19x19 FOR BEGINNERS SERIES
THINGS I WISH I KNEW WHEN I WAS 25 KYU
PART 5: THE STAGES OF THE GAME
[A quick disclaimer for any readers who are SDK and up: this series was written by a DDK for TPKs, with the goal of making the process of “losing their first 100 games quickly” less confusing and more productive. I say this because I’m aware of how many elements and strategies I’m leaving out given such a broad topic. My reviews of the three games included here will omit a lot of detail, and feature my subjective conclusions - which often work figuratively rather than literally.
If you would like to fill in those gaps and expand on the topic, I welcome your input. However, I humbly ask you to take the time to craft your explanation with the 25 kyu player in mind. Chances are your contribution may include elements that are obvious to you, but far from obvious to the beginning player. If you can break your explanation into parts that TPKs can understand, you will be doing both them and me a great favor.]
LONG ARTICLE ALERT: At 43 pages - this is my longest article yet. It also has more illustrations, and will encourage active participation from the reader. Since it’s split into 3 sections, I encourage you to read each one separately, and study the attached game until you understand how the stones reflect the narrative of the story, before diving into the next one.
Welcome back! In my last 4 articles, I’ve been zooming in on individual elements of the game, and taking them apart to help you see how they work. Now, it’s time to zoom out and take a look at the underlying mechanics which shape the whole game - from start to finish. As with so many aspects of Go the connecting theme is efficiency. Because both players have been given an impossible task and not enough resources, the one who can forge the most efficient path through this battlefield will come out slightly ahead. It’s important to remember that, in this game, you do not need to capture, dominate, or destroy your opponent - you just need to be one teensy little half point ahead.
I’m going to circle back to something I said in my introduction. In order to make sense of Go - you will need to learn a new language about how the stones interact in space and time - absorb a body of experience for what things are possible and what things are impossible - and begin to see how the various strategies of Go emerge procedurally from those interactions.
If you’ve read Part 3: Playing a Balanced Opening, you’ve seen one example of how the physical properties of that space (points enclosed by N stones arranged in 2 walls > 3 walls > 4 walls) shape the efficiency priorities (corners > sides > middle). In this article, I will try to show you how the focus behind those priorities shifts and changes to inform the other stages of the game.
Unfortunately, the game of Go is complex enough that focusing on efficiency alone will not be enough to guarantee you a win. If you play someone much stronger than you, they’ll have the experience to play in ways that don’t seem very efficient(1), but are still quite adequate to win(2). However - while you are playing against beginners - until you understand the priorities which inform each stage of the game, you will continue falling behind players who do.
(1: creating eyeless sticks)
(2: denying your eyespace and capturing your groups)
FOCUS ON SHIFTING GAME PRIORITIES
The stages of the game only have 3 names, - Opening, Midgame, and Endgame - so you often read about them as 3 different entries in a tutorial or wiki page. When it comes to playing in real life, I feel it’s more realistic to approach the Midgame in two fairly distinct phases. As such, I’m going to discuss the priorities of the game in four stages:
1: The Opening:
- Claim freely available potential in order of corners > sides > middle
- Create large extensions / prevent large extensions
- Create small extensions, prevent small extensions
As soon as all the edge potential on the board is claimed, we move onto the next set of priorities
2: Early Midgame:
- Expand your claimed potential away from the edge and into the middle
- Push into the middle from both sides of one extension to create a moyo
- Prevent your opponent from creating a moyo
- Claim any unsecured edge or corner potential
As soon as half of the remaining potential is claimed, we switch to
3: Late Midgame:
- Expand your frameworks into whatever unclaimed areas remain
- Push tendrils from living groups into poorly secured section of your opponent’s potential to reduce eyespace
- Close up the poorly secured sections of your own frameworks
- Connect any weak groups at risk of getting stranded without making two eyes
As soon as all of the major battles on the board are resolved, we switch to the final set of priorities
- Close up the remaining gaps in your borders in such a way that you retain sente for as long as possible (even move sequences) rather than exit in gote (odd move sequences) and pass sente to your opponent
- Play on any points near your borders that could be used to make eye space by your opponent
- Play moves which force your opponent to fill in their eyespace
- Resolve any ko threats
- Fill in any empty dame points (points which do not change the score either way) to make score counting easier
- Both players pass and accept the results
(Side-note: I’m not going to discuss seki (mutual life) at this time - I feel like that’s more of a 10-15kyu topic.)
DIRECTION OF PLAY
If you’re reading the above list, wondering “How the heck am I supposed to do all those things?” - let me discuss another tool for moving around the board: using direction of play to your advantage.
Direction of play happens whether players are aware of it or not - it is just a neutral description for how the walls between opponents grow over time.
Think back to Inego and Wesley dueling at the cliftop ruins. While the local exchange of thrusts and parries moves the fighters around, both use their situational awareness to angle that battle to their advantage - by pushing someone into a tight corner, or forcing them to jump backwards over a broken bridge.
Rather than passively following the contact fight around the board, I’m hoping to show you how you can utilize direction of play by
- Being aware of where the next most valuable areas are located
- Angling your attacks so that the walls between groups expand in those directions.
3 EXAMPLE GAMES: BEGINNER, INTERMEDIATE, ADVANCED
I’ve talked before about how difficult it is for beginners to wrap their heads around Go. When I was just starting out, I often got the unpleasant feeling that I was trying to play one game, and my opponent was playing a completely different one. When I review my old games - or watch the games of 20-25kyu players OGS - I see folks falling into the same traps, and losing in ways that feel awfully familiar. What is it that keeps beginners from seeing the game the same way an experienced player would?
One common theme is that beginners often start the game with an appetite for destruction - they begin contact fighting early, and either continue one long fight around the board, or follow wherever their opponent tenuki’s and keep it going there (a.k.a. Puppy Go). I keep wondering, what makes this play-style so prevalent?
There are a few potential culprits. One might be what I call the “Chess Mindset” that many bring to the game. Since Go provides a way to gain an advantage by capturing, it becomes a familiar starting point - which eclipses all other game priorities. Another may be the fact that many come to the 19x19 game by way of the 9x9, and are trying to apply the tactics they learned there - i.e. grab the middle and contact fight to the end. It may even result from beginners trying to follow examples from the SDK and dan games we observe, which often begin by chasing weak groups, and focus on advanced tactics like denying eye-space or capturing large dragons.
In this article, I will try to show you the different ways in which beginner, intermediate, and advanced players utilize the stages of the game to their advantage. Hopefully this will help you see how this game responds to your play-style, and understand how the shifting efficiency priorities inform each stage of a game, regardless of what style the players throw at it.
GAME 1: BEGINNER
Here’s how we’re going to do this - please open up this game in a different window and follow along as I talk you through it: https://online-go.com/game/28334440. For this example, I’ve chosen a game of my own - one of my first handful of games on OGS. I was playing Black as a 24 kyu against a 22 kyu opponent. At the time, I knew very little about making living shapes, or contact fighting, or the shifting game priorities I am discussing now. I had a very vague awareness that;
- Approaching many corners in the Opening was better than fighting for a single corner
- The importance of corners > sides > middle in getting an early lead.
I want to show you that - even with this minimal understanding - Black was able to gain an advantage by using the stages of the game and direction of play.
Before I begin, I want to acknowledge that this was a 2-stone handicap game for Black. I had created this ranked challenge not knowing that the OGS algorithm would automatically assign a handicap based on our kyu difference - I was simply ignorant of the proper buttons to push. I’ve always wondered if White’s ferocity had to do with starting the game down by 2 corners - but the fact that White tenuki’ed only once makes me think it might have gone this way even with a more standard opening. Anyway, let’s get down to it. (Please click through the following moves in your other window and then come back!)
White starts off attacking Black’s upper right corner. Black tries to defend but gets 2 stones captured. At Move 22, Black feels like they’ve secured that corner enough to approach the upper left (D16), but White is determined to dominate the right (R12). Direction of play for both is shown in RED.
White presses the attack, exploiting Black’s lack of liberties with atari threats. Black zooms out from the corner fight and tenuki’s to H3 (enclose a corner). When White plays T15, Black realizes they’ve lost the corner, and tries to rescue the remaining 5 stones by running them to the loose framework along the top (create a large extension).
White pursues the local fight and traps 5 Black stones at the top (for now), creating a wall on the L line. When White secures the corner with R18, Black tenuki’s to approach the last unclaimed corner with Q4. Now - at this point - if White was following Opening priorities they could still approach a corner (C14, C6, O3, R3) or prevent a large extension (D10, M4, Q7, etc). Instead, White doggedly attacks the upper middle border between groups - like a sword fighter who is always running towards their opponent swinging wildly.
Black fights back - capturing 5 White stones in exchange for losing 3 of their own. Seeing that White is determined to drive this fight into the middle, Black tenuki’s at P10. If White was paying attention to Opening priorities, they could still approach a corner or prevent an extension (marked with O’s). However, Black has already moved on to Early Midgame priorities and is expanding their potential away from the edge into the middle. With the wall on the L line reaching 9 lines away from the board Black feels like they have secured the shaded area well enough to count it as “territory” (i.e. it would be difficult for White to invade and make a living group).
While White is focused on attacking and capturing stones, Black is happy to continue the contact fight away from the upper edge - thanks to direction of play, Black is getting a nice solid wall and a decent corner moyo.
Black and White resolve the lower left corner. Black manages to keep their stones from getting captured and creates a nice wall on the H line - hoping to exploit the existing large extension and develop another moyo along the bottom.
When White pushes down with K6, Black cuts with K7. Black and White both have weak groups trying to live inside the other’s potential, splitting the direction of play in two. White is running into a classic problem with contact fighting in tight spaces - making your stones heavy - or over-concentrated. After closing off a few of their own liberties due to atari threats, White has 14 stones clumped together. At times like these, the only way to survive is to capture the cutting stones - which White could still have done if it had played Move 177 at K8. Unfortunately, White lost track of their liberties, and Black captured those stones with Q5 for move 178.
Black loses 6 more stones negotiating the right side. This is cold comfort for White who takes a moment to catch their breath, hits the OGS score estimator, and gets this discouraging picture. White may be ahead in captures (24 to 21) but Black is leading in territory. Even without the 11 points the OGS auto-score is denying in the upper left in case of a 3-3 invasion, Black is still ~80 points ahead. Usually we would be in the Endgame stage, and start closinging up our last border gaps.
Instead, White tries one final invasion to turn the game around. Unfortunately, one cut later (G15), they find themselves over-concentrated and lose 5 stones. White keeps attacking with their remaining 3 stones, but gets caught on the wrong side of their own Ko threat (moves 245-250). This time Black is more methodical, and makes one long group (D15-G18) to keep their cutting stones from getting captured. Here’s the board right before Black played H17 to capture White’s heavy group with move 276. White resigns.
If this type of game feels familiar to you, I am hoping this replay helped you see it in a different light. It’s not that either side was better at contact fighting - it’s that White was trying to play one game - and Black was playing a slightly different one. (Side note: just so you don’t think I’m patting myself on the back, here’s a different game I lost a few weeks earlier, where I made very similar mistakes: https://online-go.com/game/4952799.)
GAME 2: INTERMEDIATE
Now that you’ve seen what happens when players don’t pay attention to the stages of the game, where can we find a counter-example: a game where players focus on the priorities at each stage rather than fighting / capturing each other? For this I relied on my old study companion - the GnuGo AI. I set its strength to 5 on a 1-10 scale (equivalent to ~12kyu), generated a few AI vs AI games, then picked one that demonstrated these concepts.
My goal here is to show you a different play-style; something commonly referred to as a Peaceful Game where - aside from a few stranded stones - everyone’s frameworks end up alive, and there are no large captures. This happens because each side takes the path of least resistance and tries to grab those next-most-valuable parts of the board which are freely available, rather than spending their time and effort fighting their opponent over mutually disputed areas. I played a lot of these games climbing up from 20kyu to 15kyu, and I feel it’s a useful exercise to help beginners see where the valuable areas are at each stage of the game.
Picture this if you will - you’re playing a dart throwing game at a carnival fun fair, and you’re competing with another player to see who can win the most prizes. The darts are being thrown at a board arranged like this, with big yellow balloons on the outside, medium orange balloons in the middle, and small red balloons in the center. Popping yellow balloons gives you the most expensive Tier 1 prizes, orange and red give you smaller Tier 2 and 3 prizes, respectively. At the beginning, both players will start popping yellow balloons because they’re big, easy to hit, and provide the best prizes. But once there are only a handful of yellows left scattered across the board, the concentrated orange balloons and Tier 2 prizes will start to look like an inviting target from an effort vs reward perspective.
Obviously, this metaphor is drastically over-simplified. The valuable territory at each stage will not simply progress inward in a predictable target-like fashion (it will in some cases, but not others!) As we go through the game, I’ll show you how these bots choose the path of least resistance rather than taking options which require more risk and effort. Think of this game as an abstraction - a perfect sphere sliding along a frictionless plane. However if you hold a grain of salt firmly between your teeth, and keep that dart game metaphor in mind, I hope this example will help you to see the order in which parts of the board shift in value over the stages of the game. As before, please open this link in another window and follow along: https://online-go.com/game/28396652
MOVES 1-28: The Opening
Black and White play a fairly loose opening - claiming edge potential quickly. I want to call your attention to a few things:
Move 5: Black plays C15 to enclose their own corner rather than approach one of White’s corners. I’ve been encouraging players to use the sente in move 5 to approach rather than enclose, but it’s good to remember that enclosing your own corner is a perfectly valid option - especially when Black can expand that corner into a large extension like they do here.
Moves 6-12: When White approaches at P16, Black pincers at M16. White pushes up the P line, using direction of play, and Black responds by creating a small extension along the upper right. White counter-pincers with J16, putting the upper middle in dispute - meaning both will have to fight for it at some point - but not right now.
It’s important to note that neither side shows any interest in contact fighting at this stage - instead both tenuki at the earliest opportunity to grab unclaimed potential while it lasts. When White counter-pincered for Move 12, Black could have extended from M16, and started a fight in the middle, distracting both away from Opening priorities. Instead, Black approached another corner with Move 13 to prevent a large extension and try to reduce the corner.
I’ve color coded the board to show you the type of analysis each player might be doing as they decide whether to continue the Opening, or start Early Midgame. The areas shaded Black and White are places where both feel fairly secure about their potential. The upper middle is shaded grey because it’s in dispute, and will take more effort to resolve than either side wants to commit to right now. There are still a few yellow edge sections, but - as you can see - the orange areas that represent Early Midgame priorities are starting to look pretty inviting.
MOVES 29-34: Are we there yet?
The transition between Opening and Early Midgame often happens in fits and starts as players dither about which priorities they want to pursue.
Move 29: Black makes a small extension to resolve the right side
Move 30: White extends into the middle to protect the R9 stone
Move 31: Black extends into the middle just in case White has started Early Midgame
Move 32: White resumes Opening priorities and reduces Black’s left side potential with a Monkey Jump
Move 33: Black encloses the final bit of corner potential in the upper right
Move 34: White extends to protect their portion of the top middle and keep the M16 stone cut off from the right corner group.
MOVES 35-44: Early Midgame: dueling moyos
Black plays G15 to expand their corner potential into the middle. This puts White in a bit of a pickle. If White wanted to prevent Black from turning their left side extension into a massive moyo, they would need to play some sort of invasion or reduction around E12. However, not only would this start a complex fight, but White might get cut off from above or below, and forced to run to reinforcements. Instead, White follows the path of least resistance and allows Black to expand their corner/side moyo in exchange for a center moyo of their own.
White uses direction of play to push into the center from their lower left group, confident that Black will respond by connecting their lower left group to the top left corner. It seems like Black is getting the better end of that deal until White pushes into the center from the other side of their extension. Rather than fighting to invade or reduce Black’s moyo, White utilizes the efficiencies available in Early Midgame to grab a moyo of their own to balance out Black’s advantage.
MOVES 45-52: Early Midgame: a (relatively) peaceful trade
Black plays J14 to expand their moyo further into the middle, forcing White to defend their upper middle framework with K15. Black doesn’t want to risk trying to save the M16 stone, because White could cut it off from any direction, so Black turns the other way and expands their moyo into White’s center potential. Instead of defending, White gives Black two sente moves and sacrifices their G10 stone in exchange for resolving the upper middle group. White signals the end of this exchange by protecting their lower moyo at J8 - “that’s far enough, thank you.”
When we turn the analysis filter back on, you can see that the orange areas have shrunk significantly. While there’s still a decent oval of unclaimed space in the middle, both sides must attend to the areas at the edges and between groups. Also, where before both have managed to avoid touching stones, now the fights are running out of elbow room. This is another sign that it’s time to transition to the next stage.
MOVES 53-150: Late Midgame
The pace of the game shifts to reflect the new priorities. Rather than grabbing large, freely available parts of the board, each side now plays contact moves in smaller areas to resolve the boundaries in between. In this particular game, both sides tenuki constantly rather than focusing on resolving any one area. This is because their estimate of the values of different parts of the board keep changing, and they’re more focused on “popping the next balloon” than finishing the current fight. They are leaving “unfinished sentence fragments” all over the board and risking the kind of trades we saw in moves 45-52: a minor reduction in one area in exchange for an advantage in another.
Human players (between 15-18 kyu) might be more likely to play in complete sentences - following a given sente attack all the way through to settling their stones in that area. Let’s zoom in on one such exchange:
102: White threatens Black’s cut point at R11
103: Black extends and makes a tentative connection with the R13 stone
104: White hanes to threaten the R11 cut point and invade the side
105: Black has to play S12 to protect they key point or White might cut at R11
106: White pushes towards the edge, reducing Black’s potential
107: Black has to connect
108: White makes their last forcing move
109: Black has to connect
110: White plays a gote move to connect, passing sente to Black.
Think of using forcing moves as having a short contact fight with a plan. If you can predict where your opponent’s responses are going to be at each point, it’s sort of like getting those moves “for free” - they accomplish their goal, and you don’t have to worry about them getting counter-attacked - as long as you can settle those stones effectively. If you can angle that fight into a direction of play that’s good for your local group, then you’re accomplishing multiple goals at the same time.
MOVES 151-196: Endgame
Here’s a map of Black’s options for Move 151. As you can see, all the medium orange sections are gone, and the borders between groups are mostly resolved. Only the little red balloons remain to be popped, as each side closes up their gaps and creates secure walls.
The trick at this stage is for each side to keep sente for as long as they can. This means using even-numbered sequences (where one side retains sente) rather than odd-numbered ones (where sente passes to the other player). We see this in moves 151-155 for instance:
Because Black’s two moves extend from already-secure stones, White has to respond locally, allowing Black to one last move at K14. (note: White didn’t feel it was threatening enough, and played a sente move elsewhere) I’d hoped to have more examples to show you, but in this game, sente kept getting passed back and forth every couple moves.
If you go to the game, and open up your right sidebar and Enable AI review - you can see how Katago analyzes this game. You’ll see that Black and White were fairly even until Moves 91-95 when Black took the lead by reducing White’s lower left. White takes the lead back with Moves 134-136 when they secure their lower center moyo. By the time Endgame rolls around - the win percentage swings back and forth (+/-95% etc) - the score so close that just one stone tips the game one way or the other. However, halfway through Endgame (Move 179) Katago calls it for White.
In the end, this is what we get:
BLACK: 92 (90 territory / 2 captures / 0 komi)
WHITE: 93.5 (82 territory / 5 captures / 6.5 komi)
Black takes slightly more than half of the board, but not enough to counter White’s advantage in captures and komi. The good news is - despite all those big land grabs, and reversals along the way - each side came remarkably close to breaking even. On the other hand - neither was willing to take the risks to gain a decisive advantage precisely because they were so focused on their pursuit of efficiency.
Personally, I always find an odd sort of beauty in Peaceful Games. Seeing the resulting shapes as something procedurally created by hundreds of separate decisions evokes two types of moss, battling for sunlight over the same rock. Hopefully, exploring this play style will help you climb from 25 to 17 kyu, as these are the ranges you’re likely to find players who haven’t stumbled onto these concepts for themselves yet.
[A quick side note: if you’ve read this far without taking a break, I encourage you to do so now. I’d rather you digest this new info and come back refreshed, than push on if you’re already feeling stuffed. The next section is quite dense - it’s your call.]
GAME 3: ADVANCED
When I was starting to learn about Go, watching high level games felt confusing and impenetrable. Eventually, as I began to understand some of the language, strategies, and priorities of the game, it transformed into a fascinating spectator sport which helped me see the game in a different way, and provided me with new goals to strive for, even if they felt far ahead of my level. Hopefully, this walk-through will help you unlock this complex and enjoyable pastime for yourself.
Playing Black is Go legend Cho Chikun
- first player to hold Kisei, Meijin, and Honinbo titles simultaneously
- inventor of the tea stealing tetsuji
Playing White is go badass Satoru Kobayashi
- a maverick known for creative and unusual moves
- once got suspended after injuring his opponent in a post-game brandy-drinking accident.
I stumbled onto this particular game via this fantastic YouTube playlist of Japanese NHK Cup matches with English translation (make sure to hit the CC button when watching).
If you have 90 minutes, I encourage you to watch the whole game, as your commentator is a Honinbo in his own right O Meien. Compared to his, my analysis of the game will be very shallow. My goal is simply to show you how the stages of the game adapt to a new strategy - each player managing increased risk by juggling multiple weak groups at the same time.
Think of it this way - in the game of Go - each player must create risk for their opponent while minimizing risk for themselves. In a peaceful game, that risk is generated by grabbing parts of the board more efficiently than your opponent. In an advanced game, a different type of risk is created by threatening the opponent’s eye space, and forcing their group to run or capture stones to live. The problem is, any moves the aggressor plays to deny that eye-space can then become weak groups themselves, and must survive when their opponent turns the tables.
Much like our first example, the contact fighting in this game starts early, and drives direction of play into the middle. But this time, each player maintains a delicate balance between attacking their opponent and utilizing the efficiency available on the board. Let’s take a look. (Please open this now and follow along: https://online-go.com/game/28425701 )
MOVES 1-8: The Opening fuseki
Black starts the game with this unusual variation of the Chinese opening. After Black and White claim all the Hoshi points with moves 1-4,
Moves 5-6: Black approaches the lower left, but doesn’t settle that stone - opening themselves up for a pincer from the right,
Moves 7-8: instead Black plays another sente move at F17, confident that White will respond by enclosing that corner aiming at a large extension on the left,
Move 9: rather than settling either approach stone, Black cashes in their sente advantage by attempting to create a large extension on the right with Q10.
In the language of Go, this is a clear message that Black wants to begin fighting early. It’s commonly referred to as “inviting your opponent into your house” and it forces White into an uncomfortable position. If White doesn’t invade the top or bottom frameworks with their next move, Black will secure one, and gain an early advantage. If White does invade, it’s likely to get pincered and create a weak group that must run or fight to survive. Rather than give White options this early in the game, Black is using their sente advantage to force White into a narrow range of daunting choices.
MOVES 10-20: staging the fight
Move 10: White invades Black’s lower framework
Move 11: Black encloses the right corner
Move 12: White encloses the left corner
Move 13: Black extends their weak group into the middle
Move 14: White knows it is pincered, and has to run up the middle
Move 15: Black makes a risky jump, trying to get ahead of White’s center group and put pressure on the White’s left side at the same time
Move 16: White expands their lower left framework, protecting the large side extension
Move 17: Black attack White’s weak group, trying to cut off their escape
Move 18: White extends
Move 19: Black cuts at K7 forcing White’s center group to turn right
Move 20: White counter-cuts at J7, stranding Black’s 4 stones in the lower left.
MOVES 21-27: the calm before the storm
Each side settles their weak groups to give them a fighting chance now that they are expeditionary forces, trapped in enemy territory. I’ve given them nicknames so we can keep track of them going forward, and outlined them in red because they are all in danger. If they don’t run to reinforcements, they could get surrounded faster than they could make two eyes. As you can see by the arrows, the various directions they need to run put the groups in danger on two fronts, meaning they must balance escaping and attacking at the same time.
When I watched these types of games as a beginner, I often wondered - are we still in the Opening? Or have we moved to Midgame - since contact fighting in the middle seems like such a Midgame element? The answer is a bit tricky - think of it as a game of Red Light / Green Light. When either player’s groups are in immediate danger, that risk overrides all other game priorities - Red Light! However, as soon as any tendril from one weak group outruns their two surrounding borders, the signal for that player turns Green and they can resume Opening priorities - securing or invading any big yellow edge sections still available.
Alpha company tries to negotiate the corner, but gets 2 stones captured. Alpha turns upwards with N7, causing Bravo company to respond with L9, making sure they can escape up if needed. Alpha is confident that they can still run up or invade Black’s framework to the right. This feels like enough of a Green Light for White to play F9 - expanding their left corner extension, and attempting to trap Delta.
Move 41: Delta splits White at E8 - turning the F9 stone into a reinforcement for Charlie
Move 42: Charlie expands upwards with J10
Move 43: Delta settles with E9
Move 44: Alpha threatens Bravo with N9
Move 45: Bravo extends with L11
Move 46: Alpha keeps chasing with N11 - but at this moment, Bravo can still get out. This means the light is Green enough for Black to go back to Opening priorities
Move 47: Black plays P7 to extend the corner framework halfway up the right side
Moves 48-52: Charlie chases Bravo up to K15
Move 53: Bravo turns right with N13, putting pressure on Alpha
Move 54: Charlie reinforces the connection with J14
Moves 55-58: Black uses their right side group to put pressure on Alpha with forcing moves on the 10 line - Alpha keeps their line secure
Move 59-60: Black peeps at H13, forcing Charlie to connect, and giving both Delta and the weak stone at F17 something to run to
Move 61: Black reinforces Bravo with M14
Move 62: Alpha is in trouble - it knows that Black can reduce the eye-space at the bottom with forcing moves - it keeps running with P12 (Red light!)
Move 63: Black has Green light and plays R12 to expand their right side extension.
If you think about it - this has been Black’s plan all along since Move 9. Black knew that White would have to create a weak group and run, and they could use that direction of play to slowly expand their right side extension from one corner to the next. Even if White lives, Black is getting exactly what they wanted.
White is banking on a similar gamble on the left and the top - it has three groups aiming at huge chunks of corner and side potential, with Bravo, Delta and Black’s two weak stones at the top to oppose them. Even this late into the game, there are still major Opening priority areas to be claimed - but with Alpha in so much danger, both sides will have to resolve a lot of Red light/Green light on the right before they can start that conversation on the left.
This is where we get into the really complex do-or-die contact fighting, so I’m going to hit the fast forward button a bit. Please feel free to lean on O Meien for the expert analysis at this point. Let’s just say this type of fighting requires extensive high level reading. Anyway, White fights back, chasing Bravo to the top of the board, thinking it’s made at least one eye with P12, then pressing that advantage all the way up to O17.
Unfortunately, White left too many cut points, and Bravo captured 4 stones to make life and stranded another 3-4 invasion stones at the top. White takes two Black stones in trade, giving Alpha at least one eye. White has a Red light on Charlie, who tried to rescue the stranded top stones, but is now facing off against a new group at the top - let’s call them Echo company. If Black can resolve those stones in dispute at the top, then Echo can link up with Bravo and try to create a large extension at the top.
Confident they can still save Charlie, White is tempted by the sudden opportunity in the lower right corner. Now you or I might look at that and see a monkey jump at most. Anything else seems like suicide right? Our stones would get over-concentrated and captured in such a tight space! But to a 9 dan professional - that’s an invadable corner. Even with Charlie in danger, grabbing corner territory away from Black takes priority. Let me show you how White pulls off this neat trick.
White is able to attack Black’s corner group from two directions - invading along the right edge and cutting Black’s extension between P7 and Q4 at the same time. I’ve numbered White’s moves to try to give you a sense of the flow. White’s first 4 moves sketch out a quick framework, and offer Black a choice - connect at P5 and allow White to live, or attack White and risk getting cut. Black attacks with S7, trying to cut White off and create eye-space, so White cuts with P5. While both are busy making forcing moves, White slyly links their stones together into this elegant shape, enclosing the corner. Black throws the S2 stone in there to keep the area in dispute, but White is confident that they can resolve it later. Where before Black could count on the whole corner, now they’re left with two groups - one still fighting to make life (let’s call them Golf and Foxtrot. White’s new corner group can be Hotel.)
Even though it feels like we should be in Late Midgame, there are still huge yellow Opening priority areas of the board to be resolved. Black wants to develop a large extension at the top, and keep White from creating a large extension on the left. Confident they can save their two new weak groups, Black turns their attention to the fleeting efficiencies still left on the board.
Moves 126-128: Bravo loses a piece of their tail when Alpha and Charlie connect
Moves 129-132: Delta pushes left, trying to prevent an extension, White responds at C11
Moves 133-135: Echo invades the top left corner
Moves 136-139: White threatens Delta and both secure
Moves 140-143: White resolves the top left corner
Moves 144-147: White negotiates the center boundary with Delta
Moves 148-153: Alpha pushes into the top right corner as far it can and Bravo counters.
And with that, we transition from our complex and extended Opening all the way to Late Midgame. Even though there are unsettled groups on the board, and Delta has Green light (it’s not in immediate danger), now is the crucial time to secure that eye-space because it is the only unclaimed area left on the board. And with the Charlie+Alpha group now alive, White can push up on Echo or down on Delta. Facing 3 different priorities, Black has to walk a very fine line.
MOVES 154-179: Late Midgame
Moves 154-158: Black pushes towards the left edge on the 9 and 12 lines, White secures
Moves 159-160: Black peeps at White’s J11 cut point, instead of responding locally, White plays F13 to trap Black’s D12 stone
Moves 161-167: Black cuts Charlie and Alpha apart with J11/H10 - now Delta and Bravo have joined up. Charlie has Red Light again, and has to run to the left or get pincered.
Moves 168-171: Both sides secure weak stones
Moves 172-176: Black moves in from both sides but Charlie makes a tentative connection with G15
Moves 177-179: Echo and Charlie resolve the top - Echo connects to Delta and Bravo.
MOVES 180-190: the final bits of Late Midgame
Moves 180-185: White pushes as far as they can into the upper right corner
Moves 186-188: Black pushes into White’s left middle
Moves 189-190: Foxtrot and Alpha secure eye-space.
And with that, the last major disputes on the board have been resolved. I’ve left out the analysis layer so you can practice with your own eyes. Take a look - do everyone’s groups look alive? Where are the last weak points? What’s left on the board to be resolved? What are the areas where sente or gote could make the difference between securing another eye or letting your opponent push one stone further? If you’re asking these types of questions, you know you’ve entered the final stage.
MOVES 190-245: Endgame - setting up the final Ko battle
Watching this for the first time, I remember the play being very tense. Both players knew it was a close game, but counting all the areas in dispute challenged even the acumen of O Meien, who looked uncomfortable being put on the spot when asked to guess who was ahead. At one point, during moves 244-263 he said, “Whoever wins this Ko wins the game.”
Katago had a slightly different assessment, that highlights some interesting things. According to the AI analysis, Black was leading by ~20 points until White invaded the lower right corner (Moves 110-128). After that, the game went back and forth until Move 196 (White secures Hotel with S3), which is when Katago called the game for White. From 196 to the end, Black was able to reduce that lead from ~2 points to only 0.5 points - but was never able to change the score enough to alter the win percentage.
White set up the final ko fight with Move 244 - trying to attack Black’s 5 stones in the upper left, while fully aware that Black could still counterattack with forcing moves to threaten the lower right corner. White pressed the attack with the cut at F18 (Move 250), but Black was able to defend effectively. White backed down and secured their lower right corner with the capture at T1, and Black ended the game by capturing at B19. White won the game by half a point.
PUTTING IT ALL TOGETHER
I hope that reviewing these examples has helped you see the game in a different light. As I said in my introduction - in Go - the one who wins isn’t the bravest warrior, or the most skilled fighter - it’s the shrewd tactician who is able to evaluate the most layers of information, and correctly pick the most important priority from the bewildering range of strategic options at each moment. Understanding the stages of the game is crucial because it allows you to put each exchange of stones in context, frames the reasons why you’re pursuing your current strategy, and offers you clues on where the next best moves might be based on the flow of the game.
I am hoping that you gained a better understanding of how much time and effort you need in order to perceive and analyze the different layers of information on the board. I’m not talking about reading out a local fight in analyze mode, to see if you can capture a single stone. I’m talking about global questions like
- Do I have sente right now or do I need to play a gote move?
- Are my stones settled?
- What are the most efficient actions I can take at this stage of the game?
- Should I keep playing in the same area or should I tenuki? (hint: you should tenuki!)
- What areas of the board are the most important at this stage?
- Can I use forcing moves and direction of play to expand my groups in that direction?
- Are my weak groups safe? Can they make life or connect if attacked?
- Does my opponent have any weak groups that I can threaten?
- Once this exchange wraps up and my opponent gets sente, where are they likely to play next?
- Once this stage of the game is over, what are the next biggest areas of the board we will be competing for and how can I get a head start at this stage?
Right now, keeping track of all those layers may seem like hard work, but once you start using this process to examine the board before every move, you will begin to see how it frames your actions at every step, and gives you clear goals to work towards. With practice, you’ll find it easier to keep those layers in mind, and catch yourself tracking them naturally because they are helping you make sense of your game. In time, the information in those layers will get richer as your analytical skills grow with experience, and more flexible as you begin to notice different opportunities available on the board depending on which options you and your opponent explore. Rather than feeling lost or overwhelmed - always running from one fight to the next - you’ll find your place in the story with a clear understanding of where you are, how you got there, and what you need to do next. Good luck and see you next time!
PREVIOUS ARTICLES IN THIS SERIES: