CrossFit Weakness Bias

Written by Dan Williams
Accredited Exercise Physiologist
BSc. BEx. Rehab. Sc. MAAESS
Originally Published in the CrossFit Journal, 2010.

The CrossFit methodology aims to create a broad, general, inclusive fitness by increasing work capacity across broad time and modal domains. Constantly varied, high intensity functional movements (and workouts) have been proposed as the best vehicles to reach this outcome. There is no doubt that standard CrossFit programming offers CVHIFM – though is this the best way to make the time and modal domains as broad as possible? What if there was another way?

What if the high intensity functional movements weren’t so constantly varied?

The essence of CrossFit lies in its ability to define the previously undefined. Definitions have been pioneered for such concepts as fitness, health and work capacity. Let’s consider CrossFit as an adjective rather than a noun. I am CrossFit, as opposed to I do CrossFit. This also deserves defining. What is it to be (Cross)fit? I propose that you are (Cross)fit if you are generally physically prepared. For what? You guessed it, the unknown and the unknowable. Everything. It comes back to that randomised physical task that you would least like to see come out of the hopper. Your performance in this task is your true measure of (Cross)fitness.

You are only as strong as the weakest link in your exercise chain. The weight hanging on the end of this chain is your level of General Physical Preparedness (GPP). The more the chain can support, the higher your GPP. If each link in this metaphorical chain represents a component of fitness (cardio/respiratory endurance, stamina, strength etc), the focus of training should be obvious. The first link to snap and drop your GPP is the weakest link. To increase GPP, our weaknesses should not simply be overcome, they should grow to match our strengths. To quote Glassman in ‘What is Fitness’, ‘You are as fit as you are competent in each of these ten skills’. Perhaps this should be narrowed to state you are only as fit as you are proficient in your weakest link.

We have a CrossFit strength bias, a CrossFit endurance bias and for arguments sake a CrossFit power bias (CF football). So why not a CrossFit weakness bias?

As with the foundational defining and measuring of fitness being the cornerstone of the CrossFit methodology, to have any scientific credence a CrossFit weakness bias also requires measurement. How can this ‘weakest link’ fitness be quantified? We measure fitness and work capacity as the ‘area under the curve’. To measure fitness as a product of our weaknesses, we may be better measuring the volume inside a circular curve. Picture a wheel with six spokes. Each of these spokes represents a ‘trainable’ (rather than ‘practicable’) component of fitness. By measuring the proficiency in each of these six areas, we are able to score each area and assign it a position on the spoke. This position indicates our relative competency in that component of fitness using a reverse scale, with one being a skill in which we are highly proficient, and ten being a skill at which we are not. The measurement of each component is an issue requiring further thought and the methods by which this can be done are numerous.

One such method may be similar to the scoring system used by CrossFit Seattle in their ‘Athletic Skill Levels’. In a simpler model, the scoring could even be made subjectively, with the athlete simply assigning themselves a perceived score. Once all areas have been scored (whatever the method), the indicated points of the spokes are joined to create our circular curve. So far, this offers us the same information as the ‘volume under the curve’ model. Where this model differs is that (according to the assumption that we are only as fit as we are competent in our weakest link) rather than joining the dots on each spoke, we simply draw a concentric circle which passes through the point on our weakest scoring spoke. The volume inside the curve thus becomes a weakness adjusted level of fitness.

The current model of broad, general, inclusive fitness increases all 10 components of fitness evenly. Therefore, relatively speaking, your existing deficiencies will still apply. Relative to your own fitness, certain components will be just as deficient as they once were – you will be better at everything, though still better at some things than others.

If standard dot com CrossFit programming chases a ‘broad, general, inclusive fitness’, and the result of this chase is the evenly distributed improvement of the ten components of fitness (read: broad time and modal domains), then how do we define the goal of a theoretical CrossFit weakness bias. Is it broad, yes – just not as often. Is it general, yes – just not as often. Is it inclusive, yes – as inclusive as standard programming. The variety still exists, though instead of all session types (strength/metcon etc.) occurring at a similar frequency, an individual’s strengths relegate the sessions they excel at to outliers (frequency wise), thus increasing the frequency of sessions containing their weaknesses. Sessions the individual specialises in (those dwelling on the outside of our wheel) become less frequent and sessions where they are weak (on the inside of the wheel) form the mainstay of programming. And so we depart from Specific Physical Preparedness, and become truly Generally Physically Prepared.

Now sure, a marathon runner beginning CrossFit would experience a drastically sharper curve of improvement in strength than in cardiorespiratory endurance, just as a powerlifter would with cardiorespiratory endurance. But regardless of this narrowing of the void between components of fitness, the void still exists. Taking a less extreme approach, the same applies. Until the programming spits out career CrossFitters (a product purely of time and with the current sponsorship trend a matter of ‘when’ not ‘if’) who begin their athletic careers in the pursuit of GPP (our current CrossFit kids), all new CrossFitters will possess some form of specialisation. This is a reflection of their previous physical pursuits, whether it be strength from a bodybuilding background, flexibility from a gymnastics background, power from a contact sport background or respiratory endurance from a track background. Even when these career CrossFitters do develop, established athletes with their established specialisations will continue to enter the program and the ‘void’ will always be an issue. A CrossFit weakness bias can fill this void.

Much has been made of the specialist athlete with a ‘10’ in one of the components of fitness and a ‘1’ in the other nine domains. There is no doubt that this athlete is un(Cross)fit. However the reverse may also ring true. An individual with ‘10’ in nine domains and ‘1’ in one domain would beat the first individual in the majority of cases, but if the weakness did come out of the hopper, they would be as deficient as the specialist. Think of Annie Thorisdottir and her muscle ups in the final event of the 2009 Games, or Khalipa in the run (even if he was just having a bad day), or the athletes DNFing due to an inability to do strict handstand push-ups in the triplet. In the majority of cases the misshapen wheel would turn further for our nine ‘10s’ and one ‘1’ athlete, but eventually, it would get to the deficiency (as is the case with the examples above) and cease to turn. The wheel of a deficient CrossFitter can never make a complete revolution. In both cases, the weakness adjusted level of fitness would be the same – that is, the circle through their lowest scores would intersect ‘1’ on each spoke (both individuals overlayed on the same graph with matching weakness adjusted levels of fitness). A specialist athlete would still get better at everything (as per Glassman’s proposition that training GPP will improve the performance of specialists more than a reliance on SPP), it is just that the rate and degree of improvement of the weakness will be greater.

A weakness bias does have an end point – the ‘smoothing of the wheel’. When the weakness bias has delivered us to a point where the wheel is circular and the resultant figures across broad time and modal domains are identical or near identical, we are then at a point where ‘broad, general, inclusive fitness’ takes over to become the dominant training methodology. The criteria for progression is a round wheel. The volume of the circle matters not – the uniformity of shape does.  Until this point, a weakness bias will strength the GPP supporting weak link. Here lies the beauty of the relationship between standard CrossFit programming and a weakness bias. A weakness bias is not a replacement for standard programming, just as supplemental strength training is not a replacement for dot com WODS. It can be used as a preparatory program, a short term ‘top up’, or at the other end of the scale, as a weakness destroyer leading up the CrossFit Games. A weakness bias smooths the circle; constantly varied, high intensity functional movements increase its size.

With the concept of a weakness bias discussed, what remains to be seen is how the programming should occur to create this bias, and deliver an athlete to the point where CVHIF movements and WODS take over. There are several mechanisms by which the programming for a CrossFit weakness bias could occur, and the beauty of this empirically driven peer reviewed journal is that if a weakness bias is accepted as a possible training method, the methodology will be organically developed by the community. I propose just one possible example.

  1. Score relative competency in each category (of the six trainable components of fitness) in a reverse scale ie: one being strongest, 10 being weakest. An example for a typical endurance athlete may be: cardiorespiratory endurance = 2, stamina = 2 strength = 8, flexibility = 3, power = 8,  speed = 8.
  2. Then add the scores for each category together ie: 2+2+8+3+8+8 = 31.
  3. In determining the programming, the score for each number reflects the frequency that that component of fitness will appear. Eg: Cardiorespiratory endurance biased sessions (a strength of our hypothetical athlete) will occur twice out of every 31 sessions, while power biased sessions (a weakness) will occur eight times out of every 31 sessions.

Granted, the individual components of fitness are by no means targeted in isolation in WODs, on  the contrary. Fitness created in a vacuum is useable in a vacuum. However, a high score (thus a weakness) in strength may increase the frequency of ‘Linda’ (heavy deadlifts, bench and cleans) or ‘Diane’ (deadlifts and handstand push-ups) type sessions, while a high score (weakness) in respiratory endurance may increase the frequency of ‘Helen’ (run, KB swings and pull-ups) or ‘Kelly’ (run, box jump, wall ball) type sessions. Some components of fitness (balance, coordination, agility, accuracy) are more neural in nature, and are therefore trained by performing movements – with less of an emphasis on intensity. Glassman attributes their improvement to practice. For the remaining components of fitness which are improved through training, the above rings true.

It is worth exploring a practical example of an individual and tracing a hypothetical weakness bias through performance, measurement and programming. Consider the following information for John, a fictional CrossFitter. All information other than Girl times (eg: age, weight, exercise history) has been withheld. From this information alone we should be able to score the individual and prescribe a weakness biased program.
Angie (with blue band assisted pull-ups): 39.25
Cindy (with blue band assisted pull-ups): 7 rounds
Fran (as Rx): 16.45 (all thrusters unbroken)
Grace (as Rx): 6.32
Helen (with blue band assisted pull-ups): 22.01
Karen (as Rx): 22.22 (first 50 unbroken then died)
Linda (as Rx): 25.37

John had respectable times on Grace and Linda. Angie was very slow, his Cindy score was low (even with the scaling), and his times were poor on Karen, Helen and Fran.

From John’s performances on the Girls, we can extrapolate that he is strong, though not relative to his body weight, and has low levels of cardiorespiratory fitness and stamina. He excels in any movements where an external load is used (Grace, Linda and the thruster section of Fran), while falling down in more bodyweight gymnastic type movements (Angie, Cindy and the Fran pull-ups). He also performs poorly in cardiorespiratory and muscular endurance based WODS (Helen and Karen).

From these results, and after a subjective scoring by both the individual and their coach, the following scores can be assigned (remember the reverse scale, 1=highest, 10=lowest). Cardiorespiratory endurance = 9, stamina = 8 strength = 3, flexibility = 5, power = 4,  speed = 7. Discounting flexibility (which is usually trained supplementally), if we add these numbers together (29), we know that CR endurance based sessions should occur nine out of every 29 sessions (30% of the time), stamina based sessions should occur eight out of every 29 sessions (28% of the time), while strength based sessions should only occur three out of every 29 sessions (10% of the time) (table 1).

Now again, I must emphasise that individual components of fitness are rarely targeted in isolation in CrossFit WODS – and nor should they be. There is no argument that Grace is enormously cardiovascular, but a strength athlete will beat an endurance athlete more often than not. These percentages simply give an idea of the types of sessions that should occur in a higher frequency.

With the current fitness levels and their relative deficiencies measured, what remains is the prescription of a weakness biased program, in this case, a bias towards endurance (muscular, respiratory and cardiovascular) and stamina with a lower frequency of strength (external load) and power based exercises. Endurance and stamina can be grouped together as similar session types cause an adaptation in each.

Table 2 contains four cycles (12 WODS and four rest days) of standard CrossFit dot com WODS (From Thursday 15th October to Friday 30th October 2009), with the individualised weakness bias programming to the right. I have endeavoured to maintain the same ‘feel’ of sessions where possible. Based on the percentages, endurance and stamina biased WODS should occur six times out of 12 sessions, there should be three speed based sessions, two power based sessions and one strength based session.

Some comments on the differences between the original programming and the weakness bias programming:

  • The majority of metcon based WODS have been retained – especially those utilising bodyweight and gymnastic based movements.
  • Strength based metcon workouts have been retained in an effort to ‘fortify the strengths’ while still gleaning an endurance/stamina benefit.
  • Low rep max WODS have been removed and replaced with a mixture of phosphagenic, oxidative and glycolytic interval based sessions and cardiorespiratory/stamina based WODS.
  • A more basic way of prescribing a weakness bias when following mainsite WODS may be to have a list of WODS that reflect an individual’s weaknesses. When a WOD comes up that is based around the individual’s strengths, it is substituted for a WOD off this list.

An athlete who CrossFits five to six times a week can justify a weakness bias. Chances are this athlete has the innate drive that personifies our sport. The drive for GPP – and a drive only quenchable by the achievement of broad, general, inclusive fitness. The individual who trains two to three times a week not only has less need for a weakness bias – they cannot afford it. A weakness bias for an individual who trains twice a week will seriously limit that individuals capacity to chase GPP. If the bias was towards strength, then half that individuals programming for that week could conceivably be 1RM deadlifts. The limits of this are apparent. If we use fitness and health as interchangeable terms (a concept towards which Glassman and the Community seem to be trending), the fitness of this ‘part time CrossFitter’ would benefit more from the randomised nature of standard dot com programming.

Traditional exercise (unlike life) favours the specialist. CrossFit (like life) creates and favours the generalist. A CrossFit weakness bias favours the specialist looking to become a generalist. It strengthens the weakest link in the GPP supporting chain of fitness.