Saturday, 12 January 2013

Oh Taleb

I started to read Nassim Taleb's Antifragile. 150 pages in, it's an amusing read. Right now I'm playing a little game as I go called "how many pages can Taleb go before sounding like an asshole?" I think the record still stands at one. A random example, his footnote on page 123 in response to those who do anything they do not enjoy:
  A friend who writes books remarked that painters like painting but authors like "having written." I suggested he stop writing, for his sake and the sake of his readers. 
Oh Taleb.

What I actually wanted to share was on page 46-47, that succinctly contain his views on exercise. Backstory: after having written Black Swan he "received all manner of threats" (aside: From who? disgruntled bankers planning to beat him up on the street? Never mind). He decided to act; he hired a 280lb bodyguard, or 127 kg (which is not 130kg, as Taleb's math seems to arrive at).

Anyway Taleb was so taken by Lenny's physique he decided to train like him.
He was into 'maximum lifts' type training and swore by it, as he found it the most effective and least time consuming [form of exercise]...The workout was limited to trying to exceed that mark once or twice, rather than spending time on un-entertaining time-consuming repetitions. The exercise got me into naturalistic form of weightlifting, and one that accords with the evidence-based literature: work on the maximum, spend the rest of the time resting and splurging on mafia-sized steaks.
Looks like Taleb discovered elementary theory on early improvements in increasing maximum lift strength. After four years he has worked his way up now to 330 lbs (150kg). Not bad, I suppose. Taleb claims to now have physique of a butcher. I smiled a bit here, as I recalled a passage from another book of mine, Keith Livingstrone's Healthy Intelligent Training, where he describes strength training for skinny runners. At his peak body weight of 140 pounds, Keith could also half-squat the same 150 kilos as Taleb ten times. He went on to say
If that sounds heavy, it is, but it is nothing compared to what a trained power lifter of the same weight can lift.
He is correct; the WR deadlift (similar records as for squats) is -for a 140-lb individual- 237kg. With a 3,000m PB of 8:04, Dr. Livingstone is pretty well-educated and well-trained individual who also, incidentally, advocates practising maximum lifts ("I'd have been better off halving the number of reps and lifting a bit heavier"). This is all to say what Taleb here claims, that training dead lifts helps improve maximum strength is true, but neither the complete truth nor not much of a discovery for someone of his education.

What Taleb's proof of practice does show is how lousy a weightlifter he is. Spending four entire years on the practice and having the "body of butcher" only gets him a measly 330 lbs off the ground? At the age of 13 (and at 144 lbs) between May and August I went from being able to clean & jerk 70 lbs up to a total of 140 lbs in my basement using plaster weights. I lost interest thereafter, I was just wanted to try to lift my own weight. I wasn't even that strong in grade 9. And considering that deadlift world records are roughly double clean & jerk, I was pretty close to Taleb as a skinny barely teen.

Taleb has in fact shown how little gain there is from concentrating on only one type of exercise. Assuming Taleb with his butcher-build weighs at least 90kg, at age 53 his 150 kg lift pales to the 50-54 masters female record; the record for a 165lb/75kg woman of that age being 170 kg. The men's record for his age, however, is a soul-crusing 305 kg. You are almost half-way there Taleb. The secret, of course, is that weight training is a complex art and you find yourself quickly maxing out unless you change stressor types. That means working on improving deadlifts one week, abs the next, squats, curls, some aerobic strength (yes, even for weight lifters), and so on. Rounding yourself a little takes care of strengthening smaller support muscles. Complex lifts like picking up a boulder will not improve max strength as much as a leg press machine, and a leg press machine will not work minor muscles as much as lifting a boulder. So you need both sorts, duh.

The other observation I had to make was his claim of earning himself the right to eat "mafia-sized steaks" after these powerlifting sessions. Let's do the math to see how many calories Taleb burned in a bout of deadlifting, assuming that is all he did.

Lifting anything against gravity, you expend (in joules)

W = mgh
Where m is mass (kg), h is height (meters) and g is the force of gravity (9.8 m/s2). Assuming again Taleb himself weighs about 90 kg and that a deadlift means brining something 1 meter into the air, and that the mass he lifts is on average 150 kg, then one lift requires about

W = (150+90 kg)*(9.8m/s2)*(1m)
= 2352 Joules/lift
= 0.56 (food) Calories

Yes indeed, a single "massive" lift will burn just half a calorie. But assume that the body requires more energy than this to operate (as lifting is not a 100% efficient task). Assuming a mere 25% lift efficiency,  then each lift is now costs 2.2 Calories. It's surprisingly hard to corroborate these numbers, but at Livestrong they suggest for body-building levels of weightlifting you burn 0.055 Calories per pound lifted per minute exercise. If one deadlift takes 5 seconds (an overestimate), and again we're lifting the same 240kg/528lbs, then we'd expect one lift would burn 2.4 Calories. Pretty close! For the record I didn't calculate that number until after I estimated mine own number above.

We also know that any lifting session requires multiple sets and reps, so let's say -to be generous- he did 8 sets of 8 reps for a total of 64 lifts (and consistent with recommendations of most strength coaches). Taleb may have burned up to a total of 140 Calories from his deadlift routine. Let's celebrate with some mafia-sized steaks!

We'll assume the mob dines with the finest, so I'll guess Taleb likes to eat the $350 Kobe steaks. Alas he has not earned an entire 8oz steak. By my calculations Taleb has expended enough energy go get himself 2 ounces worth of Kobe steak. Those though guys, they must be watching their weight more than I expected. But the glutton I take Taleb to be I bet he'd eat the whole thing anyway faster than a fat kid a McD's.

To sum, it is obvious that a handful of power lifts earns you next to nothing either with your appetite or your strength. But if you want to weigh as much as you can lift, go ahead and keep picking rocks and eating massive amounts of beef. Cheers.


  1. I'm not done reading the book yet, but I'd say your argument at the end is doing exactly what he talks about with the bed of procrustes. You're assuming the body acts as a simple machine and then using that model to define how the body will process food calories. His premise is that academics etc, are guilty of taking convenient models and trying to apply them to the complex world in ways that don't work. Just struck me as funny that you did that to prove that his approach is wrong "in theory".

    1. Which one are you not done, Bed of Procrustes or Anti-fragile? Not important, I never read the former. I'm glad to hear there was something funny at all about this post! Humour isn't my strength. But yeah, there's the unfortunate problem that everything needs a model. A paradoxical self-referencing issue to be sure, but at the root of logic itself. We've all heard of Russell's paradox. But more importantly I'm not convinced why I should abandon the energy model I used. It works pretty well for calories burned in a marathon: a one-line formula can show why you bonk at 30km. And using an almost as simple model I can predict ground impact forces that match well with empirical data ( Seems reasonably robust given the stupid-easy physics I'm starting with. Not sure I follow your logic. It hints along the lines of "some models are false, your approach uses a model, hence your model is false" Or is it that all models are false. Which itself must then be a model to explain the universe... Help me out here.

    2. It's really simple. I'm just saying the body is more complicated than a machine and how it absorbs food may not be as simple as calories in - calories out = calories retained. That model is an example of one of his main complaints in the book anti-fragile which is intelectual tendency to come up with a model and then pretend the world fits it. He refers to this mistake as the Bed of Procrustes throughout anti-fragile. I don't believe all models are wrong, physics can be modeled well. Economics and Biology less so. The fact that the calories is able to predict a marathon bonk doesn't mean it's a good model for other aspects. If this idea annoys you, you should avoid reading the book.

      I don't know what Russell's paradox is, I'll look it up.

    3. The coolest thing about Russell's paradox is that it's only a few lines of logic, it undermined years of his work, he discovered it by himself and despite this didn't try to hide it.'s_paradox
      But about mechanical (and other) models used to represent the human body, there are indeed many. I found several that work on a short timescale where metabolic considerations are negligible. But for long-term equilibria you must consider those pathways. Did you by chance also read my post about how using oversimplified models then leads to absurd answers (and hence why calorie counting beyond a day's work is rather silly):
      Keep in mind this post used some rather basic empirical proof that Taleb's exercise regimen is poorly designed. I simply found people half his weight who can life more than he, and equal his weight who can lift more than double. His bragging about a 330lb deadlift is (within the universe of weightlifting) about as impressive as a runner bragging about a 4-hour marathon.
      The point is that Taleb is full of hot air, a crank. I read his book cover to cover and Black Swan. Though he is entertaining, I could easily say to have found more useful information from one page of Origin of Species, Guns Germs and Steel, Godel Escher Bach, Fractal Geometry of Nature, or Mismeasure of Man. (perhaps not coincidentally they all explored the usefulness or misleading nature models). Antifragility was an attempt to find a grand meaning in his musings, but he undermined his arguments in the early chapters. This post mean only to explore his crassness rather than fallible logic.

  2. Taleb "full of hot air" - what BS is this! Taleb's efforts are a practical,usable and honest appeal to all of us to be careful about what we think we "know" and even more careful about all there is that we don't know that could hurt us. I don't know how you found more "useful" information in one page of Guns Germs and Steel - I suppose it depends on what your "uses" are - perhaps you are an academic? Look your criticisms that Taleb is arrogant, a crank, and full of it on excercise topic are irrelevant almost; you don't really devastate his main arguments. Even if he doesn't make it to your pantheon of 'useful books' doesn't mean he doesn't have a place; doesn't mean that his books don't tackle another subject more usefully than the ones you named.

  3. this retard cites Livestrong while trying to critique taleb lol

    1. Interesting line of reasoning. Have you considered sharing your analysis more widely?