Strength Standards Explainer

What a beginner, intermediate or advanced strength standard actually is, whose data it comes from, and why no governing body publishes one. This page computes no standard, because any table it printed would be somebody’s percentile of somebody’s population presented as a fact about lifting.

What the four federation documents read for this batch contain

What a bodyweight multiple quietly assumes

Whose data a strength standard is, and why no federation has one

This page is about the strength sports — powerlifting, the squat, bench press and deadlift, and weightlifting, the snatch and the clean and jerk. They are different sports under different bodies, the International Powerlifting Federation and the International Weightlifting Federation, and almost nothing carries from one rulebook to the other. Every figure below that belongs to an edition, a rulebook date or an Olympic cycle is printed with it, because both bodies revise them.

The finding behind this page, stated plainly: no governing body read for this section publishes a beginner, intermediate or advanced strength standard, and the tables circulating under that name are commercial products built from their publishers’ own data. Four federation documents were read for this batch — the IPF’s 2026 Technical Rulebook, its GL coefficient sheet of May 2020, its evaluation report of March 2020, and the IWF’s Technical and Competition Rules & Regulations 2025 with its November 2025 annex — and not one contains a table of that shape. What they publish instead is a bodyweight-adjusted score, a weight class, a qualifying placing and a results archive. So this page explains the tables rather than reproducing one.

What the circulating tables actually are. The best of them say so plainly. One publisher states that its figures come from “real powerlifting competition results, millions of them, via the OpenPowerlifting dataset”, that each number is “a percentile over raw lifters who stepped on a platform and had their lifts judged”, and that “intermediate is the median competitor, advanced is the top 20%, elite is the top 5%”. Every word of that is a choice the publisher made: the data set is a volunteer-run results archive rather than a federation’s, the population is people who entered competitions rather than people who train, and the tier boundaries are percentiles somebody picked. The site is explicit that its standards therefore run stronger than a gym population’s. Another publisher names no data set at all, no number of lifts, no method and no body, describing its tables only as “a reference point built from a defined, published methodology” while conceding that they are “built from available data, which may under-represent some ages and populations”.

So the figures are not fabricated, and they are not what readers take them for either. “Intermediate” is not a property of a lifter; it is a position in a distribution that one publisher assembled, with a label that publisher chose. Two sites can call the same 140 kg squat novice and advanced without either being wrong, because they are describing different populations. That is why this page does not print a table: a table here would inherit somebody else’s population and present it as arithmetic.

The honest limit of the finding. This is an argument from absence across four federation documents, not a proof. The IPF’s 2026 rulebook was truncated at section 6.3, so its sections on referees, the jury, records and coaches were not read; the IWF’s rules were truncated inside the rule 3.3.3 tolerance chart. A table of tiers could in principle sit in a part that did not come through, although none of the four documents’ contents pages suggests one. Two further fetches were refused and nothing is claimed about them: ExRx.net’s strength-standards table returned a robots error and strengthlevel.com’s permission request went unanswered. This page reports what was read and what was not.

What a lifter can honestly compare against. Two things, and both are published. The first is a bodyweight-adjusted score: IPF GL points is the IPF’s own current one, where 100 is the elite reference level the curve was fitted to, so a GL figure places a performance against a stated standard rather than against a guess. Wilks and DOTS compute the three other powerlifting formulas in live use, and the Sinclair coefficient is weightlifting’s. The second is the results themselves: both federations publish full scoresheets, and a total next to the totals of everyone in the same class at a real competition is a comparison with no interpretation in it at all.

One trap in the numeric form of the same idea. A bodyweight multiple — “a double-bodyweight squat”, “a triple-bodyweight deadlift” — is linear in bodyweight, and every published formula in these sports is not. A men’s Wilks coefficient is 0.86617 at 59 kg and 0.55881 at 140 kg, so a twice-bodyweight total scores 102 Wilks points for the light lifter and 156 for the heavy one. The rule of thumb makes exactly the error the coefficients exist to correct, and it makes it in the direction that flatters heavier lifters.

Where the line runs between this page and the health section: health owns the lifter’s body and this category owns the competition. A one-rep max estimated from a submaximal set is a training calculation about a person and lives at the one-rep max calculator, with body fat percentage and BMR and TDEE. Here is arithmetic on numbers a referee wrote down: a total, a score against a published curve, a weight class, a loaded barbell. Nothing on this page prescribes training, advises a weight cut, recommends a bodyweight or makes any medical claim — getting into a class is not this site’s business, and the classes are a published fact about a rulebook.

Finally, two things this page has no view on, by design. It does not say whether any particular performance is good, and it does not say what anyone should be able to lift. Those are coaching questions, and a calculator that answered them would be doing the thing this page exists to decline. For the arithmetic of the lifts themselves, the powerlifting total and attempt selection calculator and the plate loading and barbell weight calculator deal in numbers a referee recorded, and the weight class and bodyweight interpreter deals in what a rulebook prints.

Frequently asked questions

Related calculators

References

1. Lift Vault. “Strength Standards from Real Competition Data”, liftvault.com (read 15 Oct 2026). Cited for the same reason and for its description of the rest of the field. Of its OWN standards it says they come from “real powerlifting competition results, millions of them, via the OpenPowerlifting dataset”, that each number is “a percentile over raw lifters who stepped on a platform and had their lifts judged”, and that “intermediate is the median competitor, advanced is the top 20%, elite is the top 5%” — a definition credited to a volunteer-run results archive and not to any federation. Of the others it says that “Most strength standards online are either self-reported gym numbers (inflated, nobody logs their missed lifts)” or “tier labels someone invented”. It names no governing body as publishing strength standards. Its own tables were not read and no figure from it is used.
2. strengthlevelcalculator.com. “How to Calculate Your Strength Level” (read 15 Oct 2026). Cited as evidence about the SOURCING of a circulating strength-standards table, not for any figure, and no number from it appears anywhere in this batch. Asked where its standards come from, the page names no data set, no number of lifts, no number of lifters, no statistical method and no governing body, describing them only as “a reference point built from a defined, published methodology” and conceding that “Tables are built from available data, which may under-represent some ages and populations”. It also states that a level “is not a medical measure or a ranking against every lifter in the world”.
3. Fetches recorded as refused rather than paraphrased. ExRx.net’s strength-standards table returned a robots-disallowed error, so nothing is claimed about what it says or where its data comes from. strengthlevel.com’s permission request went unanswered and was not retried. The IPF’s women’s equipped detailed scoresheet on goodlift.info was likewise unanswered, which is why the GL reproduction rests on the men’s equipped sheet only and no women’s GL row was checked against a published figure.
4. International Powerlifting Federation. Technical Rulebook 2026, the file dated 01.03.2026, powerlifting.sport (read 15 Oct 2026; about 50,000 characters, truncated partway through section 6.3, so sections 7 to 10 — referees, jury, records and coaches — were not read). Source for: “Each competitor is allowed three attempts on each lift”; “The lifter’s best valid attempt on each lift counts toward his competition total”; “If two or more lifters achieve the same total, the lighter lifter ranks above the heavier lifter”, and on an equal bodyweight too, “the lifter making the total first will take precedence”; the bodyweight categories, printed as intervals to the hundredth (“83.0 kg Class from 74.01 kg up to 83.0 kg”); “the progression must be at least 2.5 kg between all attempts” and “the weight of the barbell must always be a multiple of 2.5 kg”, with a record bar “at least 0.5 kg in excess of the current record”; “Three unsuccessful attempts in any lift will automatically eliminate the lifter from the overall competition”; and the barbell — “Weight of the bar and collars are to be 25 kg”, collars that “Must weigh 2.5 kg each” and “Shall always be used”, the disc face values with their maxima and minima, the tolerance “within 0.25 percent or 10 grams of their face value”, and the colour code “10 kg and under – any color, 15 kg – yellow, 20 kg – blue, 25 kg – red”. The bar and disc diagrams and the knurling distances are images and were not read. The GL formula is not in this rulebook at all: it names an “IPF formula” for the best-lifter award and prints neither equation nor coefficients.
5. International Powerlifting Federation. IPF GL Coefficients for Relative Scoring, powerlifting.sport (read 15 Oct 2026; about 8,000 characters). Headed “ACTUAL FROM MAY 01, 2020 TO DEC 31, 2023”. Table 1 gives A, B and C for eight categories; all 24 values came through as text and match the IPF’s separate evaluation report digit for digit. Says the coefficients come from regression on elite “Golden Standard Samples” from IPF and European competitions from 2011, that a coefficient is computed “with an accuracy of 6 decimal places”, and that it is “reasonable to adopt a four-year cycle”, recounting them “one year before the World Games year”. Both printed formulas are images. Formula (1) came through as the bare label “(1)” and Formula (2) as “⋅⋅ (2)”, with every operand, constant and exponential missing; the prose names the parts (“Bwt – body weight; A, B, C – constants; e is the base of the natural logarithm; 100 – normalization factor”) and not their arrangement. Its two worked examples give inputs only — their results are absent — so neither was used as a check.
6. Kopayev O, Onyshchenko B, Stetsenko A. Evaluation of Wilks, Wilks-2, DOTS, IPF and Goodlift Formulas for Calculating Relative Scores in IPF Powerlifting Competitions, March 2020, published by the IPF at powerlifting.sport (read 15 Oct 2026; about 20,000 to 25,000 characters). Its Appendix 1 prints the coefficients of all five formulas side by side and all of them came through as text. Its conclusion ranks GoodLift first overall (23), then DOTS (44), Wilks-2 (50), Wilks (57) and the IPF’s then-official formula last (66), with GL first in six of eight categories. Two extraction defects were found in it by arithmetic rather than by reading. The men’s DOTS row came through with no minus signs, giving a coefficient of 0.239 at 50 kg where the women’s row on the same page gives 1.253. The women’s Wilks fifth-order coefficient came through as −0.0000009054, which makes the coefficient negative above about 55 kg; the correct figure is ten times smaller. Neither defect is in the published document — both are in what the fetch delivered. The five equations themselves are absent.
7. International Weightlifting Federation. Technical and Competition Rules & Regulations 2025, the file “as of 05 November 2025”, iwf.sport (read 15 Oct 2026; roughly 90,000 characters before the fetch truncated inside the Regulation to 3.3.3 tolerance chart, so later sections were not read). An absence worth naming: no “in force as from” statement appears in its readable text beyond its own 2025 title and the 1 August 2026 dates in section 1.2. Source for: the bodyweight categories of section 1.2, which print both the set in force at publication and the set effective 1 August 2026; ranking rule 6.8.2, under which equal results are separated by who achieved them earlier in the calling order and bodyweight does not break a tie at all, with “the athlete(s) who competed earlier in time” ranked higher across groups and rule 8.11 giving a tied record to “the athlete who completed the lift first”; rule 8.9, that “World Record(s) can only be broken in the bodyweight category and the age group where the athlete is entered”; and rule 3.3.3, specifying the discs as 25 kg red, 20 kg blue, 15 kg yellow, 10 kg green, 5 kg white, 2.5 kg red, 2 kg blue, 1.5 kg yellow, 1 kg green and 0.5 kg white, the men’s bar at 20 kg, the women’s at 15 kg and two collars at 2.5 kg each. The Sinclair formula is nowhere in the text read.
8. International Powerlifting Federation. IPF Qualification guideline for The World Games 2025, powerlifting.sport (read 15 Oct 2026; about 6,500 to 7,000 characters; no publication date is printed on it). Cited for what the IPF publishes INSTEAD of a strength standard: places in the eight IWGA weight classes go to the lifters finishing first to third at a named 2024 world championship, with GL points used only to separate lifters of the same country in a category, to fill remaining slots and to order the reserve lists. There is no qualifying total in kilograms and no performance tier anywhere in it.