The Hacker's Diet: Losing weight the hacker way | Hacker News Reader

The Hacker's Diet: Losing weight the hacker way

fourmilab.ch

50 pointsjjguy53 comments

I followed this Diet back in 2003 and lost 40 Kilos (about 88 pounds) over 6 months, most of them in the first 3. The key things for me were:

Giving up alcohol for two months Not eating much at all Doing the fitness regime to boost metabolic rate

By the time I stopped I was on level 45 of the ladder, had never felt fitter in my life and wasn't ripped per se but was certainly toned.

Since then my weight has gone back up largely due to work and old habits coming back - I've taken up cycling which is definitely having a positive effect, but I'm not sure it'll be as effective without consideration to other areas (crap food, booze etc.)

Four years after reading this the first time, I still remind myself of one insight: engineer vs. managers and "fixing problems" vs. "managing problems".

You can find this discussion under the heading "Problems: managing, fixing, and solving" here http://www.fourmilab.ch/hackdiet/e4/eatwatch.html

It's pretty brilliant. Even when my (manager) head knows I should be just managing problems, my (engineering) heart really, really wants to just fix them.

Has anyone tried the Shangri-La Diet? Seems too simple not to try it. From wikipedia:

"The diet itself consists of taking 100–400 calories in the form of extra-light (not extra-virgin) olive oil or sugar water per day, either all at once or spanned throughout the day. This must be consumed in a flavorless window, which is at least one hour after flavors have been consumed, and at least one hour before flavors will be consumed.[4] The consumption of these flavorless calories supposedly lowers the set point, and therefore, lowers weight."

I just mention this because I've read the authors paper "Self-experimentation as a source of new ideas: Ten examples about sleep, mood, health, and weight" and it seemed really interesting. Available here:

http://sethroberts.net/science/

> Has anyone tried the Shangri-La Diet

No, but I've tried the Get Off Your Ass and Don't Eat More Than 2000 Calories Per Day diet, and it works really well.

And by 2000, I presume you mean 1500 or less. Most people's bodies can probably easily adjust to 2000 calories at something near their current weight. In fact, for most people who want to lose weight, it would probably be easier just to keep calories/day to "about 1000" rather than trying to hit a precise target like 1850 or whatever their height and build should require.

No, those numbers are way off. If you eat 1000 calories a day and are male, you will almost definitely be in poorer health. You will lose muscle and gain fat %, and while you may be "lighter", you will almost definitely be fatter.

Use fitday.com or something to ACTUALLY TRACK (yes, actually, as in EVERY DAY) the calories you eat. You can't "guess" something like this. Furthermore, google for 'daily caloric expenditure calculator' and find out exactly how much someone of your height and build is burning, and eat accordingly.

Stop guessing! It obviously doesn't work.

Of course you'll be in poorer health. You're losing weight, and if you keep it up you'll eventually starve. The process of "losing weight" has to be temporary. However, if you could determine the calorie count that would keep you stable at your desired weight, you will never be able to reach that weight on that calorie count (all else equal: no extra exercise, etc), because your body will lower metabolism enough that you'll level off above that.

I have lost weight in a sustained fashion to accomplish a goal just once in my life so far (~310 to 187 at my lowest; let's agree that I wasn't fatter at 187, eh?), and it would have taken years to drop it had I just started eating 2000 kcal a day. Even 2500 kcal is probably enough to maintain 300 lbs if you don't exercise and don't actively try to build muscle.

> if you don't exercise and don't actively try to build muscle.

Which is why I said "Get Off Your Ass".

I parsed that as a modifier to the bit that followed, rather than a separate prescription. :)

Yes (I bought the book, which is really just a lot of fluff around the central idea as you summarized), though I didn't stick with it long enough to make it stick, it definitely started a downward weight trend.

His original insight (discovered while traveling in France) was worthwhile, though, and confirmed through testing, and confirmed by thousands of people commenting on his blog: apparently, the body will somehow store calories more when it detects incoming things "it" likes, and won't otherwise (thus the sugar water/flavorless olive oil idea, which provides the calories without the "likes").

Yeah, I tried it two years back and lost about 15-20 pounds over the course of 8 months. Regardless of whether the reduced hunger is a real effect of the oil or just a placebo effect, it did work, albeit slowly. If you're going to try it, I suggest the oil route, since it also had a noticible positive effect on my skin and complexion.

But like any regimen, once I stopped taking the oil, I regained the lost weight. You'd have to keep taking the oil forever or gain some self-control without that crutch.

I'm on this "diet" and have lost about 45 pounds so far.

Getting skinny always seemed like black voodoo or something. Looking at it the way I would look at hacking together a system made it really, really, really easy.

The hacker's diet is a perfect example of a watched metric improving just by being watched.

When I needed to lose weight recently, I watched what I eat (in a google docs spreadsheet) and weighed myself every day, and lost 10kg in about 10 weeks.

Can you see yourself doing that for the next 30-40+ years?

No, but the principle of the hackers diet is to keep doing it.

I'm hoping that something I read recently about people likely having two different weight equilibriums (can't find a link) holds true.

Your weight equilibrium is definitely not fixed. It depends almost entirely on your hormone levels. If they are brought in to balance, your weight equilibrium will fall somewhere in a healthy range of 14%-20%. The good thing is it will stay that way. You'll have to keep your blood sugar and other things like cortisol under control, but you won't have to count calories.

Why not? It doesn't really take much time to note down what you ate, or step on a scale. I brush my teeth every day and that takes much longer.

Seconded. When my weight starts ticking up and I can't bring it under control, my "nuclear option" is to record my weight every morning and write down everything I eat. At the end of the day I put everything on a blog. Even though the blog doesn't bear my name, doesn't have any readers, and can't be connected to me in any way, it hurts like hell when I have to record a bad food day. It hurts even worse than eating right. It's my bogeyman.

The link is interesting - it's a diet by the guy who founded AutoDesk (AutoCAD), and has some good discussion on plotting trends of weightloss, taking into account large noisy fluctuations in water weight.

There's an unofficial site you can join and plug data into to get those graphs here: http://physicsdiet.com/

He's right about the fixing versus managing. Although for sanity's sake you really should think of it as permanent lifestyle changes, not "management." And also you should try to optimize for overall health with your lifestyle changes, not weight loss (which comes with improved health anyway).

One place this approach falls down is that it doesn't distinguish the effects of various types of calories. Just pursuing "variety" isn't going to cut it. You can eat a ton of fiber (with high caloric content) but you'll actually absorb very little of it. If you eat a ton of sugar you'll crash your metabolism, your insulin will skyrocket, and you'll put on pounds even if you started out burning more than you consume (plus you may induce diabetes).

Really, a hacker approach will acknowledge that what you're really trying to reduce is fat, not weight alone (is anyone here trying to lose muscle?) And calories input do not go directly to fat. There is a sophisticated set of hormone feedback systems that govern whether fat is stored or broken down for energy, and until you hack your hormone system you may well be chasing your tail.

The simple formula of 'Calories in - calories out = change in weight' has always been interesting to me. More recently i've been wondering if there is some way to measure calories burned accurately using some sort of skin-contact device. I've seen the "bodybugg",[1] but I have trouble believing a few sensors and a microcontroller is worth $250+.

Anyone have any experience with this sort of electronics? The "bodybugg" website claims that it does its measurements with an accelerometer, a 'heat flux sensor', a temperature sensor, and a sensor that measures galvanic skin response.

The more I look around, the more this looks like a market ripe for a quality start-up.

[1] http://www.bodybugg.com/science_behind_bodybugg.php

Suunto T6 owns this market. To use it right, tho', you need to have a VO2max test done.

Also, it's not as simple as that. You might think "I'm eating less so my body must burn fat" but your body is thinking "shit, I'm starving, better conserve energy" and turns down your metabolism, prioritizes fat storage, etc. A whole bunch of survival mechanisms kick in. Cutting more than 500 calories/day doesn't lead to sustainable results. There's no quick fix.

If you have the dedication to run, or do some other cardio exercise, for hours a day you can lose fat quickly.

I'm having poor google-fu at the moment, but I recall hearing about a study which suggested that weight training was 7 times more effective than aerobics, so just do 20 minutes of weights a day and skip most of that cardio.

"Cutting more than 500 calories/day doesn't lead to sustainable results."

Sure it does, if you KEEP IT low (enough). That's the problem with viewing a diet as something you do for a while, and the return to what you've always eaten. If you want to follow a diet to get permanent results, you need to do it for the rest of your life.

The point of the advice is that if you cut too much your body starts catabolism (destroying muscle tissue). Ever kg of muscle you lose reduces your basal metabolic rate, and so if you go on a really restrictive diet you will find that you can't go back up to what was previously your maintenance level without gaining weight again.

Yes, you need to keep to A diet for the rest of your life to avoid adding weight back, but if you cut reasonably low amounts of calories of your maintenance level you will be able to go back to a much higher level when you've reached your target weight than if you cut a high amount of calories to get your weight loss.

As somewhat of a sidenote:

The currently most hyped diet in Sweden (and possibly other parts of the world too?) is LCHF - Low Carb High Fat (a variation on the Atkins and GI theme if I've understood it correctly) claims to be different from other diets.

From what I've understood the idea is to eat practically no carbs, which will trigger something called ketosis in the body, which in turn supposed to be good.

Read all about it at http://en.wikipedia.org/wiki/Low-carbohydrate_diet

Whatever you do, don't start a diet debate with LCHF fans, they tend to be fairly fanatic :) (although in a curteous way, they'll overwhelm you with references to various studies, papers and whatnot).

This is why you must continue weight training while cutting, so that your body knows it needs to keep the muscle, and you can lose more fat than muscle. Your maintenance level will drop, but once you cut low enough, you can bulk back up and pack on more muscle than fat, which brings your maintenance level back up as well. Repeat this process until you are at the weight and bodyfat level you desire, then just maintain.

You right. So you do not need to eat less. You need to eat less calories, right? Eat meat instead of burger - you will be satiated for much longer time with much less calories. Also you need to eat vitamins - they will help organism to convert fat into energy when necessary and you will be less hungry in average.

It's not even a case of less calories. Let's say you need 2000 calories a day. You could eat that in two larger meals with lots of fat and refined carbs, or you could eat in in six smaller meals with protein and complex carbs. Diet 2 is massively better for you.

> Also you need to eat vitamins

Since when did "vitamins" replace "fruit and vegetables"?

While on diet, you need more vitamins than usual.

Fruits and vegetables has calories too. Look at tables: http://www.fatfreekitchen.com/calories.html

1 apple is about 200g (five apples per 1kg), thus it is equivalent to 2 tablespoons of sugar.

You can measure calories burned pretty accurately by measuring CO2 exhaled. Measure the concentration of CO2, the length of the exhale, and the force of the breath and multiply to get a quite accurate view.

Hmm, you might also be able to use a pulse oximeter to measure dissolved CO2, plus a chest band to measure respiration. Dissolved CO2 should more or less match CO2 concentration in breath, so you would not need a visible cranula on the nose.

Carbon is not the only thing the body burns, hydrogen too, which makes water which can't be used to measure calories, but because the majority of what we eat has about the same ratio of hydrogen to carbon, measuring CO2 should be enough.

This Diet is way to complicated. A better way is simply:

* Don't drink alcohol. Alcohol has calories like fat. * Don't drink stuff with calories. The best is plain water * Cut back on carbs. Don't eat outside. Most food you buy is not suited for dieting * Eat your veggies. Eat at minimum one pound per day. It is a must. It is healthy and fights your hunger. * After you are down to the desired weight, don't get back to normal. Still eat your veggies and you are fine.

Maybe that works for you, but most of us in the real world have been trying that for years, and with little result. You cannot manage what you cannot measure. A simple, easy to understand, but certainly cumbersome at times system like the hacker diet works well.

Also, you'll stick to your diet longer by eating smaller portions of the food you like than by eating large quantities of food you hate (I'm looking at you, salads). You're better off being hungry than dreading meals.

As always, YMMV.

> but most of us in the real world have been trying that for years, and with little result

Oh please. If you followed all those points and didn't lose weight, you're defying the laws of physics.

Use fitday.com to actually count your calories. If you're doing all those things and STILL can't lose weight, the simple fact of the matter is you're eating TOO MANY CALORIES.

You're probably one of those people who has no idea that a small cheeseburger, fries, and a coke represent over HALF of the calories you're supposed to eat in a single day.

The diet isn't really too complicated, it just puts the emphasis in the wrong areas. What you eat is actually slightly more important that how much you eat.

Your recommendations are a good start, but you should add that 50-70% of each meal should be protein. Protein is the fuel for your metabolic fire (long term), and will cause your calorie burning to outrun your caloric intake in general terms.

Part of why the veggies are such a good idea is that they can be mostly fiber, which are calories that are not absorbed at all, but help you absorb the other, more important nutrients (protein, minerals, vitamins).

And to maintain this you must view it as a lifestyle and the goal is overall health, not a certain weight. The right weight will happen as the result of being healthy.

Forget the protein. It's ok, but you don't need. I included the veggies, because the don't have much calories and they are fighting your hunger. At the end they are healthy too, but that is not the point here.

You don't need protein? That's an odd statement to make.

That kind of protein intake is absolutely nuts.

On average, the Japanese get over 50% of their calories just from rice and over 80% from carbohydrates. They also have the longest lifespans in the world and far lower obesity and heart disease rates than similarly developed places.

You're probably talking about the Okinawa Diet (http://en.wikipedia.org/wiki/Okinawa_diet). It contains 25% of the average Japanese intake of sugar and 75% of the average Japanese intake of grains in general. It is very low calorie.

To get the long lifespan, low obesity, and low heart disease of an Okinawan you'd have to truly live like an Okinawan... eating exactly the kinds of foods they eat and having little stress, similar sleep habits, and similar genetic dispositions. People have not been successful in replicating the approach elsewhere.

The approach I described is the way to eat to stoke your metabolism and keep your blood sugar low. It's good for fat loss, muscle building, high energy levels, and general good health. There's more to it than I described (such as eating 5 or more times a day), but those are the basics.

I followed this diet 3 years ago and it did have great results. I think that the exercise had the biggest effect on me, as I was completely flabby. I'm planning on starting again as I have reverted to my old chubby self the past couple of years.

I think he's optimizing the wrong parameter. What's the ultimate point of hacking weight? For me, it's having good health. Health is not just about weight, health is composed of quantity and quality of food, frequent aerobic and anaerobic exercise, and adequate sleep. Hacking the diet is necessary but insufficient without the other two parts.

The main beef I have with his "hack your weight" approach is that he's strictly counting calories, and not the kind of calories.

Yes, it'll "work," but if you use McDonald's calories instead of Trader Joe's calories (you get the general idea), your body will be much worse off with the junk food, even if your weight is down.

I'd like to try this. It seems like he only has spreadsheets available for Excel though. Anyone tried to make something similar for gnumeric or maybe google spreadsheets?

OOo and Google Docs can open most Excel spreadsheets..

He also has a webapp that does the same thing:

https://www.fourmilab.ch/cgi-bin/HackDiet

I stopped drinking a 6 weeks ago and I've lost 15 pounds.

You might be surprised how much alcohol (not just the calories) is contributing to your fat ass.

This is a lifestyle thing. It's not just alcohol, but yeh, out partying and drinking leads to 2am takeaway, hangover foods and the calories in and of itself in alcohol are bad.

Like everything, you will find people who tell you "the best way to lose weight is X" ... employ a multi-faceted approach and you'll get the best results. There is no silver bullet. Some people will not stop snacking, some people will not stop eating greasy takeaways and some people will not stop driving 3 blocks instead of just walking. Try "Being healthy" and if you find it difficult to do everything at once, take baby steps. Eat healthy food on Tuesdays and Thursdays only, building up one day at a time every two weeks. Sometimes it can take a while to get used to not being full of MSG-ridden salty foods and stuff like removing caffeine from your diet can be difficut. Getting used to the taste of wholegrain this and low fat that takes a while. It's better introduced slowly... you'll find it more difficult to keep to some sort of strict that requires that you eat stuff that you don't like 3 times a day.

It is important to do as much as you can to increase your metabolism. This means not being sedentary! Get out, go for walks, exercise, go for a jog. Whatever. Measure, rinse, repeat.

Take a sensible approach to changing your lifestyle and it can definitely come about. Fad diets, exercise machines, stomach crunchers etc. are useless without you recognising that you want to be healthy. It's not just about losing weight and remember that! There are people of average weight living extremely unhealthy lifestyles who will die younger too.

> This is a lifestyle thing

"Losing weight" is a lifestyle thing... people with healthy lifestyles aren't fat, no matter what. I believe this is one of those rules to which there aren't any exceptions.

Beer and the soda sugar (added to liquors) have a lot of calories and carbs.

You didn't say what you were drinking. You can get a lot of the benefit by switching from beer to Gin and Tonic / Vodka and tonic / Whiskey / Bourbon / any alcohol without the added carbs.

> You didn't say what you were drinking

Yeah, well, part of the problem I was drinking anything I could get my hands on.

Reply on news.ycombinator.com