It is currently Thu Oct 29, 2020 10:33 am

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 213 posts ]  Go to page Previous  1 ... 7, 8, 9, 10, 11  Next
Author Message
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #181 Posted: Thu Jan 16, 2020 10:45 am 
Lives with ko
User avatar

Posts: 192
Liked others: 39
Was liked: 28
Rank: EGF 2k
KGS: MKyle
Of course! 19-18 is illegal! Translating between thoughts and coordinates continues to be tough!

13(N)-19

Trigger: if b 11-19 then
19(T)-17


Comment
Oops! Must watch out for these transcription errors!
Nothing has changed, my plan continues. Depending on blacks threat I may need to count the size of this ko, think about how the approach ko really works and maybe count threats..

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #182 Posted: Thu Jan 16, 2020 11:29 am 
Lives with ko

Posts: 130
Liked others: 1
Was liked: 16
Rank: AGA 4 dan
Universal go server handle: telegraphgo
Trigger accepted. B plays 11-19, white follows with 19-17. I then play
2-6
B6

Captures: B: 9, W: 5


Finally we've gotten to the interesting part. I'm definitely hoping that he makes some mistakes here. But we'll see.

I know I said I was going to play the 5-14 ko threat. The problem was that if he ignores it, I capture, and he extends on the side with something like 3-11, it's hard for me to make up the remaining points difference. He would have 13 points underneath the upper right, plus 30ish on the right side, 5 in the middle, 10 in the lower left, another 8 in the upper left, and komi. That's like 75 points, and I'd have only the top and some aji on the left to show for it. It's just not big enough to spend two moves capturing those stones right now.

That would change if I prevented him from making a comfortable extension on the side. So I'm playing 2-6, with the intent of continuing 2-7 if he plays 3-7, and continuing again at 3-5 and 3-4 to try to secure myself before playing out the center fight. It looks pretty scary for him, to be honest.

I have a feeling he will back off again at 2-5, which will allow me to claim that I played a good ko threat and come back to reclaim the ko.

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #183 Posted: Thu Jan 16, 2020 11:32 am 
Lives in sente
User avatar

Posts: 992
Liked others: 84
Was liked: 354
Black has 10 captures, White has 5.

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #184 Posted: Thu Jan 16, 2020 2:28 pm 
Lives with ko
User avatar

Posts: 192
Liked others: 39
Was liked: 28
Rank: EGF 2k
KGS: MKyle
2(B)-7

Trigger: if b 19-16 then
9(J)-1


I haven't done my ko homework properly yet. As far as I can see in the corner if I settle the ko then I get 15 points in the area but if black approaches and settles then he has 10 points, so the full swing is 25. I think some people would multiply by 1/3 or 1/4 or 2/3 or something to give threat values or values per move and such but I think these methods miss a lot of the subtlety and/or I don't understand them, so I'll have to work more on first principles or more realistically dodgy intuition.
Black's threat looks like it would be a serious headache to ignore and it feels like a basic mistake not to give the strongest response, despite the follow up threats. My threat is just a move I'd like to play. It keeps blacks group technically one eyed.

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #185 Posted: Thu Jan 16, 2020 4:30 pm 
Lives with ko

Posts: 130
Liked others: 1
Was liked: 16
Rank: AGA 4 dan
Universal go server handle: telegraphgo
Trigger accepted. 19-16, T16, is answered by 9-1, J1. Then I play:
7-2
G2

Trigger: If white plays 19-17, I play:
3-7
C7
And captures will be B:11, W:6. Thanks jlt!


Comments:
I deeply considered playing away from this threat. If I had played away from it, then I would have wanted to play 3-7 before retaking the ko, so that he can't answer with 2-8. But I don't think I'm going to play away from it. It's the same problem as before - things simplify too fast. If I was ahead or the game was very close, then I probably would have let him have my stones.

But I think I can force him to find some scrappier (read: losing) ko threats. His response at 2-7 gives me several more ko threats locally, while also getting a slightly poor result. My plan is that he will get another ko threat at 8-1, and then I will get another ko threat with my atari at 4-6. Then I see only one decent threat for him at 4-19. I still have some big ko threats that aren't too bad, so it's looking like the ko fight might be a bit hard on him. Playing this way should keep him under pressure.

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #186 Posted: Thu Jan 16, 2020 5:23 pm 
Lives with ko
User avatar

Posts: 192
Liked others: 39
Was liked: 28
Rank: EGF 2k
KGS: MKyle
Trigger accepted so w19-17, b3-7

2(B)-5

Trigger: if b 19-16 then
3(C)-10


Comment
It's not an approach ko! I'm being an idiot! My cut at 16-18 was just really really bad.
There's no doubt that black has more threats so I'll see if he will just let me make two nice moves here.

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #187 Posted: Thu Jan 16, 2020 5:54 pm 
Lives with ko

Posts: 130
Liked others: 1
Was liked: 16
Rank: AGA 4 dan
Universal go server handle: telegraphgo
Trigger accepted. 19-16, T16, is followed by 3-10, C10. I then play:
4-6
D6

I think he realized the ko fight might be trouble and has decided to try to take two big moves, losing the ko to win the game. It's a good strategy, but I'm not sure I agree with the choice of 3-10 as the biggest move. I guess it's up to him to justify it.

I'm going to be taking 14-19 next and winning the ko, but first I want to make this exchange. If I didn't, then it's very possible that he would play first at 4-6, which is totally unnecessary to allow.

Since I'm winning the ko without ignoring a critical ko threat or making a bad ko threat, it seems I have made up for my dumb mistake at 17-19. That's good news, but I think that there's still quite a few points to be made until I equalize.

He should play 3-5 and then I'll take 14-19, but I'm not going to make a trigger in hopes that he makes a mistake! :D

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #188 Posted: Thu Jan 16, 2020 6:04 pm 
Lives with ko
User avatar

Posts: 192
Liked others: 39
Was liked: 28
Rank: EGF 2k
KGS: MKyle
3(C)-5

Meant to say thanks too to jlt! if we get to count then we may well need a referee :tmbup:

Comment
Solid here at least for probably 9 points. black is still a tiny tiny bit harassable on the basis that there aren't two explicit eyes yet.
Will black now finish the ko?

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #189 Posted: Thu Jan 16, 2020 6:25 pm 
Lives with ko

Posts: 130
Liked others: 1
Was liked: 16
Rank: AGA 4 dan
Universal go server handle: telegraphgo
14-19
O19

Finishing the ko.
Atari.

He didn't mess it up. Oh well. Time to give him sente!

I'm gaining a deeper understanding of the alphabet-number relation by playing this game :D

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #190 Posted: Thu Jan 16, 2020 7:36 pm 
Lives with ko
User avatar

Posts: 192
Liked others: 39
Was liked: 28
Rank: EGF 2k
KGS: MKyle
2(B)-8

Comment
I hope this move isn't too slow. It seems like nice endgame.
Now the right is settled I count 29 points for me. I previously counted 9 points for the bottom left and I was thinking that the upper left should be more than 10 now that I have that extention. Saying 2 points for the centre would give a conservative estimate of 50 points on the board for me plus komi. Blacks points are a lot harder to count. I found about 35 totally safe points very close to the edge on the top side and I think another 5 seems like a minimum. But then the other two groups won't make many points. I thought maybe a large endgame point might pick up enough points that black might struggle to make it up in the centre? It's very tough for me to judge that potential though. I guess black will take a lead as we play out the endgame

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #191 Posted: Fri Jan 17, 2020 1:06 am 
Lives with ko

Posts: 130
Liked others: 1
Was liked: 16
Rank: AGA 4 dan
Universal go server handle: telegraphgo
9-12
L12

Two-space jump.

I'm pretty sure his move was much too slow. I don't think that my bottom left group should be very easy to attack profitably, so he ends up gaining a couple of points and a little bit of security with his 3-10 stone. I think I can make him regret that :rambo:

My move is just playing at the point of greatest contention. I said it was small earlier, showing how it was worth on ~15 points to the four stones ~20. But 15 points is more than 2-8 made, and more than anything else I can see on the board. Also, I think it sets up a sort of miai for reducing his upper left at 4-11 or 5-14, but maybe I'm just being overly optimistic. Don't forget, as well, that his group on the bottom is not yet technically alive, so I may be able to squeeze out some good endgame there.

The midgame is soon coming to a close, and I'm not in time trouble, so I'll count again.
W: bottom left - 12, upper left - 12, center - 5, right side - 30, komi - 7.5, total of about 65.
B: bottom left - 3, bottom right - 2, upper side - 8 + 10 + 30 is about 48, captures - 5ish, total of about 60.

Now this was an optimistic count for black, so I bet I'm more like 10 points down than 5, but it's still an improvement for sure. I still need him to make a mistake or two, but it's definitely not yet become too simple.

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #192 Posted: Fri Jan 17, 2020 1:17 am 
Lives in sente
User avatar

Posts: 992
Liked others: 84
Was liked: 354
9-12 is not the same as L12.

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #193 Posted: Fri Jan 17, 2020 1:24 am 
Lives with ko

Posts: 130
Liked others: 1
Was liked: 16
Rank: AGA 4 dan
Universal go server handle: telegraphgo
Gah! J12. The alphabet 9p is far too strong for me :oops:

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #194 Posted: Fri Jan 17, 2020 8:53 am 
Lives with ko
User avatar

Posts: 192
Liked others: 39
Was liked: 28
Rank: EGF 2k
KGS: MKyle
8(H)-12

Comment
It feels like my rationale behind my last move should have lead me to play around 9-12 but I wussed out because it's a little bit fuzzy and the move I played felt more concrete. It looks great now black has played it!
Since black is strong enough I'll lean here. If I get a few moves around then my distant distant attack on the lower left black group might just come back around again.

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #195 Posted: Fri Jan 17, 2020 10:33 am 
Lives with ko

Posts: 130
Liked others: 1
Was liked: 16
Rank: AGA 4 dan
Universal go server handle: telegraphgo
8-13
H13

Hane.

His move is nice - since I'm strong, he's building both of our strengths and his is more useful. But I'm kind of insolent - I'm planning on tenuki next move, and if he spends a whole move accomplishing just a little over here, it's going to be too slow again.

I used to never play hane and tenuki, but AI tends to like it much more than I used to, so now I play it more. Sometimes I still miss it, and sometimes I play it when I'm not supposed to - so at least it's a more even distribution of error.

The other moves I considered were backing off and hane inside at 8-11. Backing off looks like it makes my upper wall inefficient, and hane inside looks like I'm fighting for no reason. Hane feels like the only move.

What will he do next? I'm not exactly sure, but he'll definitely try to build his left side a bit- and I'll definitely stop him :razz:

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #196 Posted: Fri Jan 17, 2020 12:19 pm 
Lives with ko
User avatar

Posts: 192
Liked others: 39
Was liked: 28
Rank: EGF 2k
KGS: MKyle
8(h)-11

Comment
There is a lot of uncertainty for me but I think this game is currently close (maybe I will look a fool for saying that?)
I'm pretty proud at getting to this stage with a game that I'm not clearly loosing to be honest.
It seems reasonable to count 50 for black at the top. Black has 5 extra captures and maybe 10 points or so with the two other groups. I have komi, 29 points on the right (sticking with that) and almost certainly the same on the left I'd say. A lot of boundaries are actually already settled but I would favour black in this game on the basis that the stronger player usually has a pleasant time in the endgame.

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #197 Posted: Sat Jan 18, 2020 3:54 am 
Lives in gote

Posts: 494
Liked others: 1
Was liked: 148
Rank: KGS 2k
GD Posts: 100
KGS: Tryss
Diagram


Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #198 Posted: Sat Jan 18, 2020 9:54 am 
Lives with ko
User avatar

Posts: 192
Liked others: 39
Was liked: 28
Rank: EGF 2k
KGS: MKyle
Thanks to those posting diagrams. Should make it easier when we look back over the game. There is also great comedy value in the red warning text
Diagram
:lol:

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #199 Posted: Thu Jan 23, 2020 3:01 am 
Lives with ko
User avatar

Posts: 192
Liked others: 39
Was liked: 28
Rank: EGF 2k
KGS: MKyle
Malkovich stuff
Since we're not so far from the end, I figured I might attempt a blind review. When we are all done I'll take a look at an actual sgf/diagram/board and/or review with AI, but I think it will be beneficial to try to analyse myself with hindsight first. I may well get move order, coordinates or even moves wrong. I'll need to split it into parts.

Part 1
B 16-16
W 4-4
B 16-4
W 4-16
B 6-3
W 3-6
Straightforward modern fast go game so far.
B 7-6
Interesting move. I'm sure AI will hate it because we played above the 4th line without a fight going on but AI misses the value of moves like this in amateur games. One of the drawbacks that AI would be particularly sensitive to is that black has spent two moves and has no base. I extracted a tiny bit of value from this (and may yet get a little more) but I would expect AI sequences would exploit this better.
W 14-3
Seems reasonable. Want to play in this space and I'm probably better not poking at the jump just yet.
B 12-3
Reasonable enough. A wider pincer still leaves space and problems with the left so why not get the right hand side a bit thicker.
W 17-3
Other options feel like they might help black fix up. AI will probably think this move is fine (maybe not top choice?) but it's more difficult for humans to handle the moyo it encourages.
B 16-3
W 17-4
B 16-5
W 16-2
B 15-2
W 17-2
All Joseki. Wasn't sensible for black to block the other side so there was almost no choice.
B 15-3
Again, bots will dislike this but for amateurs surely it's fine. The drawback seems to be that it leaves black a little bit heavy. Although I certainly didn't exploit this correctly, playing this way did make my results on this bottom side possible and I was pleased with how this went.
W 17-6
B 15-7
This move looks a little bit strange when black plays the 16-6 exchange which I guess is why I forgot it was there! This is probably the right move though.
W 10-4
I think I should have done something more technical here. Maybe directly poked at the thin jump on the left and/or attached to the pincer stone directly
B 10-6
Surely good. Tough for me now. On AI analysis I'm sure there are some wild swings here as we will both make mistakes in this sequence.
W 14-2
B 14-1
W 13-1
B 15-1
W 14-5
B 16-6
W 17-7
I'm happy to fix problems on the right but I overlooked how comfortable this makes black in the centre. Very curious what AI will suggest around here.
B 12-5
In hindsight perhaps a little thin again, but it looks good so it doesn't feel like I can say it's a bad move.
W 9-6
Slightly desperate thrashing around. Almost certainly bad.
B 9-7
W 11-3
B 13-2
Perhaps 12-2 would give a couple more forcing moves but possibly leave less effective aji? Idk
W 11-6
B 11-5
There were moves that I was really concerned about here but I can't seem to read it now. Maybe just 10-7 leaves me with no real forcing moves and lots of problems?
W 10-5
B 10-7
W 12-4
B 13-4
W 12-2
B 13-3
At the time I think I thought this was the mistake but now I'm really unsure. Perhaps capturing at 11-4 for ko is reasonable, but my local threats seem okay and when I win the ko black may be in danger. Tough choice.
W 13-5
B 14-4
W 12-6
I was probably too happy about this capture. I'm still not alive here in the endgame.
B 8-4
Maybe black could have captured with 9-5 to make this group thicker towards the centre?
W 8-5
B 7-5
W 12-8
At the time I think I was thinking about this move as a nice honte settling luxury move and definitely thinking about tenuki for the top side. Now this seems pretty huge for the relative strength of groups. I'm glad I played it

Top
 Profile  
 
Offline
 Post subject: Re: MikeKyle v. TelegraphGo (Blind)
Post #200 Posted: Fri Jan 31, 2020 7:38 am 
Lives with ko
User avatar

Posts: 192
Liked others: 39
Was liked: 28
Rank: EGF 2k
KGS: MKyle
Continued Malkovich stuff
Review part 2
B 6-17
W 3-14
I considered playing 4-14 instead which just seems weird to me now. I think I was more afraid of the centre black stones than I am now. Low seems fine. I don't think a shoulder hit works well at all for the immediate future as black isn't quite alive anywhere.
B 4-2
W 3-3
Again, I briefly considered other moves here (once I correctly read what blacks move was) but in hindsight this seems important for the relative strength of groups.
B 4-17
W 3-17
B 3-18
W 5-17
I didn't give this much thought but maybe this is the right ocasion to play 2-17 or another of these modern things that is more likely to take the full corner back? I think my move is basically fine but it does work towards the left hand side which is difficult to develop due to black's centre stones
B 4-18
W 5-16
The alternative 5-18 looks interesting to me now. Maybe it would just invite complications that I didn't need.
B 5-18
W 2-17
B 6-16
W 7-18
I'm not so sure about the timing of this peep now. when black has a hoshi stone in the oposite corner this pattern makes a nice formation across the top. It's probably playable for white but I might prefer to just approach at 14-17 in hindsight.
B 8-17
W 8-18
Maybe here is another reasonable place to tenuki and do something in the top left.
b 3-16
w 2-16
b 10-17
w 7-17
b 4-15
w 5-15
b 8-16
w 6-18
b 9-18
w 2-18
b 7-16
I can't see any good branches from this sequence. Seems like a one way road.
w 17-17
Not sure if it's inconsistent to tenuki now - maybe I should have tenukid earlier if I didn't want to save the cutting stones right away. This move seems reasonable though. I might also have considered 17-14.
b 16-17
w 17-16
b 17-14
w 17-15
b 16-15
w 18-14
b 18-13
At the time I thought this black move was unusual but now I think it's just my lack of recent playing experience. I think b17-13 would have been fine.
w 17-13
This is still the best looking move I can see. I don't think atari the other way and hane connect does much for me. This feels like a weird little blind spot and I'll be interested to investigate when we're done.
b 16-14
w 18-12
b 17-18
If black had played 16-13 here I think I would have had to play 17-12. Surely then it's a big mistake not to make this exchange?
w 19-13
(the move that my oponent thought was 3-19 i believe. I can absolutely understand mistakes in reading the coordinate system)
b 18-17
w 18-16
b 19-16
w 18-15
How to judge this result!? I think it's fairly good for white. White has okay prospects on the right and is fairly thick. localy white should get the corner as that ko is so risky (although globaly black is so confident in the ko)
Black is kind of thick too. If he plays around 15-12 then the moyo looks really good.
b 12-11
This is exciting but I really feel like the more straightforward 15-12 does a better job? This takes less teritory at the top but I guess is slightly better for left centre development? I wonder if black is trying to build an attack on the bottom side white group? If so this feels like a very minor threat to me.
w 15-12
I guess if I was more worried I could have gone a bit deeper but this seems fine.
b 13-13
w 14-13
Urgh. start of my terrible aji keshi stuff
b 13-14
w 13-12
b 12-12
w 14-14
Argh!! this is stupidly early. This should be my sente at almost any time and black is a long way away from considering playing anything reverse sente here. I think I just played it so soon because I didn't seem like there was really any aji (I really didn't think black could play the ko) and the game feels more comfortable if things are settled. There's a chance that theres a big AI % drop at this move but it probably depends if AI thinks black can fight the ko or not. There's also a chance that AI unrealistically thinks I'm far ahead here because the AI has less respect for moyo.
b 14-15
w 10-8
I thought this was really slow around the time but it seems okay to me now. Not great but reasonable enough.
B 15-10
I guess maybe this is a response to my move in the sense that it stops moves around 10-2 being sente? At the time I wished I'd played on the right but it feels slightly slow now?
W 5-3
Rubbish! My most stupid feeling move. With b 15-7 on the board it just doesn't work but even if it wasn't there I'm not sure it's a very good idea.
B 5-2
Black is very happy to do this. I think I'm correct in believing that black would have to play 6-2 without b15-7 on the board?
W 6-4
B 7-4
W 7-3
B 6-2
W 8-2
B 9-3
W 10-2
Not sente against right group :/
B 9-2
W 15-5
Thought b15-7 wasn't there => thought black couldn't link up and this was the killing move. At this move I guess I could have played something around 15-8. I'm not 100% how this would go but maybe I could force black to have to go back and play 15-5 to live and take a nice right hand side? Not much different to the result I got though maybe.
B 14-9
Seems to leave aji at 13-10. Not sure. Maybe it's safe?
W 16-7
B 15-6
W 8-7
Here's where I tried to play on top of my opponent's stone at 15-7! When I discovered that there was a black stone here I trusted that black was connected up fine and played away. Now I'm wondering if 13-10 works. I can't see a refutation. With b14-11 w13-11 b14-12 w13-9.. black can capture 2 stones but then white can still cut?
Will continue reviewing in a later post.

Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 213 posts ]  Go to page Previous  1 ... 7, 8, 9, 10, 11  Next

All times are UTC - 8 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group