in the case of countermovement squat jump). 

The difference between highest position of the barbel and the catch position might be very important in eccentric stress and injury potential (IMHO) 

Thus cleaning 60kg (not with a cloth though) is a lot safer than squat jumping 60kg IMHO. And a lot of researchers doing squat jumps with a lot of kilos uses Smith machine with pins and/or special breaking devices. In my opinion it is a lot safer to use hex barbell jumps than squat jumps. If you use them (especially with a lot more weight than 30% squat 1RM) it might be wise to do it in a cage with safety pin and doing it only concentric wise. But this again depends on the athlete/sport.

This is me doing pretty high concentric squat jumps from safety pins. GymAware uses angle corrections

Thursday, March 21, 2013

The problem with [Peak] Power [calculus] – or why I don’t believe in this Sacred Cow

I am going to say it straight ahead – coaches who believe that they are training peak power at certain percentage of 1RM are simply misguided.

There are two important logical errors behind it [IMHO]. First one is the belief that there is a magic bullet – a certain % of 1RM intensity that will result in peak power output and in return training with that peak power output will result in great transfer [power and explosiveness] to all other zones or related activities (e.g. improving power using jump squats will improve sprint speed, jump height or even 1RM). Using that intensity you are training [mythical creature in Platonistic world of motor abilities] POWER. Using any other intensity you are not training power.

Second one is the belief that there is peak power. This peak power is TOTALLY dependent on the way/method it is being calculated. There could be 100+ methods of calculating it and all of them will result in different % of 1RM (Range: 0-80% 1RM). Let me expand. To get power, one needs two variables: instant velocity and instant force.

When it comes to velocity one could measure velocity of COM [Center of Mass] of both the athlete and barbell or only velocity of barbell. This will create different velocities. This could be also measured using 1-LPT [Linear Position Transducer], 2-LPT or 3D video analysis.

When it comes to force one could measure it using force plate directly or estimate it by using reverse dynamics [speed is derivative of position, acceleration is derivative of velocity and force is acceleration times mass involved] from positions derived by using LPTs. Also, estimating force from acceleration demands multiplying it with the moving mass. What is moving mass? Is it barbell only or body weight and barbell? Someone use 90% of body weight, some use 100% and some use 0%.

When we decide on all of these then we have range of movement. Are we going to use full concentric ROM or only part of it where acceleration is >0 [Fmuscles = Fgravity]? Or when Fmuscles start to actually break the movement (which actually happens with lighter loads – you engage the antagonists to stop the movement) and that the point where P [power] < 0.

Are we going to take into account average power over ROM or peak power over ROM?

Taking all of these into consideration it could be confidently said that the peak power concept is measurement/analysis dependent. And even if it existed I doubt that training at that intensity will provide a magical transfer to everything else that demands power output.

Another thing I actually hate is seeing this in studies. Researchers measured 1RM and power output at 50% or any other percentage of that 1 RM. Training over time improved 1RM. Then they measure power output at the same 50% or any other percentage of the new 1RM. Depending on the method of measurement (see the above) they might see statistically insignificant (don’t get me started on this B.S. concept) change in power output or even a drop. So they conclude that improving 1RM doesn’t (statistical) significantly improve power output. Well, duhhh. What about measuring it and comparing it at the same absolute weight, like P60kg (power output with 60kg)?

So, after all this rant what is the solution? One should do needs analysis and check what are the loads and resistances one needs to move during one’s sport. Then one needs to increase the SPEED and EXPLOSIVENESS at which one is able to move those around (including bodyweight). Improving this demands working over the full force-velocity continuum and applying different methods of training and not only trying to find peak power intensity that is going to magically improve everything else. First of all this isn’t going to happen and second it is measurement dependent and really hard to find. One exception to this rule might be cycling. They measure external power and they can manage the gears to take advantage of it. This peak power is usually around 80-100RPM (if I remember correctly), but the cyclist also train at cadences above/below that. 

To conclude - One needs to train to be more powerful (faster, more explosive) over a range of intensities  instead of focusing on one small zone.


Tuesday, March 12, 2013

[Research] Velocity Loss as an Indicator of Neuromuscular Fatigue during Resistance Training

[Research] Velocity Loss as an Indicator of Neuromuscular Fatigue during Resistance Training

Med Sci Sports Exerc. 2011 Sep;43(9)
Sánchez-Medina L, González-Badillo JJ.
Faculty of Sport, Pablo de Olavide University, Seville, Spain.

This study aimed to analyze the acute mechanical and metabolic response to resistance exercise protocols (REP) differing in the number of repetitions (R) performed in each set (S) with respect to the maximum predicted number (P).

Over 21 exercise sessions separated by 48-72 h, 18 strength-trained males (10 in bench press (BP) and 8 in squat (SQ)) performed 1) a progressive test for one-repetition maximum (1RM) and load-velocity profile determination, 2) tests of maximal number of repetitions to failure (12RM, 10RM, 8RM, 6RM, and 4RM), and 3) 15 REP (S × R[P]: 3 × 6[12], 3 × 8[12], 3 × 10[12], 3 × 12[12], 3 × 6[10], 3 × 8[10], 3 × 10[10], 3 × 4[8], 3 × 6[8], 3 × 8[8], 3 × 3[6], 3 × 4[6], 3 × 6[6], 3 × 2[4], 3 × 4[4]), with 5-min interset rests. Kinematic data were registered by a linear velocity transducer. Blood lactate and ammonia were measured before and after exercise.

Mean repetition velocity loss after three sets, loss of velocity pre-post exercise against the 1-m·s load, and countermovement jump height loss (SQ group) were significant for all REP and were highly correlated to each other (r = 0.91-0.97). Velocity loss was significantly greater for BP compared with SQ and strongly correlated to peak postexercise lactate (r = 0.93-0.97) for both SQ and BP. Unlike lactate, ammonia showed a curvilinear response to loss of velocity, only increasing above resting levels when R was at least two repetitions higher than 50% of P.

Velocity loss and metabolic stress clearly differs when manipulating the number of repetitions actually performed in each training set. The high correlations found between mechanical (velocity and countermovement jump height losses) and metabolic (lactate, ammonia) measures of fatigue support the validity of using velocity loss to objectively quantify neuromuscular fatigue during resistance training.

This is VERY interesting study, especially taking into account my recent interest in velocity-based strength training (LINK, LINK, LINK).

The authors did a bunch of novel things:

Established velocity-load profile

Tested for 1-RM, 12-, 10-, 8-, 6- and 4-RM

Utilized 3 sets of Bench Press or Squat of different relative intensity over 8 weeks (3 × 6[12], 3 × 8[12], 3 × 10[12], 3 × 12[12], 3 × 6[10], 3 × 8[10], 3 × 10[10], 3 × 4[8], 3 × 6[8], 3 × 8[8], 3 × 3[6], 3 × 4[6], 3 × 6[6], 3 × 2[4], 3 × 4[4])

Performed Vertical Jump for the Squat group and bench/squat at load at estimated 1 m/s before and after 3 set protocols to assess neuromuscular fatigue, along with measuring lactate levels and ammonia.

Neuromuscular fatigue and velocity loss over 3 sets of 12 @ 12RM. Taken from Med Sci Sports Exerc. 2011 Sep;43(9) 

What the authors wanted so see is how different reps combinations (how close to failure – a concept explained by excellent system by Mike Tuchscherer) affect neuromuscular and metabolic fatigue.
Here are some of the findings:

...To the best of our knowledge, this is the first study to analyze the acute response to manipulating the number of repetitions actually performed in each training set with regard to the maximum number of repetitions that can be completed

...Our results indicate that, by monitoring repetition velocity during training, it is possible to  easonably estimate the metabolic stress and neuromuscular fatigue induced by resistance exercise.

...The present study confirms that the magnitude of velocity loss experienced during RT gradually increases as the number of performed repetitions in a set approaches the maximum predicted number.

...A finding worth noting is that greater MPV losses were experienced for BP compared with
SQ for all protocols analyzed

... In the present study, very high and significant correlations (r = 0.91–0.97) were found between the three different types of mechanical measures used to assess neuromuscular fatigue (Figs. 2 and 3A, B). These relationships are an important finding for the quantification and monitoring of training load during RT. The fact that there exists such a close relationship between loss of MPV over three sets and loss of MPV with the V1 mIsj1 load in two exercises as different as SQ (Fig. 2A) and BP (Fig. 2B), as well as between both variables and loss of CMJ height in the SQ group (Figs. 3A, B), is a novel finding that emphasizes the validity of using percent loss of repetition
velocity within a set as an indicator of neuromuscular fatigue.

... The relationships observed in Figure 2 also mean that, for a given percent loss of velocity within a set, the degree of fatigue incurred during RT is very similar irrespective of the number of repetitions the subject is able to perform (shown in different colors in Fig. 2), at least in a
range from 4 (~90% RM) to 12 (~70% RM) repetitions.

... Lactate increased linearly as the number of performed repetitions approached the maximum predicted for each type of REP (Table 1) and showed extremely high correlations (r = 0.93–0.97) with loss of MPV over three sets (Fig. 4A), loss of MPV pre–post exercise with the V1 mIsj1 load (Fig. 4C), and loss of CMJ height (Fig. 3C).

... Lactate increased linearly as the number of performed repetitions approached the
maximum predicted for each type of REP (Table 1) and showed extremely high correlations (r = 0.93–0.97) with loss of MPV over three sets (Fig. 4A), loss of MPV pre–post exercise with the V1 mIsj1 load (Fig. 4C), and loss of CMJ height (Fig. 3C).

... Although some studies have reported the point within a set where a significant reduction in velocity (18) or power output (1,26) was observed, the optimal time to terminate a set before reaching failure has never been clearly established.Although the present study does not come up with a definitive answer to that question, it does, however, provide us with some valuable information that may indicate when it could be appropriate to end a set. According to our
results (Table 1; Figs. 3 and 4), a maximum MPV loss of ~30% for SQ and ~35% for BP could be established to prevent blood ammonia to significantly rise above resting levels.

... Monitoring repetition velocity during resistance exercise seems important because both the neuromuscular demands and the training effect itself largely depend on the velocity at
which loads are lifted. A velocity- or power-based approach to RT is not entirely new, and authors such as Bosco (5) and Tidow (37) already provided some initial guidelines for putting it into practice. However, the role placed by movement velocity has not been sufficiently investigated (28). The findings obtained in the present study strongly support the use of velocity monitoring to control the degree of incurred fatigue.

Furthermore, the immediate velocity feed feedback the athlete receives during each session may increase the potential for adaptation. With this training approach, instead of a certain amount of weight to be lifted, strength and conditioning coaches should prescribe resistance exercise in terms of two variables: 1) first repetition’s mean velocity, which is intrinsically related to loading intensity (15); and 2) a maximum percent velocity loss to be allowed in each set. When this percent loss limit is exceed the set must be terminated. The limit of repetition velocity loss should be set beforehand depending on the primary training goal being pursued, the particular exercise to be performed, as well as the training experience and performance level of the athlete.

... In conclusion, the present data show that the relationship between the number of repetitions actually performed in a set and the maximum predicted number that can be completed
is an important aspect to take into account when prescribing resistance exercise because the velocity loss and metabolic stress clearly differ when manipulating these variables. The high correlations found between mechanical (velocity and CMJ height losses) and metabolic (lactate,
ammonia) measures of fatigue support the validity of using velocity loss to objectively quantify neuromuscular fatigue during RT. The nonlinear response of blood ammonia to loss of repetition velocity could perhaps be used as a reference to indicate the point within a set where the exercise
should be terminated when the main training objective is to improve movement velocity or maximal power production.

This reminds me very much of the Prilepin table

Using these recommendations power/velocity loss during the set won’t be bigger than 30-35% as suggested by this article.

I describe one potential approach of prescribing exact % 1RM to be used, but instead of prescribing exact number of reps one could use reps zone (to allow for day to day variability in readiness), but also, if one is equipped with system like GymAware, use quality threshold (QT) – or a maximum percent velocity loss to be allowed in each set.

This quality threshold is very much in line with Mike Tuchscherer’s RPE system (Fatigue Percents and Critical or proximity of failure – of course taken into account that all reps are done with maximum effort.
There is one study done by using 80% quality threshold on experienced bench pressers with great results compared to the group that done sets to failure and without compensatory acceleration.  Results are so great that they are a little bit ‘fishy’. Anyway very interesting study and training concept indeed.
I think with the help of systems like GymAware we will tend to see coaches devising and utilizing more examples of velocity- or power-based strength training.

I am ‘experimenting’ with this at the moment. Just today I performed Bench press at 85% of 1RM with quality threshold set to 80% of the first rep. My first rep average velocity was 0.37 m/s and I set up the alarm on GymAware to 0.29 m/s. I end up doing 10 sets of 2 reps with 2 min rest. One could stop doing sets when one can’t maintain minimum velocity for the first reps with set rest time, or when one exceeds allotment time per exercises (this idea by Charles Staley although I saw Joe Kenn and Mike Tuchscherer using the same; this way productivity is a lot bigger, along with controlling larger group of athletes in the gym).

There are certainly different methods to do this and this might be just one of them. Anyway, one could create different stress and workout goals by prescribing different intensities and quality thresholds. For example intensive workouts should prescribe a bit heavier %1RM, lower QT (keep the quality) and longer rests. Volume (or extensive) workouts should prescribe a little lower %1RM, bigger QT (allow higher fatigue) and shorter rest. One could play with these variable to create different types of workout depending what is more the goal of it, what load aspect is being stresses (quality, grinding, intensity, intensiveness, volume, etc) or what of cycle is being utilized.

I believe velocity- power-based strength training will gain more popularity especially in power and mixed sports in the time to come.