basic opening principles -- white playing j2 Hex, Havannah
14 replies. Last post: 2016-08-31Reply to this topic Return to forum
wccanard at 2016-06-08
I’ve taken a break from theorising about hex and have tried to devote some time into becoming a better player. I want to understand the opening better. I still see experts playing moves that I think are very strange. Here’s an example.
Maciej plays 8 j2:
Until this point I felt that I had some sort of idea what was going on. The first four moves (a3,i10,k10,k12) look like some standard opening. After that black plays 5 e9 making some weak connection to the bottom edge (it seems that nowadays j12 is preferred but this was 2010). After an exchange near the bottom edge which again looks reasonable to my inexpert eyes, white then plays 8 j2. This is presumably a defensive move but it seems so weak — it is so far from the black stones and also so far from the white edge in the sense that it’s so near the black edge. If white wants to play defensively whilst maintaining some sort of attempted connection to the right edge they could play something like i5 perhaps? On the other hand I’ve seen this j2 move played in other games between experts — what are its merits? Is the point that it stops moves like black j3 and h4 from being certain connections?
Would an expert now play something different on move 8? Are there other reasonable moves or are we still in the sort of position where experts can see that there are only one or two reasonable options?
HappyHippo at 2016-06-08
I used to be pretty good until I stopped playing for a few years, so I’ll try to weigh in until someone actually good gives it a shot...
Often in Hex a play comes down the short diagonal and is blocked into a ladder heading towards the acute corner (something like this, although there are many variations). The reason is that if you make a play on the short diagonal your opponent can’t stop it with close blocks, otherwise you can just connect: http://www.trmph.com/hex/board#13,i5i4j4j3k3k2l2 Nobody plays that out of course, but the point is the threat of it means the defender is forced to block a step away, which gives you the ladder (even further from your edge, where there probably wouldn’t be a true ladder, I still feel the situation is advantageous to the attacker).
At this position (before j2) Black would probably like to play a central move that would connect to the bottom and then use the a3 piece to escape the ensuing ladder when blocked down the short diagonal (a3 can help third row ladders like on this page, although I would probably make a few changes to that page). By playing at j2 white effectively shortens black’s top edge (notice how there isn’t even enough room to fit the third row template IIIa in to the right of j2). Now, to black, the short diagonal has been moved to the left three columns. You can even see on the next move that Black plays e6, which lies on this new shifted short diagonal. But now it’s much further from his pieces that are connecting to the bottom than it would have been had he been able to play on the true short diagonal.
On the next move white plays e7. Notice again that if you play out the naive close block (f7, f6, g6, g5...) white can use the j2 piece to get himself a fourth row ladder which is escaped by i10-k12. So black is forced to block at g6. Then that allows to white play g9 which is pretty strong towards the right and threatens to connect to either e7 or e11.
Anyway I hope some of that makes sense. There’s probably a lot of details I’m missing. However I would also offer some general advice which I apply when trying to improve in any game: try to look at games of people who are no more than a few hundred rating points above you. There’s just too much going on to try and figure out the motivation behind the moves of players who are so much further ahead of you – they seem mysterious. Games by players who are just a tier ahead of you have lots of lessons to teach that you can integrate into your current knowledge, and once you absorb those lessons you’ll be better prepared to understand games at the next tier.
wccanard at 2016-06-09
Bill LeBoeuf ★ at 2016-06-14
My thanks as well HappyHippo, very nice explanations!
Arek Kulczycki at 2016-06-17
In fact there are two separate questions here: Why did white play there 1) in local terms, 2) in global terms.
1) J2 is a spot that is furthest from white edge whilst being connected to it (there is nothing black can do to prevent a ladder from j2 to bottom-right corner). Above “being furthest” is good because of kind of expansion of white territory – white has many more directions to go.
2) This is a much harder question – question of what area to play in. The strange move e11 didn’t offer any local continuation, so for #8 there was an obvious top-left long diagonal area to play and also some interesting moves in center. To be honest I consider the j2 move an escape from lack of other ideas because it is very safe, you can use it with both diagonals and center. When there is too many to choose from it’s hard to pick the best option, but I think usually there is a better option then j2. Don’t take the easy way, be creative! :)
wccanard at 2016-06-18
Oh! I hadn’t appreciated that j2 was connected to the right. But now I see it is. I guess this is a standard fact and probably explains why the i10 k10 k12 moves are so common — this position is a ladder escape for 2nd, 3rd and 4th rows.
Why does Arek think 6.e11 is strange? What would Maciej play now? Has opening theory changed a lot since 2010? I don’t know much about chess but my impression is that not much has changed over the first 6 moves since 2010 (although I appreciate that there are fewer moves in chess and it is played a lot more, for some reason, probably just historical ;-) ). Whereas in newer games opening theory seems to change a lot — I used to be quite good at 5x5 dots and boxes and I think one of the reasons I was so good was that there was essentially no opening theory at all when I used to play it! But loony tells me that now new ideas are appearing in dots and boxes opening theory.
I see a3 i10 k10 k12 so much in games between good players, I would imagine that there are standard 5th moves and responses nowadays, but so many people now play 5 j12 that it’s hard to get information about what to do against other moves.
scrampy at 2016-06-18
Thanks for this thread! As usual, I’m struggling to follow along, but I’m trying.
Thank you HappyHippo for the clear explanation – there is a lot of dialogue on this forum, but the majority of it is of poor instructional quality.
Is it correct that J2 is only connected to the right side because of K12? Otherwise http://www.trmph.com/hex/board#13,a3i10k10k12e9e11g10j2k3l1m1l2m2 this would happen. Is it correct that without the support of K12 or an equivalent piece, J2 (on its own) is not connected to the right?
scrampy at 2016-06-18
Also – if J2 is good, is it obviously better than other moves on J? I tend to stay closer to the centre when I play – are J4, J5, etc. obviously inferior?
HappyHippo at 2016-06-18
scrampy – yes, k12 serves as a second row ladder escape, and along with i10 forms the third and fourth row ladder escapes.
wccanard at 2016-06-19
Scrampy — here are my possibly amateurish thoughts about your questions. First, as HappyHippo says, yes absolutely K12 is needed for that J2 connection. As for other J moves — J2 cuts off more of the top row, making it harder for black to connect to the top. The move makes a lot more sense to me now I’ve realised the (in retrospect obvious now) fact that J2 is connected to the right. I guess the way to think about it is that in some sense any move on the J column up there has pretty much equal influence on the right hand area (because the ladder can be as long as it needs to be), but J2 has a lot more influence on the top. I guess the thing about J5 is that it is actually connected to the right independent of the pieces in the bottom right, but I think a general principle is to make as weak a connection as possible, and that’s just what J2 does. White has already invested two pieces in the bottom right and it would be a shame not to use them by making a strong connection like J5 — it almost makes them pointless pieces. Maybe that’s one way to think about it. A move like J4 is so strong that you’ll guarantee a second row ladder, but J2 only guarantees a 4th row ladder — however a 4th row ladder is all you need, so again it’s following this basic principle of making the weakest connection. If K12 or I10 were not there it would be a very different story. For example if white I10 were not there and white just had K12 then I suspect J4 would be a more sensible choice than J2 because it guarantees a second row ladder and then K12 can be used as the escape.
Arek Kulczycki at 2016-06-21
scrampy, yes, j2 is better than other moves on j.
There are many ways of thinking about it, one way is that making “weak” connections is better, as wccanard mentioned. Kind of true.
My way of thinking about it is different. Suppose there is one point on the board that I want to connect to right edge. Then I imagine directions (straight lines) in which I can go from the point to the edge. The wider is my range of directions the harder it is for my opponent to prevent the connection. Hence j2 is the strongest choice. Well, at least to me it makes sense.
Arek Kulczycki at 2016-06-21
wccanard, e11 is a move of no practical purpose. Black only has to care for possible 8.e10 and it solves all sorts of problems.
lazyplayer at 2016-08-31
i think HappyHippo explained it best when he said "By playing at j2 white effectively shortens black’s top edge (notice how there isn’t even enough room to fit the third row template IIIa in to the right of j2).". But naturally, this argument is valid only insofar white has valid replies against potential black K2 or black L2 (L2 is more interesting than it looks, and it leads to complications like j2l2k3l5l3j3i3k2i4j4i5k5). So naturally Arek is also right when he says that that the connection to right is important. Other common replies to J2 are J3 and I3, and J3 is often followed by I3. If I3 at this point is also connected to right, it’s naturally much better for white than if it is not. Finally another very serious reply to J2 is simply J4. The strength of J2 is basically that beside these 4 moves, black has no other LOCAL choices that are likely to do better than these. Its main weakness is that if black can connect stones in the top-right zone to the top via top-left (for example, via 9 C5), then J2 becomes practically a completely useless stone. The secondary weakness is if black can cut the “invisible link” between J2 and bottom-right that keeps J2 connected to right. This is hard to do because it is in the J column, and so black has to work hard to achieve this without simply losing a tempo.
And by the way, i invite players to play “infinity” here, there is also hex 11x11, where one can see some basics elements...
lazyplayer at 2016-08-31
And beside 9 C5, other obvious moves that could be interesting are 9 C9 or 9 C8 because they take advantage of distance of J2 from the left side AND also of its “orientation” toward center and bottom-left. Surely i don’t like 9 e6 because 14 f4 just kills it as Maciej has proved very well.