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

High Anxiety (raganwald on Go and learning new things)

github.com

21 pointsColinWright22 comments

According to a recent theory[1], there are three affect systems in the human brain: - Threat/protection (Safety seeking) - Drive/Excitement (incentive, resource focussed) - Soothing/Comforting (Affiliative)

Anxiety problems are caused by the first one. The theory holds that problems with the first system are best tackled using the third one (whereas I imagine that most HN readers are more likely to assume that everything should be done using the second one).

[1]http://apt.rcpsych.org/content/15/3/199.full

NB I can't claim that I have any evidence for this theory, as I have only know about it for a short time. Anxiety has ruined my life so far; this is merely the first thing that looks like it might really help.

thank you for sharing this.

So I recently decided to learn Go myself, after decades of hearing about it from fellow techies. It's incredibly difficult. I read, downloaded apps, and then signed up on DGS as 30kyu (the lowest rank) and have lost every game except for one I won on timeout. I stink. Playing Go is a paradigm shift. Knowing (and being good at) other games is almost irrelevant at my level. This is a little like coming to pure functional programming after decades of kicking ass in the imperative programming world. And like that, smarts and experience will give me a leg up. But later.

So many times, as a smart and capable person, I can see the light at the end of the tunnel in a new endeavor. Even when I don't really see it I trust it's there if I can grasp even a little. But with some things you just feel lost, and that's very difficult for smart people.

This is so different from learning a game like backgammon. One summer in my mid-teens, a friend and I played many times daily over a summer. I've rarely played since, but the strategy is easy enough for a smart person that I can hold my own against anyone except world class players. That's a bit to one extreme, but it's closer to the average experience of a smart developer.

Finding things that I just can't grok after a few minutes of thinking about them (and there are many) has made me more accepting of my limits, and raised my awareness that some things take a mental shift and/or really are complex.

It's incredibly difficult. I read, downloaded apps

What apps?

http://senseis.xmp.net/?GoPlayingPrograms

I found the link raganwald gave. Of those, I found SmartGo Kifu to be excellent, in that it has a tutorial that walks you through many basic concepts in Go. That I have not become proficient is not their fault.

I don't think it's that strange. Ultimately you have to play in order to read ahead in the game. People who think they can sit down and learn to read ahead will struggle. But if you start actually playing, a lot of the basic reading ahead will become second nature and then it's on to harder/deeper reading. The same is true of chess. Someone given the rules and a set can't hope to get it in any meaningful way though it is theoretically possible. For Go, if you can find people better than you to play against and keep it up for a year you'll look back and realize that even your initial worries were just a misunderstanding. DGS is not a good place to start. You need to play in volume. Bulk practice.

Go is worse in this respect than chess.

In chess, in principle, you can sit down and think through every possible combination in a simple position and figure out tactical traps. After doing this a few times you learn what sort of things to look for.

In go, even for simple problems, there are too many possibilities to think through all possibilities. Until you can start to recognize key points, nobody can even explain to you how to think about problems.

I'll take your advice regarding volume/bulk. It seems sound to me. I took up chess a bit more than a year ago, and tactics practice and non-stop playing have done more for my game than any reading.

sorry, by 'read' I meant calculate. Calculation is the chess term, reading is the Go equivalent.

If you're losing that badly, you need a bigger handicap. A lot bigger.

Yes, that would help. There are other things that would also help. Just now I'm in a holding pattern on learning Go. When I come back to it I'll go over basics again, play on small boards, and play more real time games. Anything else you'd recommend for me?

The biggest tip is that for a while you'll be playing high handicap games. Don't try to attack. Just try to pick what to defend. If while your attacker carves out a chunk you manage to defend a bigger chunk, you'll win. As you improve in defense, your handicap comes down. After a while you'll find that the skills you're learning in defense also apply to attack.

And the very first rule to learn is that if you create a solid wall and leave someone with little space, their stones will either die or be left alive on little territory. Which happens doesn't matter to you, and you don't have to figure out the details - just keep on limiting them and things will work out. If ever it comes to direct conflict, you'll give them opportunities - that is the game they need you to play to have a chance. So don't give them the opportunity.

Oh, and always start a wall at least 1 space away from them to limit their tactical chances.

Go is a fascinating game but it requires a lot of effort to learn to play well.

I've found that after a hard day of programming the last thing I usually want to do with my free time is solve what is essentially another set of logic problems. I'm much happier if I can do something that lets me turn off and rest that part of my brain for a while.

Many won't get/grok the following statement and I suspect those who don't will have a hard time with go.

You cant learn go, you need to feel go.

It's not a coincidence that some eastern philosophy Tao, Zen, others have similar concepts at their core. You can't learn enlightenment. You shed the unenlighted until you are. Approach go with an empty mind. Think not of moves. Look for flows, patterns, feelings. The OA had some of this, "her position looked strong, mine weak an isolated". Seeing that is go. Attempting to formalize why you feel that is not go, at least not until you are a master.

Many hackers are optimizes. They want to learn all the parameters and rationally arrive at the optimal solution. Go is highly resistant to this.

Go is indeed like functional programming in that it's a paradigm shift. It will teach you a new way of thinking, a new way to see and understand the world. I highly recommend playing go. Just don't try to learn it.

I’ve heard this many times, but it’s orthogonal to the point of the essay. If we take the words “know” and “learn” and replace them with “feel,” I can write the same essay. I felt anxiety because I didn’t have any feeling for the game. Every time I did something, I felt punished for having the wrong feeling, but I didn’t have any feeling, much less a wrong feeling.

You can say there are no wrong feelings, yet the game has a winner and a loser and is very much scored like there are right and wrong feelings. My problem then was what it is now: I have no feeling for the game, and there is no feedback loop to help me feel what is going on.

Which is not to say that you aren’t wrong about how to approach the game, but my essay isn’t really about not being able to become a good player, it’s about not being able to do anything because I had no “feeling” for what to do.

I've been there. It's easy for me to ignore the "right and wrong, win or lose" mentality of certain situations, but there are some times when I feel disengaged or nauseous about being in a situation where I am not expected to perform well.

The solution, IMO, is to remember that you aren't the feelings in your head. There was a time you would have had a feedback loop for the game regardless of winning or losing, and you've lost touch with it due to age. Your remedy, which is the correct one, is to persevere beyond your banal mental fortress until a new feedback loop is created.

By the way, I'm a huge fan of your work. Good job!

Previous discussion: http://news.ycombinator.com/item?id=904765

That would have made a nice tweet.

To the down voters: perhaps you are unaware that https://twitter.com/raganwald/status/222061086348820482 resulted in two submissions to HN just now. Almost any raganwald post will result in this, as you and raganwald well know. That this is old and had previously appeared here would have been welcome news. That said, I am glad to spend my karma saying true but unpopular things. Probably the best use for karma that exists.

Indeed, and I searched for and found that, but discussion there is closed, it was nearly three years ago, and HN has a lot of new users since then who may choose to share their experiences.

I have now edited this title to be more descriptive, and to match that previous submission.

Added in edit: I also see that this has been flagged, so I guess it won't last long, which is a shame. I think there's a lot to learn from this article, but you need to find the lessons for yourself, so many might not see them.

If you want to teach someone go, give them an insane handicap to start with. I am only a mediocre player, but I'll play on a 9x9 board and hand out a 9 stone handicap. More than half the time I'll win, then I'll give them a couple of tips. (Unless you're connecting your stones in a solid wall, play 1 space away from me to avoid tactical tricks. Also don't try to kill my stones, just try to build a wall that I can't get through.) Then they usually win.

People often resist accepting the handicap. So I point out that the game has a more complex strategy than anything they have ever seen. The handicap is a necessary part of the game. If they can't accept this fact, then they will never learn or enjoy the game.

Once they are beating me easily with a 9 stone handicap, we start dropping the handicap down one stone per game. Somewhere about 5-6 stones they will hit a wall again. Then I can give more tips and we go to the "adjust the handicap after 3 victories in a row". Over time they start seeing more of what is going on.

Once they get down to a 3 stone handicap on a 9x9 board, then I give them a 9 stone handicap on a 13x13 board. Now they can feel that they are really playing, and start to see how varied it is.

If you're trying to teach yourself, I recommend taking a game like GNU go and doing it yourself. Crank up the handicap to 9 on a 9x9 board, and try to win. Slowly drop the handicap. If you can get someone to give you tips, do. You will improve and can see it in the handicaps you need.

In chess there's usually a reluctance for strong players to play weak players, unless it's for mentoring. That Go encourages people of different skills to play a somewhat even game through handicaps seems very sociable and healthy to me.

Reply on news.ycombinator.com