The Wilk’s Formula is Too Wrong to Use

medal

I love powerlifting because of how objective it is.  We don’t have any of the debates one might find in gymnastics or figure skating or bodybuilding – sports in which the end result appears to be much more subjective to the observer.  In powerlifting you either lifted the weight to a certain standard or you did not.  And while every sport has its judging issues, for the most part with the respected federations those mishaps are few and far between.

All of that is well and good but when it comes time to pick the best lifter, now we have a problem.  The most commonly used formula is the Wilks.  It can be the deciding point if you can compete in a prestigious competition or not – for example you need a 430 Wilks formula to qualify for the Arnold.  And put simply, the Wilk’s formula is too wrong to use.

Let me use an example to highlight why this is the case.  I recently assisted a friend of mine in hosting an informal bench press competition at a military base.  He had weight classes but he wanted to give a “best bench” award.  These soldiers were not seasoned lifters and most were lifting 225 to 275 at normal body-weights, but there were benches that were clearly at the top of the field.

 

Those lifts were:

 

Lifter A: 175 lbs bodyweight and benched 330                    Wilks: 102

Lifter B: 285 lbs bodyweight and benched 405                     Wilks: 104 – Best Lifter

 

Wilk’s says that Lifter B is the best lifter.  I don’t know how you feel, but to me one of those lifts is clearly more impressive than the other, and it ain’t Lifter B.  The first lifter is benching close to double body-weight and he is benching 155 lbs over his body-weight.  The second lifter isn’t even hitting 1.5 x body-weight and he is benching only 120 lbs over his body-weight, and he is heavier.  Now I am not suggesting we use the old pound per body-weight rule where you take the total weight lifted and divide it by the body-weight, as everybody knows that is slanted toward lighter lifters.

The total pounds you are lifting over your body-weight should normally increase as your weight increases.  In both of these examples the lifters are lifting at least 1 lb per body-weight and that is a standard any high level powerlifter breaks in every lift.  Yet here one athlete was allowed to gain 110 pounds of body-weight and the lift only went up 75 lbs over his competitor’s bench.  Do you see the issue with that?

 

Lifter B: 285 lbs                  Lifter is lifting 1.42 lbs per bodyweight

Lifter A: 175 lbs                 Lifter is lifting 1.89 lbs per bodyweight

110 lb weight difference

Lifter B is 110 lbs heavier than lifter A, but he only lifts an additional 75 lbs of weight

75 / 110 = His extra weight is only adding .68 lbs lifted per bodyweight

 

As the lifters body-weight increases, he only has to increase his lift by .68 lbs per body-weight gained, which is a ratio that doesn’t match anything we see with lifters in real life (even the biggest, heaviest lifters – the best ones – are usually lifting close to 2 lbs per body-weight in the bench and well over 2 lbs per body-weight in the squat and deadlift.

For the formula to allow a lifter to gain 110 lbs of body-weight and with only 75 lbs of added strength and then it ranks that as more impressive, that is a big problem.  That is just not a little off, as though one discovers a team’s 3-point line is 2 inches closer to the basket than another teams.  This is like discovering that their 3-point line is 4 feet closer, it is immediate problem that needs to be fixed.  That kind of problem would ruin the game.  The good news is there is a solution.  That solution is to calculate a lifter’s AS.  That stands for allometric score, but AS is more user friendly and easier to say and from now on that is how I will refer to it.

My friend Greg Nuckols over at his excellent site Strengththeory.com addressed this issue as well.  He suggests, and I agree, that we should instead use allometric scaling to compare the lifts of one lifter to another.  This formula has a huge amount of theoretical evidence behind it in all sorts of applications and most importantly it simply works when you apply it powerlifting.  Here is the link to his full article, he explains the math in greater detail here.

You will still want to use a “best lifter” calculator as the formula is probably too complex to do in your head, luckily I have provided that for you here.  It is just as easy as it was before, you simply input the weight lifted and the bodyweight to find the best coefficient for that individual, and whomever has the higher coefficient wins.  When we calculate their AS and compare our lifters we get the following information:

 

Lifter A: 175 lbs body-weight and benched 330 lbs                              AS = 10.55

Lifter B: 285 lbs body-weight and benched 405 lbs                              AS = 9.40

 

Lifter A clearly has the more impressive lift.  I am not saying that Lifter A is stronger – strength typically refers to absolute strength and obviously Lifter B did indeed lift more weight than lifter A did.  But we have weight classes and we have formulas for a reason – people are interested in relative strength.  They want to compare one lift to another across body-weights, if this wasn’t important we would not have weight classes in the first place.

The current standard is way off. Wilks isn’t just a little bit off.  This isn’t a “well, it isn’t so bad let’s just keep the status quo” type of problem.  Wilks is fatally flawed and it should no longer be used to determine best lifter status.  Too much is at stake for it to be this wrong.  It is unfair for lifters to put forth all of that time, energy, and devotion to the sport only to be jilted come competition time.

It is time to get it right.  It is time to use Allometric Scaling to find a lifter’s AS and settle this once and for all.

Know someone who wants to get strong?...Share on facebook
Facebook
Share on twitter
Twitter
Share on google
Google
Share on tumblr
Tumblr
Share on reddit
Reddit
Share on email
Email

22 thoughts on “The Wilk’s Formula is Too Wrong to Use

  1. benchie

    In most credible bench press competitions, lifter A and lifter B would not be competing against each other for the same “best lifter” award. They would be split up in a lightweight/heavyweight set up where the divide would between around 200lbs bwt. There’s a lot of data out there that would support that the normal powerlifting wilks wasn’t made for bench only competitions, it becomes more accurate when dealing with 3-lift totals, but I think the problem here was oversight in set up.

    1. Tim Henriques

      Thanks for the post. There are some competitions that do separate out a heavy weight and a light weight best lifter but I would say that is a rarity and not the norm. A best lifter is usually just that – the best lifter at the meet. The point of this example was just to compare 2 lifts, Wilks doesn’t care if it is a squat, bench, deadlift, or total. The flaw in Wilks is once you get to a heavy bodyweight you only have to add a small fraction of strength for it to be considered a better lift.

      Here is another example.
      At a bodyweight of 220 if you lift 550 lbs that is a Wilks of 151.9

      At a bodyweight of 315 if you lift 605 lbs that is a Wilks of 152.9

      Both of these people are over 200 lbs and thus would be in the same category in your example. The 550 lb lift is obviously better but Wilks gives the nod to the 315 lb lifter. That lifter added 95 lbs of bodyweight and only 55 lbs of strength.

      If we use the AS the first lifter gets a score of 15, the second 13, the first lift is clearly better.

      It is not a lift dependent thing, it is just that once you hit about 275 Wilks is super skewed in favor of the heavyweight lifters.

      1. Carlos Matesanz

        While I might agree with wilks not being any good of a measure the problem in your reasoning is your statement:
        “The 550 lb lift is obviously better”

        Well, you find it better, or more impressive, and thus you find a formula that favors it to be more appropiate. Who decides which lift is better? It was for that reason that formulas were created. According to wilks lift B is better, according to AS lift A is better, according to “Matesanz’s magic formula” they might be tied.
        In that example you put out what happens if they both lift 10 more pounds?. How many pounds must guy B add to his total for every pound lifter A adds?

        No matter which formula you use, you’ll find examples someone might find unfair. That’s why noone should care about points.

        1. Tim Henriques

          I agree that no current formula is perfect, but I don’t agree we shouldn’t or can’t care about points or the idea of relative strength.

          I think it would be near universally agreed that a lifter 200 lb lifter benching 400 lbs is better than a 405 lb lifter benching 405 lbs.

          I also think it would be near universally agreed that a 150 lb lifter benching 300 lbs is worse than a 300 lb lifter benching 650 lbs.

          Both of those are discovered by “formula” so now the question becomes two fold for me. Which current formula is the best one we have available (I would argue the AS is) and what can we do to make it better?

  2. Pasquale

    I agree with you. Moreover, it doesn’t consider how tall you are. 200lbs at 1.60m and 200lbs at 1.95m are not the same. Obviously the second athlete has less musclar mass and strength.
    And we are still not considering arms and back length, which are smaller in the 1.60m athlete. So, I think it should be considered something more accurate.

  3. Miklos

    While the article makes a valid point, let me offer this:
    In conversation with Mr. Wilks he himself mentioned the Wilks formula is no good to calculate single lift results (bench only) as it’s way off. It was calculated and intended only for three lift competitions. Check three lift results and you’ll see the formula’s correctness vastly improves. Wilks said he never had the time to figure out a new table for single lift meets, so put the project on the back burner.

    1. Tim Henriques

      Thanks for the comment but I still don’t agree. For example if you have a 198 lb lifter that lifts 1700 lbs total and you compare a 275 lb lifter that lifts 2000 lbs total, the Wilks formula gives the nod to the heavier lifter by a significant margin (571 to 492). The 198 lb lifter is lifting 8.5 lbs per pound of bodyweight; the 275 lifter is lifting is lifting 7.27 lbs per bodyweight.

      The heavier lifter has added 77 lbs of bodyweight and only 300 lbs to their total (that isn’t even 4 lbs per pound of bodyweight). For them to have an equal score the 275 lifter only needs to beat the 198 lb lifter by 200 lbs in the total. I don’t think that is fair or accurate.

      1. Alexander Sharman

        I agree with the point the previous poster was making.

        Wilks is intended to be used across al three lifts, it was created by the President of the Australian Powerlifting Federation, Ray Wilks, precisely for this purpose.

        A scientific investigation was done into the Wilks formula around 20 years ago .. and it found that the formula had no bias in bench press, favoured lighter women in squats, was unfavourable for heavier men in deadlift, and that overall it favoured lifters with a medium weight (not light or heavyweight).

        Personally, my point would be that lighter lifters tend to lift much heavier weights compararatively in Deadlift, than in Bench and Squat, where bodyweight seems to have a much greater advantage, so if Wilks favours a heavier lifter in your example (bench), this is offset in deadlift where the formula favours lighter lifters.

        A second point is that in calculationg how “impressive” a lift is for you, you have mentioned how many times a lift is as a % of a competitors bodyweight.

        The reasons formulas were developed are because it has been shown time and again that there is not a direct correlation between bodyweight and how much someone can lift, every extra kilo more of body doesnt mean you can lift 1 kg more of weight.

        The current Raw World Record Totals show this, for the lightest male category, the current record is 1339lbs at less than 123 bodyweight … (10.9 times bodyweight) while in the super heavyweight category it is 2513 at 342 pounds (7.3 times bodyweight) … this is why there is a fomula .. because you cannot simply compare how many times their bodyweight someone is lifting, because that will always favour lighter lifters.

        It does seem that there needs to be revision of wilks or a new formula created .. at the moment the formula seems to favour women (last weekend 2 or 3 female lifters broke the all time wilks record at the US open) .. and people have commented that the formula was created based on a list of weight class world records or equipped lifters in the late 1970s .. and things have changed since then, especially amongst the performance of women.

  4. Dennis P L

    Here is another example. If you look at the all time Raw total record without knee wraps. The all time best at 198 is Jesse Norris with a 2033 total. The superweight total record with knee-wraps is by Ray Williams with a 2436 total.

    https://www.powerliftingwatch.com/records/raw/world

    I believe Williams weight in at 403 pounds.

    Us the wilks calculator

    https://www.lift.net/wilks-calculator/

    Norris wilks score is 589.348

    Williams Wilks score is 593.248

    Both are great lifters and obviously off the charts strong. Williams obviously is stronger in absolute terms but is he really pound per pound stronger than Norris? I can’t see how yet according to the WIlks formula above he is.

  5. Pingback: Powerlifting Formulas – Is Wilks Best, and What Are the Alternatives? | Taylor's Strength Training

    1. Tim

      Hi Joe – thanks for your comment. So just to be clear you are arguing that weight classes don’t matter and we should just look at absolute strength when talking about how much weight was lifted?

  6. Matthew Beeler

    This is an old thread so someone probably will not see this, but the relationship between strength and body mass is not linear. Astrand and Rodahl already showed this. Therefor, trying to look at two lifters performance by just comparing someone that is lighter and someone that is heavier is not possible and is a moot point. It is mere speculation. You need a regression equation to compare the two, which is what Wilks did. He did not publish his work, but he did a 5th order polynomial regression. The allometric scaling has been found to favor middle weight lifters, while underestimating lifters in the light and heavy categories. Its called Heteroscedasticity. Also, allometric scaling that has been used cannot generalize over all lifts, each lift needs its own scaling factors. It is also only relevant for the subject population that was studied. Clearly, a large scale study still needs to be done to generate a better formula, however the Wilks has been proven to be valid:

    Results of data analysis regarding the Wilks formula indicate that: 1) there is no bias for men’s or women’s BP and TOT; 2) there is a favorable bias toward intermediate weight class lifters in the women’s SQ with no bias for men’s SQ; and 3) there is a linear unfavorable bias toward heavier men and women in the DL. Furthermore, the allometric approach indicated a bias against light and heavy men and women which may be considered acceptable given that half as many lifters are found in the lightest and heaviest weight classes as in the intermediate weight classes.
    In VANDERBURGH, PAUL M.; BATTERHAM, ALAN M.

    Medicine & Science in Sports & Exercise
    Issue: Volume 31(12), December 1999, p 1869

    Conclusion: As used currently (BP and TOT only), the Wilks formula appears to be a valid method to adjust powerlifting scores by body mass.

    1. Tim Henriques

      Thanks for the post. I agree that strength doesn’t increase linearly in proportion to bodyweight, but use a simple formula to find a flaw.

      Just take these simple numbers.
      Lifter A is 200 lbs and lifts 400 lbs
      Lifter B is 300 lbs and lifts 455 lbs

      Both of these lifters are lifting well over 1.5 x bw, and yet the Wilks favors adding 100 lbs of mass for only a 55 lb gain in strength (that is now .5 x bw or the new strength is less than a 1/3 of the ratio of what they could do before). That is ridiculous. Clearly a 200 lb lifter lifting 400 lbs is a better lift than a 300 lb lifter lifting 455, yet Wilks says otherwise.

      And it looks like the IPF agrees as according to this they plan on no longer using the Wilks formula in 2019 and beyond: https://www.reddit.com/r/powerlifting/comments/9mecd6/wilks_formula_to_be_replaced_in_the_ipf/

  7. Justin

    Who cares about Wilks. Lifter B won, end of story, if lifter A doesn’t like being weak then put on 110lbs and see how good your bench is. Usain Bolt is the fastest man alive cause he runs the fastest, not because you put his times and weight into a formula.

  8. Van Wolfe

    The only thing that matters is who can move more weight, not who can move in relation to body size. No one cares what the proportion is. The strongest is the one that lifts more. End of story.

    1. Tim Henriques

      It is fine to say “you” don’t care who can lift more in relation to body size, but it is silly and inaccurate to say “no one” cares. Powerlifting and Olympic Weight Lifting have weight classes. Most competitions award a best lifter and no federation simply uses the most weight lifted to calculate that. If a lifter weighed 450 lbs and benched 400 lbs, it would seem odd to me that lift would be more impressive than a 130 lb lifter benching 390 lbs. Ed Coan isn’t considered the GOAT because he could lift more than anyone else; but because he was so dominant in his weight class, because he could hang with the SHW despite being much lighter than they were.

  9. Barbara Haiden

    I have just been comparing wilks scores with raw, equipped, and multi-ply lifters. In the case of lifters having the same totals the wilks scores were the same weather they were equipped or not. Example my last raw score was 1025 as a 69 year old female, and my wilks 2020 was 527. something. When I changed the rankings to single ply and multi-ply of approximately the same totals 1025, the wilks 2020 scores were 527. something, the same as raw. I can understand the body weight vs strength argument but totally disagree with raw and equipped scores. A person totaling 1025 multiply could not come near that as a raw lifter. I really believe that wilks scores should be calculated within the same category. I could loose a best lifter award from someone who lifted 10 lbs more than me as a raw lifter, by someone in a single or multi-ply suit who’s total may only be 700 to 800 if they competed raw? Maybe I am wrong, I don’t know, but the way I am looking at it it doesn’t seem that I am. Let me know. Thank You

    1. Tim Henriques

      I agree with you, I think it would be a mistake to use the same formula but apply it across raw, single ply, and multiply platforms. We don’t have enough data to discern exactly how much gear adds to one’s performance to make that reasonable, and even more significantly sometimes gear helps someone a lot and sometimes it doesn’t help them that much.

      1. Barbara Jean Haiden

        Just a quick reply. Many years ago, back in the 80’s I did compete equipped and for me it did make a lot of difference in my lifts, maybe up to 150 pounds on my total on a good day. I only do raw now for the past 8 years. Equipped allows me to increase the weight on my lifts to a point that it just hurts my old joints and bones to much any more. It’s not worth the pain and maybe hurting myself really bad just to get larger numbers. By the way, going to a meet and compete tomorrow at 71 now and plan to be over 1000 again. Thank you for your reply Tim.

  10. Pingback: The Confusing History of Strength Co-Efficients - Physical Culture Study

Leave a Reply

Your email address will not be published. Required fields are marked *