Review request 19x 19 Tpk vs tpk

Please review game


Very peaceful game.

Black lost a little advantage and eventually made a small mistake that could have changed the outcome.
After white 123 (L4) Black should’ve connected.
White J3 could capture two stones and reduce a little.

But white didn’t notice it. Lucky Black. :grinning_face_with_smiling_eyes:

I would just suggest to try and invade those huge moyos. 4-4 corners can be invaded. If you aren’t sure of the outcome, it could be worth a try anyway.


Two-stone game (23k–20k), reviewing for White although with many comments on Black’s play as well.

(interesting, btw, that Sensei’s Library has apparently no dedicated article on two-stone handicap)

Move #(s) Comment(s)
1 4-4. The 4-4 point is also called the star point or hoshi, but “4-4” technically only refers to those four star points which are in the corners, and not to the five other marked points that aren’t. The 4-4 takes its name from being the fourth point from either edge. It is usually considered an influential move, although that is an oversimplification.
2 3-3. Also called the san-san, which is just Japanese for 3-3. It’s a territorial move.
3 Shoulderhit, also called the kata (shoulder) or, more pedantically, katatsuki (shoulder connection). An influential move. See this page for discussion of the local position.
4 The only move. There is no difference between this push and D17 since the whole-board position is symmetrical.
5 This should extend to D15, preventing Black from putting White in a hane-on-two, that is, a hane at the head of two stones. Allowing this hane would violate the proverb that five liberties are needed for tactical stability.
6 Note that Black does immediately make the hane. It reduces the liberties of your stones, as said, and also increases the height and strength of his group.
7 Making the best of a bad situation.
8 Black can consider a double hane, which can develop into a position like in Diag. 1. However, that is needlessly complex and not demonstrably any better for him, so the simple extension is perfectly fine.
9 Again making the best of a bad situation. It would, however, perhaps have been better to simply fix at F16, making a tiger’s mouth or, as I would say when D17 is part of the shape, a crocodile’s mouth. This would avoid pushing Black along the fourth line, which is usually (but not always) a questionable strategy, due to the large amount of territory which Black can attain that way.
10 Black is entitled to this spirited hane, since although he has cutting points you have your own cuts as well. A safer strategy for him would’ve been to just extend, but this move can’t be criticised. Black is attempting to raise his already high position onto the fifth line.
11 White can’t cut due to the E16 defect. You could still consider fixing at F16.
12 Black finally backs down. He could have played an extension or even a double hane, but this solid move is reasonable.
13–14 This atari is a natural sente exchange.
15 I still recommend fixing at F16. Your shape isn’t strong enough to resist the E16 cut.
16 Just a big star point.
17 Checking extension (tsume). Still doesn’t address the cut.
18 The armpit hit is rarely a good shape. In most cases the subattachment (here H17) is better. This shape leaves Black with bad aji at L17.
19 The natural and only move.
20 Slow. There’s still aji at M17 and a shape play for White at K15.
21 With this, the cut is finally protected by Diag. 2.
22 A bit loose. The fourth-line extension would’ve been a more natural and solid shape.
23 Just a big star point.
24 Fairly large territorial move.
25 This first-line hane aims at a hane-connect (hanetsugi). Hanetsugi become sente in the endgame, but at this point there are many big points available all over the board and Black is thus under no obligation to respond.
26-7 The hanetsugi begins to be played out.
28 Not even making a hanetsugi fix, which would be made with either the solid connection at K18 or the tiger’s mouth at L18.
29 Taking the tactical opportunity. Note, though, that this whole discussion is rather small and should be reserved for the endgame.
32 Losing some points by asking White to connect, which you’re happy to do. Later, a Black hane at M17 can then be answered by White M18, capturing the hane stone – without this exchange, White M18 could be answered by Black K18, beginning a ko (although, since Black would have many local threats, a ko in name only).
34 Black protects, giving White sente.
35 It’s not clear what the purpose of this stone is. It doesn’t really begin to surround territory, except somewhat inefficiently with a follow-up in the local area like J13. Better to begin playing on the open bottom side with F3 or K4, or similar moves.
36 Note that Black does immediately play on this key bottom side.
37 Another checking extension.
38 Again a bit loose and high as compared to its equivalent on the fourth line.
39 I still don’t see why you value the centre highly in this position. How about R14, developing your right-side position? Or invading at the 3-3 point in the lower left?
40 Note that Black required an additional move in this area to enhance his flimsy first.
41 Good! This is a big point.
42 Too cagey. In the process of clinging to his corner territory, Black strengthens your approach stone, reducing the aji he had to later invade at R12.
43-4 Seems to be a fairly neutral, fairly sente exchange. An SDK player would likely question whether it is entirely sente, but the follow up of C11 isn’t small, so there’s no point in criticising it too much.
45-6 I was going to claim that this exchange wasn’t sente. It was only when making Diag. 3 that I realised my mistake!
47-8 Neutral exchange. Not sente, though – it’s early endgame.
49 At least the issue of the E16 cut is decisively and simply settled. To spend sente here, though, was too much.
50 Should be at Q14, a bulge over White’s stone.
51 Consider pushing up as in Diag. 4.
52-3 You’re still playing for the centre. This, at least, cooperates well with (51), making the big bulge o///o double keima shape.
54-5 Bad exchange by Black, who is becoming overconcentrated. Your connection / sealing shape is good.
57 Slightly bad endgame since Black can continue with the F14 cut – no matter from which direction you atari, he can connect back to H14.
58–60 Fairly neutral exchanges.
61 Big endgame block, but still only endgame.
62-3 This exchange wasn’t sente.
64 Pretty big move, if a bit endgamey. Black is playing a quiet game.
65 Again, the subattachment is better than the armpit hit. If Black tries to cut then you can escape with Diag. 5., ruining Black’s position in the process. He should, instead, hane on the inside.
67-8 This attempted seal allows Black to jump into the empty gap left, the so-called elephant eye of the “elephant jump”. A better local shape would be the double keima at L6 – however, Black has his own defect at K5, which you exploit, so you don’t really have an issue.
69–71 Sealing.
72 Small endgame.
73 Not necessary. You do get ensure slightly better endgame around J8 and L2, but it’s too small to play at this point. Even after all this time, it’s still biggest to play to enclose the corners.
74-5 Not a sente exchange.
76-7 Not a sente exchange.
78 Not small, but playing around the two lower corners is still bigger.
79 This induces Black to extend to R12, into your territory. Better to just clamp there yourself.
81-4 You played good endgame here.
85 Slow – it’s only a stone.
86-7 Not a sente exchange.
88 Very slow – again, it’s just one stone.
89 Great! Really, though, a tighter enclosure of the corner is called for so as to prevent a 3-3 invasion: R6 or, even better, R5.
90–91 There’s enough aji around L4 that I can’t complain about you answering here.
92 Very slow. You can now play in the lower left.
93–123 Thirty moves of rather inconsequential endgame.
125 Due to Black playing away on the previous move, you now had the J3 atari and Diag. 6. The bot suggests that this would’ve been enough to win by a small margin; however, I’m not sure whether or not it expects you to still consider playing in the lower left.
126, -7, -9 Dame. These moves don’t score in a Japanese-rules game (or under other territory-scoring rulesets).
130, -31, -33 More dame.

Diag. 1


Diag. 2


Diag. 3 (White captures)

Diag. 4


Diag. 5


Diag. 6 (White 3, Black 4)


Some take-away points:

  1. Try to recognise and avoid playing endgame moves too early, since they’re gote and the opponent can take a big point elsewhere.

  2. Research 3-3 shoulderhit and invasion joseki, shape and tactics on Sensei’s Library.

  3. If you’re having difficulty surrounding territory, ask yourself whether you’re habitually attempting to do so in the centre, where there are four initial sides open to incursion; rather than on the sides and in the corners.

  4. Stand from a kick, as after (42).

  5. Remember that just because you’ve claimed the territory around a corner, the corner itself can still be susceptible to invasion, especially until you improve your tactical understanding. Of course, permitting invasion and attempting to kill the invasive group is also a way to learn, but enclosure can be more pragmatic.

Even though this review may have been too detailed, I hope that my comments were useful to you.


Thanks @Lys

1 Like

Thanks @bugcat

1 Like