A Powerlifter's Guide to VBT Pt 5: Future Directions for Auto-Regulation Training

[Update 5/9/2017] The manufacturers have responded pretty robustly and directly. Some features I say manufacturers didn't have they actually did have - either on their premium account features or through some other way. Other features were already in the development pipeline.

Most notably, Beast reached out and told me they do provide some velocity-time graph features for the first rep of every set - which is useful for analyzing opening attempts or 1RMs. You can export on Beast - again possibly with a trainer account which I don't have. OpenBarbell also debuted a new export to CSV option. Who even knows that OB is coming out with in their version 3 release, if any changes at all.

I'm not going to edit this as every advance in the technology or science happens. I'm hella lazy. So just keep that in mind. I also apologize for the horrible formatting I did. I've tried to correct that.

[Update 5/17/2017] Both PUSH and OpenBarbell seem to have developed through different avenues. PUSH with seemingly more outreach/education and OB with a new build and promise of an updated app. OB is exporting to CSV after a few hiccups with the initial rollout, and it very clearly stores the information on a linked Google Drive account upon successful export.

INTRODUCTION

There are two avenues I'd like to see developed. The first if the most practically oriented, and that's things manufacturers and practitioners can do to bring the technology forward. The second is possible future directions for the research. For the folks on the front line, the obviously more actionable one is on the manufacturer and practitioner end. Without the research to validate our experiences though, we're essentially working by trial and error. I don't mean to fetishize what the academics are doing, but if we were limited to exactly what the practitioners were doing, VBT would probably be limited to the secret strength and conditioning manual of Missouri Football (jokes). There is some overlap between the manufacturers, practitioners, and the research. This is obvious with folks like Dan Baker working with PUSH, Bryan Mann working with GymAware, and Kabuki teaming up with OpenBarbell.

MANUFACTURERS - THE EASY SOLUTIONS FIRST

This is probably the easiest part to sort out. The private sector tends to be rapid about sorting out solutions fairly quickly. The main issue is holding some of these things back due to novelty and propriety. One of the most interesting recent advances in VBT devices that I expected more out of was Open Barbell being open source. This was different than previous devices that tried to meet the demands of the loudest and highest paying voices in the industry. These were namely strength and conditioning in sports, mostly money sports (which powerlifting is not). My original hope was that Open Barbell would have diverse application development that would run awry, much like the Google App store. This has not been the case. Meanwhile, other manufacturers seem to be porting out features that have no direct application to powerlifting and pertain more to speed and power sports.

So what do I hope to see/expect?

Possible Hardware Configuration Changes or Methods of Measurement

This section is full on nerd. You might not appreciate it, so skip to the next subheading.

This is possibly more pertinent to accelerometer systems, but it would make sense to me that additional sensors could increase the accuracy of accelerometer VBT devices. Form Lifting seems to be beating down the right track by incorporating a barometric altimeter. PUSH, on the other hand, tries to navigate the measurement by using accelerometers and gyroscopes. Using this approach, if you can get a signal on the orientation and the acceleration of the sensor, you can derive velocity data by normalizing to individuals with some simple inputs (height, weight, etc) following a basic template that teaches the data processor how to interpret the information. I haven't had hands-on experience with Form Lifting, but it would seem to me their sensor has an advantage that PUSH doesn't: movements where the rate of change in orientation of the sensor and velocity is very low. This is most pronounced in the deadlift, where the rate of displacement of the sensor is slow and orientation does not change to a significant degree. A sensor like Beast struggles with deadlifts for one common reason that many accelerometer systems do: the jostling of the bar after dropping and vibrations. The difference with Beast is it allows the user to purge repetitions. This solution is easier with faster movements like the squat and straight guesswork with a heavy deadlift.

Accelerometer systems only have so many ways to navigate this issue. One potential fix is to change sensor location. This seems to be an area where Beast has the most versatility, advertising itself as both a magnet mounted unit and a wrist-mounted unit. PUSH also uses this method by relocating the sensor for jumps (waist belt mount) and pull-ups (upper arm). Another sensor, BarSensei, seems to build their units to be more purpose driven, with accelerometer systems that are both bar mounted and medicine ball internal. 

The next solution ties into the following to some extent. One issue with the systems tends to be that they are unable to tell their position relative to the whole system that's moving (the body and the barbell). One possible fix is to give it more inputs specific to the user. For example, if the system understands orientation about the human body and how it accelerates at different angles, it could be pertinent to teach it the exact position of those pesky joints. For example, a system that assumes your trunk is 30% of your height, leg is 23%, and arm is 32% could gain from knowing what the actual lengths are so as to know how they are supposed to accelerate about the y component when functioning as a whole system. Form gets to sidestep that issue by utilizing their altimeter, but other systems have the advantage of additional inputs, like asking users to calibrate the system through measuring body segments. This is in contrast to using generalized proportions. This is most important when anthropometrics can garner an advantage in a sport, like long arms aiding deadlift mechanics.

Now that we're headed down the path of calibration, we should talk about calibration. If anyone remembers early GPS systems, they required you to do a calibration process whereby the user spun the unit across two planes and raised and lower the device. The counter-quip to this is to build a product that's fully calibrated and holds it to begin with. This might be a sound philosophy when it comes to smartphones becoming outdated every year, but I generally expect my gym equipment to last longer than that. If I had to replace my iron every year, I'd rent it through a commercial gym rather than owning a home gym. An ancillary issue is the price and/or supply can fluctuate wildly. In particular, Beast has taken forever to create an American distribution location, the price of PUSH has only increased due to increased features and development, Open Barbell sells out of stock inside of a day, and GymAware cost roughly the same as a cycle of steroids so why not go that route anyways (kidding, USAPL). Long story, short: it might be helpful to have a calibration feature built-in on the operator end.

Load-Velocity Mapping Built In

PUSH essentially has this built in, but it doesn't really make it useful. For whatever reason, you're perfectly able to run a 1RM-VBT estimate through an app feature, but that information just stays in your dashboard being as useless as calf raises for your bench press. Instead, you just get this feature that lets you dictate a really broad range of velocities for a specific training quality (strength, speed-strength, etc). Below is an example of PUSH's built in 1RM estimate test.

I would like to see a feature that allows you to map Load-Velocity. Ideally, I would like to see the app conduct a reverse lookup on a build-up set (as you're trying to find your working weight) and return an approximate estimate of your %1RM. Even better would be the ability to set a %1RM and have the device alert you when you're within acceptable tolerances of the corresponding velocity. With this one feature, half the reason to use templates like I do disappears. Chances are, if it's something I do in MS Excel, there's a really easy way to make it happen in an app.

I don't shame people often, but the industry deserves to be shamed on this fact. Some VBT devices have been on the market for years now. The research and practitioners have been pretty clear on the need for L-V tables to make it individually actionable. Instead, manufacturers have sidelined this simple implementation for others. This is a letdown, especially for manufacturers that have had e1RM features since day one. Failing to capitalize on this relegates some VBT devices to cocktail party tricks that have little to no effect further down the road.

Update 8/28/2017: Even though I've railed against video analysis systems in the past, the Apple app Powerlift actually does this. Furthermore, it has been validated by Carlos Balsalobre-Fernández. There are some limitations, but it's still good in a limited way. To set up your load-velocity profile, you have to know the length of the movement. To get this, I cheated and used the readouts from Open Barbell's ROM measurement (btw: my squat is 666 mm). You can build out your load-velocity profile and then measure the velocity of individual reps. If I were using it in real time, I would probably just ramp up to 80% actual 1RM, measure the velocity, and use the estimate + basic algebra to figure out the day's 1RM and the working weight for the day. This might not be optimal given recent findings, but I could see this working it's way into the app shortly.

Also, I wouldn't get hung up on the fact that Powerlift's velocity differ from another device if you use one. What's important is reliability, not so much accuracy. The key part is to try to conduct your 1RM predictions from the same camera angle, height, and distance as your load-velocity mapping session.

Load-Exertion Mapping Built In

Likewise, a Load-Exertion (L-E) or built in rep to failure (RtF) table would be just as useful. While I tend to follow velocity loss, there are practical limitations to this simplistic approach. A built in L-E table would give you a good indicator to follow in real time, alerting you to when you're 1-3 reps away from failure. A L-E table establishes what your last rep in the tank, 1-3 reps left in the tank, and so on. Rather than working based on velocity loss, which gives us a very rough guideline, we can dial it into alerts based on how far away from failure we want to be: 1-4 reps depending on the training goal. Granted, RPE ratings essentially do this for us, but this feature alone could increase the cross talk between RPE and VBT. This provides a very clear objective read on reps in reserve that can be used in real-time for VBT users or to validate RPE's assertions on their ratings.


To make L-E feature built in, all that would be required would be RtF sets at two intensities, like 75% and 85%. Those RtF sets could also feed forward into e1RM tests to establish MVT and a more accurate estimate of 1RM.


How about this for a really easy feature: any time you set a new record low velocity, ask if the user wants to designate that as their MVT. Make it something that can opted in or out of, just in case it wasn't a full rep, it didn't meet the criteria for a technically proficient lift, or was purposely moved slow according to tempo prescription.

The "Velocity Zones" Sucks

This one thing is the most important factor IMO to make VBT data fluid and actionable.

Firstly, the velocity zones for are very dependent on the exercise and athlete, depending on the training outcome. You could probably find ideal zones for different exercises for different training qualities for different types (most specifically heights) of athletes. This sounds very time-consuming though and doesn't really help the user. KISS: keep it simple, stupid. That, or I guess you can exploit the fact you have a user database with self-selected data, do a database analysis controlling for height and exercise classification, and multiply it by the squared cosine of making a simple solution harder than it needs to be.

This one thing is the most important factor IMO to make VBT data fluid and actionable.

Granted, there is a way to do this through program creation with some apps and webportals, if you can get over the fact that every system that has a program creation guide as rigid as frozen dog poop. Here's an idea: let me designate the velocity zone myself while using the app. Specifically, let me choose a velocity, plus or minus a given percentage of that velocity for an initial target, and an acceptable stopping point (either a specific velocity or a certain percentage of velocity loss).

Surprisingly, Beast for all it's
lackluster does have this simple
feature. Not exactly on the mark, but
why is this not common to all?

It seems pointless, but when you consider that rows and pull ups have a high MVT, according to the velocity zones, you can hit a 1RM without ever really working within the "absolute strength" zone. If we had this one feature, you wouldn't have to see the screen for real-time feedback. This one thing is the most important factor IMO to make VBT data fluid and actionable. I'll keep saying that in hopes that some manufacturers pick up on the point that this needs to be a basic feature.

Athlete Monitoring - Daily Readiness Testing

Take for example a traditional Sheiko program. Something that has you squat for a certain number of reps, gives you a brief reprieve by benching, then has you do an assistance version of the squat. If you have a very, very clear training priority in the day (squat, in this example), you want to know how beat up you are. One of the more popular ways of checking your daily readiness is by doing a squat at a standardized load. I usually use 60% of my last official 1RM (not my daily 1RM). I call this my "cold start" since I do it first without any warm up or practice reps (maybe air squats or some body weight variant). If that weight moves slower than my L-V tabled velocity, I know I'm carrying a bit of fatigue around. This doesn't stop me from training, but it prepares me for what to expect. If it moves faster, I can probably stack more plates on the bar that day.

Not terribly recently, Dan Baker suggested a standardized load of 80% 1RM IRC. The smallest worthwhile difference in determining deviation from readiness was 0.04 m/s difference from baseline. If you can move that faster, you're stronger (at least that point in time). More than 0.04 m/s difference in terms of drop in velocity means fatigue is inhibiting your readiness. This seems to come from his coaching experience and possibly data from his previous studies (the Oceania strength and conditioning community is awesome like that). YMMV

With enough measurements, you should have a baseline reading (think of it like a rolling average). With that baseline, you'll also have a gauge of what is abnormally above or below your baseline. This is the same kind of principle that Heart Rate Variability uses on a global level, but this gives you a direct read on local neuromuscular readiness. Again, this is something you can do through Excel using something like basic Z-scores or percent difference from baseline. Here's a fictional example:

Standard traffic light system. Green means increase training
load, yellow means normal training load, red means decrease
training load. Arbitrarily set at a difference of 10%

The hardest part of athlete monitoring is finding a way to make it unobtrusive and have additive value to the training process. It's already a bother to make an excel sheet to accompany an app to accompany a device to accompany a barbell. How is this not a thing? Manufacturers that want to sell their product like it pertains athlete monitoring should put something in the fine print that says, "but you figure that part out on your own, hoss." 

Graphs Against Time

Given the amount of samples most of these devices collect per second, it would be really helpful and plausible to have velocity-time graphs or displacement-times graphs much like what is shown here:

These seem minor, but in powerlifting it really helps validate struggling areas of your lift. For example, a change in the slope of displacement over time can show you the slow parts of your lift. That slow part of your lift is likely going to be a weak point or sticking point. Depending on the location and your development, you can overcome that sticking point by "strengthening the zone" by focusing work directly in that weak range of motion. Another way to breaking past it is by developing acceleration before that point so more acceleration can be developed to "power through" that sticking point through inertia. 

Granted, this is something you should see during the course of training but gives you a ruler to measure it by. Objective feedback is one of the selling points of VBT. There are likely some practical limitations here like the throughput of information that can be sustained through Bluetooth, but I see this information be utilized more after the fact during set review. This does not need to be a real-time feature, making this still a plausible feature.

Added Value: Tandem Video Recording with Data Overlay for Coaches

I have to admit this is something GymAware already does to the best of my understanding, along with graph overlays. It seems like online coaching with set review is fairly common. Advances in the internet and mobile devices have facilitated the process. I'm all for it, within reason. For coaches that see added value in the velocity data, this helps centralize all the important information by giving the mean velocity of each lift as the lifts are performed. This would be especially helpful for 1RM attempts, RtF sets, and L-V mapping sessions. I highlight those because those three things could throw off the whole accuracy of V-%1RM plots and potentially over/underload future training sessions. Below is an example of video synced with mean power from Dan Baker:

Even on the individual level, this allows the athlete to understand potential causes of velocity inconsistency, such as intervals in which they brace, range of motion contributions to velocity, the magnitude of simple form breakdowns, etc. Unless GymAware has copyrighted this specific tool, it seems easy to implement. This is something people commonly do through two different applications and the feature would merge the process and consolidate the information.

Cut the Fat: Microcycle Management for Coaches

No, I don't think this being able to create and push training plans to athletes is worth a damn. It seems like everyone except for the barebones manufacturers want to provide this. However, it's limiting to the point of being useless. A good example is one manufacturer automatically prompting you to follow the training plan as soon as it starts. Needed a build-up set to figure out what load is appropriate for your working sets? Too bad, that single plate warm-up just counted as one of your 6 working sets. There are plenty of scenarios where this level of inflexible management just performs poorly. Any workout program that's so rigid that it doesn't allow you to take a dump mid-workout to prevent pooping your pants during your squats is a bad workout program - I don't care what it adds to your total.

One example of program creation from Beast

Part of the whole point of auto-regulation is that you're not following a play by play script. These program features attempt to constrain the training process that is meant to be boundless. So I'll make it simple: you don't need to remove this feature, but there are zero reasons to buff it out. If there are recreational users that think it's useful, freeze it's development because it's a waste of time and resources. 

The market is already glutted with program creation and distribution products without VBT. No one's buying the product strictly for this point alone. Recognize it for what it is: fluff. I rate this feature a flat out zero fucks given. Here's a shorthand program that will get me through a session with VBT:

Comp Squat: 6+/2@0.42 m/s, stop @ 0.32 m/s

Close Grip Bench:  @0.54, stop @ 0.30 m/s

Pin Squat: @ 0.6 m/s, stop @ 0.33 m/s

(4x10 follows)

Ab rollout, shoulder external rotation, pull ups

That alone gives me 90% of what I need to know and the other 10% I want isn't supported by any program management tool I've seen.

Air Cast and Mirracast

[Edit 8/8/2017: If it works with Apple, you can cast it. Just make sure you change your screen time out. That said, this actually hasn't been super helpful now that I've taken it for a test drive. I much rather prefer to have some sort of mounting system. Especially when it comes to bench press and I'm staring straight up.]

This is not really a feature that needs direct support. As far as I know, you can do this on both systems, possibly with a little bit of ingenuity. Screen-casting (Air Casting on Apple and Mirracast on Android) solves the simple problem of using the information in real-time. This might seem ridiculous, but it is quite an obstacle to finding a clear way to suspend your iPod in plain view during your squats and bench presses. Granted, a Perchmount makes this easier, but the only way I've gotten this to work with squats has been good ole 2x4" ingenuity. This is probably limited to users that lift in home gyms, but it's something to consider for people that struggle with using real-time feedback to regulate their sets. 

This whole casting idea means less if manufacturers make initial velocity and end velocity prescription more accessible, but there is some additive value of buffing that up with something like delay video casting with metrics overlay on the video. Video demonstrates a simplistic system:

Data Export (Plus Import) and Cloud Based Syncing

This is something some manufacturers already provide. PUSH exports to CSV, which functions the same way as Excel. This seems like a minor thing, but depending on what you're using your device for it can be annoying to manually transcribe information. PUSH is unique that it gives the extraneous details, like time spent concentric/eccentric, force, and other things that aren't available in the user interface. As far as I can tell, Beast doesn't do this. OpenBarbell just made an update that uses your Google login to sync, but as far as I can't tell it's not syncing anywhere where a user can retrieve it and the only viewable data is stored locally - a problem across multiple devices.

Export as shown on PUSH - probably the most underrated feature

The most interesting idea would be removing long-term data analysis from the realm of manufacturers' web portals and allowing hand off to others. There are plenty of platforms that currently do this, like AthleteMonitoring.com and Fusion Sport. Some of these platforms have opened avenues of cross integration to allow automatic syncing (rather than manual inputs) of pertinent athlete monitoring metrics, like heart rate variability. Reactive Training Systems has developed a system that performs according to the metrics Mike T thinks are important, like resting heart rate and subjective questionnaires. It's helpful to note that this service is free to use regardless of whether you work with RTS or not. It does simplify the process. This saves you the trouble of developing your own data collection system and figuring out the analysis yourself. The problem with TRAC is it's developed with RPE in mind. The only other game in town that's powerlifting specific is MyStrengthBook, which is more training analysis than it is athlete monitoring. Regardless, neither currently seem to think of training in VBT terms. If you buffed out your VBT process to include multiple RtF sets alongside referencing velocity to %1RM, you could potentially create a VBT to RIR translator, making TRAC a usable training log.

The truth of the matter is there is no good VBT training tracker for powerlifting. This is to include among the manufacturers of VBT devices. Manufacturers need to abandon the idea of managing their own data when there are plenty of others whose sole attention is managing the training that can do the job better. The first training analysis to take VBT seriously and integrate or import velocity data and provide effective long-term analysis is probably going to a get a lion's share of the VBT consumer's market. The flip-side of that market is it's likely smaller than the RPE market. It would seem to me the easiest way forward would be to exploit the features of MSB/TRAC and translate VBT data into RIR as a proxy for RPE, then release the extensive data under the handle twitter handle Guccifer to satisfy our Russian periodization overlords.

Here's a walkthrough for data export on GymAware's portal:

And finally here's a video demonstrating how you export data on Beast and Open Barbell:

[update: since the original post, Beast and
Open Barbell has added this feature. #TomHanksVBT_Hero]

RESEARCH AVENUES

This isn't nearly as easy. It's easy to find the shortcomings of manufacturer's for current consumer requirements. It's easy to list out the things we don't know. It's hard to list out the things we don't know that we don't know. One of those Donald Rumsfeld known knowns, known unknowns, and unknown unknowns things.

So here are the obvious things we can talk about that we need to pursue on the research side of the house. Most of this is from the interest of strength training and powerlifting, not power development which seems to be better covered in the literature. For the sake of simplicity, this is the short list of things I'm selfishly interested in seeing studied for current needs in powerlifting and strength training.

1) Inter-individual Differences: determine velocity characteristics and differences across heights, limb extremities, anthropometrics, gender, etc

This one seems obvious. Taller folks or people with longer extremities are moving over a longer range of motion. This probably has some influence on their maximum and minimum velocity under load. That part seems easy, but the more complicated part would be examining the effect on sticking points during common movements. 

It'd be interesting to see how well our understanding of the literature holds up for training women and lighter class lifters in terms of velocity. It seems to be internet slapstick to cite that women and lighter lifters have differences in training load and reps at a given %1RM. Given the granular detail VBT provides, it could be helpful to see if we can characterize these training differences in velocity terms (velocity loss, mean velocity, MVT, etc).

2) Intra-Individual Differences: changes in velocity over time from different training modalities

We kind of have a working idea of this through the force-velocity curve, but this doesn't typify the differences in training that any practitioner has seen.

Credit to Haff and Nimphius

Everyone that's worked with a new athlete while running VBT (even if only for observation and not auto-regulation) have typically seen changes in MVT across time, usually corresponding with a change in strength (1RM). I've also seen this difference in speed trained athletes that tend to have higher MVT's than traditionally strength trained athletes. There's probably some interest in finding some middle of the road types of folks, whether that means a type of athlete or a type of training. In that selfish powerlifter way, I wonder how a conjugate method F-V curve compares to other powerlifting training modalities.

Furthermore, a mechanistic view on MVT would be interesting to see. It would seem there are many factors that could contribute to MVT dropping over time, such as changes in form, bar path, and other kinematics, but part of me wonders what the absolute floor to MVT is and whether there is a way we can predict the bottom end of minimum velocity threshold or the absolute minimum that minimum velocity threshold can move. There's probably some energetic or fiber type contribution that would also aid in explaining that, but I feel like that's a far reach and would not significantly impact how I use VBT in training.

3) Characterize particular movements: DEADLIFTS, rows, overhead presses, pull-ups, and variants of the primary lifts

[Edit 8/8/2017 - The Spaniards covered 4 different bench press types: regular bench, bench throws, bench press from pins, and bench throw from pins. They appear to be doing another article on the subject.]

[Edit 8/28/2017 - This has been fleshed out even more with different squat variations, plenty of stuff on pull-ups, and prone rows. The general trend seems to be that when the eccentric and concentric part of the movement is separated, there's increased accuracy and reliability. For many of these things, it's just a matter of the research catching up to what the practitioners know to be true. There have been circumstances where it appears researchers have really challenged our understanding of VBT.]

First of all, we need to recognize that deadlifts quantifiably behave differently. Deadlifts are noticeably absent from the research when it comes to VBT strength training. I get it. You can't publish an article with no real results to show for. It might be more productive to make comparisons across multiple exercises, both in the context of low-moderate intensity for power development and high-intensity for strength development.

Rows and pull-ups (for devices and methods that support body weight movements) are another confounder. When I did measure pull ups, one thing I found was that the velocity floor seemed higher (MVT around 0.42 m/s, far enough within accelerometer performance limits) for rows than it was other exercises. This seemed to be the case for seal rows without resetting to the floor, Yate's rows, and Pendlay rows. Normally I would contribute this to my lack of emphasis on upper back strength (I know I'm not only PL that seemingly negates upper back in favor of comp lifts), but according to folks like Dan Baker, this tends to be the case with his Rugby players that spend a lot of emphasis doing bench pulls.

I can tell you bench, squat, and deadlift general MVTs. I can't tell you overhead press MVT's and only have a rough understanding of velocity behavior across a set. In particular, the first rep and any "broken" reps (short stops at the bottom on successive reps) tend to be lower velocity. I generally don't care about this too much, but it needs to be said since there's some carry over to WL.

A hypothetical graph of %RM-V behaviors

Having a general understanding of how to treat assistance movements would really helpful. For example, if you knew that close grip bench press and Spoto presses were only mildly different from regular bench press, it would be helpful to characterize the magnitude of difference in the same way we generally say bench/deadlift and squat MVT's are 0.15 m/s and 0.30 m/s. It would seem to me that the main difference in the %RM-Velocity plot could be the slope of the line, the intercept, and the MVT. A picture of this would be appropriate and is shown on the right. This is a simplistic explanation, and it's likely certain factors tends to change different parts of the relationship.

This is generally what I find for me and what I've found with a few clients when using VBT as an observational metric, not an auto-regulation metric.

[Edit 8/28/2017 - Actual graphs of variants of the bench press, pin press, bench throw, and bench throw from pins.]

[Edit 9/5/2017: The same graph as above recreated with overlapping trend lines using Desmos online graphing calculator]

4) How rep count schemes and rep tempo modifiers affect concentric velocity

This is somewhat covered in the non-VBT scheme of research. I mentioned in the past the idea of using cluster sets to maintain velocity at high intensity and increase set volume. Myo-reps are another method which acts roughly the same way but is used to increase volume for hypertrophy/work-capacity appropriate loads. The mechanistic approach is helpful, but a good question would be whether this training modality is meaningful.

One interesting development in VBT was a study that focused on a VBT program vs an eccentric VBT program. Spoiler alert, there were no significant differences in 1RM improvements in the squat or bench press when comparing training modality. The Spanish VBT proponents have also played with rep modifiers, finding that imposing a pause between the eccentric and concentric improved reliability (something I have found useful). Anecdotally, I'm wondering if this utility can be expanded to include other tempo schemes. VBT has one added benefit to traditional methods of controlling tempo that it can give instant feedback as to whether the stated goals of the protocol are being met, and thereby what the practical limitations are.

Until we have a clear understanding of these things, we're shooting in the dark or we're going to need other methods to pick up the slack (like RPE).

[Edit 8/28/2017: We're getting a little of that now. This shows rest time equated cluster sets and traditional sets.]

5) Comparison showing the obvious superiority of VBT to that of traditional percent based training and rating of perceived exertion

Not necessary, but if anyone wants to beat down that path they're welcome to it. Different strokes for different folks.

CONCLUSION

Most of the things that manufacturers can do to bring the technology and practice forward are very simple. All of the concepts are not new, and many pre-date many of the VBT devices on the market. Bringing VBT to bear in use in the general public necessitates this process. Simple changes like %1RM-Velocity mapping, Load-Exertion mapping, and others would streamline the process and help athlete management. Many of the features that have been fleshed out apply to those that work with "velocity zones" which are too wide to manage the fine processes of strength training. And when it comes to powerlifting we are working with a narrower range of the velocity spectrum than power athletes. This is not because VBT does not lend itself to strength athletics, but because VBT manufacturers have not made their products nearly as accessible to strength sports.

It is my opinion that manufacturers should focus on how to manage the training session, not try to fill the role of presenting ways coaches (or self-coached individuals) should attempt to manage training across time. Rather than attempting to do two jobs poorly, they should focus on auto-regulation mechanisms internal to the system (via the app) and outsource mesocycle data management to those that have the potential to focus on it better. I've linked to multiple places, but for powerlifting specifically, this is best handled through MyStrengthBook or TRAC. 

The science still has a few questions to answer. Most of these relate to how we understand velocity as a marker of intensity, exertion, and fatigue across more movements and variations of movements. Giving us an understanding how this varies across different individuals or across an individual's development would better help us understand VBT beyond specific practitioners' experiences.