High Anxiety (raganwald on Go and learning new things) | Hacker News Reader

High Anxiety (raganwald on Go and learning new things)

github.com

69 pointstptacek46 comments

I've taught Go to over fifty people over the years - and i've seen this over and over again: some people have shaking hands putting down the first stone in their first game ever.

For some reason they assume they ought to be good from the beginning. Even in their first game they try to be original. They're afraid to imitate moves they saw someone else play. It feels like watching someone trying to learn a language by not imitating the sounds they hear when others speak it.

But those that i saw who became really good, were always on the other end of the spectrum. Those who jump right in and played /really/ bad without any anxiety or pressure (no thinking at all and mostly just stubborn imitation of moves they saw someone else play before) and fast (many many many games, often not even to the end), almost always sticked to the game long enough to learn to appreciate and enjoy it, and sometimes even managed to excel at it.

There's something magical about the fearlessness of just playing. Pure curiosity, maybe even quite a bit of ambition, but especially the absolute surrender to repeated and premeditated failure.

I wonder which type of Go player I would be. Sometimes I do something like this with development, where I spend too much time feeling ignorant and try to ramp up my knowledge by reading. Once I jump into the coding I still feel underprepared as things stick better when you are intimate with the context, and I end up feeling like I should have just jumped into coding first. I think most of us have an easier time learning by doing.

Since you teach Go -- you probably teach it well, but I thought I'd share my experiences as a Go student who gave up after a couple of years.

My problem, and Raganwald's, is that I can generally read the rules of any new game and, based on those rules, formulate a strategy. I might not expect to win, but I at least have enough understanding of the mechanism of the game that I can pick a direction and run with it. (I was going to say, "that I can play passably well", but that really isn't it at all -- it's the formulation of strategy that's important to me, and, I suspect, Raganwald.)

So what happens in Go? Well, you're told the rules, and then you're given a few hints that don't make any sense to you, like the 3-4 stone. Either way, it's impossible to form a strategy. You might as well stand above the board and let stones loose from your fingers to fall randomly upon it.

And my teachers, bless their hearts, took it upon themselves to win by extremely large margins repeatedly, which I would not have taken personally if only I had finished each game with the feeling that I had learned something.

In the end, I became an "eh" player, but found myself dreading Tuesday night Go club, found it occupying too much of my mind space, and found that it wasn't really giving me anything positive in return.

There were very very few long term new recruits in the local club.

Well, i've seen this too, not only with Go but with Programming and a whole lot of other things. They're busy showing off how good they are, instead of helping you learn.

However, once you know the trick, you don't need anyone to teach you, but just someone you can imitate. Just get them to be quiet and play lots of games with you, or ask them questions when you really need to know something. Be the driver of your own learning, then your teacher's teaching skill matters less.

one of the stronger players i know did this -- he just played blitz with a 7d over and over and over in a coffeeshop, didn't ask many questions but just imitated first & drew conclusions second.

Yeah, your rules->formulate strategy->develop tactical plan->iterate is usually a great approach! Sadly, Go's rules are so simple it's really tough to guess what good strategies are from a reading of the rules. I think this is because so many of the strategic concepts -- power vs. influence, chasing, pincers, leaning attacks, etc., are the emergent properties of tactical features -- eyes, life & death, ko fights -- which are themselves emergent from an unbelievably simple ruleset. It's similar to looking at DNA and figuring out protein shapes. It's just CGAT, how hard could it be?

Relatedly: So i've heard of two different ways to teach go -- the Eastern way and the Western way.

The western way is to be told the Rules, the Object, and the Strategy and Tactics that will get you there.

The eastern way (this is hearsay) is that a teacher shows you the rules -- and then leaves you to memorize 50 games in their entirety, on your own.

For someone who's done all his learning the western way, my first reaction to the eastern way was "that's stupid."

What's not said in the eastern way is that you learn in reverse. First you have to puzzle out the Tactics -- why did they do this and not this? Why did they take 10 seconds for these 5 moves and 5 minutes for this one? Then you can start to figure out the strategies and feel the flow of the game. Then, maybe -- only maybe -- can you then guess at the actual object of the game.

After memorizing 50 games, then a teacher might actually help you. How dumb! how silly! And yet, this way, the teacher already knows that the student is willing to work.

Anyway, that kind of rambled. I'm sorry you had bad teachers. There's lots of them, probably myself included. Go is really, really hard to teach. I still don't know what the object is, just how to count the score at the end...

One of the difficulties for me here is that I have extremely poor long-term memory, probably due to extended (and now habitual) sleep deprivation.

I simply cannot commit a large enough number of games and sequences to memory.

I do however have a better-than-average working memory, which allows me to compensate in most games by visualizing lots of different game sequences simultaneously.

Go is not one of those games.

I'm with you. I know people who can remember every game they've played and i'm just not one of them. I learned "the western way" myself. I've noticed, though, that i don't remember the recurring patterns or sequences, i just kind of feel them. It sounds weird, but your brain gets quite good at the pattern recognition if you let it.

I would say that being able to visualize lots of different game sequences is crucial in go to learning tactics. You'll do great! Give it another shot :D

This is the most provocative comment I've read on my post. Reading it, I find myself thinking I will never be good because I have some sort of emotional disability that gives me shaking hands compared to some other people that are imbued with the magic gift of curiosity and fearlessness.

Logically, I consider the possibility that such things can be learned or cultivated, and that perhaps I can one day be fearless and curious. But emotionally, there is something pessimistic inside me that believes I will always be this way.

It's quite a disturbing thing to contemplate.

I don't think your problems with Go are at all a given. Playing repeatedly and fast is a choice you can make out of pure logic.

Go is a language and should be tought like a language. You don't explain phonetics to a child, but instead let it immitate sounds, scream and giggle. Try explaining phonetics to a six-month-old child.

There's no reason why you couldn't just play a game, play fast without thinking, and immediately start another one, maybe even before finishing the first one. What you're doing that way is not looking for ways to win, but for the responses you get for your actions. It's like babbling 'apdy' at your father and getting back a 'daddy'. Thinking about how to use your vocal chords isn't going to get you there, what you need is practice. It may not look like you're learning, but well, you do.

View it as learning the sounds of a completely new language, and imagine you're a baby trying to appreciate the noises someone else makes and then try to immitate them. There's really nothing about you that prevents you from doing this except a choice.

You have control of your actions. Even if you're scared of the outcome, if you still practice despite that fear, you will improve.

When I first started training BJJ, I was indeed scared about being "stuck" in bad situations when rolling. I trained anyway. What I tell the beginners is that I'm not the most athletic or the most talented, but I've trained regularly for years. In fact, I feel that I'm actually a slow learner in jiu-jitsu. Showing up to practice despite my fears is why I've learned what I have.

I don't like being controlled by fears, so I find I'm drawn to the activities that scare me. The standup portion of fighting - boxing and muay-thai - scare me. I'm not good at it. But when my jiu-jitsu training partners spar, I spar too. I'm scared, but I don't like backing down because of that fear.

I can't help but recall Thomas Edison's arduous yet ultimately fruitful work to invent the light bulb.

He says, "If I find 10,000 ways something won't work, I haven't failed. I am not discouraged, because every wrong attempt discarded is another step forward."

I'm just discovering Go, and I am terrible! But, hey, I don't know what I'm doing and it's certainly not going to do me any harm to fail until I get it right!

Developers are great at figuring problems out, but there's no replacement for experience. I frequently create little "spikes" to figure out how to nail down a concept... and sometimes I just outright fail. I don't mind, it's a learning experience.

Take a step back from the game, mentally, and put your focus on the metagame; track your progress over many games (might be easier with some consistent opponents), tracking area and game length as well as your recognition of the various Go forms (Seki, Double Atari, etc).

If challenging yourself doesn't work in one aspect, try another way.

You're definitely an inspiration to me, so I hope something here inspires you to overcome your challenges!

Matt

I tend to be afraid of learning new skills. This is quite inconvenient, because I love using the skills and improving on existing ones. So I'm pushing myself to learn the skills anyway.

What I'm finding (and this may change, as I've only been doing this a few years) is not that the fear of new things (and being bad at doing new things) goes away, but that I can recognise it as something that will pass, and ignore it until it is no longer there. Because I've seen the pattern enough times to know that that sort of fear always goes away. And I have some shiny new skills to show for it.

This advice is applicable to so much more than Go or programming. For any endeavor, you have to be willing to fail for a while before ever gaining any kind of competence. Writing, sports, picking up women.

There's a little more to this than just being willing to fail. With Go, I don't even know that I'm failing as I play.

Writing is a good example. Go feels like trying to learn to write with the instruction to place letters on the board one at a time, but I don't know any words or why there are these funny squiggly marks and dots next to some of the words, or why some of the letters are bigger than others, or why there are blank spaces between some of the groups of groups of letters.

And that isn't the point either, because I'm not afraid of learning to write by myself. The fear is a social fear.

The biggest problem I've found is that the feedback loop is too long if you play a whole game and then go back to discuss it (especially when playing in person without a record). It often works better to stop the game and talk about why a sequence of moves was particularly good or bad, and play out a few variations immediately. Depending on how well the student is doing, either continue the game from the original position or the best alternative.

That social fear is what keeps plenty of people from just jumping in and failing at plenty of other things. Of course there's more to success at a thing than willingness to fail. But out of fear, people sabotage the endeavor before it's begun. Fear that people won't like what I've made. Fear that what I've written isn't good enough. Fear that I won't be good enough compared to my perception of their expectations. Fear of making a mistake. Fear of rejection.

That's the thing. Of course some people won't like what you've done. Of course you'll make a mistake. Of course you'll get rejected. Being able to accept that will at least put you past the self-defeating stage. You miss 100% of the shots you don't take.

I've experienced more tension playing go than any game other than poker.

I think initially go particularly punishes people who are very strong at thinking their way through combinatorial problems.

If you're good at strategy in general, and used to quickly and deliberately thinking through decision trees, I think it's only natural to build up a sense of confidence of being able to be good at nearly any game.

But go utterly defeats this, as the initial few moves of the game involve very vague intuition. Certainly experienced players have a library of standard opening formations and patterns they rely on, but more importantly, they've played the countless games necessary to intuitively apply that knowledge effectively. There's a saying that if you learn the standard patterns (joseki) you become 2 stones weaker. It's true. It takes many games to integrate that knowledge into actual play effectively. Trying to learn it before you're ready is counterproductive.

The other proverb is great advice: lose your first 100 games as fast as possible. I'm still on this path, and have gone from understanding little to playing even against 14k players in the local club.

Keep courage, the anxiety may not go away, but the enjoyment of the game will rise to meet it.

When I was eight years old, my father taught me to play Go. My father is terrible at Go, probably the worst player I've ever played. So, I entered college thinking I was fairly good.

I got completely, embarrassingly, destroyed in my first game. And the next two.

After that, I read a book on it, probably the same one the author read (Learn to Play Go by Janice Kim) and didn't really follow at all. Still getting destroyed.

Then I did what the author didn't do: I played at least twenty games of Go a day. I played a Windows program called Igowin, I played a really bad Palm app while standing in lines or on the bus, I played GNU Go on my laptop.

At first I lost every time, then about half the time, then almost never, at which point I played a human again.

And got destroyed, because I had learned tricks to beat the computer, rather than actual good Go techniques.

I guess my point is that this taught me two things: when learning something, there is no substitute for doing it twenty times a day; and that no matter how much you think you have Go figured out, you probably don't.

I've been playing go on and off for over ten years now and _heartily_ recommend anyone remotely interested in learning to sign up for kgs (http://www.gokgs.com/) and download the java client. There is a fantastic community there, including a few beginner rooms. Often, people will stay after and review the game with you, which is basically a free lesson. People are consistently polite and respectful on kgs, which I can't say of other go networks (igs, for example). Additionally, it's really easy to get a quick match whenever you want with someone appropriately ranked, and it figures out your rank as you play people.

If you're going to play in person, try to play smaller boards, like 9x9, so you can get a sense of tactics and local patterns. Mark_h said it already in this thread, but I'll repeat: lose your first fifty (or more) games as quickly as possible. When you're first getting started, the goal is just to get a handle on how play progresses, not to win. Getting comfortable losing makes it easier to learn.

Web resources I can recommend: http://playgo.to/iwtg/en/ and http://senseis.xmp.net/

Please feel free to find me on kgs, username jrothstein. I'd love to help share my love of the game with other hackers. Also feel free to contact me (info in profile) with any questions about the game or for a review (I'll try my best).

I'm blackthorn on kgs... we should play some time. I'm about 5k, though I haven't played in a while.

People are polite on KGS because politeness and topicality are enforced by the moderators.

Many of the techinical people I know play go, or atleast know of it. I'm not sure if there is a direct correlation, but generally the stronger they are at Go, the better of a programmer and learner they are in general.

If anyone here is interested in learning or playing, I am about 2d on KGS (nickBlake) and would be willing to play or teach anyone.

If anybody in San Francisco wants to play in person, I'm usually up for a game after work. I've only been playing seriously for a few months (current rating 25.5k), but I'll gladly play anybody, including teaching what I can to beginners. Send me an email if you're interested in playing.

This post makes me want to start a blog, just to answer some of these questions.

Go is really hard to teach. I try to avoid it whenever possible. This isn't trying to be a wiseass; i find this actually has some great effects. I've found teaching go has to be a student-driven process to be successful; the onus is on the student to generate questions, and, to go another step, the student must seek & have the experiences that will generate the questions. This almost always means playing go for yourself.

This implies that the student has already conquered a lot of (very real) internal fears; the fear of performing incompetently, the fear of being thought dumb. The fear of failure, basically. An interesting thing to run across on a blog for entrepreneurs.

The saddest part about that article was that so many of those sentences are "I" sentences; you write "I feel" this, "i think" that, "I am [x], i won't be [y], i will never be [z]". Even the comments about your friend are defined in terms of yourself. The problem is that probably no one else is as concerned about your performance as you are.

This is something that could be incredibly liberating from the right perspective. Who is judging you here? Not your friend, certainly! It's not a reflection on your character unless you make it one -- and you don't have to make it one.

Your conclusions at the end are so close! "The courage to play incompetently is a cure for fear of incompetence" this is true, but it's almost a tautology. Perhaps a more useful formulation is this: "The courage to play incompetently is a cure for incompetence." Maybe it is the only cure.

Relax, enjoy, & Dare To Suck -- your performance at something you've never done is not and will not be used as a measurement of your worth as a person. That gift of "curiousity & fearlessness" is not a personality trait; people get nervous when they feel like they're being judged. You're not -- at least, not by anyone but yourself.

See also: http://senseis.xmp.net/?FearOfLosing and the go proverb "lose your first 100 games as quickly as you can"

relax & good luck

-seigenblues

P.S. I will happily answer any go questions. Send me a link to your game & questions, (perhaps on eidogo.com), and i will try to answer to the best of my ability (4d kgs).

Thanks for your thoughtful comments. I just want to point out that ""The courage to play incompetently is a cure for fear of incompetence" is true, and obviously a tautology. However, it is taken out of context. The full line reads:

Competence is not a cure for the fear of incompetence. The courage to play incompetently is a cure for fear of incompetence. Put in contrast, I think it has more value than standing alone.

quite true. I'm sorry for taking you out of context. It is indeed more valuable that way.

I would ask the next question; which is the more important problem we are trying to "cure" -- the lack of competence or the lack of courage? One can lead to the other, but not vice-versa.

Lack of courage. I really have no incentive to play Go competently except or the fear of being mocked for not playing this game that my peers venerate as being the domain of deep thinkers with fully functioning right brains.

So if I lose the fear, I'm good. It's then irrelevant whether I ever become competent, except possibly that my playing at a certain level of skill would make it interesting for my friends to play me.

But otherwise, it's just the fear that concerns me.

Raganwald,

You are being so honest about your ego that there seems little doubt it is starting to dissolve and you will regain the lightness of touch and mood that betokens huge creative power.

I see this often in Brazilian Jiu-Jitsu, where beginners not only have no idea what they're doing, but they don't even realize how little they know. And being incompetent in BJJ means you're going to spend many minutes with someone literally on top of you. Some people are so overwhelmed by this experience that they don't come back.

But, my experience has been the people who don't come back are those that thought they would do well - the ones that figured "I'm a tough guy, I can hang." When they do just as one would expect a beginner to do, their ego is hurt so much they just don't want to go through it again. I think that for some, being a "tough guy" is part of their identity, and being confronted with the reality is so jarring they'd rather not even learn.

The ones who stick around are the ones who aren't upset about losing.

There's actually a Go proverb (there are lots of these, and they're quite good!) along the lines of "lose your first 50 games as quickly as possible".

Having said that, I struggled to do this (even against a computer I hated to lose, which impeded my progress for ages).

I don't know much about the Go game, but this is such a beautifully written article about the general anxiety one encounters while learning new things. I have this exact feeling in my effort to learn programming (despite working as a programmer for 10 years now). Every time I encounter a new concept (possibly very basic), I criticize myself for working as a programmer without knowing such things.

One interesting tidbit about Go is that it's one of a very few games where humans are still better than the computer.

http://en.wikipedia.org/wiki/Computer_Go

I've learned many abstract strategy games. I play some fairly well, others not so much so.

When I approach a new game, I assume it will take many hundreds (or thousands!) of plays to become truly good at it. I assume that I will play poorly for my first N games -- so, it's best to get them out of the way as quickly as possible. I try to think in terms of, "did I learn something from that loss?" -- if the answer is yes, it was still a worthwhile experience.

I think this is a good lesson for learning anything new. I think my first N programs were extremely poorly written. There was so much that I didn't know when I was starting out with programming! Now, 10+ years later I feel like I have a pretty good handle on things, but there are still areas of programming that I would like to know better.

I expect I'd see the same pattern learning anything new. If I were interested in drawing I'd expect the first hundred or more pictures to be lousy. If I wanted to become a writer, I'd expect the first hundred stories to be boring or full of misspellings and poor grammar.

For me, the question is: Does this knowledge make it easier or more difficult to start a new pursuit?

You are so great!

Sorry, I meant to give the opposite impression. Everyone sucks when starting out with something new. Maybe acknowledging that can help with "first move" anxiety.

Either I didn't express that very well, or it's just not that insightful.

With go there are three very different stages of learning (has something to do with the fact that there are different sized boards). First you learn the rules - how to place the pieces, when is a group alive etc. Then you learn tactics: how are pieces 2-3 spaces apart linked, common patterns, counting etc. Finally you can make use of the 19x19 board and learn strategy - like how it's better to control corners rather then edges and edges rather then the center.

At each step you will get totaled by people who are ahead of you because each step gives you a completely new way of seeing the game. And you can't cheat and "learn ahead": strategy without tactics is useless.

Also, if you hate losing, for the love of god never play online. Sometimes it looks like half the population of Korea is online playing the other half, and when they classify themselves as "beginners" it usually means they can beat you while watching TV.

Rather than merely resolving to overcome the fear, I'd make myself imagine what's supposed to be so bad about the thing I'm afraid of (not being top-.001% gifted? being destroyed by an expert? that both occur in view of others?), and reassure myself that the feared outcome is not actually so terrible.

However, making a public resolution is also a good approach.

If you don't like what I said, tell me why.

The word "platitude" springs to mind, as well as the phrase "it's easier said than done." Everyone knows the basic processes of cognitive behavioral therapy; it's applying them when you're actually feeling anxiety that's hard.

>Everyone knows the basic processes of cognitive behavioral therapy

(cough) Um, yeah, of course. Of course we do. (cough) [darts off to wikipedia]

I agree that it's not easily done. If thinking about the fear abstractly doesn't work (it probably won't), actually allow yourself to imagine/feel the negatives you believe as concretely as possible. I don't see how you can approach knowing what's really going on in your own mind if you're unwilling to observe without trying to suppress and control.

I can accept that reasoning away the specific concerns/self-beliefs isn't always possible. Some things really are hard and require practice, even after obviously buggy internal thoughts (usually exaggeration of the negative consequences you risk by trying) are refuted.

I think the quote should be "lose your first 50 games as quickly and thoughtfully as possible."

Ideally, you are losing to progressively better opponents. This gives you the opportunity to ask for help and learn better technique as you are losing. Online Go is great because you can record your games and get others to critique them.

I love Go. I love that there are implied threats, feints, and sacrifices. I love that seemingly random moves can be so critical in later stages. Go is a game of experience. You really have to enjoy the process of unraveling the mystery of the game and not worry so much about winning.

My older brother taught me to play when I was 8 years old (a long time ago!) - now I can give him a 9 stone handicap :-)

Once, in the late 70s, I got to play the womens world champion when she was in San Diego - awesome experience.

Go is a beautiful game, but not for everyone.

When I was an undergraduate living in a dorm, I and several friends all learned Go together. Nobody knew what they were doing, and so no one was particularly nervous. Even keeping score was an exercise in passing around examples from the instruction book.

Not the most efficient way to learn, but among the least stressful -- if you're patient enough with one another.

Reply on news.ycombinator.com