Blog / Autoregulation

1RM vs RIR: When Each Is the Right Autoregulation Tool

July 16, 2026 · 9 min read

Ask how heavy your next working set should be and you will get two kinds of answers. One camp anchors everything to your one-rep max: 5 sets of 5 at 80%, waves at 85%, singles at 92%. The other camp barely mentions load and talks about effort instead: stop at 2 reps in reserve, whatever weight that turns out to be. Both camps are using a real tool. The mistake is treating them as rivals, because each one is measuring something the other cannot see.

Two numbers, two different questions

Your 1RM (one-rep max) is the heaviest weight you can lift once with acceptable technique. It is a capacity measure: it describes what your body can produce on the lift, independent of any single session. It moves slowly, over weeks and months, which is exactly what makes it useful for judging whether training is working.

RIR (reps in reserve) is a per-set effort measure: the number of clean reps you still had available when you racked the weight. Finish a set of 8 knowing you had 2 more, and that set was 2 RIR. It moves fast, set to set and day to day, because it reflects your current state: sleep, stress, accumulated fatigue, and how the specific weight felt in your hands. The full scale and how it maps to RPE live in the RIR guide.

So the two numbers sit on different clocks. The 1RM is the slow clock that tells you where your strength is trending. RIR is the fast clock that tells you what today can absorb. Most load-selection mistakes come from asking one clock the other clock's question.

The problem with loads planned weeks in advance

Percentage-based programs prescribe load from the slow clock alone: week 4 says 80% of max, and 80% is what goes on the bar. The number is precise, but it was computed from a snapshot. Your actual capacity on the day drifts with readiness, and across a block it drifts with fitness itself: if you are progressing, a percentage of an old max quietly becomes lighter than intended; if fatigue has piled up, the same percentage becomes a grind. The prescription cannot see either shift; it was written weeks before the day it prescribes.

Effort-anchored loading flips the direction: fix the target effort, let the load float to meet it. The head-to-head evidence is more modest than either camp likes to claim. Over 8 weeks matched for sets and reps, RPE-based and percentage-based loading both built strength, with a small, non-significant edge to the effort-based group in most lifters (Helms et al. 2018); a 12-week trial found a significant squat advantage for RIR-based loading (Graham and Cleather 2021). Both work, and the effort-based version fits the day better, which is the whole case for autoregulation.

Estimated 1RM: the strength signal without the max-out

Both camps can agree on one thing: you do not need to test a 1RM to know it. Any recent hard set implies one. The Epley formula, the one Anneal uses, is: estimated 1RM = weight × (1 + reps / 30). A 200 lb set of 5 implies about 233 lb. An estimate computed the same way after every session gives you the slow-clock trend for free, off the training you were already doing.

The estimate has a working range, and it is worth respecting. Rep-based formulas are tightest when the set has 1-5 reps, still usable to about 10, and unreliable past 10-12: the load-to-reps curve bends at higher reps and a linear formula keeps drawing a straight line through it. Prediction research has been consistent on this for decades, and different lifts miss differently (deadlift estimates tend to run low). This is why Anneal simply refuses to compute an estimate from sets above 12 reps rather than log a number it cannot stand behind. If you want the estimate to mean something, feed it a heavy set of 3 to 5, and treat the result as a trend point: it tracks direction, not what you could grind out today.

The same calculator runs as a standalone page in the free training tools, alongside its inverse, an RIR-to-load calculator that starts from a target effort and hands back a working weight.

When testing a true 1RM earns its place

A true max has real uses. If you compete in powerlifting or weightlifting, meet-style singles are specific practice: openers get picked from them, and the skill of straining against a limit weight has to be rehearsed somewhere. Testing is mostly a competitor's tool.

For everyone else, the accounting rarely favors it. A limit single is the highest injury and fatigue cost per data point in training, it disrupts the week around it, and it returns one number that an estimate already approximates. If you want the occasional reality check, a heavy single around RPE 8-9 (one or two reps still in hand) once per training block, roughly every 8-12 weeks, buys most of the calibration at a fraction of the cost. Skipping even that and letting the estimated trend do the work is a completely valid choice.

When RIR is the right tool

RIR does its work inside the session. Pick the effort a set should land at, commonly 2-3 RIR for most working sets, then choose the load that produces that effort today. On a strong day the same target puts more weight on the bar; on a flat day, less. Both are correct, because the stimulus you were aiming for, proximity to failure, stayed where you set it. Rate honestly, and calibrate your ratings against occasional sets taken to failure.

RIR assumes you can feel where failure is, and that is a learned skill. New lifters tend to stop several reps short of true failure while believing they are close, so an early RIR log reads harder than the training actually was. If you are in your first months of training, anchor to a fixed rep target with conservative loading instead: hit the reps cleanly on every set, add a small increment when all sets are clean, in the pattern of double progression. The RIR scale becomes trustworthy with exposure, and it does so faster if you occasionally find out what failure actually feels like on a machine or cable movement where finding out is safe.

Use both: e1RM for the trend, RIR for the day

The division of labor is clean: track your estimated 1RM to know whether you are getting stronger, and use RIR to decide what goes on the bar today.

This split is exactly how Anneal is built. Your best qualifying set each session (12 reps or fewer) updates a per-lift estimated 1RM trend, so the slow clock maintains itself. The next session's weight is pre-filled from your last performance, reps and logged RIR included: hit the top of your rep range with reps to spare and the prefill steps up; grind or miss and it holds or backs off. And the two clocks get read together where it matters most: a stalled e1RM trend on a muscle that is working harder and harder, with logged RIR sinking at the same loads, is one of the signals that fires a deload suggestion. The decision rules, if you want to apply them by hand, are in how to know when to deload.

Common questions

Is an estimated 1RM accurate?

Accurate enough for what it is for: tracking your strength trend. Rep-based formulas like Epley are tightest when the set you estimate from has 1-5 reps, still usable up to about 10, and unreliable past 10-12, which is why Anneal stops estimating above 12 reps. Different lifts also err differently (deadlifts tend to be underestimated). Treat the number as a session-to-session strength signal, and expect the estimate from a heavy triple to be closer to the truth than one from a set of 12.

Do I need to test my 1RM to train with percentages?

No. You can estimate it from any recent hard set with a rep-based formula (Epley: weight times 1 plus reps divided by 30) and take percentages off the estimate. A 200 lb set of 5 puts your estimated 1RM around 233 lb, and every percentage prescription can anchor to that. As new hard sets come in, the estimate updates on its own, which a tested max never does.

Is RIR-based loading better than percentage-based loading?

The head-to-head evidence is close. Over an 8-week block matched for sets and reps, both approaches built strength, with a small, non-significant edge to the effort-based group in most lifters (Helms et al. 2018), and a 12-week trial found a significant squat advantage for RIR-based loading (Graham and Cleather 2021). The practical advantage of RIR is that it absorbs daily readiness swings: the load lands where the intended effort lands on the day, rather than where a spreadsheet guessed it would weeks earlier.

How often should I test my true 1RM?

For a lifter who does not compete, rarely or never. A true max carries the highest injury and fatigue cost of any set you can do and returns one data point. If you want a heavy calibration single, cap it at one single around RPE 8-9 once per training block, roughly every 8-12 weeks, and treat even that as optional. An estimated 1RM from your working sets tracks the same trend every session for free.

Can beginners use RIR?

Eventually, yes, but the ratings are unreliable at first. New lifters tend to stop several reps short of true failure while believing they are close to it, so an early RIR log reads harder than the training actually was. A better starting anchor is a fixed rep target with conservative loading: hit the target reps on every set, add a little weight when all sets are clean, and let the RIR scale calibrate in the background as you learn what failure actually feels like.