Super Fun Game!

He Who Laughs came over on Monday, and we played this super-fun game. He was white.

18 - I made an empty net, just to make sure he couldn't kill the four black stones.
44 - It hurt me to make that shape, and I said as much. I've come a long way since the last time I played against He Who Laughs.
45 - Playing in the elephant's eye. Asking move. So, which direction do you want? I chose the wall side.
57 - This was a dirty trick. He even said something like "oh, well at least I can still make eyes," and I totally fell for it.
72 - Thought about connecting the wall of five. Took a gamble on a tenuki. That was a mistake.
84 - Creating a bamboo joint would have been a stronger move, to keep white from creeping up.
104 - Really passive, but I wanted a strong defense of the corner.
130 - He called that move sneaky, but clearly it wasn't sneaky enough or he wouldn't have seen what I was up to.
131 - Took the threat out of my sneaky move by putting my two black stones in atari.
138 - I really wanted to find a way to win back that corner, but I didn't know how yet.
143 - We had it check the score, which, if I remember correctly, was how He Who Laughs became aware that that group of mine was kill-able, since the AI was counting them all as prisoners.
144 - Big mistake. Didn't look at the board carefully or long enough to see what He Who Laughs was up to.
145 - With that group gone, the whole lower left corner is looking pretty hopeless. At this point, I'm fairly certain that mistake cost me the game, and I'm looking to reinforce what I have.
166 - Still going after that corner. There has to be a way, right?!
168 - Turns out, yes. Yes, there is a way.
175 - And I won by 1/2 point. Huzzah!

“Don't start from the assumption that you are weak, or you will never learn to be strong.”

This is the continuation of a discussion I alluded to in this post.

“Don't start from the assumption that you are weak, or you will never learn to be strong.”

The hypothesis was stated in the context of altering one's play style as a reaction to the perceived skill difference between you and your opponent, which we already concluded was a bad idea. However, the statement itself still stands.

My first objection was in defense of accurate self-awareness. If I don't know my strength (or, rather, weakness), in the first place, how will I know when I'm getting stronger, or even what I need to do to improve? I need to be able to see where I am to know where I'm going. The first step to attaining competence is understanding one's incompetence.

I suspect the hypothesis is built on the assumption that someone who believes they are weak will take fewer risks than someone who believes they are strong. That's a critical falsehood. A person who believes they are weak but isn't afraid of failing will take plenty of risks, while the person who believes they are strong but is afraid of failing will take fewer risks. Fear of failing might be amplified by holding “strong” or “smart” as part of one's identity. As soon as failure threatens identity, a mistake is no longer just a mistake, it's an identity crisis. Higher stakes usually mean taking fewer risks.

There is also a question of which approach a player takes to their learning. In The Art of Learning, the author discusses the critical difference between entity and incremental theories of intelligence. In short, an entity theorist considers ability inherent and ingrained - “I am naturally good at x” - while an incremental theorist considers ability an achievable product of hard work - “I am good at x because I worked hard.” In this case, it is the difference between believing I am doomed to be weak forever and believing that weakness is simply the starting point on a continuum, and that getting stronger is totally possible if I just put in the work. This is exactly why the first thing my favorite professor in art collage did was rip apart the concept of “talent.”

There's also something in here about the perception of risk. There's a close enough link between experience and skill that I'm going to group them together for the sake of this argument. With experience comes the ability to read out further and more accurately. A weak player might think something is risky (and the best move), based on what little they can see, and tackle that instead of the thing a more experienced person would know is actually more important and/or in more danger. I wouldn't say the weak player chose poorly in that case, because they chose based on what information they had and the best of their ability to parse it. What more can you ask? If we aren't talking about a game of Go, you could ask them to ask for more helping parsing the situation. In that case, I'm strongly against “don't assume you're weak...” because a weak person trying to get stronger is more likely to ask for more help than a person who already believes they're strong.

I think a better motto for me might be something as simple as “do what you'd do if you weren't afraid.” My problem isn't that my opponent is stronger (the handicap is supposed to balance that anyway), it's internal; I'm letting fearful self-limiting beliefs narrow the range of options I consider viable. The trick is being able to see clearly enough through the fear, or dismiss it (easier said than done) enough to know what the best move would be if I weren't afraid. Fear also tends to make people think locally, short-term, instead of globally, long-term, which may be an advantage in crisis, but not in a Go game.

This is one of the things I adore about Go: whatever issue you're tackling in life, it's gonna come out on the board.

Fun with editors.

I wrote a commentary in the editor, couldn't get it to show up here, and then accidentally lost it. The gist was that I'd probably learn more from commentating on games I lose rather than win. Oh well, maybe next time.

Tomorrow I'm getting on a plane to Minnesota for the holidays. I anticipate many AI games in transit.

Also: last night I had a dream that I was playing Go with green and gold mini gift bows instead of stones. Happy Holidays!

Conclusion.

There's how the game against He Who Tells Stories came out. The mid-game partial analysis is in this post.

I got pretty slaughtered, but that's okay.

Things I did poorly:

- The aforementioned mistake on the right side, where I tried to prevent eyes on the edge side rather than corral either from the middle or below his invasion.

- The aforementioned weird 1-space extension uselessness on the 4th line, top center left. I think those moves were when He Who Tells Stories gained the advantage to creep down the left side and push out into the center. I also lost some groups in that affair that I just didn't see coming.

- I think it was a reading error that allowed the big white section in the bottom center to survive and ultimately expand. Unfortunately I don't have a complete game record of this one. Next time.

Things I did well:

- Top left corner. I'd read it out accurately, and knew I was safe several moves before he was certain.

- I had more awareness about who had sente and what advantage that provided.

- We both had the same most critical move in mind when we started from where we left off. Unfortunately, it was his turn.

A silly thing I do that I should not do:

- I get pretty slap-happy in the end game and play silly moves which could benefit me, if and only if my opponent makes a fairly obvious mistake. Playing the silly move at all is sufficient to draw his attention to whatever I see, so I rarely benefit from the attempt. I suppose this is harmless enough if he has to answer and I can keep sente, but if not, then it's just wasteful.

Setting a low bar.

Been feeling pretty down lately, almost didn't post today at all. Then I thought, "okay, I'll aim for a really achievable goal, and at least ship that." The plan was:

1. Play and record just one game I wouldn't be ashamed to post in its entirety to the internet.

2. Learn what I need to learn to get that game from SmartGo Kifu on the iPad to the blog. 

That's it. No commentary. No philosophical ruminations. Just push through the tech barrier.

As it turns out, "a game I wouldn't be ashamed to post in its entirety to the internet" is synonymous with winning when I'm already feeling sad and vulnerable.

The bad news: I'm not winning games tonight.
The good news: SmartGo Kifu ranks me at a 2-stone handicap, so the games I'm losing are at least different than the 4-stone handicap I'm stuck on in regular SmartGo Player.
...
Whilst heating myself some comforting hot chocolate on the stovetop, I continued losing until it gave me 3 stones instead. I'll take it.
...
Shit yes. I even floundered through enough hoops to get this thing visible. It's not pretty, but it exists. Enjoy.

Mid-Game Review.

He Who Tells Stories and I have been playing on Wednesday nights for a couple months. Games have been getting longer and longer, so we're trying something new. Rather than let the game drag on past the point when most of my brain cells have gone to bed, we paused, took a picture, and plan to finish this Wednesday. Here's the current state of things:

I'm black, he's white. Even match.

I remember doing a lot of reactive damage-control, so at this point, before reviewing it, I feel pretty dejected. Maybe fresh eyes will help. My memory is hazy, so these notes are more about what I see now than what was actually going through my head at the time.

White 2 - Seems like a conservative opening to me.

Black 5 - Making an enclosure.

Black 9 - I wonder if I could have gotten away with something more aggressive there. 8 was a pretty big jump.

White 10 - White's claiming the fuck out of the top left corner. And why not? I'm not putting pressure on. 

Black 11 - Okay, that's enough of that. Time to start shit before white digs in. White is playing all low, aside from the 4-4. Then again, so is black. We both want edges, and we're both willing to play low to make it easier. Are neither of us confident enough to play high?

Black 13 - A choice between up or down. The corner is worth more and his friend is further away.

White 14 - White has to respond. Defending the corner and making it harder for 13 to get to the edge.

Black 15 - Not sure what the hell that's about. Trying to split 2 apart from 8? I bet I wasn't clear on my intent, then, either. Am I trying to win the corner? Am I trying to corral white into the corner and build a wall for myself? Who knows.

White 16 - Support for 8, trying to make a base. Both white's here and black's at move 11 and 13 are 1-space extensions on the third line rather than the standard 2-space extension, which I just realized in grabbing the link for "base." Huh.

Ugh. This whole exchange is really bad for black. I don't know now what I was trying to do then, and I bet I didn't really know then either. 

Black 17 - Possibly trying to connect under and make it harder for 16 to get to the edge? Too many stones too loosely connected, with no clear intent. Danger! Danger!

White 18 - And the cutting. 

Black 19 - And the extending. Attempting to protect the weaker stone.

White 20 - And the blocking.

Black 23 - I have to take something out to save the 3 black stones. "White, who's it gonna to be?"

White 24 - He chooses well.

White 26 - Securing the corner.

Black 27 - I think I thought it was important to take out that white stone so I could make an eye with those three.

Black 29 - Feeling panic-y about the groups on the right. Should have been more worried about the two stones on the left.

Black 31 - Well, I can't save it. Maybe I can extend help to the straggler and corral white into the corner at the same time. Maybe?

White 32 - So, white gets the corner and black gets potential for an eye and influence toward the center.

Black 33 - Killing 22, like I wanted to. Probably should have kept sente instead.

I'm going to stop reconstructing this in my pretty Photoshop file now because it's too time-consuming and not clear enough. I'll upgrade to an app that'll let me save and export games after I'm done studying what happened in this one.

Other thoughts on this game:

If we scored it right now, I'd be winning, but I still have a ton of weak points and groups that aren't connected. He might also be able plunk down an invasion out in the center there and live, eating up a good amount of territory. Which weaknesses are the most critical?

White sprawl happened along the right edge because of a mistake where I tried to prevent eyes on the inside rather than corralling from the outside. 

The 1-space extensions along the fourth line at the top were pretty useless. Not sure what I was trying to do there. The first one was peep-ing myself, and the second one was even weaker than that.

Lost some groups in the upper left quadrant because I just plain didn't see the danger until it was too late. Pay more attention?

I need to play more intentionally. Know exactly what I'm trying to do, even if I'm not skilled enough to pull it off every time. It might help to list off all the things the last stone is doing, so I can answer with all the things I want my next stone to do. 

I still don't know how I feel about this game. Maybe that's okay. At least now I know what I'm going to do next. Stay tuned for the thrilling conclusion. 

Go Seigen v. Hosai Fujisawa.

Monday night I went to another one of Nick's classes at the Seattle Go Center. I foolishly didn't bring anything to take notes, so I ended up with illegible scrawl in borrowed pen on the back of a Home Depot receipt. It was a nice little collision of my past and current lives.

The lesson was in memory of Go Seigen, who just died on Nov. 30, and was "considered by many players to have been the greatest Go player in the 20th century." Nick talked us through this famous game which basically ruined Hosai Fujisawa's career in March 1953.

That was the first pro game I'd seen. The most striking feature, for me, was how much air was still on the board at the point when Nick stopped playing it out because everything was more or less settled. It makes sense. A skilled player can look at a loose configuration, see the various ways to connect it, and what each of those moves is worth, approximately. Armed with that much information, yeah, a lot of those local moves just aren't as important as something larger elsewhere. As an unskilled player, I bet I play relatively densely and slowly because I just can't see enough to know whether my local fights are resolved and I error on the side of caution.

List of hits:
Every stone must serve multiple purposes. Practice identifying all the things a stone does will help me understand the board more holistically and be more efficient. 

Balance slow and fast. Playing a slow move gives you the power to play a fast move in that area next. If you play a fast move first, you'll need to come back and play a slow move to balance it. This, of course, relies on the ability to properly distinguish slow from fast moves. 

Safety first. I don't fully grasp all the implications yet, but his example was something about setting up the potential for shape or life before getting crazy with riskier moves.

Anchor for invasion. I tend to think of invasions as air-drops behind enemy lines. The stone Nick was talking about (move 30) was out in the middle, near enough several openings that it could be used to support an invasion. Presumably, the invasion could run out and connect, cutting the opponent's wall, or the anchor could simply apply pressure from both sides. Neat.

It might be time soon for me to start cracking open joseki. Even reading about them makes my head spin, which will be fun and beneficial to push through.

New terms:
Avalache - a joseki with several variations that follows from this starting sequence. See also: large avalanche and small avalanche.

Shoulder Hit - a "sly" diagonal reduction move on the fourth line that is difficult to catch or cut off. 

Armpit Hit - the opposite of a shoulder hit. Usually a bad move. 

Tesuji - "a clever play, the best play in a local position, a skillful move, a special tactic," Appears to be a broad complimentary label for about a million different things, many of which have their own special names. 

Miai - two options wherein if one player takes one, the other player can take the other, amounting to an approximately even trade-off.

Jubango - a 10-game challenge match.

How I started playing with humans.

I tend to form strong associations between games and the people I learn them from. Unfortunately, if my relationship with that person goes sour, so can my love of the game. So when I was first learning Go, I avoided playing with humans because I wanted to solidify how I felt about the game itself before forming any strong association between it and third parties.

About 6 weeks after I started playing, I posted this piece of art on Facebook:

(More information about the art is available here.)

People responded. Unbeknownst to me, I had a variety of friends who wanted to play, used to, but didn't have anyone to play with anymore. He Who Tells Stories was so excited that he brought his board and stones along to a beach bonfire. Suppressing the social anxiety, and extremely aware that I had no idea what I was doing, he and I hunkered down in the sand and played my first 19x19 game. Despite not knowing what we were doing, either, the rest of our bonfire party gathered round, creating a tableau of people more interested in the game a few feet away than the slowly dying bonfire we were all nominally there to attend. One of our friends had played against an AI for a while, oh, about 4 years ago, which made him the local expert and commentator. Others struggled to remember what they'd picked up from Hikaru no Go. We welcomed tasty interruptions in the form of charred brats and s'mores from the spectators. As the sun set over the ocean and the night grew steadily darker, our game carried on under flashlights and headlamps.

Sometime in the mid-game, I tried to stand up and stumbled, kicking sand and scattering the stones. Nooooo! Only then did we snap out of it and return to the fire at hand. Even so, I don't think I could have asked for a more idyllic setting or better company for my first real (half) game.

Just play your game.

Played a really great "lesson game" with He Who Gives Nothing Away yesterday, which is to say, he got to talk through copious alternative moves, branching possibilities, and high level strategy, while I was somewhat more transparent about my thought process, both voluntarily and prompted.

13x13, 7 stone handicap. He won.

I'm chewing on an interesting phenomenon, which is that I adjust my style of play depending on the perceived skill of my opponent. You know that tendency to give up too fast? That, amplified. Which also explains why I was utterly helpless against Nick Sibicky in the ad-hoc simul he staged while waiting for his real class to begin. Mentally, I'd given up before we'd even started.

It wasn't nearly that bad with He Who Gives Nothing Away, but I did hurt myself with the extremely passive opening moves I made before I gained that awareness. Things got better when I started asking "what would I do against He Who Tells Stories?" Not sure yet how much of it is the perceived skill difference or how much of it is simply the degree to which my opponent is an intimidating motherfucker. The former, ergo the latter? I need more data.

I was also thrown off by the enormous handicap, which initially made me think more in terms of how I'd play on a 9x9 with a large handicap than what I really ought to be doing on a 13x13. I'm still frustrated by 9x9 AI games, so choosing to play in the style of the board size I'm frustrated by and stuck on rather than the board size I'm rather enjoying and making swift progress on was simply a terrible idea.

It occurred to me several hours later that I also don't have a mental warm-up. It usually takes a bit to settle in, which does me no favors if my weak opening moves put me at a disadvantage for the rest of the game. In an even match, neither of us is likely to start shit right away. Claim your corners, sip some tea, we'll get to the fighting when we get there. In a large handicap game, white needs to come in teeth and claws, and I need to be ready for that.

One of the things I adore about Go is the philosophical implications of the abstract principles, as applied to other walks of life.

One such example, which I'm debating the merits of offline, is the idea that you shouldn't assume you are weak, otherwise you will never learn to be strong. I think that idea is flawed, if not outright false, but interesting, none the less. Perhaps more on that later.

Another is Go as a meditation on jealousy and privilege. This came up in a discussion about how one must continuously fight back the frustrating truth that one can never hold the entire board. My answer to that, which maybe comes more easily for me that for him, is to let go of the desire to control the whole board. If jealousy is defined as a fear of having what you consider yours taken away, the game becomes really interesting really fast (not that it wasn't already), especially in the context of a base assumption that the whole board ought to be yours in the first place. That's just not going to go well, in life or in game.

One of my moves was a small instance of that. I wanted to defend two different areas. Instead of picking one and making a strong move, I made a weaker move in the middle, which only sort-of defended both. I didn't want to let him have either, and in so doing, left weaknesses in both.

Unreasonable late game invasions also came up in the discussion about whether or not to alter one's playing style based on the perceived skill of one's opponent. On the one hand, yeah, the stronger player might be able to get away with it. On the other hand, if the stronger player is losing by enough that an otherwise unreasonable invasion is what they need to win, then they actually fucked up a long time ago, either when the weaker player got the edge or because they waited too long to invade.

We concluded that it's better to just play your game as well as you can at all times; avoid altering it based on the perceived skill of your opponent. For the weaker player, getting intimidated will surely only weaken your game. They might well be able to pull off some ballsy shit, but they will definitely get away with it if you don't even try to stop them. For the stronger player, the move you're contemplating when you start with "I wonder if I can get away with this?" is almost certainly weaker than "what's the best move?" Taking advantage of a less skilled player's incompetence simply doesn't up your game as fast.

New terms:
Yose - a move that approaches fairly stable territory, enlarging my territory while reducing my opponent's

Oyose - a large yose. Also refers to the early endgame, when those moves are the most common. The board is fairly well settled, but plays are still worth a decent amount of points.

Victory!

I won! I won! I finally won on a large board! 23.5 - 31. 19x19, even match, against He Who Tells Stories.

We started by claiming the top and bottom halves of the board. I invaded his upper right 3-3 point and lived with 3 eyes. He had a long wall containing my group. A bit later, I extended a different group into a counter-wall along the outside of his, suffocating it. Corner for me.

Later, he got a little base in along the bottom with only one eye and struggled to extend it out into the middle. I stopped trying to kill it too quickly. There was a point in the game where I could have prevented the second eye shape, but didn't, because I was more concerned about connecting my groups to contain his sprawl, rather than trying to kill his whole group.

Bottom left and right corners were basically life and death puzzles. He died on the left and lived on the right (just barely).

Upper right irritated me to no end. He had a lot of space, well defended, and there just wasn't room for me to invade or reduce effectively. That didn't keep me from trying (a lot), which was either valiant or stupid, depending on your outlook.

I think he was surprised that I'd won when we counted out at the end. For one thing, he had several dead groups still on the board, which makes for a muddy quick-read. For another, it was not that long ago that he was still offering me a handicap. Yay! Progress!

Good game, He Who Tells Stories. :)