Introduction To The Lynch Formula

(WEBMASTERS COMMENTS:  The following story was given to me by Roger Davis some time ago.  I just “found it” again as I was cleaning off a portion of my desk looking for something else.   As most of you know,  the Lynch Formula is the formula we use in the USAWA & IAWA to make the correction adjustments for lifters of different bodyweights.  I’m not even sure what publication this article is from, but it does outline the ideas Ian Lynch had when he developed the Lynch Formula.  It appears to have been written in 1988, which is about the time we started using the Lynch Formula.  I’m still thankful to Roger for keeping this information with the original Lynch Formula so the Lynch Formula Factors could be extended to lifters of higher bodyweights, which we did a couple of years ago.  I won’t rehash the fairness of the Lynch Formula at this time.  That was covered in depth a couple of years ago, and for those of you interested, those blogs are still on the website.  This story should have been ran at that time, but I’m doing it now so it will be saved on the website before I lose this paper again.  I know it is a little boring if you are not the mathematical type, and if you don’t want to read all of it that is fine.   Come back tomorrow and I promise to have some entertaining  ”feel good” piece full of fluff by Thom Van Vleck!!)

by Ian Lynch (October 1988)

The O’Carroll Formula is familiar to most Guild members, and is used to handicap lifters of different weights.  In devising the Formula, Mike O’Carroll used both statistical and physiological evidence to arrive at a fairly complex mathematical function:

Y + (75 – 35)1/3  / (B – 35)1/3

Y is the O’Carroll coefficient used in tables
75 is to make the coefficient 1.000 at this weight.
B is the assumed weight of the lifter’s “non-muscular” mass, eg. bones, brain, etc.

Lynch Formula Graphs

There are many other Formulas, eg Austin, (the original one used in Britain) Lietzke, Vorobyev and , of course, Schwartz.  The strength of the O’Carroll system is that Dr. O’Carroll looked at the physiological reasons to arrive at a basis for the mathematics then cross-referenced this against actual performances to arrive at the constraints, eg. 35 kg.  The Formula is tried and tested and was worked out before the drugs era, which is important since it is difficult to say how “Smartie-taking” might change the physiological factors on which the Formula is based.

Unfortunately for the Guild, however, the Formula was designed before even a 52 kg. weight class appeared, and is very unreliable at weights below this.  This is because the 35 kg. “non-muscular” component of a person’s body is not, in practice, constant.  If it was, anyone weighing 35 kg. would be a totally non-muscular skeleton like me or, in Cookie’s case, a 35 kg. tub of lard, and indeed, no one would weigh less than 35 kg.  In the Guild we have opened up competition to more ladies and younger people who, invariably, are lighter than the weight classes provided for by the formula.  To combat this I have taken a small liberty with the Formula.  Instead of assuming that the non-muscular weight is constant, I have assumed it to be a non-linear function.  The particular function was chosen because it means that the final curve produced fits very closely to the O’Carroll curve at greater weights than 52 kg., but at lighter weights produces realistic allowances down as far as we likely to need.  See Fig. 1 .  There is no deliberate physiological reason for choosing the function I have used other than it fits experience and data so far available.  I suspect that there are too many other factors such as age, sex and such like, to arrive at a simple system that is perfectly fair to everyone, but I feel we should make every effort to develop good practice  to cater for as many as possible.

For those who like maths, I have replaced the constant 35 by 39.53 – (300/w) – (3000/w^2), in order to preserve the 1.0 coefficient at 75 kg. and modify the curve as illustrated in Fig. 1.  Fig. 2 shows how what was previously a value fixed at 35 kg. varies with the weight of the lifter.  In practical terms it means that a 40 kg. lifter is assumed to have a non-muscular weight of about 30 kg., a 75 kg. lifter 35 kg. (as in the old system) and a 120 kg. lifter about 37 kg.  This marginally helps lifters heavier than 75 kg. and marginally hinders those less than 75 kg.  I have stuck to kgs, but it would not be difficult to convert this to pounds if required. 

*** Ed’s Note:   Curious isn’t it, that someone like Ian who, you must agree, exhibits at the very least a modicum of Intelligence, has a non-muscular constant (head) of 2.25 kg.  However, this was mostly bone as brain mass wasn’t discovered and the question was mooted that perhaps he was a Scots Powerlifter.  Close.  Still, when his teaching days are over he always has his legs to fall back on as those who have enjoyed the dubious pleasure.

Right.  now we are even on insults.

Seriously, we cannot thank Ian enough because his expertise will give us a greater platform on which to base the accuracy of our results in the coming years.  We have used the O’Carroll Formula in every aspect of our activities and now we will use the”LYNCH FORMULA” with the same degree of confidence and to the same satisfying effect.  For the moment – at least – I am only publishing the new figures in Kilos.  If it becomes a trial to those applying the new system, then I’ll publish in Imperial Pounds, but I’d prefer if everybody used the metric Kilos from now on – as a matter of course – for uniformity and ease of application

Is the Lynch Formula Fair??

by Al Myers

There has been “lots of talk” regarding the Lynch Formula recently.   Most of this centered around the fact that the Lynch Formula has just been expanded to contain factors for lifters that weigh over 138 kilograms.  Now the Lynch Chart goes to 180 kilograms.  The Lynch Formula has been the “adopted formula” of the USAWA and the IAWA since the early 90’s to calculate adjusted points in determining weight lifted to bodyweight comparisons in scoring.  The Lynch Formula creator, Ian Lynch, developed and modified his formula to apply to the lifts done in All-Round Weightlifting.  As far as I know, no other lifting organization uses the Lynch Formula.  So, you could say, that we have a Formula that tailors to our specific lifting sport – All Round Weightlifting!   I have never really heard the reasons how the Lynch Formula was derived.  Most other weightlifting formulas are derived from a data set of numbers, usually records or performances of lifters of different bodyweights.  I know this is how the Sinclair Formula was derived  in Olympic Weightlifting.  It has even been changed and modified over time when it is “re-evaluated” using new data, and new factors are created to maintain the fairest formula possible.  However, this is easier to do when you are analyzing only two lifts (the Snatch and Clean and Jerk) than when you are looking at over 200 lifts, like we have in All-Round Weightlifting. I find it hard to believe that Ian Lynch used any data involving All-Round Lifts when he developed his formula.  Afterall, what data involving All Round Lifting was available 20 years ago?

The big question always arises, is the Lynch Formula fair?  I have several larger lifters in my gym who feel that it isn’t, and that the Lynch Formula favors the lighter lifter.  But then I hear from light lifters who say it favors the heavier lifters.  And when the fact is pointed out that the  past several years  the Overall Best Lifter at the IAWA World Championships has weighed over 105 kilograms,  they have a good argument.  I always try to be as open-minded as possible, and I like to have the FACTS before I form a hard opinion on something.  This is why I performed my own self-study on this – to answer that question to myself.   In no way is this information I am presenting you a scientific study that has any statistical significance.  I am making that disclaimer LOUDLY, so my statistics friends like Tom Ryan (who is way smarter than me in matters like this)  won’t point out my deficiencies in the methods of my study.  This study is entirely just a compilation of data that must be taken on surface value.  But it is still VERY INTERESTING and should provide the best factual support  regarding the fairness of the Lynch Formula that has ever been available.

Study – Determining the Fairness of the Lynch Formula

Objective: The objective of this study is to evaluate the fairness of the Lynch Formula in regards to correction factors for bodyweight adjustments.

Design: The USAWA Record List will be used as the data source of information that will be evaluated.  The USAWA record list has accumulated information on records in various lifts for over 20 years.  Twenty lifts will be selected (the Heavy Lifts will be left out).  The lifts selected will be the ones that have the most records established in them through all weight and age classes. Three weight divisions will be arbitrary selected – lightweight lifters (80 kilogram class and below), middleweight lifters  (85-100 kilogram classes), and heavyweight lifters  (105 kilogram class and above).  The best record according to Lynch Formula will be selected from each weight division.  These three divisions will then be ranked according to the best lifts according to the Lynch Points, and all points will be added up to determine which weight division has the best ranking, and thus assumed to receive the biggest advantage from the Lynch Formula.

Assumptions: Since individual bodyweights are not known from the USAWA Record List, the weight of the weight class will be used in calculating Lynch Points.  Lifters in the 125 kg plus class will be assigned the Lynch Correction for 130 kilograms bodyweight. This may be an underestimate of the actual bodyweights of superheavyweight lifters, and if so, would provide numbers that would artificially elevate the lifts of SHW  lifters in regards to Lynch Points (NOT an advantage for heavy lifters).   Also, the assumption is made that the record lifts are representative of the average lifting ability of all lifters in these bodyweight classes. By picking the 20 lifts with the most records, it is assumed that these are the 20 all-round lifts that are performed the most, thus providing the best data base of numbers available from the Record List for evaluation.

Results:

Lift Lightweight

(80 K class and below)

Middleweight

(85 K to 100 kg class)

Heavyweight

(105 K class and above)

Bench Press

Feet in Air

320# – Smith

(70K)

LP – 320.0 points

480# -  Succarote

(100K)

LP – 406.6 points

441# – Meek

(125+K)

LP – 327.2 points

Clean&Jerk

Right Arm

132# – Zaremba

(75K)

LP – 132.0 points

160# – Bryan

(85K)

LP – 148.4 points

175# – Burtzloff

(125+K)

129.8 points

Clean&Press

Heels together

226# – Hirsh

(80K)

LP – 217.2 points

248# – Bryan

(85K)

LP – 230.0 points

300# – Meek

(125+K)

LP – 222.6 points

Cont Snatch 220# – Waterman

(70K)

LP – 229.9 points

248# – Bryan

(85K)

LP – 230.0 points

265# – Ciavattone

(125+K)

LP – 196.6 points

Continental

to Chest

325#- Waterman

(70K)

LP – 339.7 points

380# – Anderson

(90K)

LP – 431.1 points

385# – Conners

(125+K)

LP – 285.6 points

Continental

Clean&Jerk

287# – Waterman

(70K)

LP – 299.9 points

320# – Bryan

(85K)

LP – 296.8 points

369# – Anderson

(105K)

LP – 304.6 points

Cheat Curl 190# – Gazda

(60K)

LP – 220.8 points

235# – Anderson

(90K)

LP – 210.9 points

260# – DelSignore

(105K)

LP – 214.7 points

Deadlift

2 bars

463# – McKean

(80K)

LP – 445.0 points

610#- Schrock

(100K)

LP – 516.7 points

600# – Myers

(115K)

LP – 473.3 points

Deadlift

Heels together

560# – Hirsh

(75K)

LP – 560.0 points

605# – Schrock

(100K)

LP – 512.5 points

650# – Myers

(125K)

LP – 491.5 points

Deadlift

Rt Arm

369# – McKean

(70K)

LP – 385.6 points

402# – Ullom

(100K)

LP – 340.5 points

562# – Ciavattone

(125+K)

LP – 416.9 points

Deadlift

TrapBar

600# – Hirsh

(80K)

LP – 576.7 points

635# – Schrock

(100K)

LP – 537.9 points

661# – Myers

(115K)

LP – 520.9 points

Hack Lift 670# – Hirsh

(80K)

LP – 644.0 points

605#- Anderson

(90K)

LP – 543.0 points

620# – Schrock

(105K)

LP – 511.9 points

Jefferson

Lift

702# – Hirsh

(80K)

LP – 674.8 points

601# – Schrock

(95K)

LP – 523.5 points

601# – Spayd

(105K)

LP – 496.2 points

Pullover

& Press

287# – Hirsh

(80K)

LP – 275.9 points

275# – English

(90K)

LP – 246.8 points

352# – Myers

(115K)

LP – 277.4 points

Pullover

& Push

331# – Crowe

(80K)

LP – 318.2 points

446# – Anderson

(90K)

LP – 400.3 points

474# – Burtzloff

(110K)

LP – 382.0 points

Snatch

Rt Arm

127# – Waterman

(70K)

LP – 132.7 points

160# – Bryan

(85K)

LP – 148.4 points

171# – Burtzloff

(110K)

LP – 137.8 points

Front

Squat

355# – Fleischer

(80K)

LP – 341.2 points

441# – Bruner

(95K)

LP – 384.1 points

495# – Meek

(110K)

LP – 398.9 points

Steinborn 325# – Monk

(70K)

LP – 339.7 points

375# – Schmidt

(100K)

LP – 317.7 points

441# – Ullom

(110K)

LP – 354.6 points

Swing DB

Rt Arm

120# – Smith

(75K)

LP – 120.0 points

120# – Schrock

(100K)

LP – 101.7 points

150# – Ullom

(110K)

LP – 120.9 points

Zercher 504# – Hirsh

(80K)

LP – 484.4 points

500# – Anderson

(90K)

LP – 448.8 points

529# – Moore

(120K)

LP – 408.1 points

NOTES:  LP stands for Lynch Points.

Summary: Overall points were scored on placings with 1 point given for first, 2 points for second, and 3 points for third.  These points were then “added up” to give total points for the 20 selected lifts, which would give the low overall score  as being  the best.  The lightweight division had 40 points, the middleweight division had 38 points, and the heavyweight division had 42 points.  The lightweight division had 6 “firsts”, the middleweight division had 8 “firsts”, and the heavyweight division had 6 “firsts”.  Also, the Lynch Points were added for each division to give another comparison.  The lightweight division had 7057.7 points, the middleweight division had 6885.7 points, and the heavyweight division had 6671.5 points.

What can be interpreted from all this??

The “total points” are really not that much different.  A couple of points either way could easily be said to be an “acceptable tolerance”.  All it would take is one of those records broken and it could “sway” back slightly the other way. The differences between the divisions (in regards to points)  are not enough that anyone could make an argument one way or the other.

My opinion is that Ian Lynch was pretty much “right on” in regards to fairness to all bodyweights using his formula.  Whether he did this using  scientific calculations, or merely having “luck” in picking the right correction factors doesn’t really matter.  The evidence of comparing the Lynch Formula to over 20 years of collected data in the form of USAWA records prove to me that his formula is very fair and one we should remain using.   Of course, it is easy to pick out certain lifters that obscure the data due to their very exceptional lifting within their class.  Bob Hirsh is a prime example as he greatly distanced himself from the others in the Hack Lift and Jefferson Lift.  His Jefferson Lift record outscored the next lifter by over 150 Lynch Points, the biggest variation of all the lifts recorded in this data set.  But there are other lifters in the middleweight and heavyweight classes who are  “in a class of their own” also.  Everything averages out.  I was also concerned that the weight classes on the fringe of the lightweight and heavyweight classes (the 80 K and the 105K) would be overly represented, and thus tend to discredit the ranges I picked for this study.  However, this was not the case as you can see from the results  that the lighter lifters (70K and 75K), as well as the heaviest lifters (the 125+ lifters) were often represented as having the BEST lifts within their division. Only one 60K lifter made the list (this is not a largely represented class at meets), and he ended up having the BEST Lynch corrected Cheat Curl.  Geoff Gazda’s 190# Cheat Curl in the 60K class outscored Antonio DelSignore’s 260# Cheat Curl in the 105 K Class, 220.8 points to 214.7 points.  One 125+ K class lifter had the TOP Lynch Score among all divisions.  Frank Ciavattone and his 562# One Arm Deadlift ranks above all the others.

I welcome any comments regarding this study of mine.  You can either address them on the USAWA Discussion Forum or you can email me directly.

Lynch Formula Expanded

by Al Myers

Chris Bass is the "brains" behind the IAWA. Chris is often the official scorekeeper at International Events, and he maintains the IAWA World Record List, which is an overwhelming task. Chris was very influential in helping solve this problem with the Lynch Formula.

A  project that I have been working on for the past several months has been trying to expand the Lynch Factor Chart to allow bodyweight adjustments for heavier lifters.  This has been a big problem with the Lynch Factor Chart in the past – it only went up  to 138 kilograms bodyweight.  This is how it always has been since the original calculations were done.  Nowadays, it is not uncommon to have lifters that weigh over this amount, and NO LYNCH FACTORS have been available for them.  It forced the Meet Director to either just use the highest Lynch Factor available on the Chart, or “make a guess” at a factor for them, either case not being the way it should be done.

Efforts were made to acquire the original Lynch Formula or extrapolate the current Lynch Factors to higher bodyweight factors.  Both of these ideas were unsuccessful – UNTIL NOW!!  I brought this problem up at the IAWA World Council Meeting in Glasgow, Scotland, and a committee of myself, Chris Bass and Roger Davis were formed to look further into it and find a resolution.  I am glad to say that Roger Davis has found the original Lynch Formula, which the Lynch Formula creator Ian Lynch had given to him many years prior.  Through the works of Roger, he placed the formula in an Excel Spreadsheet and generated the new, higher numbers, which are needed to make the Lynch Factor Chart complete. The chart now goes to 180 kilograms bodyweight, with all the new numbers generated from the same formula that the previous numbers were.  For those interested, the mathematical expression of the Lynch Formula is:

=((75-(39.53-(300/BW)-(3000/(BW*BW))))^(1/3))/(BW-(39.53-(300/BW)-(3000/(BW*BW))))^(1/3)

This is indeed a much more complicated formula than I thought it would be originally!!  I have expanded the USAWA Lynch Formula Chart with the new values.  It will always be  available on the USAWA website – under the  page titled  “Scoring Information”.  I am the type of person who likes to solve problems quickly, and then immediately take action to implement the solution.  So, I presented this to the USAWA Executive Board for approval, which would allow the USAWA to implement this new Lynch Chart right away.  The USAWA Executive Board, of myself, Denny Habecker, Dennis Mitchell, Chad Ullom, and Scott Schmidt,  approved it unanimously.  This means the USAWA, as of now, will use this new Lynch Factor Chart in our scoring.  The IAWA will need a membership vote, which only occurs when there is a World Council Meeting.  This means it may be next year before the IAWA approves this new Lynch Chart, unless a Council Meeting is called for at the 2010 Gold Cup next month.  Regardless, the committee of myself, Roger, and Chris will present it for IAWA approval at the next Council Meeting.

The new Lynch Factor Chart  – Lynch Factor Chart

I want to especially thank the efforts of Roger Davis and Chris Bass.  Without the efforts of these two, this BIG improvement in our scoring system  would not have been realized.

History of Formulas used in the USAWA

by Al Myers

When you KNOW it's time to blame the formula!

This past week on the USAWA Discussion Forum a lively debate got started on the fairness of using Formulas in comparing athletes for overall placings. The USAWA is unique in doing this compared to other lifting organizations which prefer to just give awards for different divisions or classes.  Sure these other organizations might use a Formula to award a Best Lifter, but the USAWA uses a formula to determine the overall ranking of every athlete in the competition, from the top to the bottom.  No OTHER organization does this!!  We now use the Lynch Formula to make the “handicap” adjustment for bodyweight differences and use a Age Allowance Percentage for Junior Lifters and Lifters age 40 and above.  I am NOT going to give my viewpoint and opinion on the fairness of using formulas  in this article, as that is better left for the Discussion Forum.  Instead, I would like to review the history of formulas used in the USAWA, and give insight to how these different formulas were derived.

The main All-Round competitions that occurred prior to the USAWA (1987)  were “odd lift” meets promoted by Bill Clark out of Missouri.  These meets were contested under the direction of the Missouri Valley Weightlifting Federation, the Region IV division of the USWF.  The formula used at that time to determine rankings was the Schwartz Formula. Numerous old Zercher Meet results verified this. The USAWA really began in the summer of 1987, with the first official records recorded for the USAWA in the fall of 1987.  At this point the USAWA adopted the use of the O’Carroll Formula for bodyweight adjustment and it was used extensively in the USAWA in 1988.  The Zercher Meet in 1988 used the O’Carroll Formula.  The National Masters Weightlifting Program started using the Sinclair Formula at this time, and even used a unique formula developed by Joe McCoy that adjusts for bodyweight AND age at the 1987 National Masters Olympic Lifting Championships directed by USAWA Hall of Famer John Vernacchio. It was called the Sinclair-McCoy Formula and ONLY applied to Olympic Lifting totals.  The FIRST USAWA National Championships directed by John Vernacchio in 1988 used the O’Carroll Formula. I did find a few old USAWA meet results from the late 80’s where the Sinclair Formula was used for All-Round Meets.  The Sinclair Formula was developed by Canadian mathematician/weightlifting enthusiast Roy Sinclair.  He used the weightlifting results from the Olympics as his data base to determine the coefficients for his formula. Another interesting formula brought forth from the IAWA in the early 90’s was the Blindt Formula, developed by British lifter Adrian Blindt.  It didn’t correct for bodyweight, but for the lifts involved.  Each lift had its own factor.  The idea was this would make it more fair, in example, to compare a lift where lots of weight can be lifted, like a Hip Lift to a lift where much less weight is lifted, like a Press.  This formula was never used in the USAWA, but was used in some IAWA competitions.  I remember it was used in the IAWA World Postal Meets promoted by the Australians a few years back.

However, by the early 1990’s most all USAWA results started using the Lynch Formula, which we still use today.  The Lynch Formula was developed in 1988 by Ian Lynch, a lifter from England. The Lynch Factors (or coefficients)  have not changed since its inception.  I know this because I found an old Lynch Chart from the early 90’s and compared it to today’s Lynch Chart and it’s the same.  No updates and no modifications.  Recently, we have been faced with a problem using the Lynch Formula, and that problem is the highest bodyweight factor on the Lynch Chart is 150 kilograms.  It is not uncommon nowadays to get a lifter that weighs over 330 pounds, and we have no way to correct for them!  In the past when this happened the meet director would either  “estimate” a Lynch Factor for them or just give the athlete the highest bodyweight correction on the chart. I don’t think EITHER of those solutions are acceptable – and thus I began to try to find the “original Lynch Formula” so the chart could be extended for higher bodyweights.  I inquired several places and couldn’t find any leads.  Finally, thanks to Tom Ryan and our IAWA President Steve Gardner, Ian Lynch was located and I could go right to the source of the Lynch Formula!  Unfortunately, the exact formula and method of reproducing it has been lost.  However, Ian Lynch was very helpful in sharing some details and information on why the Lynch Formula was developed and used in All-Round Weightlifting.  The Lynch Formula is very similar to the O’Carroll Formula with the differences being that the Lynch Formula is slightly more favorable for lighter lifters. In a sense, it “leveled out the curve”  on the light end of the  O’Carroll Formula.   Both the Lynch and O’Carroll Formulas were derived using body factors whereas the Schwartz and Sinclair Formulas are based on Olympic lifting performances or World Records for Olympic Lifting.  I think this makes the Lynch Formula more applicable to All-Round Weightlifting. It sure doesn’t make sense to me to use a formula based on the two Olympic Lifts, and then expect it to correlate to over the 200 lifts we do in All-Round Weightlifting!  Ian Lynch had this to say about the development of the Lynch Formula from the O’Carroll Formula considering body factors , ” The O’Carroll Formula assumed all lifters non-muscular weight was constant, ie bones etc. That didn’t work well particularly for lighter lifters.” That must have been the reason for the points being adjusted slightly in favor of lighter lifters.

Just out of curiosity I “recalculated” the results of last year’s World Championships using the Sinclair and O’Carroll Formulas.  This is how it would change “the top ten”:

Lifter BWT Lynch Pts. Sinclair Pts. O’Carroll Pts.
Mark Haydock 122.9 764.3 1005.2 771.1
Al Myers 114.7 763.1 978.5 768.6
Chad Ullom 104.3 749.8 936.2 758.9
Roger Davis 81.6 736.6 896.9 737.0
Denny Habecker 86.1 661.1 805.3 661.9
John Monk 79.8 658.3 802.1 658.4
Bill Spayd 107.9 655.3 825.0 659.1
Scott Schmidt 119.7 598.6 779.3 603.6
Art Montini 78.2 588.9 718.3 589.1
Josh Haydock 66.9 582.2 724.6 565.7

This group of lifters make up an interesting data base for this comparison, because lifters of different body weights are represented (from 66.9 kgs to 122.9 kgs).  Not much changes in the placings between the three formulas being applied except for Bill Spayd.  He placed 7th overall using the Lynch Formula, but would have been 5th using the Sinclair Formula. You can really see how the Lynch Formula favors lighter lifters compared to the O’Carroll Formula.  Just compare Mark Haydock (at 122.9 kgs) to Josh Haydock (at 66.9 kgs). Mark’s Lynch Points are LOWER than his O’Carroll Points while Josh’s Lynch Points are MORE than his O’Carroll Points.  John Monk, at 80 kilograms, has the bodyweight that gives the same points using both the Lynch Formula and the O’Carroll Formula.  The Sinclair Formula MUCH favors heavier lifters.  Just look at the top four placings where the bodyweights of the lifters decrease with each placing. It is  pretty close using the Lynch Formula, but with the Sinclair Formula the point gap widens much more with each subsequent placing.

This doesn’t answer the long debated notion that “formulas are not fair”, but I hope that it provides some insight to how the formulas work and why we use them in the USAWA.