Velocity-Based Training: Device Testing and Review
Velocity Based Training – or VBT – has slowly been gaining popularity as athletes and coaches look for the most cutting edge methods to train their athletes. With VBT, instead of training for maximal strength, athletes train for maximal speed. In theory, this trains more fast-twitch muscle fibers which translates more directly to the athlete’s sport. We have begun implementing VBT with some of our more advanced athletes, and you can read more about it in our VBT articles: Part 1 and Part 2.
But how do we measure velocity? Sure it’s easy enough to just “feel it out” and do some lighter weights as fast as you can, but how do you know how fast you’re actually moving the bar? Well, there’s the Tendo Unit, which is the gold-standard, but retails for $1600. VBT guru Bryan Mann swears by the GymAware, but that costs $2200.
Fortunately, a wave of much cheaper IMU based sensors are being released, namely the PUSH Band, the Beast Sensor, and the FORM Lifting Collar. Instead of using a cable to measure velocity, these sensors use a series of gyroscopes and accelerometers to determine barspeed, rather than a string attached to the barbell like the Tendo and GymAware. Most importantly, they’re all priced in the $249-$289 range, which offers a much cheaper alternative to the wired units.
But do they work?
We took it upon ourselves to carry out a small in-house study comparing these 3 sensors to the gold-standard of VBT, the Tendo. Our primary measurement was average velocity, which is the primary measurement used in VBT. Each sensor was compared individually to the Tendo, and 20 athletes performed 10 reps of bench press each. We looked at the average absolute difference between the Tendo and the sensor for all the reps, as well as the mean absolute percent error, or MAPE (MAPE is a measure of accuracy, lower percentage is better), in order to compare the sensors.
Let’s not waste any time, here’s the results:
Just by looking at the table, the obvious winner is the PUSH Band. It had the lowest average difference, and the lowest MAPE. Purely based on accuracy, it’s the obvious choice.
It is worth noting that although the overall difference of the sensors is pretty good, there are some trends to point out. The PUSH Band consistently read both negative and positive results, so you wouldn’t necessarily know if any particular rep was slower or faster, but the overall average for the workout was pretty accurate. The Beast Sensor, on the other hand, read almost always slower than the Tendo. Your actual speed would more often than not be 0.2 m/s faster than what the sensor recorded. FORM was similar to the PUSH in that it read both negative and positive results.
But there are a few caveats that I’d like to address.
First of all, yes, the FORM Collar looks pretty bad. There were, however, multiple misreads. For example, I doubt that one of our athletes did a rep at -3.86 m/s. When we eliminate all of these outliers, we wind up with a much more reasonable result.
They still lag a bit behind the PUSH and Beast, but the results are much more competitive. Standard outlier calculations were performed to create this table.
I’d also like to briefly talk about peak velocity, as some methods of training use it as the recorded metric. FORM does not measure peak velocity, so only Beast and PUSH were compared. Here are the results:
Both the average differences and the MAPE were identical to those for average velocity, with the PUSH slightly edging out the Beast Sensor. Beyond peak velocity, power calculations were rather inconsistent in all 3 sensors, and due to its limited use in the VBT training world, I have omitted it.
The final thing I want to discuss is the usability of the different sensors. All three of them are Bluetooth enabled, use a phone app, and were pretty easy to set up and use. PUSH and Beast both have a web portal in addition to the app, while FORM only exists in the app. All 3 apps are pretty easy and intuitive to use, although going back and looking at past sessions was a bit confusing in all of the apps. The portals for PUSH and Beast were much better for looking at past training sessions, offering rep-by-rep tables and all the data you could possibly want. The PUSH Band in particular had a very nice and user-friendly portal (although it costs a small monthly service charge to access it).
Wrapping it Up
The disparity between the Tendo and the IMU-based sensors seems to come from their method of measurement. Maybe the discrepancies are due to the fact that PUSH and Beast are placed on the users forearm/ wrist, and not the bar. Maybe the athlete bounced the bar off their chest and the sensors were prone to jostled. My guess, however, is that the IMU technology is just not yet accurate enough to compete 1-on-1 with the Tendo or other similar VBT devices. That’s not to say these devices will never compete, and I’m sure technology will continue to advance and these types of sensors will get more accurate.
So if I had to pick one, I’d pick either the Tendo or the GymAware. Ok, if I didn’t have $2000 I’d probably pick the PUSH Band. It’s not 100% accurate, but it’s pretty close, and definitely gives you a good idea of how fast the bar is moving. It’s easy to use, and the app and the portal were probably the best out of the bunch. Just make sure to account for the fact that velocity readings may not be 100% accurate.
This article was written by Joseph Marsh, Lead Product Engineer at Driveline Baseball.
Ricky Norton -
How about Bar Sensai? Should add that one to the list also.
Garrett Boyum -
I would also add OpenBarbell to the mix although they are hard to get a hold of… Their price is $250 which comparable to Push and Beast. But OpenBarbell is a drawstring based unit similar to a Tendo with Bluetooth app capabilities.