Russell Berger vs. Mark Rippetoe on ESPN


“You don’t get to mis-define a thing, and then on the basis of that mis-definition, say that that thing is bad … it’s incumbent upon you, if you’re going to make a criticism of something, that you actually have at your command a working definition of the thing you’re criticizing …” – Mark Rippetoe, 2008.

I recently appeared on ESPN’s Outside the Lines to “react” to Mark Fainaru-Wada’s theatrical and inaccurate critique of CrossFit. I appeared opposite Mark Rippetoe, a former CrossFit subject matter expert in the barbell lifts, who left our organization a few years ago due to personal disagreements. You can watch our conversation here.

Our discussion quickly strayed from Mark Fainaru-Wada’s segment and became focused almost entirely on Rippetoe’s personal issues with CrossFit, which he has written about on T-Nation’s website. Rip’s issues stem from his own misunderstanding of the CrossFit program, which he perceives to be random, and therefore equates with spontaneous human activity (exercise), rather than planned, goal-driven activity (training). Based on his assumption that CrossFit is “random,” Rip claims that CrossFit is simply “exercise.”

Yet Dr. Lon Kilgore, Rip’s co-author and one of the most respected academics in the field of anatomy and physiology, disagrees with Rip’s relabeling of these terms,

“It’s sort of a semantic thing that he was writing about there. Because, previously, he and I had talked about the discriminating between physical activity, which is unplanned, spontaneous human activity… and then exercise was planned movement in order to improve fitness.” 

Dr. Kilgore went on to say that, “He’s using different terms and a different definition. And that was a new use for him, in my past history.”

Regardless of how we define these terms, Rip’s argument is based on an inaccurate assumption about CrossFit programming. Rip defines training as “…physical activity done with a longer-term goal in mind, the constituent workouts of which are specifically designed to produce that goal.”

Based on this definition, it is demonstrably false to claim that CrossFit lacks longer-term goals. CrossFit’s goal is increased work capacity. This is a clear, measurable goal. CrossFit methodology also programs specifically to improve this goal, and does so by making use of constant variance.

“Random” programming would almost certainly fail to produce the greatest work capacity, as true randomness contains clusters and groupings that might cause athletes to miss certain stimuli for long periods of time. For example, 30 randomly picked workouts might only include 1 heavy day and multiple long-distance runs, simply because it’s random. On the contrary, CrossFit programming requires intentionally tracking the variables of training and purposely modulating them to attain the broadest possible training stimulus. This prevents athletes from intentionally, or unintentionally neglecting one type of stimulus in favor of another.

Think of the Hopper Model of fitness. The Hopper contains all of life and sports’ many skills and drills, and is a theoretical and statistical test of fitness. But it is not a model for programming. A truly random program would draw from the hopper with no hierarchy of movement selection. Softball throws would be programmed as often as squats, and neither frequently. CrossFit doesn’t excludes fringe movements like ball throws, shoveling, or even back rolls to support. In fact, CrossFit encourages that athletes regularly seek out new sports and challenges, but CrossFit does emphasize a core group of functional movements for consistent practice and training. Think of how often below parallel squatting movements appear in CrossFit’s benchmark workouts. This is not an accident.

Now Rip may not agree with our equating work capacity to fitness, he may not believe constant variance is the most effective way to improve work capacity, he may just simply not like CrossFit and not like me, but as convenient as it is for him to claim, CrossFit isn’t random.


  1. J

    It’s a shame they cut it off so quickly, was heating up to be a good discussion.

    I think ‘increased work capacity’ is just too general a goal to be considered training. For instance, if somebody walks into a gym and their goal is ‘Increased strength’ and they do a couple sets on most of the machines and then next time do a couple sets on some different machines, popping on whatever weight they think sounds about right with no logical progression, by your understanding they are now training, as they will be a bit stronger. If they continue, it is very likely they will continue to ‘increase strength’ for quite a while, especially if they are fairly unadapted.

    Training is more like I want to add x weight to my squat and I will do that by squatting x on monday, add x weight and squat x on wednesday with the goal of being able to squat x on friday. You should be able to know where you’re at at the moment and have a pretty good idea of exactly where you’ll be over the next while. Somehow I don’t think the people at the commonwealth games have training goals like ‘increased strength’ or ‘increased speed’.

  2. Alex

    Why didn’t Ripptoe just say how he really feels?

    “For casual exercisers, CrossFit-types and the like, the calculation is a bit different. The vomit I see on the internet – complete lumbar flexion, everything pressed out, everything intentionally rebounded from the floor, all done under the watchful eye of some moron saying “Nice!” – makes me of two minds.

    Part of me hopes the fools hurt themselves badly (after all, orthopedic surgeons gotta eat too), and part of me hopes their incompetent, stupid-ass coaches all die in a great Job-like mass of infection (boils, abscessed hemorrhoids, lungs full of fluid, etc.).”


    Context/Source: (wfs? leaning toward no.)

  3. Tom

    “At first glance the template seems to be offering a routine or regimen. This may seem at odds with our contention that workouts need considerable variance or unpredictability, IF NOT RANDOM, to best mimic the often unforeseeable challenges that combat, sport, and survival demand and reward.” (Capitalization added)

    “Random” programming would almost certainly fail to produce the greatest work capacity, as true randomness contains clusters and groupings that might cause athletes to miss certain stimuli for long periods of time.” -Berger

    “In fact, it is mathematically likely that each three-day cycle is a singularly unique stimulus NEVER TO BE REPEATED in a lifetime of CrossFit workouts.” (Once again, capitalization added for ease of use)

    Not counting your quotation…the other two are from:
    by Greg Glassman

    • Tom,

      Very good finding the word random in a quote by Greg Glassman. Had you also included this excerpt a few sentences later, and confusion about what CrossFit teaches would be easily cleared up:

      “The template is engineered to allow for a wide and constantly varied stimulus, randomized within some parameters, but still true to the aims and purposes of CrossFit as described in the “What is Fitness?” issue. Our template contains sufficient structure to formalize or define our programming objectives while not setting in stone parameters that must be left to variance if the workouts are going to meet our needs. That is our mission – to ideally blend structure and flexibility” – Greg Glassman

  4. Tom


    Thank you. Except, maybe you should read the words surrounding the part I capitalized; if you were to only notice the word “random” in that sentence, that would be disappointing, and……of course it’s at odds with your contention. All the second quote does is to easily add to the confusion about what CrossFit teaches, not clear it up.

    And the latter quote? It’s a response to your writing, and it doesn’t contain the word “random,” happy?

    • Glassman’s point in the article is very clear. Constantly varied training can appear, and should sometimes have elements of randomness to it. This randomness tests athletes in the unexpected, and a good example of it would be the Hopper workout programmed at the 2007 CrossFit Games.
      To say from this quote that “CrossFit programming is random” is to flatly ignore the rest of the article, which is a guide on how to plan training by modulating and controlling the variables of movement choice, load, repetition, intensity, and time domain.

      Your second quote, about the possibility that a three-day cycle might never be repeated, has nothing to do with randomness, unless you are defining the term as Rip seems to, something like “not periodized = random.” This is of course a false dilemma, since there can exist other types of programming that are both no periodized and not random.

      • Tom

        “our contention that workouts need considerable variance or unpredictability, IF NOT RANDOM”
        Great, so the rest of the article basically argues against your/CrossFit’s contention. Do you understand contradiction? You can spin words all you want, but “considerable unpredictability” is pretty much “random.”

        Sure, I’m not talking about randomness here, but I zoomed in a little:
        “miss certain stimuli for long periods of time” –that’s supposed to be a bad thing and….
        “likely that each three-day cycle is a singularly UNIQUE STIMULUS NEVER TO BE REPEATED”

        Perhaps there is a contradiction? Your strawman is quite good, but actually I think Rip and I would argue that periodized training is a type of non-random programming, Whether or not Crossfit is can also be determined, but only after you clear up some contradictions.

      • If I’m understanding you correctly, you are claiming the article you have quoted is an example of Greg Glassman contradicting my statement, because:

        1. Glassman writes of the benefits of unpredictability and randomness within certain parameters of training, within the context of, and in balance with the clearly intentional programming template he is teaching.

        2. I said that CrossFit programming is not random.

        This just shows that you are misunderstanding Greg Glassman’s point, and you are misunderstanding the characterization of CrossFit programming that I was defending against, which is that all workouts are essentially drawn at random from a large hat.

        Now if I had said “CrossFit has never used elements of randomness in workout design or in choosing workout variables” then that would be a contradiction… but I didn’t, and wouldn’t say that.

      • Tom

        Randomness confined into certain parameters is it? Parameters being modalities (gymnastics, metcon, and weightlifting) as well as priorities (element, task, and time), which determine work recovery character? So I did misunderstand that part. CrossFit is not random, but planned workouts, which are within said parameters, random.

        Workouts drawn at random out of a particular number of pre-selected large hats of specific shapes and colors in a particular order. LOL I concede to your defense.

        Oh, and what about the second part of my comment regarding “missed stimuli?” If you never repeat a stimulus, doesn’t it mean that you miss it for a long period of time, or perhaps forever?

      • You’re still mischaracterizing the quote. I can program a heavy squat workout, and after the workout, I can draw from a number of randomly selected skills and spend 10 minutes practicing whatever skill is chosen. This is perfectly consistent with how the article is suggesting “randomness” be used as a possible tool to achieve variance in training. You’re still suggesting that Glassman is encouraging trainers to blindly combine all elements of a workout without any consideration for the training stimulus created, or without any consideration for how that training fits within the broader goal of improving fitness. That’s simply false. It’s never been taught that way, by anyone, at the L1 course that Greg Glassman began teaching over a year prior to this article being published…

        Which means there are two possible explanations of your apparent contradiction-

        1. Greg Glassman has cryptically contradicted himself in this article, and contradicted everything taught about programming in the CrossFit L1, and L2 courses since 2002. He, and everyone else working for CrossFit, have also failed to notice this contradiction in 12 years.

        2. You’re misunderstanding the quote and attempting to bend and exaggerate its meaning in order to win an argument on the internet.

        I could know nothing about CrossFit and suggest that you might want to look at #2 first.

      • Tom

        Now you’re mischaracterizing my understanding. You program a heavy squat by randomly choosing it out of a hat consisting of weightlifting exercises at a particular rep and intensity scheme that facilitates a particular work-recovery character. Now you go into another hat, and randomly pick a selected skill to practice. The resulting combined stimulus is supposedly planned for so as a three-day cycle is a unique stimulus never to be repeated again, your contention (I hope that he defines uniqueness not by degree i.e. load, time for the same). The unique stimulus, non-random in that it will never be repeated, but random in every other way. If the stimulus was repeated frequently…well you would have something akin to periodization.

        Regarding your two possible explanations:

        I took your suggestion, read the 2nd explanation, and found it ironic. Crossfit prides itself for its pursuit of constant variance, a “big picture” theme throughout its methodology and worldview. However, it has found that structure is unavoidable for progress, so it decides to come in with a structured framework, and in the framework, it randomizes pretty much everything in order to have structure be congruent with its worldview. Sure, that is not a crime to do so, however, the resulting effect is to have little frequency in skill based tasks, such as the olympic lifts (look at the mainsite programming). Of course to address that concern, there are two choices. One, to have people do more work on modalities that they are weak on along with the mainsite programming, and two, to disregard the mainsite programming and program more important modalities with higher frequency. These two options, you will argue are still Crossfit, and I would be hesitant to disagree, except that the latter option probably moves towards periodization for particular modalities. The first option however, even though still crossfit, raises the question of changing the mainsite programming. The lack in frequency of particular exercises on the mainsite means that it is suboptimal because it will inevitably predispose people to weaknesses in those areas if all they do is the mainsite programming (crazy genetically endowed beast-like specimens may possibly be excluded). Why then has the mainsite programming not changed for the better? Is it combining together randomness and structure through sheer stubbornness? Even in your community, there are better ways to program which involve more structure and less randomness, but accepting that seriously hurts a central tenet in Crossfit worldview.

        So, actually I like your first explanation. I’m glad that you’re the first person to notice it. Progress after 12 years is expected, is it not?

      • I’m trying to make sure I understand what you did here. I gave you an example of how an attempt at randomness might be worked into a typical workout to help improve GPP. After performing a scheduled heavy squat workout, you might randomly choose a higher skill activity to practice technique on before calling it a day. So in response you insist that the squat workout I’m referring to must have been randomly chosen, and you go on to talk about how CrossFit doesn’t include enough repetition of skill work…

        Again, I’m just checking, because if that’s what happened it’s kind of funny.

        CrossFit programming isn’t random. You can keep erroneously suggesting that we randomly program workouts, or that there is some gross internal inconsistency in CrossFit because Greg Glassman used the term in an article about how to intentionally vary programing, but it isn’t.

        As for your new claims about CrossFit and programming. How often does the Clean and Jerk appear on mainsite? What is the threshold for how often this exact movement should appear for athletes to continue to improve fitness? You seem to be very aware of the answers to these two questions. If you aren’t, you’re just narrating an opinion for which you have no practical data to support. You are correct on one item however- CrossFit mainsite programming exposes weaknesses. Where you go wrong again is in somehow blaming CrossFit for causing them. When a naturally strong and powerful athlete who excels in the olympic lifts compared to his peers struggles badly with basic bodyweight movements, do you blame the mainsite for not programming enough pull-ups and push-ups, or do you recognize that individual differences in athletes can lead to the need to target those weaknesses in programming? CrossFit has historically encouraged the use of skill and weakness development in warm-ups, as well as scaling and modifying workouts daily. That is “changing mainsite for the better” as you say.

      • Tom

        So how are you choosing to put the squat in a particular workout? I thought you have a variety of strength exercises, and you choose them at random. If this is wrong, than you’re saying squats are programmed at a set frequency compared to presses and pulls? So it isn’t random, but closer to periodization? The skill work can be random if you want, although, you’ll agree that you’ll have to focus more on weaknesses. Seems that structure is winning….

        21-15-9 and Double Gracie programmed 2 days apart over the course of 20 days. hang power snatches was 10 days beforehand. So for clean family exercises that’s on average 1 per week, alright, but for full clean, 1 per 10 days, and for clean and jerk 1 per 20 days? I’d probably program the clean and the jerk at the very least once per week. However, even if I didn’t know this, I would be narrating for others who do have such data.

        Actually, not expose weaknesses but rather creates relative weaknesses. (Doing the mainsite stuff definitely improves baseline metrics, not going to deny that). Is there a better way to baseline program as to not create these relative weaknesses (or expose them, as you would say)? Do you use the mainsite programming? If so, good for you. If not, why not? What general weaknesses does it not address? Is the mainsite programming to develop a base level of fitness and to diagnose weaknesses in fitness parameters, so that they may be improved later? (What’s the best way to improve them, constant variance, periodization?)

        Regarding the question: “do you blame the mainsite for not programming enough pull-ups and push-ups, or do you recognize that individual differences in athletes can lead to the need to target those weaknesses in programming?” Actually, I do both if the mainsite programming has the tendency to result in a relative weakness in basic bodyweight movements. Those two things are not mutually exclusive, false dilemma much?

      • ” I thought you have a variety of strength exercises, and you choose them at random. If this is wrong, than you’re saying squats are programmed at a set frequency compared to presses and pulls?”

        I know you think we program randomly, that’s why I have told you repeatedly we don’t. Movements are chosen for a variety of reasons, but suggesting that they are either random or “programmed at a set frequency” is the same informal logical fallacy of a false dilemma that you accused me of making, and Rip committed as well. CrossFit is neither, but uses a third option that you aren’t acknowledging- intentionally planned variance.

        “even if I didn’t know this, I would be narrating for others who do have such data.”

        I’m not going to let you slide by appealing to mystery people with mystery data. Either you are aware of data that shows a relationship between training frequency of the C&J in constantly varied programming and fitness or you aren’t. I think it is obvious that the C&J is an important movement to be in a fitness program, but exactly how often it must be present in order to gain the most work capacity across the broadest time and modal domains is up for grabs. To suggest that the mainsite has it wrong places the burden of proof on you, so please provide evidence of your claim.

        Finally, in my hypothetical to you (that you refer to as a false dilemma), I am simply copying your previous claim and substituting movements. You fault programming for not using the C&J enough, based on what I assume is an anecdotal example of someone who is weaker in these lifts relative to other aspects of his or her fitness. I replaced C&J with bodyweight gymnastics to see if you feel the same way. It sounds as if you do as long as there is an undefined “tendency” for this to happen… which takes us right back to you needing to provide evidence that the mainsite has such a tendency.

  5. The beauty to me of CrossFit as a program is that it shamelessly adapts into whatever program best achieves its non-specialist objective. Rip is right that CrossFit has changed over the years – there always seem to be new elements making appearances in the programming, from specialized lifting drills to mobility exercises. CrossFit as a program is a Survival of the Fittest model, and while Rip is probably a great coach, his contributions were simply unable to prove through results that they were accretive to the program.

  6. I am a fan of Rippetoe’s work, generally speaking, but I think Berger is spot on with this one.
    It seems like “random” is a convenient synonym for “constantly varied.”
    Basic movements are more than adequately covered in most CrossFit programming, but the order in which these movements appear, as well as their varied paring with other movements can give the misperception of randomness, which is failing, in Rippetoe’s own words, to “command a working definition of the thing you’re criticizing.”

  7. Griccini

    First I think it’s funny ESPN promotes the games then also tries to call out crossfit. That in itself just shows how they play both sides and are trying to gain viewers. Secondly I think Russell and Rip make valid points. If I were rip I would be more concerned about individual affiliates than mainsite. I’ve had the pleasure of training at a few boxes where things were random there was no design or program template where boxes created wods the day of. Again both sides have valid points but fact is people have been able to branch off I to areas of fitness they wouldn’t have previously explored without crossfit. Ain’t that a catch 22.

    • You are falling into the same mental trap that Rip does. Simply because a gym programs a workout on the day of training does not make it “random.” Perhaps reviewing the commonly accepted definitions of the term “random” will help.

      Here is one: “proceeding, made, or occurring without definite aim, reason, or pattern.” That’s not CrossFit. CrossFit aims to improve work capacity (fitness) measured across broad time and modal domains. It ensures this goal by intentionally programming to optimize intensity and variance.

  8. Tom

    “CrossFit does emphasize a core group of functional movements for consistent practice and training. Think of how often below parallel squatting movements appear in CrossFit’s benchmark workouts. This is not an accident.”

    So, what are the functional movements? Is there a list and a description, somewhere? Is doing bodyweight squats the same as doing barbell loaded squats with increasing weight each session in terms of increasing my work capacity?

  9. Pat McCarty

    “I don’t think there is a CrossFit gym anywhere in the world that programs every single day by randomly drawing workouts from the stack like the 2007 CrossFit Games Hopper event.”

    Well, that’s not what I asked. It was “Is it possible that many are literally programming with no real plan?” Those concepts are different. And I would submit there there are many that have no plan.

    • actually that is what you asked, but it was only half of what you asked. I ignored the second part of your question because I don’t have any idea what you mean by having a “real plan.”

  10. Andrew Shaw

    I believe by real plan he meant, they add any exercise combination whether it is a couplet or triplet to the WOD of the day. Could be an AMRAP or Rounds for time. Call it programming without any thought to Crossfit model for programming.

  11. bdh

    How about a basic thought experiment consisting of two individuals, identical twins if you will, who are assumed to be of equal ability and strength at baseline.

    Individual A runs a basic 5×5 3 days/week, let’s say Texas Method, for 6 months. Consisting of squats, deadlifts, power cleans/snatch, press and bench press.

    Individual B runs CrossFit WOD’s for 6 months.

    After six months, who do you expect to squat more? Deadlift more? etc. Basically, who will *stronger*?

    You may say this is an unfair comparison as CrossFit’s demonstration of work capacity is not best represented by the core barbell movements alone. What then is your test? And who do you think will be stronger? I submit that Texas Method will produce a *stronger* athlete almost invariably.

    • If you define strength as the ability to squat, deadlift, and press, then yes. If you start including other tests of strength like the muscle-up, strict pull-ups, olympic lifts, and various others it is less likely that athlete A will come out on top. All of this also misses the point that CrossFit is not a strength program, it is a GPP program, and as such our goal is not to make people as strong as they can possibly be, but as strong as they need to be to increase their work capacity across broad time and modal domains.

      • Andrew shaw

        You can’t compare Texas method to crossfit. Both are different with different goals and outcomes. Crossfit is great for GPP. I think that to be good at crossfit you have to a good strength base. Rip’s program does that so you can compete.

      • bdh

        Fair. I would argue that the muscle-up, strict pull-ups, and olympic lifts are all debatable as the 5×5’er programs cleans/snatch weekly (more frequently than WOD’s? this I’m unsure of) and presumably pull/chin-ups weekly as an assistance lift. Because athlete A gets stronger than B, it is reasonable to say that strength training is more effective/efficient in achieving this goal of increased work capacity across broad time and modal domains. (I believe this is the essence of Marktoe Rippestein’s message that CrossFit’s programming, or lack thereof in his mind, is inferior)

        What are the specific ways that work capacity is measured in CrossFit? I mean no offense but “work capacity across broad time and modal domains” sounds like marketing spew, it’s too vague to hold water when claiming there’s a long term objective at hand.

        PS I’m not a CrossFit dissident, merely learning more about its principles.

  12. “What are the specific ways that work capacity is measured in CrossFit”?

    “Cut your time in a WOD if half and you have doubled your power output, no ifs ands or buts” – Greg Glassman 2006

    If you come to me with the ability to do 45 barbell thrusters and 45 pull-ups in ten minutes and one year later you can do it in five minutes, you have doubled your work capacity. “work capacity across broad time and modal domains” is the opposite of vague, it is precise.

    • JM

      “If you come to me with the ability to do 45 barbell thrusters and 45 pull-ups in ten minutes and one year later you can do it in five minutes, you have doubled your work capacity. “work capacity across broad time and modal domains” is the opposite of vague, it is precise”

      But still, wouldn’t “training” to accomplish that goal be more effective?

      Allow me to explain. The sport of Crossfit, as in the events in the Crossfit games, require proficiency in many different things. From what I can tell the athlete would need solid aerobic conditioning, anaerobic conditioning, skill in gymnastics and weighlifting, a decent amount of strength as well as few other components (sprinting, high LT etc..)

      The “process” of training, at least in other sports, requires building the energy system/skill in isolation because proficiency is only improved with consistent practice. Of course PLANNING trainng makes sense when you consider the fundamental stress/adaptation model. It does’t have to be “block periodization”, (it can be concurrent), but it still requires structure in order to for it to be optimal. Specificity (a timed WOD is specific for example) is important but not nearly as building of all the components. For example: If you have a middle distance runer who does nothing but run 800 meters in training, day after day as fast as possible, he will likely LOSE when pitted against the runner who did lots of aerobic training (jogging), speed work, lactate threshold training, technique work, practicing finishing kicks etc.. Most athletes never actually “compete” or try for a PR until the day of competition. Crossfitters do it every single workout.

      Mainsite Crossfit is the middle distance runner who ONLY goes to the track everyday to run at race pace. They only emphasize specificity (to their sport) while ignoring all else. You can say “but they do train in isolation”. Which is true, but because they try to stick to the “constantly varied” model, it just happens to infrequently to be considered worth anything.

      • JM,

        You’re using Rip’s definition of “training,” which we have demonstrated is just a semantic ploy that allows him to categorize CrossFit as something inferior, all based on his false claim that it is “random.” If you don’t believe us, as Rip’s co-author Dr. Lon Kilgore. You then provide an analogy that I think is nonsensical. You attempt to compare someone doing CrossFit to a sport-specific runner who runs the same distance a max intensity every day, but you fail to recognize that constant variance in CrossFit training makes this comparison completely backwards.

      • JM

        “You’re using Rip’s definition of “training,” which we have demonstrated is just a semantic ploy that allows him to categorize CrossFit as something inferior, all based on his false claim that it is “random.”

        My comments are based on fundemental training principles that apply to any sport. That has nothing to do with Rippetoe.

        I am recognizing Crossfit as a sport using the games events as the objective. Perhaps I should have used the comparison of Strongman because that would be a closer match. Strongman do event trainng, without a doubt, but they also train the various qualities needed to complete the event at high proficiency in isolation. They CAN do nothing but event training, which would be constantly varied, just like Crossfit, and they would get better UP TO A POINT. But it isn’t optimal. Planning, using the progressive overload, stress/adaptation model is what is needed to excel in any sport. Again, this isn’t Rippetoes definition, it is a fundemental principle.

        Let me illustrate:
        If you have an alhlete that has a 1 RM in a squat thruster of 90 lbs and he is training for an event that includes FRAN. he also doesn’t know how to do a kipping pull up. Obviously he won’t be able to use the prescribed Fran weight (95 lbs thrusters for many reps) when starting out. You have this athlete follow mainsite and this will magically improve his 1RM with the thruster and ability to master a kipping pull up? how, exactly will that occure with constant variation? because you repeeat a thrusters and kipping pullups at random intervals in WODS? Wouldn’t setting out a plan- a repeated routine of skill work on the kipping pullup, squat thrusters twice per week using more and more weights etc.. get them to the goal (and beyond) faster?

        If you think constant variance is truly a superior training protocol to the tried and tested periodization model, that if fine! I am always willing to learn new things. BUT PLEASE explain the how! Rippetoe can at least articulate his point when he thorws out an objective. He draws up a clear roadmap that methodically gets you to that goal and it makes sense. Crossfit hasn’t done that. I am glad you are “defending the brand” but you are only doing so by pointing fingers as opposed to defending it with substance. Rippetoe wins the arugument by saying nothing because his publications did all the work. Anyone with a half a brain can see this.

        Years ago when I became interested in Crossfit I stumbled upon this peice from Glassman and it MADE SENSE:

        I went to the mainsite, entered the last 6 months of workouts into Exel and found that the mainsite didn’t follow the “programming” at all. Nothing was consistent. So what IS the programming? What is the PLAN for improvement? How do you get athletes moving more weights, faster (beyond the novice stage)? Afterall, isn’t that the objective?

        For the record, I am not picking a side. I like Crossfit, I really do. I just wish CF was more open to change. Crossfit football has a good thing going as does things like Golden Empire. I at least applaud CF for NOT governing the constant variation, routine is the enemy philosophy across all affiliates.

        Thank you for your time.

      • In your previous post you put training in quotation marks and implied that CrossFit is not training. To do this means you are either using Rip’s definition of training, or something similar. You are also assuming that CrossFit programming is random, but it is apparent from your post that you are mistaking unpredictability for randomness. This is a common conflation of the two terms. CrossFit programming is not random.

        You also seem to be confused on a number of other issues. CrossFit is not a sport. The CrossFit Games is a sporting event that allows athletes to compete against each other to see who has the most work capacity (Fitness). The nature of fitness makes it such that planning for specific events is not possible, as the majority of events are known only days or hours before performing them.

        CrossFit is a training methodology, which does not aim to improve performances in specific, known tasks (like Fran). It aims to improve work capacity across broad time and modal domains, which means the regular use of a handful of core functional movements and their derivatives, varied in rep, load, time domain, and combination. This improves performance in specific workouts like “Fran” while also improving capacity in a widest possible variety of other workouts. The adaptation of increased work capacity across broad time and modal domains is a quantifiable goal, and it is a completely different goal from Rip’s which is strength on a few core barbell lifts.

        You may not realize it, but the burden of proof behind your argument is on you. CrossFit aims to improve fitness, and we have offered a methodology that seems to do so very effectively. If you think traditional periodization is better at improving fitness than constant variance, please provide evidence. So far you have provided none, and shown to misunderstand the methodology you think is inferior. That may be part of the problem actually.

        Also, here is a breakdown of mainsite programming that shows exactly the opposite of what you claim:

      • JM

        Thank you for your time, Russel. It does clear up a lot of confusion from my end. Just to be clear, I never said anything about random in any of my comments. I am not sure where you heard that (from me).

        If Crossfit is not a sport (which I was mistaken, I thought it now was and they do all call themselves athletes), then there is nothing to “train” for specifically. If you are saying that it is a methodology you use that “trains” one to become faster at performing functional tasks (what CF defines as functional, I don’t disagree for the most part), then I can live with that explanation. If that improves “fitness” better then other, planned approaches is probably up for debate.

        What are your thoughts on the boxes/programs that deviate from the main site model of “constantly varied”? Some boxes specifically offer planned progression programs. I have ever heard some of them speak against the main site model for many of the same reasons Rippetoe does. How can they consider that Crossfit?

      • JM,
        This is a quote from your previous post: “how exactly will that occure with constant variation? because you repeeat a thrusters and kipping pullups at random intervals in WODS?” This made me think you are assuming constant variance is just randomly picking movements every training session.

        Regardless, we have no problem with affiliates experimenting with different forms of periodization within their programming. We are not against periodization, but we disagree with the idea that classic linear progressions and periodization work best for all athletes in all sports/professions, especially not for improving GPP. Greg Glassman used to joke about periodization being like throwing a pair of size 8 shoes into the crowd. They will fit some people just fine.

      • Tom

        I think maybe this discussion of GPP vs Strength-first is probably at the core of this disagreement. If I’m following what he says, Rippetoe’s point would be simply that *strength* is where just about everybody – but especially those who are nowhere near their genetic potential for strength – should start. He argues that you cannot do it all at once — trying to focus everywhere means you plateau in terms of improvement much earlier and if you just keep upping the intensity, especially in a competitive (as opposed to incrementally progressive programmed) way, you are asking for injury and/or burnout. Strength-first is not “a size 8 shoe for everyone!” – it’s “when you are preparing for a long journey by foot, the first thing you probably want to take care of is your feet (i.e., get some good shoes, or get your bare feet adapted to the stress by taking short trips, first)”.

        Is he right? Well, he’s got some fairly sophisticated arguments involving metabolic pathways, and a great deal of his own and the combined experience of his coaches that back this thinking up for a huge number of cases. He really doesn’t seem rigid — gruff, yes, but not rigid in terms of accepting evidence that contradicts his current take on things. He has shown learning just in the well-explained changes that have been made in his flagship book on the lifts as he has continued to make adjustments with each edition.

        I think Crossfit has also shown an ability to learn and adjust and I think Crossfit is definitely the bigger player, here, by almost any measure. The “danger” for Crossfit, to my mind, is not whether it is right or wrong on this particular point, but whether it is developing a culture of defensiveness that doesn’t allow it to continue to improve.

      • The notion that CrossFit is rigid and unwilling to adapt our program is historically inaccurate. Perhaps you all weren’t around in 2006 when Rip, partnered with Greg Glassman, authored this article announcing the use of the “CrossFit Total” in mainsite programming-

        Again, the burden of proof that Rip’s methods are better at improving fitness than CrossFit’s falls on Rip and his followers. As nice as his hypotheses might sound, he has provided no evidence that this is true. The real problem is that there is no shortage of persons following this style of training, so we should expect to see evidence if it was true. CrossFit Football, for example, leans heavily towards the type of training that I would expect Rip to approve of, and yet it does not seem to produce high levels of fitness as compared to CrossFit. This is not “defensiveness” on our part, this is asking for evidence that these claims are true, which perfectly reasonable before discussing a wholesale change to CrossFit’s programming model.

      • Tom

        The “reply” links are starting not to make much sense, but I’m responding to the use of the phrase ” …burden of proof …falls on Rip and his *followers*.” (emphasis mine) I don’t think this is a battle between sects, and I really don’t think Rip sees it that way. As far as the point about the methodological debate needing evidence-based argumentation, I think we can all agree to that. It seems to me that the first step would be getting clear on terminology and agreeing what would constitute evidence for one claim or another. If that evidence doesn’t already exist to everyone’s satisfaction in completed studies/systematic comparisons, perhaps some of those could be arranged? I know there are issues with the science that might be relevant to these questions, but it seems to me that Rip and his coaches are pretty conscientious about examining it and publishing their commentary. If the relevant studies are missing, why doesn’t Crossfit encourage them in a way that would be broadly perceived as fair with respect to outcome?

      • JM

        “Again, the burden of proof that Rip’s methods are better at improving fitness than CrossFit’s falls on Rip and his followers. As nice as his hypotheses might sound, he has provided no evidence that this is true. The real problem is that there is no shortage of persons following this style of training, so we should expect to see evidence if it was true. CrossFit Football, for example, leans heavily towards the type of training that I would expect Rip to approve of, and yet it does not seem to produce high levels of fitness as compared to CrossFit. ”

        The burden is on Crossfit! Periodization has been around for a long time; before Crossfit was even a thought. Rippetoe’s books are very clear and comprehensive explaining exactly how the model works. Rippetoe didn’t even invent it; it is a well-studied concept long before his work. In the late 90’s, Crossfit comes along with a new model that they claim is superior. When we ask about this methodology that CF uses you say “to improve work capacity across broad and modal domains”. Well, sorry to tell you but that sounds very vague. It just doesn’t hold water compared to the scientific explanation of periodization. When given the opportunity to explain itself, Crossfit spends most of its time being defensive or accusing people who do not understand the model as simply being ignorant. Perhaps we are being ignorant but the burden is on YOU to explain yourself. Rippetoe as well as Strength and Conditioning coaches going back decades have already done that. You have not. What you have explaned is what is commonly understood as “general exercise”.

        Also, it appears you have your own definition of fitness. If being able to complete unpredictable tasks across broad and modal domains is fitness, than I agree, main site programming will get a person fit; however, that is not THE universal definition of fitness. How are you measuring “fitness” to make the statement that Crossfit football has not produced the same results as main site programming?

  13. bdh

    I just want to be stronger than my peers who do CrossFit, which at this point is going very well as two of them are rehabbing from shoulder surgery.

  14. Tom,
    I agree that terminology is important in discussing these topics. The majority of the time our critics aren’t actually interested in improving fitness as we define it, making comparison of methods meaningless.

    work capacity across broad time and modal domains IS our definition of fitness. It is not a vague concept, but a quantifiable metric of average power output (force multiplied by distance over time) that is then tested over the broadest possible selection of skills, drills, and periods of time. The result is a collection of data points that can be graphed, averaged, and used to calculate the integral between power and time, giving a mathematical measure of an athlete’s fitness.

    When CrossFit began claiming that it was capable of producing elite levels of fitness, there was no quantifiable definition of fitness. The term was often measured through loose correlates like V02 max or body fat %. Since then the program has produced results that are undeniable, and at the highest levels we have a significant group of athletes that are simultaneously performing at high levels in dramatically different modalities and time domains. The thought of seeing an athlete snatch 300 pounds and then run a sub 20 min 5k is now common place. The evidence that our method works is ubiquitous and every one of our 10k affiliates is a data point representing tens to hundreds of athletes that return each day because they are convinced CrossFit works for fitness.

    On the other hand, this specific debate was spurred by the idea that periodization works better for fitness than CrossFit does. This is your claim, not mine. I am defending against that accusation, which strictly speaking puts the burden of proof on you. Wikipedia has a nice workup on this concept, which should also help you understand that you are making what is known as an argument from ignorance (not my terminology).

  15. B'ilsabab

    First, I can propose a very, very simple way of establishing that CrossFit is not random:

    Explain the methodology by which Mainsite WODs are generated, the rationale behind the order in which they are executed, and the mechanism used to ensure that they consistently provide sufficient practice for development and maintenance of technical skills. Not in vague generalities, but in specifics. Take six weeks (or however long is necessary) of Mainsite WODs, and elucidate the logic behind how they are ordered and structured, and how they contribute to the programming goals of CrossFit, in keeping with what is known about how the stress of training induces adaptations in the body.

    That’s really all you need to do.
    If you choose not to do so, perhaps claiming that these methods are proprietary, then you have no choice but to cede the ground, because your argument is nothing more than assertion.

    Second, to your point above, Rip has NEVER contended that his methods are a better way to get people FIT.
    He has said that they are a better way to get people STRONG. Which is hardly contentious.

    • Russ Greene

      That’s not true. He has repeatedly stated that he thinks CrossFit is less effective than his methods for achieving general physical preparedness.

    • JM

      “If you claim that CrossFit is random, the burden of proof is on you to show that it is”


      Just look at any set of workouts. Random, constantly varied or whatever you want to call it, please explain the methodology. Using your “hopper” model, I will give you that it isn’t random, but is there any plan in place at all? Do you at least map out a month saying that we will do hip extension exercises on days xx and pulling on days xx etc. ? Or, does that not matter at all?

      To the average person it looks like you just have a collection of movements that you mix and match from day to day and assign various time schemes (AMRAP’s of various lengths vs. rounds for time etc..). Part of the criticism is that you use very complex lifts in doing so and for a lot of people we just wonder why, especially when there are safer alternatives that accomplish the same goal (I can only assume when you are doing something for time the goal isn’t to actually get better or stronger at that lift).

      What I did learn from this discussion is that Crossfit is not a sport, so nobody is really training for anything other than improving general fitness, which is defined by Crossfit as being able to improve your capacity to exercise faster.

  16. Pingback: CrossFit Athletes Dominate IMG Academy Testing, Revisited | THE RUSSELLS

Comments are closed.

%d bloggers like this: