From zero to chess champion in 30 days
November 17, 2017 12:22 PM   Subscribe

Max Deutsch is an obsessive learner. Last year he embarked on a quest to master twelve incredibly difficult skills in twelve months. This is story of his final and most challenging month of the past year: Defeat world-champion Magnus Carlsen at a game of chess .
posted by noneuclidean (57 comments total) 36 users marked this as a favorite
 
hatereading this so hard
posted by blue t-shirt at 12:48 PM on November 17, 2017 [17 favorites]


FAQ #4: Are you working on this project full-time?

No. I have a full-time job and a decent commute from San Francisco to Mountain View, which take up a good chunk of my time.


If he doesn't drive to Mountain View, this might give him a solid couple of hours to himself every day, especially if he takes a company shuttle. When I did the same commute, even by BART and Caltrain, I managed to keep a lot of personal projects alive. Ironically (but not surprisingly if you know me), most of those projects went by the wayside when I started working close to where I live. Sometimes I consider looking for another job with a long (public trans) commute for that reason.
posted by treepour at 12:49 PM on November 17, 2017 [4 favorites]


Honestly if I could just achieve one of his minor background requirements he wants to keep up while on his quest, sleeping 8 hours a night, I'd count that as a huge win.
posted by Jon Mitchell at 12:51 PM on November 17, 2017 [12 favorites]


I think the writer of this article is an idiot, and Max seems overly self-centered. But as a chess expert I never really considered whether it might be possible for a human to achieve a high level of play by memorizing and executing a non-tree-search chess algorithm in realtime. That's an interesting idea* so I'm glad he explored it for a month.

*Although wow, would it kill competitive human chess.
posted by value of information at 12:53 PM on November 17, 2017 [13 favorites]


(After reading his blog, my guess is that such a thing is possible, but would require at least man-years of human effort being spent on tuning the model in order to be competitive, and be hard and unpleasant to execute. See his posts here and here for a few details of what he was working on, since apparently the WSJ would rather write pages of bullshit than any actual information. Here is some recent work on deep learning in chess that is a long way away from a small enough model for his approach to work.)
posted by value of information at 1:02 PM on November 17, 2017 [3 favorites]


"Keeping him a little bit nervous for at least eight moves was a minor victory"

I hope Magnus gets a reasonable appearance fee every time some moron like this wants to touch the hem of his garment.
posted by fredludd at 1:06 PM on November 17, 2017 [22 favorites]


Inspirational. In the spirit of one-upsmanship, I plan to attempt a similar program of accelerated learning, except I will do each of the tasks in only ONE WEEK:
1. Do ten push-ups in a row.
2. Gather ten thousand dead leaves into a pile.
3. Build a killer robot (software only).
4. Climb a mountain (or at least a hill).
5. Bake a pie.
6. Become the world's best Starcraft 2 player.
posted by sfenders at 1:07 PM on November 17, 2017 [26 favorites]


I found this story really engaging and compelling. It was a fun read. What struck me most is that I don't think there is anything exceptional about Max but his circumstances. I think the writer of the article tried very hard to convince the audience that Max has an unusual ability to learn new skills. He failed to convince me. Max seems like someone who has had many advantages growing up including a good education, a high-paying job that paid enough for him to actually quit it and stay financially sound, and most importantly, a world that took the existence of his skills on faith, partially because he's a cis man.

Others may take that differently from me (especially since I have benefited from many of those advantages), but I actually was pleased by that. I'm a great believer in growth mindset. I think Max achieved 11/12 of his goals because he a) spent a lot of time working on them (not the bullshit 10k hours, though), b) researched them and listened to experts rather than assuming he already knew how to do them, and c) wasn't deterred by failure. I am not a very good baker, I have a couple of weeks off in December, and I'd like to learn to make bread from scratch. Wish me luck?
posted by capricorn at 1:08 PM on November 17, 2017 [23 favorites]


I don't get the hate. It's basically just a year-long variety-show version of NaNoWriMo, isn't it?

Granted, I read some of guy's other "thing in a month" articles rather than the chess one, because I have no interest in chess, but he seems to be quite aware that he's setting specific parameters for success and -- at least in the ones I read -- neither acts as though he's a super genius nor as though these are things that everyone should be expected to be able to do.
posted by inconstant at 1:11 PM on November 17, 2017 [7 favorites]


Good luck, capricorn! Being able to make bread sounds like a very delicious skill.
posted by inconstant at 1:12 PM on November 17, 2017 [4 favorites]


I enjoyed reading through a few of his Month to Master stuff. Also don't get the hate. He's a young guy who likes to push himself and learn new stuff outside of his comfort zone - that seems pretty healthy to me.

When reading through some of the other Month to Master entries, I came across this brief mention in the Learning to Backflip month:

In fact, for nearly six years now, I haven’t consumed any coffee, alcohol, soda, desserts, sweets, or junk foods. Instead, my diet mostly consists of chicken breast, turkey meatballs, salmon, spinach, string beans, quinoa, and potatoes.

That says a lot, to me, and is inspirational as well. I've been trying to succeed at making big changes to my consumption of alcohol, coffee, and junk food. I've been failing pretty hard. This was a good kick in the pants.

Thanks for the post!
posted by lazaruslong at 1:26 PM on November 17, 2017 [5 favorites]


That says a lot, to me, and is inspirational as well.

One man's inspiration is another man's horrorshow.

But, nonetheless, good for him. If you have the time--which most people don't--why shouldn't you push yourself? His approach might be a little excessively programmatic, especially in the way that he defines his goals, but...
posted by praemunire at 1:36 PM on November 17, 2017 [3 favorites]


I've spent seven months watching and writing about every single Weird Western ever made, where's my book deal?
posted by maxsparber at 1:39 PM on November 17, 2017 [8 favorites]


Well, I thought the story of the chess match was interesting. They said Magnus agreed to do it essentially as a lark set up by the journal, and it was fun to see his enthusiasm at the concept during his post-game breakdown. Plus, this guy tried to develop a machine learning model and teach it to himself - that alone is pretty fascinating.
posted by Think_Long at 1:43 PM on November 17, 2017 [4 favorites]


I'm also not terribly interested in chess, so I read the one about learning to draw a self portrait, and I thought it was really great. I really like his methodology, and although I only have a fraction of his free time, I want to try the same thing myself.
posted by lollymccatburglar at 1:54 PM on November 17, 2017 [1 favorite]


I remember a gymnast telling that the backflip was pretty easy and she'd taught her brother how to do it. It's kinda like jumping onto a table, pretty much anyone in good health can do it but it's clear that if you miss you hit your shins and wow does that hurt.
posted by sammyo at 2:15 PM on November 17, 2017 [1 favorite]


I enjoyed reading the article but what’s up with the naming of clothing brands? The jeans and the fleece jacket... Really puzzling how the brands are named out of the blue with no reason, I had to re-read those bits twice to make sure there wasn’t some joke I had missed. Especially the jeans thing, I cannot believe someone embarking on such a challenge would pay attention to a brand of clothing as some sign from above or something...
posted by bitteschoen at 2:27 PM on November 17, 2017


This and "develop perfect pitch" are the two that stand out. The others I look at them as a non-expert and say "ok, that sounds hard, but there are lots of people who have learned how so I guess it's conceivable that he could too".

Curious to read the "perfect pitch" entries to figure out what's fudged there. Definitions of "perfect pitch" are pretty slippery.
posted by floppyroofing at 2:31 PM on November 17, 2017


To develop "perfect" pitch as he designated it (identify 20 random, presumably A440 chromatic, pitches without a reference tone) you essentially need to establish a reference tone in your head and be able to discern intervals. So you memorize one note and then use that to discern what other notes are.

After having written that, I read his method, and that's what he did. Smart!
posted by grumpybear69 at 2:45 PM on November 17, 2017 [4 favorites]


It's interesting the bars he set himself. The rubik's cube one was sub 20 seconds but the chess one is beat the world champion? To beat the rubik's cube world champion you'd need to be down in the 5-6 second range. 20 seconds is the equivalent of beating some random guy at your local chess club. Still, I enjoy people trying this sort of thing.
posted by markr at 2:48 PM on November 17, 2017 [3 favorites]


I don't get the hate. It's basically just a year-long variety-show version of NaNoWriMo, isn't it?

Eh, it's kinda like every other click-bait article on someone doing something weird and difficult? It's like every year someone is trading a paper clip for a house, or doing 100 days of rejection therapy, or recycling everything they use. Some of these also become TED talks, which are also have a mixed reputation around here.
posted by FJT at 2:54 PM on November 17, 2017 [4 favorites]


This was much less annoying than I thought it would be (which probably says as much about me as it does about him) and I like that he mixed it up and put in physical and artistic challenges as well as more obviously nerdy ones.

But unless he keeps them up (and he probably won't, because they all require lots of time to practice and reinforce), he's going to be the guy who did a backflip once or used to be able to speak Hebrew or whatever. I sort of wish I'd spent my copious free time in my 20s doing this, but then, I did. I was an enthusiastic juggler and stuck with it for years and got quite good at it and thanks to those years of practice I'm still better than just about everyone despite the fact that I haven't done it seriously in over a decade.
posted by It's Never Lurgi at 2:55 PM on November 17, 2017 [4 favorites]


I think the author mentioned the jean brand because they are a sponsor of Magnus, implying that the player in the park may have been a low-key pro or hardcore fan in incognito mode
posted by Think_Long at 3:07 PM on November 17, 2017 [1 favorite]


To develop "perfect" pitch as he designated it (identify 20 random, presumably A440 chromatic, pitches without a reference tone) you essentially need to establish a reference tone in your head and be able to discern intervals. So you memorize one note and then use that to discern what other notes are.

I figure that's the method he had to be using, but I don't think of perfect pitch as being defined in a functional sense, but rather in terms of sensory affordances: if you have it, you don't need to establish a reference pitch and relate successive pitches to it. Also, I think it implies a lot about intonational abilities, in that people with perfect pitch are usually capable of hearing tuning differences on the order of cents. I think that's one of the things people are finding frustrating about this; when you say "I learned perfect pitch in a month!" but what you actually mean is "I learned [to fake one necessary but not sufficient aspect of] pitch pitch in a month!"...you're gonna get some eye-rolls.

Too this reeks of Silicon Valley's particular style of anti-intellectualism, where all it takes to achieve a level of effectiveness in a domain equivalent to that of all those poor saps who actually spent years training in it is the ability to program intelligence, moxie, and a rebel spirit. It doesn't look as ugly as the Trumpist variety at a glance, but it shares common causal factors. I don't think he's actually pushing that directly here, given his acknowledgement that a lot of these challenges involve developing somewhat brittle heuristics to approximate mastery in a limited subset of the problem space, but it's inherent in this framing. I'd be a lot more interested and less annoyed by this if it were framed that way explicitly, using each challenge as a jumping off point to explore the ways in which that heuristic can fail and what that shows about the difference between true mastery and the kind that you can develop in a month.
posted by invitapriore at 3:08 PM on November 17, 2017 [30 favorites]


I really enjoyed the WSJ article, especially the interaction between the two people. But do the blog entries go into explaining the machine learning algorithm that was being attempted?
posted by polymodus at 3:41 PM on November 17, 2017


yeah, that version of "perfect pitch" works for a pianist maybe but not a violinist

I guess "good enough pitch" doesn't have the same ring to it
posted by idiopath at 3:53 PM on November 17, 2017 [4 favorites]


After 8 moves, Houdini or Stockfish says Yeah, white still has it's opening statistical advantage. Journalist: Max is winning!

Sure, why not?
posted by Dumsnill at 4:19 PM on November 17, 2017 [4 favorites]


It is probably pretty clear that this sentence from the article is not correct:
After eight moves, using his own limited chess ability, the unthinkable was occurring: Max was winning.
but in case anyone is still not sure: no, Max was not "winning". Carlsen played a slightly suboptimal opening on purpose in order to mix things up and Max just hadn't made any mistakes yet.

Also, the article claims that the game lasted 39 moves, but it was resignable after 14. Carlsen just took his time finishing it off.

On another topic,
I don't think of perfect pitch as being defined in a functional sense, but rather in terms of sensory affordances: if you have it, you don't need to establish a reference pitch and relate successive pitches to it.
Yeah, I hear a C# and it sounds like a C#, not like a major third above some reference A. It is just like when you see green you see green, rather than something 50nm longer in wavelength than your reference blue. However, if someone identifies pitches that way, I am happy to credit them with perfect pitch, even if they accomplish it in a less "natural" way than I do.
posted by dfan at 4:29 PM on November 17, 2017 [9 favorites]


I also found the tone of the second piece (written by Ben) somewhat obnoxious but I enjoyed Max's writing/blog itself. I've always wanted to learn to draw portraits but thought I just didn't have the skill. It's really cool and inspirational to see what a person of average talent can learn to do with some time and focus.
posted by Emily's Fist at 4:30 PM on November 17, 2017


By the way, based on the single game, which of course is a ludicrously small amount of data, if I had to guess Max's USCF chess rating, I'd guess maybe 1400? Which would place him below the median of adult tournament chess players, but not by a ton. If an adult reached that level in a year starting from scratch I'd be pretty impressed, and there are plenty of people who take the game pretty seriously who never get better than that. So I don't want to diminish his accomplishment in getting there in a month.
posted by dfan at 4:38 PM on November 17, 2017 [2 favorites]


However, if someone identifies pitches that way, I am happy to credit them with perfect pitch, even if they accomplish it in a less "natural" way than I do.

I don't think accomplishing it the way the author does here is bad or lesser in any way than your perfect pitch (it's what I rely on, after all), just that the immediate representation of absolute pitch as qualia in the mind of someone possessing perfect pitch is definitional.
posted by invitapriore at 4:44 PM on November 17, 2017 [2 favorites]


yeah, that version of "perfect pitch" works for a pianist maybe but not a violinist

yeah, yeah, but can he play in the pocket? i've seen musicians who probably had perfect pitch and excellent technique who just couldn't make the music come alive

you can buy something at a music store that has perfect pitch for under 20 bucks these days - that doesn't mean that everyone with 20 bucks is going to be a great musician
posted by pyramid termite at 5:06 PM on November 17, 2017 [1 favorite]


He could study full time for 20 years, Magnus could spot him a knight, and he would still lose.
posted by starman at 5:07 PM on November 17, 2017 [2 favorites]


I'm sure most of them are amazing but I only checked out two that share my interests and wasn’t as blown away as I expected to be. The self portrait from a photo ..... I guess he learned something about proportion and value but there wasn't much development or exploration of structure in it or much of a sense of what lines can do. If he'd even done it from life, I think he could've learned a lot more. The chinups...WTF? I'd always been told you had to do each one from a dead hang!
posted by bonobothegreat at 5:07 PM on November 17, 2017 [1 favorite]


Good lord, what a tough crowd. The kid is trying to do a new thing every month for a year in a systematic way and talk about it. This is the internet, after all, but also MetaFilter - if you don't like a thing and just want to shit on it, please skip the thread?
posted by lazaruslong at 5:10 PM on November 17, 2017 [11 favorites]


This reminds me of Magnus secretly playing German GM Jan Gustafsson (one of the best 100 or so in the world) in a game of blitz. Guess what happens..
posted by starman at 5:12 PM on November 17, 2017 [3 favorites]


I like the whole project, and I'm glad Magnus agreed to meet him even if it was just a stunt. I want to learn how to juggle.
posted by Dumsnill at 5:13 PM on November 17, 2017


A friend of mine made an art project out of trying to master a standing backflip. She did not succeed because all of the jumping practice detached her retina and she had to quit or go blind, so I'm gonna have to call out that "backflip (is) pretty easy" bit from your gymnast friend, sammyo.
posted by queensissy at 5:17 PM on November 17, 2017 [4 favorites]


Isn't the whole self portrait thing basically "drawing from the right side of the brain"? That book is cool but hardly new or revolutionary.
posted by ch1x0r at 5:46 PM on November 17, 2017 [1 favorite]


'when you say "I learned perfect pitch in a month!" but what you actually mean is "I learned [to fake one necessary but not sufficient aspect of] pitch pitch in a month!"...you're gonna get some eye-rolls.'

It's also an approach to ear-training that a lot of people fall into because it seems simple and obvious: "first I'll learn to identify each pitch, then I'll be able to hear *anything*!" Except it turns out not to work that way.

It's a bit like if someone set out to understand French by spending hours drilling identification of each possible French phoneme before ever listening to a sentence. I don't know, maybe that's a poor analogy.

It also bugs me because there are some snake-oil salesmen in this area. You'll do well to ignore any ear-training course claiming to teach perfect pitch.
posted by floppyroofing at 6:15 PM on November 17, 2017 [3 favorites]


Perhaps Max Deutsch just needed more chesstosterone.
posted by clawsoon at 6:21 PM on November 17, 2017


I'm just going to call bullshit on "learning" the sitar in 15 minutes.
posted by misterpatrick at 6:32 PM on November 17, 2017 [4 favorites]


Dumsnill, learning to juggle is great! Some people manage to do it in less than an hour, but it took me dozens of hours of stubborn practice over the course of a few weeks before I could juggle three balls consistently. I kept at it though, and after several more years of daily practice I managed my ultimate goal of juggling five balls. I got into it enough that I went to meet-ups and festivals, and for a brief period was the editor for juggling.org, which, OMG, it's still online and still looks exactly like it did 20 years ago. Anyway, there's a ton of information on that site (or at least there was) and these days there's probably even more on YouTube and the like.

The great thing about juggling is that it really rewards practice. There's always a new trick that is just within reach, but no matter how much you learn there's always something left to try.
posted by mbrubeck at 7:08 PM on November 17, 2017 [4 favorites]


> I'm just going to call bullshit on "learning" the sitar in 15 minutes.

I dunno. Figuring out the modal noodling part of playing an unfamiliar stringed instrument in 15 minutes, and then improvising little melodies, is something some people can do and other people can't. It's a fun party trick whether or not it's difficult because people are entertained by pleasant music.

Playing a raga from scratch, on the other hand, would be pretty unlikely.
posted by ardgedee at 7:15 PM on November 17, 2017


I like this circumscribed approach. I feel like long-term goals like "learn an instrument" or "write a book" are difficult because there is (at least in my brain) an unspoken addendum that to finish these things, you must be able to perform them competently. But for so many creative activities, there's no end to the new things you'll still be bad at since you've developed new critical faculties.
These set goals instead suggest that you should learn -just enough- to accomplish some aim (arbitrary or not). And you can then build on that.
posted by solarion at 7:16 PM on November 17, 2017 [4 favorites]


I love challenges like these, quests of "I am gonna do this thing for X days even though I know not much of it!". I've attempted this in the past though I have noticed that documentation is the most boring part.
posted by divabat at 9:47 PM on November 17, 2017


At least from my musical education some years ago- you "learn" how to have good relative pitch. An adult without absolute/ perfect pitch can't suddenly teach themselves.
posted by NorthernLite at 10:01 PM on November 17, 2017




Each life is a game of chess that went to hell on the seventh move, and now the flukey play is cramped and slow, a dream of constraint and cross-purpose, with each move forced, all pieces pinned and skewered and zugzwanged.

. . . But here and there, we see these figures who appear to run on the true lines, and they are terrible examples. They're rich, usually.

-- Martin Amis, Money, 1984
posted by lazycomputerkids at 12:42 AM on November 18, 2017 [4 favorites]


He's certainly mastered the art of self-promotion.
posted by davebush at 4:32 AM on November 18, 2017 [3 favorites]


I'm glad he lost. I didn't imagine he could win because, come on, but I'm glad he lost because I'm small minded and well
Hey here, here's something for you to learn Max Deutsch - you wanna show how cool you are? How much you can do? Make elections fair in America. Ok? Or how about actually provide a fair and equitable health care system? Huh? Can you do that? Huh? Ok, so why Dont you go ahead and do that and then get back to me. Yeah so until then Mr. Quirky Party Trick, you know, like, have a nice day.
posted by From Bklyn at 4:45 AM on November 18, 2017 [2 favorites]


Such sarcasm.
posted by Dumsnill at 5:39 AM on November 18, 2017


Good lord, what a tough crowd. The kid is trying to do a new thing every month for a year in a systematic way and talk about it.

This is a fair point, and while I don't necessarily think another go round on the anti-tech-bro-Ferriss Wheel is helpful or productive, so many terrible, terrible things have come from the same wellspring of confident relentlessly self-improving engineers who think they can systematically shortcut their way to a better tomorrow that enjoying this - which I did! He seems thoughtful and engaging and a thoroughly good sort! - does feel just a touch like trying to appreciate the cuteness of a baby velociraptor.
posted by Jon Mitchell at 2:11 PM on November 18, 2017 [6 favorites]


From Bklyn is my hero.
posted by Don.Kinsayder at 3:47 PM on November 18, 2017


I thought I would hate this a lot more than I did, but after reading through a lot of his blog, i'm ok with it. Did I fail Metafilter? He sets pretty clear goals on each thing that aren't "learns a language in a month", and always admits up to where he's starting from. For the Hebrew one, for instance, he's already studied it in college for years and has lived abroad. For the freestyle rap one, first he admits he did it specifically because it's scary for him, and then he analyses himself to choose the style that would work best for him.

I mean, I personally study a foreign language (for an American, in and of itself a revolutionary act I think) and I draw and I hang out with my cat.... And I also call my senators and phone bank and register voters. I'm not sure why self improvement is thought to exclude these things?
posted by tofu_crouton at 6:28 AM on November 19, 2017 [3 favorites]


I think the writer of this article is an idiot, and Max seems overly self-centered. But as a chess expert I never really considered whether it might be possible for a human to achieve a high level of play by memorizing and executing a non-tree-search chess algorithm in realtime.

Well, there's Bernard Parham's Matrix system.

Anyway. These guys, always with the world champion. Holding a draw against a National Master would be a very ambitious goal for one month of study. Very. But that won't do, gotta beat the World Champion. They always turn up on chess message boards, 21 year old beginners announcing that they will become Grandmasters in three years. It's like buying a set of golf clubs at the thrift store, and telling everyone you'll be on the PGA tour next year.
posted by thelonius at 7:24 PM on November 19, 2017 [1 favorite]


I looked at the art one and it confirmed that just about anyone can be reasonably good at art if you learn a few observational and physical techniques and put the practice in (to become really good requires a lot of effort and unlearning and relaxing... and to be a genius requires genius)

Taking on Carlsen is a total joke... I mean Carlsen played Bill Gates with a ginormous time handicap and still utterly crushed him.

Still it inspired me to have a go at chess again myself... wonder if I'll keep it up this time or quit when I get to the 'this is too much like hard work' learning barrier.
posted by fearfulsymmetry at 8:12 AM on November 22, 2017


« Older DnDish: simplifying D&D (you know, for kids)   |   A Public Comittment To The Future Generation Newer »


This thread has been archived and is closed to new comments