"Did you PR?" usually gets answered with a weight. More on the bar than last time, yes or no. That's one way to get stronger, but it's not the only way, and treating it as the only way makes a lot of real progress invisible. Squeeze out one more rep at a weight you've been stuck at for three weeks and the honest answer is still yes, you PR'd, even though the number on the plates didn't move at all.

Two ways to get stronger without adding weight

A load PR is the obvious kind: more weight than you've ever lifted for that exercise, at the same or fewer reps. It's the one that shows up on the chart as a new peak and the one people mean by default when they ask about a PR.

A rep PR is the quieter version: the same weight you've lifted before, but more reps than you managed last time. Bench 80 kg for 8 reps one week and 80 kg for 9 the next, and nothing about the number on the bar changed, but the set got harder to fail and easier to complete, which is exactly what getting stronger looks like from the inside. Ignore rep PRs and a training block that's actually working can look stalled on paper for weeks, just because the weight column stayed flat while the reps column kept climbing underneath it.

How the app scores it, not by the number on the bar

Zirv doesn't keep a separate record for "heaviest weight" and "most reps." It tracks one personal record per exercise, and updates it by comparing estimated one-rep max, not raw weight, using the same hybrid formula covered in how Zirv estimates 1RM: Brzycki for 2 to 10 reps, Epley for 11 and up, switching seamlessly at the 10-rep boundary where both agree.

Concretely: if your logged PR is 82.5 kg for 5 reps (estimated 1RM around 92.8 kg) and next session you do 82.5 kg for 6 reps (estimated 1RM around 95.2 kg), that replaces the old record. Same weight, one more rep, and the app logs it as a new PR because the estimated 1RM went up, which is the same logic it uses when the weight itself goes up. A rep PR and a load PR aren't judged by different rules. They're judged by the same rule, and a rep PR just wins it a different way.

Why the app doesn't wait for the weight to move

This matters for programming, not just for how a chart looks. Zirv's progression logic watches the top of your working rep range: hit the top of that range and the plan bumps the target weight for your next session, then drops you back down to the bottom of the new range to start the climb again. A string of rep PRs at a fixed weight isn't a plateau from the plan's point of view, it's the buildup phase before the next jump. The weight stalling is the expected shape of that pattern, not a sign it isn't working.

That also means a stretch of sessions where reps keep climbing at the same load is worth reading as progress on its own terms, not as a waiting room for the "real" PR. Progressive overload doesn't require the weight to move every session, it requires the stimulus to keep increasing, and an extra clean rep at a weight you already own is a legitimate increase.

When a stalled weight really is a stall

None of that means every flat week is secretly progress. The tell is whether reps or effort are actually moving. If you're stuck at 82.5 kg for 5 reps for session after session, no extra rep, no easier rep, and your logged RPE at that weight is creeping up instead of holding steady, that's a real plateau, not a rep PR waiting to happen. Reading the trend takes both numbers together: weight and reps tell you what happened, RPE tells you how hard it was to get there, and a set that's technically unchanged but rated harder than last time is going backward, not sideways.

The distinction is worth checking before assuming a stalled weight means stalled training. Look at the reps column next to the weight column for the last few sessions on that lift. If reps have been inching up while weight sat still, you've been PR'ing the whole time, the log just needed reading the right way.

The takeaway

A load PR and a rep PR are both real progress, scored by the same estimated 1RM comparison, not two separate categories competing for attention. When the weight on a lift stalls, check whether the reps have been climbing underneath it before calling it a plateau, because Zirv's own progression logic treats a run of rep PRs at a fixed weight as the normal lead-up to the next weight increase, not a sign that nothing is happening. The number that matters is whether the set got harder to fail, and that number moves well before the plates do.