“Tact in audacity consists in knowing how far we may go too far.”
TTP and Conditioning Research reader Noel sent this erudite email recently to both Chris (of Conditioning Research) and me, in response to the ongoing and hotly debated SS/HIT vs. (among a plethora of other related issues) “functional training” discussions. If you’re arriving late to the game, you can get a feel for what’s being said here and here. Anyway, I wanted to pass along Noel’s comments, as they are intelligent, informed, and lend much to the overall discussion on this topic. Also included is my short reply.
The recent posts about the BBS/super slow movement have irked me to
the extent that I am compelled to write this email. While on the macro
level I don’t really care what they espouse, I don’t think their
claims are getting the critical examination they deserve.
Let me crudely characterise the debate as consisting of two sides:
machine based, super slow, one set of failure based on published
research (Dr McGuff) versus free weights, 5 reps or less, multiple
sets based on coaching experience (Rip). I’m not suggesting Dr McGuff
and Rip are in direct opposition (I don’t know if they even know of
one another), but I want to use two exemplars to discuss this issue. I
hope we can all agree these guys are experts, and they hold viewpoints
that are contradictory. No one has the time to be knowledgeable in all
fields, so normally we defer to experts. When the experts disagree it
is time to examine the primary evidence more closely. The key thing
here is our standard of proof: how strong must the evidence be before
we accept it as true?
Now the BBS guys lean on the published literature. I went to PubMed,
did a search for “resistance training one set failure” and the first
relevant hit I found was:
METHODS: Twenty-one women were divided randomly into 2 groups: Group 1
(n=10) performed a single set of the leg press exercise once per week,
while Group 2 (n=11) performed a single set of the leg press exercise
twice per week for a period of 8 weeks. Throughout the duration of the
study, an amount of resistance was utilized that allowed for a single
set of 6 to 10 repetitions to muscular failure.
This seems to back up their claims — but it is extremely weak
evidence! 21 people is tiny, and 8 weeks is very short. Consider this:
would you use a drug that had been tested on 21 people? I did a search
for drug trial sizes (and perhaps Keith can say more here) and it
seems a small trial is of the order of 300 people. Technically the
statistical power of this study — that is, its ability to show an
effect if there is one — is very low.
To see the problems this study might have, imagine you had two groups
of people, both of whom can lift 100kg on some exercise, with a
standard deviation of 5kg. Imagine one group trains with protocol A,
and the other with protocal B. After a year the group on A can lift
250kg +/- 12.5kg and the group on B can lift 200kg +/- 10kg (so group
A gained 1.5x the strength and group B gained 1x the strength). The
difference in average strength is well outside 3 standard deviations,
so this should be a very significant result.
Now what do you see after 8 weeks, assuming linear gains?
Group A: Mean = 100kg + [100kg * (8/52) * 1.5] = 123kg
Std. Dev. = 5kg + [5kg * 8/52 * 1.5] = 6.2kg
Group B: Mean = 100kg + [100kg * (8/52) * 1] = 115kg
Std. Dev. = 5.8kg
The difference in means is well within two standard deviations — not
a significant result. So see how the short duration of the study has
made a significant result seem insignificant.
(This is fairly informal. If someone wants to calculate the actual
p-values assuming, say, a population of infinite size [and therefore
the t-distribution becomes the normal] that would be informative and
more persuasive than my argument.)
Also, see this:
This quantitative review indicates that single-set programs for an
initial short training period in untrained individuals result in
similar strength gains as multiple-set programs. However, as
progression occurs and higher gains are desired, multiple-set programs
are more effective.
I’m not a researcher in the field of exercise science (or kinesology
or whatever you want to call it) but a lot of the published research I
have seen is of this type. This does not meet my standards of proof.
Now consider the evidence Rip has. It would be rejected by the BBS
guys as it doesn’t meet the criteria for publication: it doesn’t
control for variability, it isn’t statistically analysed and so on.
That doesn’t mean it isn’t evidence though. From his writing Rip
strikes me as a very methodical and very experienced guy. I must admit
I am more inclined to believe him, based on his experience training
hundreds of people over long periods, than I am to believe claims
based on what I perceive as very weak published literature.
Finally, I want to address Chris’ interview with Luke Carlson.
“Q: What do you make of Crossfit?…
A: It is entertaining to me that the three movements that all humans
allegedly engage in just happen to be historically popular Olympic and
Two things. First my understanding is that the term “functional
movement” is used by Crossfit to mean a movement that carries over to
other activities. It doesn’t mean that movement mimics other
I would think anyone could see that deadlifting and squatting are core
movements. I guess Luke has never picked anything off the floor, or
taken a dump.
“The vast majority, if not every HIT advocate that I know utilizes
twisting/rotational movements. We use the MedX Core Torso Rotation
machine – a $7,000 machine that targets the muscles involved in
rotation of the trunk. This exercise is included in the working
scripts for all of our clients.”
The Crossfit orthodoxy here is that training midline stabilisation —
the ability to resist twisting — is key. I did a little test with
myself, throwing punches. It seems that I flex my obliques to avoid
twisting so as to better transfer power from my hips to my upper body.
I’m not trained at punching, but this way felt much better than
deliberately twisting my midsection out of line with my hips.
This response is also highlights an issue that I haven’t seen anyone
address yet — these guys are not impartial. I don’t need a $7000
machine to train my obliques, but the equipment manufacturers and the
gym owners that have invested in them would like me to believe I do.
In fact this is one of the primary reasons I dislike the BBS movement
— they want to make the trainee dependent on the gym to workout. In
contrast a barbell set is dead cheap, and free weights, be they iron,
a rock, or a baby, can be used anywhere. I’d rather be self sufficient
and teach people to be the same.
Finally, let me address safety:
First, there is lack of evidence to support machines being safer than
Let’s also look at the injury rate, from the same site: 0.0035
injuries per 100 hours. Imagine I’m a real gym rat and hit the gym 5
days a week for an hour. That’s 5 hours a week, or 260 hours a year,
or 13000 hours over 50 years. With that injury rate I would expect
0.455 injuries over my lifetime of training. Worried?
Well, I’d better do some real work.
A nice bit of thought here, to say the least. And what follows was my emailed response to Noel:
Thanks for your intelligent email.
My approach to all things physical culture (and to life in general) has always been to give well-conducted scientific studies serious consideration for real-world implementation, but I always defer to empirical evidence when choosing what modalities/protocols to utilize when “in the field”. And herein lay the age-old disconnect between what “works in the lab and in theory” and what produces “results in the field”. An athlete in the real world bears little resemblance to “case studies” (be they human or theoretical) in the lab. And, too, within any pro-con argument, the “purists” on either side will cede little ground to the “opposing” view/theory. Live in general, and training in specific, is rarely black or white; it’s more analogous to yin and yang. My take is that for each specified weakness, there is a good-better-best continuum of options related to how to “fix” that weakness. This holds true for diet as well as exercise.
Well, there you are folks; weigh-in, comment and kick this one around a bit. I’m anticipating some good discussion out of this bit of informed criticism. And thanks again to Noel for allowing me to use his thoughts here on TTP.
Note: I’m sure that Chris will soon have a separate discussion running relative to Noel’s comments over at Conditioning Research, and I’ll link directly to that post as soon as it’s available. Noel’s put out a good bit of opinion here, and I’m quite sure the discussion will splinter into many directions on each site.
Late entry: Here’s the link to the Conditioning Research post.