Rob's squat programme called for five sets of five at 140kg on a Tuesday in March — the same weight he'd hit cleanly for three weeks running, on paper an easy, unremarkable session. He'd been following the same twelve-week block for two months, adding two and a half kilos most weeks without much drama, the kind of steady progress that makes a programme feel bulletproof. His daughter had spiked a fever overnight, he'd slept maybe four broken hours, and he'd skipped breakfast rushing out the door. By the third rep of the first set the bar felt like it was loaded with wet concrete, but the sheet said 140kg, so he ground through all five sets anyway. He noticed how badly the bar was moving, said nothing to his training partner, and told himself the next set would feel better once he was properly warmed up. Two days later his lower back locked up so badly he couldn't tie his own shoes.
Nothing about Rob's programme was wrong. The weight was appropriate for where his strength stood that month, the volume was sensible, and a coach looking at the spreadsheet would have signed off on it without hesitation. What the spreadsheet couldn't see was Tuesday itself — the sleep debt, the missed meal, the low-grade stress of a sick kid at home. A number on a page can't adjust for any of that. You can, if you're willing to rate what a set actually felt like instead of just ticking it off.
What RPE actually measures
RPE stands for Rate of Perceived Exertion, and in strength training it's usually paired with its mirror image, RIR — Reps in Reserve. The scale runs from 1 to 10, and in practice almost nobody uses the bottom half of it; a working set worth rating sits somewhere between 6 and 10. Mike Tuchscherer popularised the modern version of this scale through Reactive Training Systems in raw powerlifting back in the mid-2000s, and it has since become the standard language coaches use to talk about effort without needing to see the athlete lift in person.
- RPE 10 — the bar stops. No more reps are coming, not with better technique, not with a louder training partner shouting.
- RPE 9 — one more rep exists, and it would be ugly. Bar speed on the last completed rep noticeably slower than the first.
- RPE 8 — two reps left in the tank. This is where most productive working sets should live, week after week.
- RPE 7 and below is a warm-up weight wearing a work-set costume, and everyone in the room knows it.
Below RPE 6 nobody really bothers rating a set — it's obviously easy, and rating it just wastes a line in the log.
The percentage trap
Your programme doesn't know you slept four hours.
Fixed-percentage programming assumes you turn up recovered by roughly the same amount every session, and for a huge chunk of a training block that assumption holds up fine. It stops holding up the moment life intrudes — a bad night, a stressful week at work, a knee that's been quietly barking since Sunday's five-a-side. A percentage-based programme is a genuinely good starting map for a lifter's first year, because it forces consistency before feel is trustworthy. Past that first year, treating every Tuesday as an identical, interchangeable unit is exactly how a manageable niggle turns into an MRI referral.
None of this means percentages are useless once you've got some training age behind you — they're still the backbone of a sane programme, and RPE works best layered on top of them rather than replacing them outright. Most competent powerlifting coaches run a hybrid of the two in practice — a percentage range set for the block with an RPE cap layered over it, so nobody ends up chasing a printed number past what that particular day's body can actually produce. But it's worth being honest about where autoregulation genuinely helps and where it doesn't. For a lifter in their first twelve months, RPE is close to worthless, because new lifters chronically misjudge how close to failure they actually are. Ask a beginner to rate a set RPE 7, then have them attempt a true one-rep max the same afternoon, and the number that shows up almost never matches what they predicted going in. Fixed percentages are the safer tool until a lifter has enough honest reps under the bar to recognise what real strain feels like versus what merely feels hard.
Learning to rate a set honestly
The single best way to calibrate is bar speed, not feeling. Film your top set on your phone once a week — not every set, just the heaviest one — and watch how the rep looks compared to the one before it. If the bar visibly slows between rep 3 and rep 4 on a set of 5, that's an RPE 8 or 9 whatever your gut told you walking away from the rack. Log the number in your phone's notes app straight after racking the bar, not from memory an hour later; memory flatters effort every single time, and by the drive home a genuinely brutal set has quietly become "fine, actually."
Adjusting the session without rewriting the programme
Here's where RPE earns its keep. Say the sheet calls for 100kg for five reps, five sets, and the plan is built around that number holding steady for the block. If the bar is already grinding at RPE 9.5 by the third rep of the first set, don't grit your teeth through four more sets at that weight chasing a number that isn't real for you today. Drop to 90kg, hit all your reps clean at a genuine RPE 7 to 8, and you've preserved almost all of the training effect without turning a routine Tuesday into the rep that wrecks your week. The following session, if sleep and stress have settled, the 100kg is usually still there waiting — sometimes it moves up, because you didn't dig a hole you had to climb out of first.
This is where a lot of men get twitchy, because dropping the number on paper feels like admitting weakness. It isn't. Missing three of five reps at a weight your body can't currently produce is the actual failure; adjusting down and completing the work as prescribed is the programme functioning exactly as intended.
The two ways men get this wrong
There are two opposite failure modes here, and both undo the entire point of rating effort in the first place.
The first is sandbagging — quietly rating every set an RPE 6 or 7 no matter how it actually felt, because a low number feels like proof of untapped potential and makes next week's planned increase feel inevitable. This turns the log into fiction and removes the one signal that was supposed to keep training honest. The second, less obvious mistake runs the other way: treating RPE as a permission slip to always leave a comfortable buffer, never rating anything above an 8, and quietly avoiding the genuine, occasional near-max effort that adaptation actually requires. A man who's rated every set RPE 7 for six straight months hasn't been autoregulating — he's been coasting, and his numbers will show it eventually, usually right around the time he wonders why the last three months look identical to the three before that.
The fix for both is the same discipline: rate the set within five minutes of finishing it, before your ego has had a chance to negotiate with the number. Rate every working set, every session, no exceptions — the moment you start skipping the sets that would embarrass you, the whole system stops telling you anything useful.
What it looks like over a training block
Six weeks after Rob started logging RPE instead of simply chasing the number printed on the sheet, he hit that same 140kg squat on a rough Tuesday and, feeling the third rep grind early, dropped to 120kg for the remaining sets instead of forcing it. Every rep came out clean. Two weeks later, once his daughter was well again and his sleep had caught up, he added ten kilos to his best top set — a number that would never have happened if the bad Tuesday had put him in a back brace instead. The programme on paper barely changed. What changed was his willingness to admit, set by set, what the bar actually felt like that day.