You are not the same person you were yesterday. A fixed plan does not know that: it hands you the same five sets whether you arrive fresh or wrecked. Training by performance loss puts an objective brake on the session: you measure something at the start, measure it again as you train, and when the drop crosses the limit your coach set, the app suggests you stop.
It is not a cutoff that kicks you out of the gym: it is a suggestion with data behind it. You are the one who stops.
Almost everyone trains between two mistakes. The first is carrying on when you should not: the last sets come out slow, ugly and with broken technique, and what is left is not training, it is fatigue that will eat the session two days from now. The second is stopping too early: you feel off, you decide today is not the day, and you leave on the table a session that was actually there.
Both come from the same place: you are deciding by feel. Feel is information, but it is a poor judge of acute fatigue. A number measured twice, on the other hand, does not argue.
Sets that no longer add adaptation but do add fatigue and risk.
Good days wasted because of a bad feeling at the start.
With no record of why you stopped, your coach cannot adjust anything.
The mechanism is the same whatever you measure it with: you need a reference and a measurement from today. The distance between the two is your loss.
The effort you compare yourself against. It can be from today (the first good attempt of the session) or from another day (your last test). If the reference is a max, the app uses your real max for that exercise, never the percentage you are training at: 85% is not your 100%.
Your coach writes how much deviation they accept and in which direction: "peak force drops 15%", "minimum SmO2 rises 20%". No formulas: a percentage and a direction.
You repeat the measurement during the session. When you cross the limit, the app tells you right there, without leaving the workout, and you choose: stop or keep going. What you chose is recorded.
The rule is the same for any measurement. The sensor changes, the concept does not.
With a dynamometer (Tindeq, Climbro): peak force, mean force, time to failure. The classic in climbing and finger work.
With a MOXY: minimum SmO2, mean recovery. How much the muscle desaturates and how long it takes to come back. Moxy guide.
With no sensor at all: heaviest weight moved, reps completed, sets. Read straight from the session log.
No numbers: "if you fail on the third set or earlier, stop". Works for any exercise and needs no device.
No sensors? You can enter the value by hand. The app marks whether the number was measured by you or pulled from the log, so your coach knows what they are looking at.
A loss rule answers one of these two questions. Both can be used at once on the same exercise.
You measure your effort on arrival and compare it with another day. If you have dropped more than your coach accepts, today’s session probably should not be the one the plan says. Your coach can mark it as mandatory: then the start button waits until you do it, or until you skip it leaving a reason.
During the session you repeat the measurement between sets. When the drop crosses the limit, the alert fires. Nothing is blocked here: it is your workout and you decide, with the number in front of you instead of by eye.
No formulas to configure. The coach picks what is measured, how much is accepted and how several conditions combine, and the rule reads as one sentence.
On unilateral exercises (two sensors, one per arm or per leg), each side is evaluated separately. The left may call for a stop while the right holds up, and that is exactly what you will see: which side gave out, not an average that hides both.
It is not a medical traffic light. It is decision support, not a diagnosis nor a guarantee that you will not get injured.
It does not decide for you. You can always keep going. All the app does is put on record what happened and what you chose.
It does not replace your coach. The limits are set by whoever programs. The app applies them and reports back what happened.
It cannot fix a bad measurement. A reference taken sloppily gives you a loss that means nothing. Measure under the same conditions.
The drop of a measurement against a reference. If your reference peak force is 50 kg and today you hit 44, you have lost 12%. Training by loss means deciding in advance how much drop you accept before changing what you were doing.
No. With a dynamometer or a MOXY it happens automatically, but you can type the value in by hand, or use rules that measure nothing at all: "if you fail on set 3 or earlier, stop".
It is the same idea, generalised. VBT uses bar velocity as the measurement. Here you choose the measurement: velocity, force, oxygenation, load or even the point at which you fail.
Against your max, whenever the measurement is a max-type one. If you are training at 85%, that 85% is not the reference: the reference is your real max for the exercise. Comparing against the prescribed intensity would give inflated or non-existent losses depending on the day.
The app asks for it and lets you work it out without an all-out test: a set of 3 or 5 reps close to failure is enough to estimate it, with the RM calculator.
Only if your coach marks the opening check as mandatory, and even then you can skip it by leaving a reason. The suggestion to stop never blocks: it warns.
Every measurement, how far it deviated from the reference, whether it crossed the limit and what you did next. Including when you skipped the check and why.
Set your loss limits per exercise and train with them in front of you, with your sensors or without them.