Durability Score Methodology

How it works

Your Durability Score measures how well you’re holding onto your strength as you age. A score of 100 means you’re right on track. Above 100 means you’re beating the curve. Built entirely from your normal working sets -- no 1RM testing needed.

Each tracked lift goes through five steps:

1. Collect qualifying sets. Any working set of 1-6 reps with weight logged feeds the score.
2. Estimate 1RM. When you log RPE (6-10), the app uses RPE-adjusted Epley to account for reps in reserve. Without RPE, standard Epley is used.
3. 28-day aggregation. The best estimated 1RM from the last 28 days becomes your current number.
4. DOTS normalization. Adjusts for bodyweight so lifters of different sizes are compared fairly.
5. Compare against baseline. Your baseline is whichever is higher: your personal peak or an age-specific strength standard.

The composite score is the geometric mean across all tracked lifts, so one strong lift can’t mask a weak one.

Two scoring paths

Tracked lifts and fallbacks

The Durability Score is built around the four competition powerlifts, which have the deepest normative data and are the most reliable indicators of overall strength retention. You choose which lifts to track in settings.

The system automatically falls back to the highest-priority alternative if your tracked lift has no logged data. For example, if you track “low bar squat” but only log front squats, the score will use your front squat data. The score is still meaningful, but it’s measuring a different movement -- your true squat retention may be higher than reported.

For the most accurate score, train and log at least one of the ideal variations for each tracked lift. The competition movements have the most research behind them and produce the most comparable e1RM estimates.

Age-specific strength standards

Standards are set at 90% of the 10th percentile (lowest 10%) of drug-tested competitive powerlifters in each age group, sourced from the OpenPowerlifting dataset -- 2.3 million competition entries, filtered to tested federations and raw/wraps equipment only.

The 10th percentile represents the weakest competitors in tested powerlifting. We take 90% of that as a conservative floor representing roughly 1-2 years of consistent recreational training. Because the data is broken down by year of age, the standard naturally reflects what’s achievable at each age without needing a synthetic age-adjustment formula.

Values are expressed as a multiple of bodyweight. Raw per-year 10th percentiles are smoothed with a 3-year rolling average to reduce noise, then multiplied by 0.9.

Overhead press is not contested in powerlifting. OHP standards are extrapolated from the bench press standard using a 0.68x ratio for males and 0.85x for females. These ratios are derived from the intermediate press standards in Rippetoe & Kilgore’s Practical Programming for Strength Training (2009): 0.75x BW for males and 0.50x BW for females, divided by our age-30 bench standards (1.10x and 0.60x respectively).

Age decline model (McCulloch coefficients)

Once your personal peak exceeds the standard, the score tracks how well you’re retaining that peak. Expected decline is modeled using IPF Masters (McCulloch) coefficients -- the official handicap system used in Masters-division powerlifting to equalize results across age groups.

A coefficient of 1.440 at age 67 means a competitive lifter at that age is expected to lift about 69% of what they could at their peak. These coefficients only apply to the peak retention path -- they are not used when comparing against the age-specific standard.

DOTS normalization

DOTS (2024 revision by Haleczko) converts a raw lift into a bodyweight-independent score, allowing fair comparison across weight classes. Both your current lift and baseline are DOTS-normalized before computing the ratio, so changes in bodyweight don’t artificially inflate or deflate your score.

Input data

Any working set of 1-6 reps with weight logged qualifies. When RPE is logged (6-10), the formula adjusts for reps in reserve:

Without RPE (common with imported data), standard Epley is used. This slightly underestimates true 1RM, but because both current and peak are underestimated equally, the ratio holds.

Sets outside 1-6 reps and warmup sets are excluded. If no qualifying set is logged for 60+ days, that lift’s score pauses until you log again.

Limitations

This is a theoretical model. Individual variation in proportions, training history, and genetics means these baselines are approximations. The score is most meaningful as a personal tracking tool over time, not as a comparison between athletes.

The OpenPowerlifting data reflects competitive lifters -- even the 10th percentile represents people who chose to compete. The 90% discount partially accounts for the gap between competitors and recreational lifters, but it’s an approximation.

Sample sizes decrease at older ages, so standards above 75 are less statistically robust. Per-year data is smoothed with a 3-year rolling average to mitigate this.

References

1. Epley, B. (1985). Poundage chart. Boyd Epley Workout. University of Nebraska Press.
2. Haleczko, A. (2024). DOTS formula. International Powerlifting Federation.
3. McCulloch, R. IPF Masters age coefficients. International Powerlifting Federation.
4. OpenPowerlifting project. 2.3M drug-tested, raw/wraps competition entries.
5. Rippetoe, M. & Kilgore, L. (2009). Practical Programming for Strength Training, 2nd ed. The Aasgaard Company. Intermediate press standards used for OHP extrapolation.
6. Zourdos, M.C., et al. (2016). Novel resistance training-specific RPE scale. J Strength Cond Res, 30(1), 267-275.