Recently there was an informal discussion on Facebook about VBT (Velocity Based Training). Carl Valle asked several of the participants in that discussion if they would formally respond to a series of questions related to VBT. Here are the answers from Mladen Jovanović. The answers from the other participants will be posted when they are made available. You may submit questions to Mladen in the comments section below.
I want to thank Carl and Chris for including me in this roundtable discussion, but I also feel a moral obligation to state my current lack of “skin in the game” (to use author Nassim Taleb’s expression) when it comes to coaching in general and using VBT in particular. Most of my thoughts come from previous observations and current tinkering; hence they need to be taken with grain of salt. I am hoping to get back to coaching and VBT, and I also plan on pursuing a PhD on this very topic.
FREELAP USA: Olympic-style lifts are very specific to body types and technique, making them more than just a simple summary of peak or average output. Besides using feedback for motivation and accountability, what else can be done to use the data beyond estimating work?
MLADEN JOVANOVIĆ: There are some “fractal dimensions” to this question, since the answer would depend on the level/scale of the implementation. In other words, it depends on who we are talking about: team-sport athletes using Olympic lifts to develop power, or Olympic lifters trying to get better at their discipline. The amount of details and their “ruggedness” at different levels makes this a “fractal problem.”
To be upfront immediately, I am by no means an expert in Olympic lifting. Kinetic and kinematic bar path traces, combined and synced with high-speed camera and even force plate—along with the experienced eye of the coach and subjective feel of the lifter—might be helpful at this level of implementation to help identify “performance limiters” and hence direct training. This way data can be used to help prescribe training, rather than just describe it.
One needs to be cautious with feedback, since not all feedback is useful and could even be harmful. For example, take average team-sport athletes performing Olympic lifting. Feedback such as peak or mean velocity might be motivational and create competitiveness in the gym, but could also be harmful for lifting technique. Assume I can clean 100kg, but now I am doing sets of three with 80kg with full clean from the ground. Now, someone attaches LPT to the bar and tells me to increase the velocity of the lift. What do I do? I probably pull more powerfully and as a result the bar ends up flying to a higher position. This eventually results in a power clean (or even muscle clean for lighter weights) version of the lift.
The “problem” with Olympic lifting and LPTs compared to, for example, jump squats or hex bar jump squats, is that it is NOT “open-ended”—the bar needs to end up at a certain height with a certain “rhythm” regardless of the weight used. Some Olympic lifting coaches might even say that the lift should look the same regardless of whether you use 90% or 60% of your best lift.
Hence the velocity variability between different loads should be minimized. This is, in my opinion, completely different than “open- ended” lifts such as the jump squat where one tries to jump as high and/or fast as possible. In this case, feedback such as height and or velocity of the lift is motivational and will increase the power/effort and the technique will not be affected as with Olympic lifting, where the goal is to perform skillful movements at a specific rhythm rather than maximally.
As I have said earlier, I am by no means an expert in Olympic lifting and someone please correct me if I have said something stupid and wrong here.
FREELAP USA: Jump testing sensitivity is not perfect from the sensitivity being limited, but more reactive options that utilize the stretch shortening cycle add more validity. Is jump training worth doing regularly, a waste of time, or perhaps valuable enough to explore?
MLADEN JOVANOVIĆ: To use jump training for estimating “readiness” or NMF (neuromuscular fatigue), one needs more sensitive tools and methods. Simple jump height will not do it because one can have high NMF and still perform countermovement jumps (CMJ) to the same height. What changes is the way that height is achieved, or the “process” behind the jump. Height is just an outcome. Please note that I am talking about estimating readiness and NMF is not a performance indicator.
I would also tend to agree that the more “reactive” the jump, the more it might tell you about NMF. Using 20kg CMJ might not be sensitive enough to estimate NMF since athletes can “grind” it through.
Speaking of using jump testing—or any other “readiness” estimate to individualize training—it shouldn’t happen acutely. In other words, we are looking for trends and creating a longer-term “prescriptive model” for an athlete based on reaction trends, rather than jumping on “red flags” to adjust daily workouts. So what if your HRV is lower by 10 than yesterday, or your jump power is lower by 5%? We need to rule out normal biological/measurement variability. Even if the real effect is the case, we want to check what caused it and what will keep causing it and will it stay for long. The goal is to modify the long-term approach rather than jumping to every twitch in the data. We also need to develop “robust” (or should I use Taleb’s term antifragile?) athletes who perform even better despite having a bad night sleep episode (new fathers will know what I am talking about!).
FREELAP USA: Submaximal loads are great for estimating repetition maximal abilities, and research is showing evidence that general exercises and lift velocity can predict what one can do if the load is heavier. One worry coaches have is that submaximal loads with maximal effort for velocity is fatiguing. What is the best way to implement one-repetition estimation with submaximal loads?
MLADEN JOVANOVIĆ: Eamonn Flanagan and I have expanded on this question in a recent Journal of Australian Strength and Conditioning article, which I believe is currently the most applied VBT article for coaches available. I suggest reading that article for further details.
Long story short, one needs to know each lifter’s MVT (or minimal velocity threshold, a fancier term than velocity at 1RM) for every lift (or use generalized velocities—they can be pretty stable across different lifting abilities). Bench press tends to be 0.15 m/s (mean velocity) and squat around 0.3 m/s (mean velocity). One can then proceed by performing at least 3 warm-up sets with increasing weights (hopefully covering a range of at least 0.5 m/s) performed with maximal effort. Using simple linear regression, one can estimate weight at MVT. This can be 40%, 60% and 80% or 1RM. This can give one a quick estimation of 1RM (i.e. daily 1RM) that could be tracked over the duration of the training block and used to make adjustments if needed, or to basically see how the athlete is reacting to the training (if the goal is to increase 1RM).
FREELAP USA: Most holistic programs in the weight room and on the field use different strength training modalities, not just one type of lift. Besides alternating intensities and volumes, does bar velocity-type tracking help with better adaptations biologically to the body? Many coaches are looking into hormonal and gene activation as part of the training process. Is this a wrong path or a good idea?
MLADEN JOVANOVIĆ: Two studies have shown that providing feedback via LPT when athletes did CMJ improves both acute power and reliability and chronic increase in power compared to a non-feedback group (PubMed Link1, PubMed Link2). Does this mean using VBT in other movements, such as squat and bench press, will yield more dividends compared to traditional percent-based approach? To be honest, I don’t know. I can only speculate that using VBT might help in more individualized prescription of the loads (especially taking into account day-to-day variability in readiness) which might allow a lifter to lift more when ready and lifting less when not, along with providing more effort in the lift by the sheer existence of real-time feedback.
But training individualization does not only mean providing “relative” load (e.g. using 85% 1RM) or using velocity (i.e. do I get more training effect if I lift in 0.5-0.3 zone or 0.4-0.2?). It also helps to select the right method to yield the highest transfer and adaptation. An interesting study by Beaven, Cook and Gill (PubMed) showed that athletes achieved the highest increase in 1RM when they performed a training protocol that yielded the highest increase in salivary testosterone. This might be sets of 10, or sets of 5. In this case, salivary testosterone serves as a “proxy” in finding the right training protocol.
In my opinion this comes back to figuring out the best prescriptive approach for an athlete based on the data (impulse vs. response modelling) instead of jumping to “science fiction” ad-hoc training modifications based on some daily measurements. I believe that we might find out different proxies to such a relationship, such as hormonal profiling or even genetic profiling. Time will tell, but I do believe that VBT might also be the step to the right direction.
FREELAP USA: Following up on genes and hormones, muscle-fiber profiles of athletes are gaining interest. Could coaches do a better job of individualizing training based on one genetic trait—specifically the amount of fast and slow fiber distribution?
MLADEN JOVANOVIĆ: I am a sceptical empiricist: if it is backed up by research and practice I don’t see why not. One thing to consider is that this ratio might be already expressed in load-velocity and velocity-exertion curves (or any other performance profiling) and might be implemented “implicitly” with some ideas from VBT. Again, individualizing what athletes can do utilizing their relative performance (i.e. %1RM, %MAS, etc.) and current readiness (i.e. velocity zones since they take that into account, HRV, and so forth) might be a step forward. But it might not represent what athletes should do to improve. Hence, even if the FT:ST ratio might be expressed in athlete performance levels (and hence implicitly expressed in relative workloads), it might also necessitate different training protocols, similar to the above scenario with salivary testosterone.
I hope one day we can plug a lot of features into predictive models and generate some inferences regarding different optimal approaches to training depending on the athlete’s qualities. But a lot of confounders might affect predictive power of this model (e.g. sleep, nutrition, and so forth). Until then, we should try to find some proxies that could be used and helpful. And as Bryan Mann said, we need to figure out the middle 70% correctly before worrying about individualization, at least in team sports.
FREELAP USA: The final need of coaches is to make training work better in reducing injuries, improving speed and size of players, and transferring to sporting actions like deceleration and jumping. How does Velocity Based Training do this with athletes?
MLADEN JOVANOVIĆ: You can still break an athlete even if you use the latest technology and gadgets. We are always looking for the magic bullet—but if you allow me to use Taleb as a source again, the best predictor of something sticking around is how long that thing has already been around. Take, for example, cooking devices and cutlery—they have been here since, well, forever, and probably will continue to stick around even if we invent Star Trek’s food replicator. Hence we need to stick to the proven basics—I see VBT only as a tool of doing basics better.
Please share this article so others may benefit.
Velocity Based Training articles available on Complementary Training.