game 1742138 Shalev vs Arek

I thought 45.  H5 was an immediate win for Shalev. What did I miss?

The game lasted 41 moves so I'm not sure which moment you mean, but at any time of the game I had this line http://www.trmph.com/hex/board#13,m1m1j9j11k11e4h6k2d4d5e8f7c7f3h4h7b10e7j7i8g7f9g8g10h9i10k10j12k12i9k6l7j8j10k9i6h8i7f8d12c9h5i4i5g5g6b7.

Soon I will post a full comment on this game.

My screen is not showing all your moves.  What do you reply to 42. H5 ?

A direct link to the game seems convenient.

direct link to my line

Very nice !

How about after 40. H5 i4 i5 b7 d7 c5 c9 b9 c11 ?

By the way, how do you set up a direct link? Does it involve one of the share image boxes?

v

Oh, then 49. D10 wins :)

There I go with the full comment.

First things first I have to say I was expecting that game against shalev would decide the tournament winner. If I could choose I would opt to make first move against Daniel Sepczuk and therefore make the second move against shalev. Things turned out opposite.

1) OK then, I opened as usual with a13/m1 and obviously shalev swapped, so I had to be playing a losing position.

2-6) I played my default moves again and my opponent made no mistake as we took over one corner each. Against m1 I prefer to play on empty bottom side because the top is "shorter" and I may need to play closer than 5th line from the edge.

7) Well, believe it or not, I'm 100% positive that position after 6 is white win and with shalev as opponent there is not much room to rescue this game anymore, so next time I definitely change a13. Knowing that I've played a losing game I already had to look for messy lines instead of solid stuff. H6 is a move that is not particulary well placed, but is threatening fast bridge connections and cannot be belittled any time soon. It's a move that I'm most proud of after this game.

8-10) Each of the moves actually deserves separate comment.

k2: I don't know what to make of this move. It is very strong, great choice by shalev... but is it THE BEST choice? On one hand it allows strong connections both from top-left and bottom-right, yet I felt white maybe lost part of his advantage. The reason is that earlier I had dillemas of how to connect to top. After k2 I got some free information: I had to play close to the top edge to maintain connection!

d4: At this time I knew how to challenge the top-left corner - I didn't need escape on top anymore, I needed most possible help to connect bottom. The idea was this (or even this if shalev was so kind to play like I wanted) which provides connection to top through g4 and also threatens black c7->b6.

d5: Surprise! shalev refuses to play "normal" e3 and possibly he is correct. At a cost that is hard to evaluate he maintains an option to play future f3. Who was winning in this position I have no idea.

11-13) Blunders. Lack of white e3 allowed me to delay my connection to top and I wanted to capitalize on this chance. Unfortunately I miscalculated and didn't notice that white h4 was direct white win. Shame on us because shalev didn't notice it either. I continued with the threats I had in mind that is c7. At this point white h4 was still playable, but black could maybe find a way out. (the last line I did not check thoroughly)

14-17) f3 is a logical move but not scary at all. If shalev was very courteous I could benefit greatly from resolving the corner. Instead I just take h4 and cherish still solid connection to top. Indeed at this point shalev tried to stop me on bottom. I don't want to go over many complex lines after a direct play, but trust me, white win that mess. I had a couple of options but b10 composes well with all I had and it effectively cancels any attempt of white connection to left other than d5. At this point I've revived my confidence that I could win the game.

18-19) Still not enough power to play directly to bottom.  If19.j7 is a winning move then I assume it is the only one, because it gets me free K8 or K6 both of which are solidifying my corner.

20-41) Let's bring it on! It was a mess so I guess shalev thought he had called my bluff, but I had a winning hand after all.

Arek, thank you so much for your awesome and also entertaining analysis of this game.  Shalev, I will look forward hopefully to your  analysis as well :)

Thanks for the great analysis.  Very interesting.

Thanks for your review Arek! I will write down some of my thoughts as well.

Ok, here's my analysis.

1-2. I always view this first move as way too strong (meaning swapping is an easy decision), and yet I always lose to Arek after swapping it. I still swapped it, but I was a little scared, having lost from this position at least twice before.

3-6. Standard stuff. I always tenuki (play elsewhere) after k11, because I prefer to leave my options open rather than finish that corner.

7 h6. Arek chose to tenuki as well. c4 is the usual move here. H6 is very strange; I wasn't sure if Arek planned this in advance from the first move, or if this is just an attempt to complicate the position.

8. k2. I wasn't sure what to do in this position - I considered h5, which blocks his h6 stone and forces him to connect elsewhere, but I thought that something like this might grant Arek's wish of complicating the position. I want to simplify things. I went with a MoHex-style move, giving me an escape in the obtuse corner.

A quick interlude into some Hex theory. In the opening, it is usually desirable to establish escapes on your sides (stones that will later help you connect ladders). Escape patterns are usually played on the acute corner, because this is more efficient. However, in this case I already have the m1 stone on the right side - this already gives me a partial escape. It's a waste to spend effort on a separate escape in the other corner. The k2 move lets me strengthen the m1 escape instead, which is arguably cheaper.

9-10. Man, I hate nonstandard corner patterns. I don't remember my calculations here exactly, but I think I decided against 10. e3 because of the lines Arek mentioned above.

11-12. As Arek points out, those were blunders. I was too concerned with how to play the bottom to notice just how strong things were at the top.

13-14. At the time I felt comfortable here, but in hindsight I'm not sure that f3 is all that great. The problem is b7 - or rather, the problem is that Arek can threaten b7, without necessarily playing it. If he played it immediately, I could respond with f2, which would be okay since it gives me some strength at the top. But Arek can wait, and only play b7 after the top has already been settled.

15-16. I still thought I was doing fine here. most straightforward paths I tried lead to me winning.

17. b10. This is a strong move, especially since it combines well with the b7 threat.

18. e7. Most calculations I made seemed to indicate I'm still fine here. I didn't have the time to fully convince myself, but it seemed like I would probably win unless Arek uncovers some brilliant path that I missed. It's possible there was something better - I didn't have the time to give this move the analysis it deserved.

19. j7. I thought this was a desperate final attempt, and didn't fully realize its strength.

20. i8. Seemed like a simple enough counter. I was sure I won.

21-38. Not much here. We played these moves quickly.

39. f8. Damn, I missed this move! It turns out I lose after all. I now wonder if there was something better at move 20... does anyone see anything?

shalev, would 38.f11 be any good?

@ypercube, I don't think so

@shalev, this is the way! it was one of the main threats throughout the game :)

@ypercube, proper realization of your idea would be e11 instead of f11, also maybe better at another moment. I was very affraid of this move, but probably it would still be losing for white.

as for #20 I couldn't see anything to rescue white. Surely at #18 and #16 there is much room to look for improvements. Maybe you had to play j12 earlier. In truth I'm the most curious if I could be ahead if played good #11. (another time I would play e6 or i4)

@Arek, oops, you're right, though it only changes the b7 stone.

I agree that I would probably need to go back to move 18 to have a chance of saving the game.

@arek and shalev: Thanks for your detailed comments on this interesting game!

Personally I also admired areks move17.b10. I was wondering at that point how shalev would counter, because I couldn't get a good result for him. I also missed the blunder at move 11 to be countered with h4, but I wasn't paying attention, in that fase of the game. I guess I would need such a blunder against the both of you, to be able to win of you two at the moment :-). Thanks for your high quality and interesting games!

Thank you very much, shalev and arek, for interesting game and thorough analysys. It's really helpful, when strong players comment such games. So I ask you a permission to post this game in Hexmasters or in Hexwiki.

I thank you both for your counterpoint analysis of this very interesting game. Shalev  I especially appreciated your comments on the concept of tenuki  as applied to  hex.   I do think that posting this game and these analyses of them will increase interest in the game of hex.

Arek it looks like  The first ranked player gets the favorite position over the second ranked player in a tournament. So in the next tournament it looks like you are again going to have to move first against Shalev.

Hm, if you're right Bill then it depends on the result of my present game against Daniel :)

Btw, apart from tenuki and joseki I found another 2 terms from Go that could be transferable to Hex: kikashi and hamete. I'm not saying these are perfect examples, but I tried to stay with the game of this thread :)

very nice  examples Arek :)

I also sometimes think in terms of liberties. For example, here I "increased my liberties by 2" in my game against Daniel. If that move increased my liberties by only 1, I might not have played it.

shalev, interesting point! I'm not a Go player so the same thing I intuitively call "cutting opponent's line" :)

shalev, this is the win you missed at 24... overall it seems you both agreed white should win, but you both missed it :)

(like you both missed h4 earlier!)

About the opening, i've a feeling that this is the right way to play after white m1 and black j9...

but i just love "tenuki" and messing up so this is just an idea that should be checked in play... :-)

On earlier moves, i think that due to the risk (for black) of this it should be better (for black) to play this...

Anyway even with this white still likely to have a win somewhere...

well, my last idea is wrong, white wins with this

anyway i think the idea is right, black should play c5 as soon as possible...

Lazy, thanks for your insights on this game.  Much appreciated !

lazyplayer, well I think you're wrong about 24!

lazyplayer.. and about c5 I think you're wrong too ;) before black plays c5 it is his choice if he wants b7 or b6. After c5 white can choose it by playing (immediately or not) either f2 or... No, I cancel it, white can just immediately play this and win.

Arek, no on 24 i'm not wrong. And the idea of c5 is to get both b7 and b6, if possible. And if it is not possible, it's lost anyway...

Here is the link on 24

Ok, my proposal on 24 doesn't work, wait a moment... i've forgot why i had concluded it was hopeless...

Probably this

Eheh no the same, sorry :D but this should work :D

I'm not sure about your objection to my early c5... what you do as white here?

c5 - continue as black

24 is complex, but to me it seems like black win

I don't see the black win in your proposal...

about my c5 proposal, look at this

then maybe late c5... Just seems black win to me, but maybe I'm wrong

on second topic: stop playing bad as white :)

you're the one playing bad as white? :P

i'm looking at your late c5 idea...

well c5 there seem to win, but h5 likely wrong there... this seem ok

On move 11... what about this?

Lazyplayer, for early c5, maybe white can counter with this? It does seem pretty fragile though - I probably missed some line for black...

late c5 again I think

move 11 I don't analyze, solution is out of our reach

early c5 refutation (real this time I hope :) )

@shalev, maybe, but i5 is a source of some problems

i5 is countered by g5

arek, there is solution to "late c5" :D

and there refutation of your refutation of "early c5" ;)

shalev, maybe something like this?

shalev, also maybe this is valid? hehe, this is crazy...

lazyplayer, wow, that looks valid. Very impressive.