Meta-analysis — 2022 — RPE & autoregulation
Accuracy in Predicting Repetitions to Task Failure in Resistance Exercise: A Scoping Review and Exploratory Meta-analysis
What did the researchers investigate?
How accurately people predict the number of reps they have left before failure, across 12 studies with 414 participants.
What did they find?
People underpredicted reps to failure by about 0.95 reps on average, with large differences between studies. Predictions improved closer to failure, with heavier loads (12 reps or fewer) and in later sets. Training experience did not change accuracy.
What does this mean for training?
'Reps left' is a useful but imperfect gauge. It is most accurate when sets are hard and reps are moderate, and it's normal to be off by about a rep.
How does FitX AI use this evidence?
FitX AI uses reps-left as feedback, not as gospel. Week 1 at RPE 7 is a calibration week, and each next target follows what you actually logged.
Where FitX AI's own choices go beyond this paper
Reps-left feedback. Reps-left is self-reported; accuracy varies between people, so one rating moves targets only a little.
RPE mesocycle. The 7, 8, 9, 10 ladder and the 5-week length are FitX AI design choices. The papers support training near failure and RPE/RIR as effort scales, not this exact schedule.
Exercise history and targets from it. Pooling sessions within six weeks of the newest, keeping 95% of the estimate after 29-90 days without the exercise and 90% after longer, and recalibrating after 90 days or with only one rated session are FitX AI choices. Strength is lost gradually with time off and by different amounts for different people; the library has no paper that sets these numbers. Sets without an effort rating are read as to failure (the most conservative reading); an effort is never filled in.
Limitations
Exploratory meta-analysis with very high heterogeneity between studies.