Wilks, Wilks 2, DOTS and GL: Why Your Score Changed and Which One Counts

You had a 402 written in the front of your training log. Same total, same bodyweight, and the calculator you used to get that number now says something else. Nothing happened on the platform. The formula underneath moved.

This confuses people constantly, and the confusion is reasonable. Between roughly 2019 and 2020 the sport quietly acquired three new scoring systems in about eighteen months, and websites, meet software and federations adopted them at different speeds and in different combinations. A score you calculated in 2018 and a score you calculate today may be measuring you with completely different instruments.

The Wilks vs Wilks 2 vs DOTS question is really two questions wearing one coat. There is a technical question about what the formulas do, and a practical question about which number you should write down. The technical answer is genuinely interesting. The practical answer is short and slightly deflating.

I will give you both, and I will tell you now that the practical answer is: pick one, stay on it, and stop refreshing.

The short answer

Four systems matter. Wilks is Robert Wilks’ fifth-order polynomial from the mid-1990s, which the IPF used for roughly two decades and which most legacy scores are calculated in. DOTS arrived in 2019 from Tim Konertz, refitting the same broad approach on newer data with a fourth-order polynomial and only two parameter sets, one per sex. The IPF left Wilks behind for its own IPF Points in 2019, then replaced those with IPF GL points in 2020, which use an exponential shape and have separate parameters for classic and equipped lifting and for full power versus bench only. Wilks 2 is Robert Wilks’ own 2020 recalibration of his original formula, same polynomial shape, refitted coefficients.

Wilks, Wilks 2 and DOTS all produce numbers in a broadly similar range, so a good lifter sits somewhere in the 400s or 500s in all three. GL points are on a different scale entirely, with world-class sitting near 100. If your number jumped from 430 to 94, you changed systems, not fitness.

Four formulas in about twenty-five years

The reason there are so many is that all of these things are curve fits, and curve fits go stale.

Every coefficient formula answers the same question: given the bodyweight of a lifter, what total should we treat as equivalent to a given total at a different bodyweight? The only way to answer that empirically is to look at what people actually lift across the bodyweight range and fit a curve through it. Which means the formula encodes the population it was fitted on.

The population changed enormously. Wilks fitted his original curve in an era when equipped lifting dominated, women’s competitive powerlifting was much smaller, and the extreme light and extreme heavy ends of the curve had comparatively few high-level performances to fit against. Thirty years later the raw division is enormous, women’s raw lifting has grown at a pace nobody in 1995 forecast, and there is vastly more data. Any curve fitted then will misprice parts of the range now. That is not a criticism of Wilks. It is what happens to every fit eventually.

So: DOTS in 2019, IPF Points in 2019, IPF GL in 2020, Wilks 2 in 2020. Four answers to the staleness problem, arriving almost simultaneously and diverging on how to solve it.

What each one actually is

Wilks, and why it lasted so long

The classic Wilks coefficient is 500 divided by a fifth-order polynomial in bodyweight, with one set of six constants for men and another for women. You multiply your total by the coefficient. That is the whole thing.

It lasted because it worked well enough across the middle of the range where most lifters live, and because it became a shared language. When someone said “he pulled a 450 Wilks” everyone in the room had a calibrated sense of what that meant. Shared vocabulary is worth a great deal, and it is the main thing every replacement has had to fight against.

Its two standing weaknesses were the ends of the curve and the fact that a single parameter set covers raw and equipped alike, which is hard to defend given that equipment does not help every bodyweight equally.

DOTS

DOTS came out of German powerlifting in 2019, developed by Tim Konertz. Structurally it is the same idea done again: 500 divided by a polynomial in bodyweight, one parameter set per sex. The polynomial is fourth-order rather than fifth, and it was fitted on considerably more recent data with a particular eye on the light end, where the original Wilks fit had the least to work with.

What DOTS deliberately does not do is split by equipment or by event. That is a design choice with a real trade-off. It keeps the system simple and federation-agnostic, and it means a raw lifter and an equipped lifter get scored by the same curve, which is arguably wrong. You will find people who consider that a fatal flaw and people who consider it a feature. Both positions are defensible.

IPF Points, then IPF GL Points

The IPF stopped scoring on Wilks from 2019 and introduced IPF Points, a formula with a different mathematical form and, importantly, separate parameter sets for classic and equipped and for the full three-lift total versus bench-only competition. Reception was mixed, partly because the numbers were on an unfamiliar scale and there was no intuition for what a good score looked like.

IPF GL points replaced them in 2020. The shape is 100 divided by a fitted exponential of bodyweight, multiplied by your total. That exponential form behaves differently at the extremes than a polynomial does: it flattens toward an asymptote rather than curving off, which is a more sensible description of what happens to strength at very high bodyweights. The 100 in the numerator is not cosmetic. It sets the scale so that a world-class performance lands near 100. Once you know that, GL numbers become intuitive quickly. Our piece on IPF GL points goes through the parameter splits properly, including the bench-only detail almost everyone misses.

Wilks 2

In 2020 Robert Wilks published a recalibration of his own formula. Same fifth-order polynomial shape, refitted constants, drawn on newer data. Databases that carry it usually label the column “Wilks2” to distinguish it from the original.

Adoption has been patchy, and I think that is the interesting part of the story. Wilks 2 arrived into a market that already had DOTS and GL competing for the same slot, and a lot of the community had already picked. A recalibration that had appeared in 2016 might have simply become “Wilks”. Arriving in 2020, it became a fourth option.

The one table you actually need

Two things fall out of that table immediately. First, three of the five live on a similar numeric scale, which is exactly why people mix them up. Second, only the IPF formulas distinguish raw from equipped, which means everywhere else in the sport a bench shirt and a pair of sleeves are being scored by the same curve.

You will also meet older systems in the wild. Schwartz and Malone were the men’s and women’s coefficients that preceded Wilks. Glossbrenner, used by WPC and AWPC events, averages Wilks with the older Schwartz and Malone approach. Reshel coefficients turn up in GPC-affiliated meets. Those are real systems with real users, and none of them are comparable to each other either.

Why your number changed, and why it probably means nothing

Here is the part that actually answers the question in the title.

Say you logged a Wilks of 402 in 2019 at 91.4 kg with a 640 kg total. You come back, put the same figures into a site that has since switched to a different formula, and it reads 391. You have lost 11 points and you did nothing wrong.

Work out what that is proportionally: 391 ÷ 402 = 0.9726, so you are down 2.74 per cent. Now do the same for your training partner, who was 415 on the old system and reads 405 on the new one. That is 405 ÷ 415 = 0.9759, down 2.41 per cent. You both dropped by a similar amount, and the gap between you moved by a third of a per cent. A percentage calculator makes that a ten-second check.

That is the general case. Refits shift the whole curve and redistribute a little between regions of it. Unless you sit at an extreme bodyweight, the shift you experience is close to the shift everyone near you experiences, so your standing barely moves. The number changed. Your position did not.

Where it does matter is at the ends. A 52 kg female lifter and a 140 kg male lifter are exactly the regions where the old fits had the least data and where different refits disagree most. If you live out there, the choice of formula genuinely affects how you rank, and you are entitled to have an opinion about it.

Worked example: two lifters and a break-even ratio

Coefficient arguments get much clearer when you stop looking at scores and start looking at ratios.

Say Owen weighs 74.2 kg and totals 615 kg. Bryn weighs 119.8 kg and totals 810 kg. Who is the better lifter, pound for pound?

Start with the raw ratio of totals: 810 ÷ 615 = 1.3171. So Bryn lifts 31.71 per cent more absolute weight. For Owen to beat him on any coefficient system, that system must believe Owen’s bodyweight is enough of a handicap to be worth more than 31.71 per cent. In other words, Owen’s coefficient has to be at least 1.3171 times Bryn’s.

Now run the DOTS numbers. At 74.2 kg the men’s DOTS coefficient works out to about 0.7224, and at 119.8 kg to about 0.5746. Owen scores 615 × 0.7224 = 444.3. Bryn scores 810 × 0.5746 = 465.4. Bryn wins by 21 points.

Check the ratio: 0.7224 ÷ 0.5746 = 1.2572. DOTS thinks Owen’s bodyweight disadvantage is worth about 25.7 per cent, and he needed 31.7 per cent. He falls short, and not narrowly.

Here is what makes this useful. The difference between how any two of these formulas value a 45 kg bodyweight span is typically a few per cent. So a system that priced the light end six or seven per cent more generously than DOTS does would flip this result. That is not a hypothetical size of disagreement. It is roughly the magnitude of disagreement that exists between the systems in this article at the far ends of the curve. Two lifters this far apart in bodyweight can genuinely swap places depending on which column you sort by, and neither ranking is wrong.

Running that comparison yourself takes about a minute. Put both bodyweights and totals through the DOTS Calculator and then the Wilks Calculator, and see whether the pair you care about changes order. Everything runs in your browser, nothing gets uploaded, and you do not need an account to check a number that ought to be free to check.

Who gained and who lost

I want to be careful here, because this is where confident wrong answers circulate fastest.

The structural claim is safe: refitting on newer data re-prices the regions of the curve where the sport has changed most. Women’s raw lifting and lighter raw men’s lifting grew enormously between the original Wilks fit and 2019, so those are the regions where old and new fits disagree most. DOTS was explicitly built with the lighter end in mind, and it is generally kinder there than classic Wilks.

The specific claim is not safe, and I am not going to pretend otherwise. Whether you personally gained or lost, and by how much, depends on your exact bodyweight, your sex, and which pair of formulas you are comparing. The differences are not uniform and they are not monotonic. Wilks 2 was contentious on release, with some lightweight lifters reporting meaningful drops relative to classic Wilks, and I would not want to generalise from that to a clean rule.

The reliable method is boring and takes two minutes: put your own bodyweight and total into two calculators and read both numbers. That answers your question exactly, for you, rather than approximately, for a hypothetical lifter.

Which federation uses what, and why that keeps moving

The IPF and its national affiliates score on GL points. That includes USAPL, British Powerlifting, Powerlifting Australia and the rest of the affiliate list. If you compete in that family, GL is your number and everything else is trivia.

Outside the IPF the picture is messier and changes. DOTS has become common at meets run under USPA, WRPF and IPL banners, and it is a standard column in the OpenPowerlifting database alongside Wilks, Wilks 2 and Glossbrenner. Some federations run Wilks still. Some meet directors pick their own for a local event, particularly for a best-lifter award. WPC and AWPC events have used Glossbrenner. GPC-affiliated meets have used Reshel.

I am deliberately not giving you a definitive federation-by-federation table, because any such table is out of date within a season or two and a wrong table is worse than none. Federations change scoring, sometimes with little announcement, and individual meets vary within a federation. Read the entry page for the meet you are actually doing. That page is the only authority that matters for your score on the day.

Age coefficients sit on top, not instead

A source of confusion worth clearing up: masters age adjustments are a separate multiplier layered over your coefficient score, not a rival to it.

The one you will meet most is the McCulloch factor for masters lifters, with a parallel adjustment applied to sub-junior and junior lifters in some federations. A masters lifter’s placing in an age-adjusted category may be computed as total × bodyweight coefficient × age factor. So a “score” can differ from another score for two entirely separate reasons.

This is why a masters lifter comparing themselves against an open lifter needs to be specific about what has been applied. Not every results sheet says. If your number seems inexplicably high next to someone with a bigger total, check whether an age factor is in there.

Pick one score and stop switching

1. If you compete, use whatever your federation scores on. That is your competitive number and there is no argument to have.
2. If you do not compete, or you compete across federations, pick one system for your own tracking. I would default to DOTS: it was fitted on recent raw data, it is not owned by one federation, and it sits on a scale most lifters already have intuition for.
3. Go back through your training log and recalculate your historic scores in that one system. This is tedious once and never again. Mixed-formula history is worthless history.
4. Write down which system you chose, in the log, in words. Future you will not remember.
5. Track the trend, not the value. A coefficient score is a comparison instrument. Its job is to tell you whether you are gaining ground on the field or on yourself.
6. Recalculate after every meet using your official bodyweight from the weigh-in, not your gym scale. Those are frequently a kilogram apart, and a kilogram moves your coefficient.
7. If your federation switches formulas, keep your personal series in the old one as well as the new one for a year. Then drop the old one.

Step three is the one people skip and then regret. A five-year log with three different formulas in it cannot answer the only question it exists to answer.

What trips people up

The commonest error is comparing across systems without noticing. Somebody quotes a DOTS score, somebody else quotes a Wilks score, and because both are three-digit numbers in the 400s nobody spots the problem. The gap between them is usually a few per cent, which is exactly the size of a real training improvement, so the comparison feels meaningful while being meaningless.

Second is the unit problem. Every one of these formulas takes bodyweight in kilograms. Feed a pounds figure into a calculator built for kilos and the result is not slightly wrong, it is nonsense. If your meet ran in pounds and your calculator wants kilos, convert first with a unit converter, then run the corrected figure through the DOTS Calculator and check the answer looks sane before you trust it.

Third is treating a coefficient score as a training target. It is a ranking tool. Chasing it directly leads people into weight manipulation decisions that cost them more total than the coefficient gains them, and the arithmetic on that trade is not obvious. Rank yourself with it; program with something else.

Fourth, and quietest: coefficient scores compare totals, and totals hide structure. Two lifters with identical DOTS can have wildly different lift distributions, and if you are using scores to work out who to learn from, that matters more than the headline. Reading published standards alongside your score helps, and our piece on strength standards by bodyweight covers where those tables come from and how far to trust them.

Frequently asked questions

Why did my Wilks score change?

Almost certainly because the site you use switched formulas, not because you got weaker. Robert Wilks published a recalibration in 2020, usually called Wilks 2, and many databases and calculators also added DOTS around the same period. Same total, same bodyweight, different coefficient, different number. Check which column you are actually reading before you worry about it.

Is Wilks still used in powerlifting?

Yes, widely, though less officially than it once was. The IPF moved off Wilks for 2019 and now scores on GL points. Plenty of independent federations, local meets and gym conversations still run on Wilks because everyone already knows roughly what a given number means. Legacy scores in training logs and old results are almost all classic Wilks, which is worth remembering before you compare them to anything modern.

Which federations use DOTS?

DOTS turns up widely outside the IPF family, including at many meets run under USPA, WRPF and IPL banners, and it is a standard column in the OpenPowerlifting database and in several meet-scoring packages. Federation choices change from season to season and some meet directors pick their own. Check the entry page for the specific meet rather than trusting a list you found online.

What is the difference between DOTS and IPF GL points?

DOTS is a polynomial refit of the Wilks approach with one parameter set for men and one for women, and it produces numbers in the same general range as Wilks. IPF GL points use an exponential form, have separate parameters for classic and equipped and for full power versus bench only, and land near 100 for a world-class performance. They answer the same question with different machinery.

Can I compare a Wilks score with a DOTS score?

Not directly, though they sit on similar scales so people do it anyway. A DOTS score and a Wilks score for the same lifter usually differ by a few per cent, and the size of that gap depends on bodyweight. Comparing across systems tells you very little. Compare two lifters within one system instead, which is what the Wilks Calculator and the DOTS tool are both there for.

Which scoring formula should I track for my own progress?

Whichever one your federation scores on, if you compete. If you do not, pick any single system and never switch, because the value of a coefficient score to you personally is the trend line, not the absolute figure. DOTS is a reasonable default because it was fitted on more recent raw data and is not tied to one federation.

The short of it

Your score changed because the instrument changed. That is the whole answer, and once you have checked that everyone near you moved by a similar proportion, the anxiety usually evaporates.

The version of this I would actually recommend: work out which system your competitive federation uses, use that for anything official, and pick a single system for your own log and never touch it again. Recalculate your history once. Then stop.

Formulas will keep being refitted, because the sport keeps changing and fits keep going stale. There will be another one. When it arrives, the useful response is not to recalculate everything in a panic; it is to check whether your position relative to the people you compete against actually moved. Nine times out of ten it will not have.