19x19 FOR BEGINNERS SERIES: Part 6: Shape - Playing with Blocks in the Sandbox

Previous articles in the 19x19 FOR BEGINNERS series:

Introduction: Making sense of Go

Part 1: Sente and Gote

Part 2: Settling your Stones

Part 3: Playing a Balanced Opening

Part 4: Joseki Basics

Part 5: The Stages of the Game

[LONG ARTICLE ALERT: This article is 46 pages long and quite dense. Rather than trying to power through all of it, I would recommend you read and absorb the first section where I talk about the various types of shapes, and then tackle the Sample Game in a different session. Enjoy!]

Welcome back! Before we begin this article, let’s take a moment and review how far we’ve come. If you’ve gotten this far, I’m hoping that you’ve absorbed each of the previous articles, and started practicing these perceptual abilities and strategic skills in your own games. By applying these abstractions to actual play, you’ve begun to understand how these elements interlock to form a coherent whole.

Your awareness of the stages of the game allows you to see where the next big moves are on the board. Your understanding of efficiency and direction of play allows you to compare these moves against each other, and pick the one that will be the most effective in pursuing those goals. Your understanding of sente and gote allows you to make moves that actualize your plans with confidence, and knowing how to settle your stones helps you to secure the potential you’ve grabbed.

There’s one final tool you’ll need to add to your arsenal before you can play the way an experienced player would - the concept of shape - or how the stones fit together to maximize efficiency. That’s not to say that once you learn this you’ll understand Go completely - far from it! You could study this game for the next 50 years and still discover new things! However, once you understand how the concepts of shape permeate and inform the game, you will have all the key ingredients in your hand. All of the other complexities you encounter will be some outgrowth of the possibilities that emerge procedurally from the interplay of these fundamental elements.

In Chess, all the possible moves are bound to distinct pieces. The Rook can move in its row or column, the Bishop moves diagonally, and so on. In Go, there is only one piece - and any non-suicide move appears possible. However, Go still has a well-defined set of moves by which any new stones either

  • make connections to existing frameworks or
  • Commit to a set of strategies to forge those connections later.

Once you learn to see the game as a sequence of these moves, it reduces the overall complexity because it helps you narrow your focus to a handful of effective options rather than trying to invent something new every time. It also ensures that whatever structures you build will eventually make shape - meaning your stones have the best chance to stay connected or secure potential as efficiently as possible.

As with previous chapters, I will call upon all the concepts we’ve discussed so far, and show you how a focus on individual moves and the shapes they create informs these elements of the game

  • Sente and gote
  • Settling your stones
  • Maximizing efficiency through the various stages of the game
  • Using direction of play strategically

I want to show you that - if you pay attention to the priorities for each stage - you will begin to see the gaps in your opponent’s shape as targets of various types of attacks, and be able to break those down to individual moves that you can plan and execute effectively.

As with so many things in Go, the concepts that underlie shape are emergent properties of the game - all trying to answer the same questions:

  • Can these stones stay connected when challenged?
  • Can these stones make eye-shape once connected?

If you step back and consider the possibilities, there are only a few ways in which any two stones on the board can guarantee a secure connection:

The Iron Pillar (Nobi): one stone group extends into a two stone group

The Diagonal (Cosumi): one space diagonal extension.
No matter what White does, they cannot cut these stones apart with one move. White can play a forcing move at A or B to challenge that connection, but Black needs only a single move to join them together.

The resulting shape - a Filled Triangle - is bad for White because Black radiates 6 liberties and White only has 2.

However, once you add a single space between two stones, you create opportunities for your opponent to cut them apart - and take the risk that the both groups might be mutually separated.

Let’s take a look at some examples with a one space jump.

Black has two stones marked with Os. White tries to challenge the cutting point, and Black connects their stones into one group. Simple right?


Now look what happens when White decides to cut:
1: White separates the 2 Black stones
2: Black protects with a Tiger’s Mouth, putting the attacking stone in atari
3: White Nobis down. Black has 2 cut points at A and B and they can only close one of them
4: When Black connects at A
5: White cuts at B (or vice versa). All the lone stones Nobi safely away, resulting in four mutually separated groups. Black’s two groups cannot reconnect unless they capture some of the White stones in between.

Beginners who approach the game with an emphasis on fighting and capturing often cut without considering the risks to their cutting stones. More experienced players understand that not all cuts are created equal, and weigh that decision based on the chances their stones face once they’re stranded in enemy territory. Let’s take the above exchange and put it in context.

Here Black approaches the lower left with F3, and White plays the pincer at H4. Black tries to escape up the middle, but White blocks (Move 1) and then cuts (Move 5). White’s two stones marked 5 and 7 are now in danger - Black might try to separate them from below - but they are free to travel and expand across the board in many directions. They also have reinforcing stones nearby they can reach for by playing the knight’s move at K6. Because Black’s lower left group has to make life in a tight space, they will probably try to make a base with H2 / H3 rather than fighting to separate White with J5, so this cut is a fairly good gamble for White.

Now take a look at this clunky and improbable exchange. When Black extends their lower right group up the side, White blocks at Move 1 then cuts with Move 5 – only to realize that their cutting stones have nowhere to go. Worse yet, they’ve lost one of their precious liberties by running into the edge. If Black plays a follow up at S9, White cannot escape.

As such, deciding whether it’s time to cut - or play a forcing move to challenge an opponent’s cutting point - depends on whether you’re willing to take that particular risk. In my previous articles, I’ve compared Go to various gambling metaphors but now we get to the real heart of the matter. Every time you cut, you are betting that you can keep your cutting stones alive better than your opponent can capture them. If your stones come under sustained attack - can they run to supporting groups or make life on their own? Can they turn the tables and become attacking stones rather than defending stones?

Rather than seeing cut points as a way of separating or capturing your opponent’s stones, I want you to see them as opportunities for expanding your groups in the best direction of play for each stage of the game. By playing moves to challenge your opponent’s gaps, and forcing them to connect, your groups can spread across the board with confidence - knowing that your opponent must play gote moves rather than sente attacks of their own.

Picture this if you will - you and another child are playing in a large sandbox. Each of you is trying to build the biggest castle you can from the available sand. As you build towers, bastions, and walls, you are filing various castle-shaped buckets and building blocks, then plonking them down to create those shapes.

Eventually, you discover that some pieces work better at the beginning, and some work better at the end. Some create long walls that cross the sandbox quickly, staking your claim. Other pieces interlock to create the towers, bastions, and keeps you need to build your eventual castle.

I often think of shape moves in the same way. While there are many moves that are used throughout the game, some moves are better for the Opening, and some are more appropriate for Midgame. Once you understand how the connections between your stones respond to the opportunities at various stages of the game, you will be able to pick the correct moves with confidence.

Some shapes are so fundamental that they are used throughout the game.

Iron Pillar / One Space Extension (Nobi)
Diagonal One Space Extension (Cosumi)
One Space Jump (Ikkentobi)
Knight’s Move (Keima)
Two Space Jump (Nikentobi)
Large Knight’s Move (Ogeima)

Think of these shapes as a gradient - going from a completely secure connection to a tenuous one. However, even though every shape below the One Space Jump can (in theory) be cut, doing so represents a risk to the opponent - by cutting, they may create their own mutually separated groups. Any connections larger than these - say a Three Point Jump - might be separated by the opponent without splitting their attacking force in two. As such - you can think of these shapes as loosely connected by the risk it would take to cut them - or connections that can be left tenuous until challenged.

The unique conditions in the Opening create opportunities for shapes that would be too risky otherwise.

Isolated stones

At any other stage of the game, you’d rarely play a stone then tenuki and leave it undefended. The reason we feel confident doing this in the Opening is our knowledge that claiming additional corner potential holds the highest reward. In certain cases, we can make more profit approaching another corner, even if our lone stone is threatened (i.e. the Knight’s move approach to the komoku). You can think of your lone stones as being temporarily safe because they’re “protected” by the bounty of other opportunities on the board.

Settling Your Stones With Large Extensions
This is also a good way of thinking about the large extensions one might see as players settle their stones in the Opening. If you review the joseki shown below, you’ll see that the first few moves are usually some combination of all-purpose shapes.


But when it comes time to settle, both opponents might risk long extensions with obvious cut points. They play these confident that there’s enough corner and edge potential left to entice their opponent into ignoring those cutting points - for now. If they can make a large extension and push both borders into the middle, they might not even need to secure those cut points later. Take a look at what White did along the bottom in this example opening:

White made a short term bet with Move 14 - expecting Black to approach the top right corner rather than exploiting the cutting points in the lower right. White then traded a smaller portion of the upper right (Black Move 21) for Move 28 on the lower left - because Move 14 was already there, White didn’t have to play a gote move to settle and could push up. If White can keep pushing those frameworks on the E and Q lines up towards the center, they can make a big moyo and gain an early advantage - all without having to secure those cut points at O4 / N4.

As the priorities of the game shift to pushing your potential away from the edge and into the middle, the shapes of the game respond. Local extensions pull in to become more secure, and isolated stones take on new meaning.

Isolated Stones
Once most of the corner and edge potential is claimed, single stones are no longer protected by the other opportunities on the board. As such, when you want to expand your direction of play into areas that are far from your frameworks, you have to rely on forcing moves designed to elicit a gote response. This allows the attacker one “free” sente move - often followed up with a securing move, but not always. The idea is to send your expeditionary forces deep into enemy territory, and give them a chance to travel back across the battlefield towards supporting stones.

The Peep
By challenging the gap in a one space jump, you can force your opponent to play a gote move to secure that connection. Having established a foothold, the attacker can play a gote move to connect with reinforcements. A peep can be anywhere that requires a single move to stay connected so look for diagonal as well as linear opportunities.

The Shoulder Hit
One way to limit your opponent’s ability to push away from the edge is to approach them just to the inside of their furthest securing stone - on “the shoulder” of their eventual framework. When your opponent secures the edge potential, you can often push in once more before playing a gote move to pull back to your own stones and stay connected.

The Moyo Limiting Move
Remember this board from the Introduction? Black realizes that White is about to make a big moyo, and plays a risky move to disrupt it. These can be the most difficult to see, because they require you to read the whole board and notice when your opponent is about to push inward from both sides of an extension. A moyo-limiting move has to push deep enough to threaten your opponent into securing from below, without going so deep that you might get trapped and capped from above. Like the shoulder hit, once you get a gote response, you can either push further in (if you have more forcing moves available) or pull back and start the process of crossing the board to your own reinforcements.

In Midgame, there are many situations when you may need to cross the board quickly. Perhaps you’re expanding your groups in the best direction of play. Perhaps you’ve played a peep and need to stretch back to reinforcements. By using these moves, you will be able to cross the board efficiently, and then build reinforcing structures from those stones.

One Space Jumps
A series of linear or cosumi jumps offer a good compromise between crossing the board quickly and staying relatively connected.
By now I’m sure you’re noticing all the cut points these moves are creating. As such, these shapes are most useful while there are no opponent stones within striking distance. As soon as your opponent approaches, these shapes must solidify their connections.

The Bamboo Joint
When Black played Moves 1 and 2, they connected those Iron Pillars securely because White can no longer cut them apart with a single move. When Black played Move 3, they created an offset Bamboo Joint which is also a surprisingly secure shape. If White tries to play in between the stones marked with O’s, Black can atari from above or below, and then secure when White extends away.

The Double Tiger’s Mouth
While three diagonal stones are a good shape in isolation - in the contact fight shown in the middle - those same stones become a very risky shape because of the two atari threats they create. However, as soon as Black plays on the key point at 7 that secures both cut points - making the Double Tiger’s Mouth a strong and useful shape.

Dogface and Horseface
When you play two Knight’s Moves side by side, you create the Dog Face. Two Large Knight’s Moves create the Horse Face. These allow you to push out in multiple directions while creating opportunities to secure those tenuous connections down the road.
These can be used in combination to build quick frameworks that can be solidified into more secure structures when challenged.

As we progress into Late Midgame, we need to remember that our ultimate goal isn’t to build walls, but to enclose as many empty points of land as possible within those walls. As such, you want to use shapes that maximize your potential eye-space while creating secure connections.

Mouth Shape
Mouth Shape shines in this context because it does so many things with so few stones. I like to think of it as two Iron Pillars with a central supporting beam. The beauty of this shape is that if White tries to play anywhere inside, it quickly gets trapped and captured. When left unchallenged, it creates opportunities for eye space efficiently.
Because Mouth Shape is so strong once complete, you will find that unfinished versions of this shape can also be very effective in the right context. As we go through our example game, I will call your attention to the times that both players use shape F - known as Table Shape.

Now - to be fair - nobody wakes up in the morning and says “I’m going to play a lot of Table Shape today!” That’s not how I want you to think about this. I am hoping that this chapter will help you start thinking of your games in terms of these move opportunities. As the board fills up with stones, and you go from simple shapes to compound shapes, being able to notice these opportunities and see how they can advance your tactical and strategic goals will allow you to move across the board with confidence. (Reminder - you may want to pause and take a break here!)

It took me a long time to pick the right game to demonstrate these concepts. On one hand, I wanted to pick examples that had strategies and tactics that beginners could relate to and use in their own games. I kept looking for the types of peaceful games I used in my last chapter - where cutting stones were sometimes abandoned and traded for moves that maximize the efficiencies at each stage of the game.

However, not only did these types of games lack the more complex shapes I wanted to demonstrate, they also left out a key aspect of gameplay - keeping separated stones alive and turning them from defending groups into attacking groups.

Go is a game of perfect information - everything you need to know is sitting there, implicit in the shape of the stones on the board. The problem is - the Beginning Player looks at the board and can only see the local exchange where the last stones were played. Narrowly focused on that local fight, they think “Situation A is happening.”

An Intermediate Player - someone who knows how to play a balanced opening and is aware of the stages of the game - can zoom out from that local exchange and put it in a broader context. Based on that, they think “Situation B is happening.”

A more Advanced Player can take all of that context, and add their experience of how such exchanges played out in the past. Even though Situations A and B may in fact be happening, they can look into the future, see how they can manipulate those elements to create a different Situation C, and turn the game to their advantage. Rather than just reacting to a given situation, they take an active hand in creating their own.

Since these articles are written towards 20-25kyu players, I’m always trying to find a delicate balance - trying to provide examples that beginners can relate to, but also pushing those boundaries and showing you the next steps down the road.

As such, we will be discussing a game I recently played against my old friend, the GnuGo AI - where both of us are playing around the 8-10 kyu level. Even though this game might be a stretch for the 20-25 kyu player, I am hoping to call upon all your previous learning, and show you how the concepts of shape can transform your strategic and tactical awareness. As we go through the game, I want you to focus on the following elements:

  • Using whole board play create weak groups
  • Using shape moves to chase weak groups in the best direction of play for that stage of the game
  • Using shape moves to turn defending stones into attacking stones and advance your groups into the best direction of play.

Please open this game in a separate window and follow along:


I am playing Black.

Moves 1-4: I start off with two komoku approaches and White responds with hoshi
Move 5: I approach the lower left
Move 6: White extends right
Move 7: I approach the lower right to prevent White’s large extension
Move 8: White makes a Dog Face to reinforce the bottom left
Moves 9-11: I play a corner joseki knowing I can expect a gote response for move 10 and settle my group hoping to create a large extension up the left side.

White wants to disrupt my plan - but what is the best way? After all, if they approach my upper left at A they might get a pincer at B.
Here’s what DIDN’T happen - White wanted to come up with a sente move that would force me to defend my lower left with a gote response, so they approach at C11 (Move 12) - assuming I will protect my base and allow them to complete their approach like this.
Move 13: Instead, I play a slightly more aggressive move and touch White’s stone
Move 14: White Nobis away
Move 15: And I play C14 to enclose the upper left corner.

My goal here is to get White to create a weak group that has to run to reinforcements. White is aware of this, but needs to focus on opening priorities since the two stones can escape, and there are still two corners that can be approached or enclosed.

Move 16: This is a smart move that does several things at the same time - it approaches a corner, prevents Black from making a large extension across the top, and gives the weak stones at C11/D11 something to run to
Move 17: If I had counter-attacked at P17, White would have had to defend that stone. However, I wanted White to tenuki - planning to come back in sente and threaten that stone later, so I played this more passive corner enclosure
Move 18: White tenukis and pincers my approach stone on the bottom.

In the past, I always used to get stressed out by pincer moves. Once I became more confident using sente and shape moves, I began playing this response, which almost feels like joseki.
Move 19: First push up to make sure the stones can run to the middle
Move 20: White plays a gote move to secure the right
Move 21: Then push in towards the corner with a Knight’s move
Move 22: White plays a gote move to enclose the corner
Move 23: This Dog Face threatens the M3 stone
Move 24: White Nobis to protect
Move 25: Black pushes one stone further
Move 26: White hanes at the head of two stones
Move 27: Black secures the cut point at N4

To sum up - when you get pincered - don’t panic. Just make calm shape moves, give your group a chance to run to the middle, stretch out in all directions, and force your opponent to respond. Once you run out of sente moves, make sure to play a gote move to settle the stones and secure your connections. Since White has not touched the stones at O5 or Q2, I can leave them for now, and connect them once they are challenged.
Move 28: White wastes no time and challenges the Q2 stone to enclose more of the corner
Move 29: I Nobi up to protect and make a tentative connection with the rest of my group. I need one more stone to close the cut point at P4 and make a Double Tiger’s Mouth, but
Move 30: White tenukis to threaten my upper right group.
Move 31: Fortunately, P4 is sente - if I atari at Q5 White’s corner is in trouble
Move 32: White plays a gote move to secure
Move 33: Now I can get back to protecting my upper right corner in sente
Move 34: White blocks
Move 35 - 36: We both secure our cut points

And here we get to one of those turning points that I was talking about - I notice that Situations A and B are possible, but I want to create my own Situation C:
We both still need to make one move each to settle our groups in the upper right. If we were focused on Opening priorities, we could both settle, and I could make an extension at either A or B - knowing whichever one I pick White will probably play the other. However, I see a fleeting opportunity to create two weak groups, and use the resulting direction of play to increase my potential in the upper left.
Move 37: The Knight’s Move puts more pressure on White’s upper right group, while leaving my corner more vulnerable
Move 38: Rather than settle the top, White risks a lone stone to threaten that corner
Move 39: I make my own risky extension at the top, preventing White from making a base
Move 40: White ignores the upper threat and settles their large extension on the right
Move 41: I fence White in - limiting the direction they can run.

White now has two weak groups at the top and will have to fight to connect or make eye shape. Because I can chase those groups from all sides, I can use that direction of play to make a moyo in the upper left corner, and drive both of my groups into the middle - which is exactly what I need to do as we transition from the Opening into Early Midgame.
Move 42: White sees the writing on the wall and starts running to the right
Move 43: I peep at the E11 cut point.
Move 44: White keeps running
Move 45: I push in, hoping to cut White apart.

Now, a less confident player might have abandoned those two stones and kept running, but White played those stones to prevent my large extension, and they won’t abandon them until that job is done. If they keep those stones connected, they can reduce my potential and turn their stones back into an attacking group. Check out the calm confidence with which White uses forcing moves to connect their stones together.
Move 46, 47: White threatens to take away my group’s base and I block
Move 48, 49: White threatens my corner, I have to protect
Move 50, 51: White threatens the D13 cut point, I have to connect
Move 52, 53: White ataris the C10 stone, I have to protect
Move 54: White pushes to the middle and makes a tentative connection with the F11 stone
Move 55: I atari the B10 stone to give my group a bigger base
Move 56: White reinforces the connection to the F11 stone and threatens my group
Move 57: I reinforce my group to keep White from invading
Move 58: White tenukis to extend their upper right group (first Table Shape of the game)

The balance of power has changed, so let’s just take a moment and do some broad-strokes strategic analysis.
When White connected their left side stones together they turned their weak group into an attacking group. Rather than having to run, they can now invade the remaining yellow area at the top and threaten to capture my lone stone at K17. But more importantly, since we are in Early Midgame - White can push in from their left side stones and the lower left group to create a massive center moyo - giving them a huge lead. How can I turn the game around? How can I protect my upper left potential and negate White’s advantage at the same time?

I do it by forcing White to make shapes that can’t connect, and use the resulting cut points to my advantage.
Move 59: I poke the Knight’s move connection between F11 and G13 with a Filled Triangle
Move 60: White tries to connect, but those 3 diagonal stones have two cut points at G11 and F10 - a dangerous shape!
Move 61: I cut at G11
Move 62: White secures
Move 63: I Nobi away

And with one little cut the tables have turned. White once again has a weak group at G12/G13 that has to focus on running rather than attacking. What more, I now have two stones near the center that can put pressure on White’s lower left group and aim for middle influence. White decides to settle their left side group before tending to the two lone stones.
Move 64: White protects the C12 cut point and ataris my B11 stone
Move 65: I capture
Move 66: White stretches towards the group from below
Move 67: I peep at the cut point
Move 68: White has to connect or it might lose all those stones
Move 69: I connect back to my two stones - getting more influence on the middle

Having secured the left side group, White can start focusing on their two lone stones. They start by making a clever peep at a diagonal cut point - can you spot it?
Move 70: E14 - right at the key point I need to finish my Double Tiger’s Mouth - nicely illustrating the Go proverb that your opponent’s best move is often your best move
Move 71: I connect and makes 2 Iron Pillars
Move 72: White tries for Dog Face, but
Move 73: I complete my Mouth Shape right in the spot White needs to connect - incidentally peeping at their G14 cut point
Move 74: White has to secure
Move 75: I Nobi up
Move 76: White connects their upper stones - loosely.

White wants to reinforce those loose connections and I want to squeeze them from the inside and the outside at the same time. As soon as I get sente, I begin to put the pressure on - let’s see how many forcing moves I can string in a row.
Move 77, 78: I hane and make a loose connection with the D17 stone, White blocks
Move 79, 80: I double-hane - White connects on the right with a Dog Face
Move 81, 82: I atari - White makes a Filled Triangle (danger!)
Move 83, 84: I have to play the key point on the Double Tiger’s Mouth, White secures all their stones together
Move 85, 86: I secure the eye at J17 and create a cut point at L15, White tenukis to reinforce their lower left group
Move 87, 88: I peep at the L15 cut point - White has to connect (with Table Shape) or they will lose 8 stones
Move 89: I connect my internal stones together, allowing me to attack the middle from above.
Move 90, 91: White tries to split my upper groups with a Tiger’s Mouth, I Nobi
Move 92, 93: White Nobis, I hane at the head of two stones
Move 94: White makes an Empty Triangle that also completes a Double Tiger’s Mouth
Move 95: I Nobi
Move 96, 97: White tenukis and attaches, I Nobi up, threatening the cut point at D3
Move 98: White makes a risky move to expand their lower moyo into the middle, and fence in my lower right group, which is still in moderate danger
Move 99: Instead of responding directly, I push down into the middle from my upper right group with a Table Shape, attempting to keep White’s two upper groups apart
Move 100: White threatens my stone
Move 101, 102: I loosely connect my two upper groups back together, and threaten White’s top group, forcing them to play M18 to ensure eyespace

Now that I’ve given my lower right group somewhere to run to, I turn my attention back to the lower left. Rather than directly fighting White’s expansion at K7, I poke at the gaps between White’s left side stones and the bottom group. White has spent a lot of time and effort keeping those stones alive. Now I get to reduce their moyo by threatening to cut them off.
Move 103: This is sente, as it threatens an atari attack at E4 or D3
Move 104: White has to connect, knowing they’re leaving a gap at E5
Move 105: I play a Knight’s move approach to White’s K7 stone, giving my attacking stones something to run to
Move 106: White ignores my attack and plays a greedy move rather than securing
Move 107, 108: I Nobi to touch White, White Nobis away
Move 109, 110: I threaten to cut off White’s left side stones, White makes Table Shape
Move 111, 112: I push right, White makes a Bamboo Joint
Move 113, 114: I push right, White makes an Empty Triangle and Table Shape
Move 115, 116: I push down, and White Nobis to connect.
Move 117, 118: I push down threatening White’s moyo, White protects

Let’s just take a quick look at the Score Estimator and compare the situation between move 102 and move 118:
White was planning to expand their bottom potential to create a moyo at the lower left
(Score Estimator: White ahead by 6.5)
By threatening those connections from above and below, I forced White to fill in their potential middle moyo with stones, leaving them only the bottom - which I can still threaten. (Score Estimator: Black ahead by 6.5)

The question now becomes - is threatening White’s bottom framework still the biggest move on the board? This is another one of those moments where Situations A and B might be happening, but is there a Situation C that I can create?
If I stayed stuck in my tunnel vision, I could keep trying to play aggressively to reduce White’s bottom potential and keep chasing those stones. However, because White can now connect at J5 / K5 / H4, their stones have turned back into an attacking group, and can threaten Black’s cut points in the middle.

However, if I pay attention to the biggest unclaimed area in the middle (in Orange) then the resulting direction of play becomes beneficial to all my remaining groups. Because White still has to play gote moves to secure the bottom, I can use that wiggle room to give myself a better position from all sides.
Move 119: I threaten the connection of White’s stone at L9
Move 120: White secures the K5 cut point to ensure that group can still attack
Move 121: I push in once more, increasing my middle moyo
Move 122: White secures the bottom and tentatively connects their groups
Move 123,124: I make a Large Knight’s move extension on the right, White blocks to secure
Move 125: I Cosumi to threaten the Q11 stone
Move 126: White Nobis to threaten my upper right framework
Move 127, 128: I atari the Q11 stone and White secures
Move 129, 130: I Nobi to threaten R8 and R10, White secures
Move 131: I Nobi to secure my upper right framework
Move 132, 133: White threatens my P8 stone, I Nobi to secure
Move 134, 135: White pushes into my corner, I secure with a Tiger’s Mouth
As you can see, Midgame is almost over - we only have a few disputed areas at the edge and that last bit of center potential in the lower right. White has sente and decides to secure their upper group’s eye-space and reduce my potential at the same time with a Monkey Jump.
Played in the middle of the board, a Large Knight’s move is a fairly tenuous connection - but played at the edge, it’s almost guaranteed to make it back to its supporting stone at M18 because if I try to block at A then White will atari me at B, chase me to M18 and capture me.
Moves 136-146: Instead, the way to handle a Monkey Jump is to calmly make moves that force White to connect back to reinforcements. White manages to play a few forcing moves of their own (Moves 140 and 144), but when we are finished the area is filled in with stones - White may have reduced my potential at the top, but at least they didn’t get any extra eyes or captured stones. White’s upper right group can now live unconditionally (i.e. I cannot kill that group).

Now I have sente and I use it to make the biggest middle moyo that I can:
Move 147: this Filled Triangle connects my middle moyo to my lower right group
Moves 148-151: White reduces my left side framework and I secure
Move 152: White secures their potential on the bottom and sets up another possible Monkey Jump at P1
Moves 153-154: This Empty Triangle peeps at the K5 cut point, White secures
Move 155: This Cosumi creates a Tiger’s Mouth, Dog Face, and a Knight’s Move - tying all the stones in that area together.
And with that - Midgame is over, and I go into Endgame ahead by about 19 points.

I feel like I’ve already stretched your patience with this article, so I won’t go into detail on the Endgame. By playing cautious moves to connect my stones, and ignoring attacks that only threaten to capture single stones rather than threatening my groups (Example: Moves 156-158) I manage to increase my lead another 12 points or so (Final score: Black by 31.5 points).

I hope that this whirlwind tour through the world of shape has helped you see the game in a different light. Like any beginning player, I spent a long time between 25 and 15 kyu playing without any awareness of shape - making awkward extensions that my opponent exploited with forcing moves at best - or killing my groups at worst. Once I leveled up to 15 to 12 kyu I got better at linking my stones together so that - even when I lost a game - at least I didn’t suffer any large captures. Once I leveled up to 12 to 10 kyu I began to see the gaps in my opponent’s shape as opportunities to make my own forcing moves, and was able to create my own story rather than just responding to my opponent.

For me the possibilities I could see through the focused lens of shape became the fulcrum point between Strategy and Tactics - my awareness of efficiency and the priorities for each stage of the game allowed me to see which Strategic parts of the board were most valuable at each stage - and my awareness of how I could exploit the shape opportunities at those moments allowed me to execute my Tactics with confidence, knowing that I was forcing my opponent to respond the way I wanted to further my goals. So, to sum up:


  • Focus on the Opening priorities and use whole-board play to claim as much potential as you can while its freely available
  • Look for opportunities to create weak groups that must run to reinforcements
  • Make sure your own groups stay connected and don’t become weak groups themselves


  • Look for opportunities to push inward from both sides of a large extension to create a moyo in the middle
  • Chase weak groups from both sides and use the resulting direction of play to push your potential away from the edge
  • When possible use forcing and cutting moves to create weak groups that you can chase
  • When your weak groups are threatened, make sure they can escape first, and then try to keep them connected so that they can still disrupt your opponent’s frameworks


  • Pay attention to where the remaining unclaimed areas are - they may be in the middle, at the edge, or even a poorly secured weak group that can be captured or threatened
  • Look for weak groups that have not yet made two eyes or connected to reinforcements to capture them or use forcing moves to create a direction of play that works in your favor
  • Make sure all of your weak groups can make eye-shape or connect to reinforcements
  • Look for poorly secured areas and make threatening moves that force your opponent to fill in their potential with securing stones
  • Look for opportunities to make Monkey Jumps and other moves that reduce your opponent’s potential at the edges of their frameworks


  • Look for opportunities to play even numbered sequences (where you retain sente) rather than odd number sequences where sente passes to your opponent
  • If your opponent threatens to capture a single weak stone but does not put any of your groups at risk, see if there are bigger moves you can make to reduce your opponent’s frameworks by ignoring the smaller threat
  • Play on any remaining points your opponent needs to create eye-shape

Thanks for reading and good luck!

If you enjoyed reading through this sample game, and you’d like to keep studying these concepts on your own, here are another two games you can review. As you go through them move-by-move, try telling yourself the same type of story I was telling here - what is Black trying to do - how is White responding? What groups are in danger? What groups are safe? What areas of the board are most valuable at each stage? How is each side using shape and forcing moves to further their goals?

Intermediate game:

Advanced game:


Wow what an epic! Are you going to publish all this?


Thank you SO much : )

The feedback I receive from other OGS forum members is always really valuable and appreciated, so thanks for saying that.

As re: publishing - yes, I’d like to pursue that at some point down the road. However, given that this 19x19 FOR BEGINNERS project is a bit of a labor of love - it has taken me a LONG time to get this far (I published the very first intro article in January 2020…)

I would like to write at least another 3 articles so that the material is more well rounded and works better as a coherent “book” so to speak. At this point, I’m thinking the last 3 articles might be

  • The various strategies that one can use in a 19x19 game, and the individual skills you need to have before you can be successful with those strategies
  • The Devil is in the Details kind of stuff: handling Ko effectively, avoiding false eyes and other traps, capturing races and seki, etc
  • Ways of studying and getting better once you’ve committed to learning this game: reading, tsumego, studying pro games, learning from AI review or AI-powered centaur Go, etc.

If my previous pace is any indication of future performance, I’m guessing it’s going to take me another year or two to get that far…

For now - I am happy to have these articles hanging out as a resource for other OGS forum members. I really am committed to the idea of onboarding new players with as little frustration and confusion as possible, so hopefully these articles have helped people wrap their heads around this game over the years.

Eventually I’d like to rewrite some of the earlier chapters so they’re not quite so OGS-specific (though the sample games will always be on OGS, and I will always try to encourage the reader to connect with OGS because it’s such a fantastic / welcoming community), and see if I can get this published as an e-book of some kind. My hope and wish is to maybe eventually find some connections with the American Go Association and see if they would be willing to endorse or even help promote this book as a means of attracting new players. But - yeah - at this point I’m still far from finishing the articles, and this is sort of a pipe dream.

Anyway - thanks for asking : )


In the context of recognising situations and learning from them, I think it would be important to share somewhere, related to this, that if you find yourself needing one of these, then you already might have missed something earlier. You ned to ask yourself “whoa, how did I let this potential moyo form so nicely for the opponent that now I have to do a risky move to counter it”.

In saying this, I am consciously echoing the words of Dwyrin, who consistently prefaces advise about “how to reduce a moyo” or “how to successfully invade” with “don’t get into that situation in the first place”. This seems like good advice to attach to any discussion of “a moyo reduction shape”.


I have tried to cover this a topic from a few angles previously - once in the Playing a Balanced Opening article, and another few times in the Stages of the Game article

The other thing I’ll say is - it took me a LONG time between 18kyu and 13kyu to learn the whole board reading / strategic understanding to catch incipient moyos before they form - so I’m trying to introduce the concept from a few different directions, because they require one to hold so many aspects of the game in one’s head at the same time (or learn how to zoom out more often to be aware of those whole-board aspects)

Thanks for the feedback : )


Well, this is a great result and with very interesting points that are nicely explained, so if/when you want to turn it into a book, let me know if you need any help in typesetting it.
I am willing to do the whole book for free. Your diagrams are all consistent in size (unlike the mess I’ve made out of the first chapters in my book :stuck_out_tongue: ), so it will be much easier to organise/complete.

After that, if you want, I can get you in touch with some of the people that helped with the translations. E.g. @jlt did the French translation for the Multilingual Go Book and he was amazing in every conscievable way (speed, efficiency, attention to detail - he was translating faster than I could typeset! :grinning: )

I am looking forward to it :slight_smile:


Wow! Thank you for your very kind and generous offer! I will definitely keep that in mind : )

1 Like

I’ll keep that in mind, that’s a way of thinking that I haven’t achieved yet. I know that’s not really the point of this article, but hey :woman_shrugging: :slight_smile:

1 Like

Hi - thanks for the feedback

Once again, I know that the games I try to describe are happening at all different skill levels - I want to check whether the stories I tell are still something that you can connect to? or am I losing touch with the audience so to speak?

1 Like

Some constuctive criticism:

  1. Longer isn’t necessarily better, edit for conciseness.
  2. Get the go content vetted by a stronger player

I disagree with this characterisation of more advanced concepts as simple outgrowth of the topics titles so far. e.g. timing, or ko.

It’s not a tiger’s mouth if white is already in the mouth and can’t be eaten. This shows a common pitfall about shape that focuses on static shapes of stones of just one colour, ignoring the presence of opponent’s stones.
5. First example of wedging one point jump black ataris from wrong side and turns a bad move into a good result for white.
6. Second wedge example black ataris from wrong side. I know you say it’s improbable, but you should have only white playing silly futile moves to show why this wedge is a bad idea, black shouldn’t return the favour and play terrible moves too but sensible moves to show why white wedge was silly.
7. Taking a step back, I question the wisdom of showing wedges of one point jumps. I understand you want to make the point that they are not strictly connected from a shape perspective, but wedges of one point jumps are usually bad and are a very common beginner mistake/intuition. You should reduce not encourage this. So if you are going to show whole board as well as local examples of it, you need to make it understandable to your weak target audience whilst making clear it’s usually a bad idea. The general them of the first example would be “bad wedge because of nearby opponent’s weakness (of other one point jumps)”, 2nd is “bad wedge because of nearby board edge giving extra support” and have a good example of one, such as along a solid corridor though qualifying this is a rare occurence.
8. The 2 diagrams of shapes in joseki aren’t joseki but contain mistakes. In 1st 5 is bad with the slide and 6 is slow. In 2nd 7 is bad shape, 8 should hane. Also don’t crop the edge of the board to make it clear these patterns are at corner not in centre.
9. The second peep example is not a peep, it’s an atekomi, and a terrible one at that. White shouldn’t submissively connect when there is no need, but atari on outside is first feeling, or laugh and tenuki.
10. Shoulder hit example is a bad one with nearby weak group which means black should not crawl but push up and separate and fight to punish this bad move. A better shoulder hit example would be one against a flat black moyo with no nearby white group.

1 Like