Why bar speed tells you more than rep count

You and your friend load the same weight on the bar. You hit 6 reps. They hit 9. They get a smug look. You start doubting your whole workout.

Here is what actually matters. Your 6 reps could have been a hard, honest effort at the edge of what you can do, while your friend stopped at 9 with plenty left in the tank and was moving the bar easily the whole time. More reps doesn't mean more strength, and fewer reps doesn't mean less. The rep count can't tell any of that apart.

Same reps. Different story.

Here are two lifters doing the exact same thing on paper: 5 reps at the same weight. Rep counter: "5 and 5, same thing." Now look at how fast the bar actually moved each rep.

Lifter A has plenty left. Lifter B just spent most of their tank, and that last rep cost far more than it looks. Same count. Completely different training stimulus, different recovery bill, different next session.

This isn't a hunch. It's measured.

This isn't a Konshu invention. Velocity-based training is an established field in strength research. A well-known 2010 study by González-Badillo and Sánchez-Medina found that bar speed at a given load tracks how hard you're relative to your max with a ridiculously tight fit (r² of 0.98). Translation: speed is a real, measurable window into effort.

And it keeps going as the set goes on. When the bar slows down rep over rep, that's your fatigue showing up in plain sight, whether you admit it or not. Bar speed doesn't lie. Your memory of a set can. So can a spreadsheet you filled in afterwards.

The metrics tier list

How we rank what's worth tracking.

Velocity loss: the receipts

Take one set of 8 back squats. Watch the speed fall off. By the end, the bar is moving about 59% slower than rep 1. The last two bars flip olive, because that's the zone where the set stops being productive training and starts being pure fatigue.

Try it yourself

Drag the slider through the set. Watch the numbers change. Watch the vibe change.

Sample set · illustrative readout.

So what does Konshu do about it?

Konshu reads the motion sensor in the fitness tracker you already wear, whether that's a watch, a band or a ring. That's the whole setup: nothing new to buy, nothing to strap on. From each set it tells you what actually happened: the reps counted, the lift named, which muscle took it, how close you came to failure, how fast the bar actually moved, and what to load next time. Support rolls out tracker by tracker.

We eliminate the logging, not the input. You'll still tell us how a set felt. You just won't be writing numbers into a notes app between sets.

And what we don't do: predict your one-rep max, or pretend a single sensor can judge your form. We'd rather be right than loud.