Hex 13x13 Openings

Thanks very much! You might be interested in the chart Alan Hensel made for Twixt, not because I want you to play Twixt, but for the way he has arranged the info. Of course a Hex chart would require half the board instead of one quadrant. Maybe someone could arrange the data in a similar fashion. How lazy can I get?

Maybe by now a 10x10 swap map has been generated. I emailed professor Hayward about it.

I'm just thinking out loud here...

A5 is very strong because it combines with b3 so it vastly improves a long diagonal play.

A9 is very strong because it combines with b10 so it vastly extends bottom side towards left.

A4 does some nasty tricks combining with either b4 or b5 - I have extensively explored it playing 11x11 on IGGC and A4 is a monster opening.

A12, apart from combining with b10 I don't know why is so strong. In my book it should be of similiar strength to A9...

C12 just sits there looking silly - I would expect lower percentage as C12 provides very little help in any imaginable line of play.

A11 I don't know why so weak. I used to think white can play 2.b12, but not really. In other case it looks pretty equal.

Others I would value A6=A7=A8=A13 slightly stronger than equal.

A3=C2 slightly weaker than equal. Notice that C2 is provably stronger than B2 and A2 as shalev has once shown.

I3>G3>H3 might be only slightly stronger than equal as well. 11x11 3rd line equivalents are all winning openings if you trust mine and lazyplayer's practical evaluation.

Ok I've uploaded the most recent changes so you can view the interactive diagram yourself now. There are two view options: with swap (win percentages are just for black, if white swaps and wins thats a loss) or without (games are as though they were never swapped, so win percentages are for the move itself, this is the diagram I posted above, except without the 20 game cutoff).

Note that you can click on a move to see the responses. Arek you can see the responses to A11 for example, without swap B12 has a 76% win percentage.

Cheers

My thoughts:

A4 and A5 are quite strong.

C2>B2>A2, and C2 is a little weak, for reasons that seem to me to be specific to 13x13. I actually expect C2 to be stronger on 12x12 or 14x14, but I could be wrong.

A3 is weak, probably for reasons that don't depend on board size. On 9x9, A3 is known to lose while A2 is known to win, so A3 can even be weaker than A2.

A11 and A12 I don't have much experience with, mostly because I assume A11 is really weak and A12 is really strong. I might be wrong though.

C12 is weak.

A9 I don't think is as strong as the data implies. I've been playing it recently.

A6, A7, A10 are all a little strong. In fact, on 13x13 it feels like all of the A edge is strong (between A4 and A10). However, the central A edge pieces are hard to play, and A10 in particular seems interesting. I might switch to it in the future.

The second row pieces D2 to K2 are all weak. The third row pieces B3 to L3 are all strong.

Do you think the difference in the weak second row and strong third row is too dramatic for either side to be useful? Obviously every opening is either winning or losing but some are so close that they make good 50% openings. How sure are we that there are no good openings along black's edge?

Also what about B11? It neighbours a very strong and very weak opening

@shalev

Third row pieces are not AS strong as some might have thought. I played them on 11x11 and they're winning but games are interesting. On 13x13 things might get even more interesting.

I'm never gonna agree a3 is very weak. A3 is my first suspect to be the most equal opening of all on 13x13. In my practice I've noticed all the A4-A10 openings are strongly winning so even though A3 is losing it might be a better choice.

Now this looks like my new quest :)

So far I was thinking black 5.e6 wins and it has actually not been played! (similar idea I played here and lost, but I'm not proud of my following moves)

From the attempts above I only like Daniel's h6 and given that he played vs gzero it's possible that he let his win go.

Another idea is black 5.b10. So there are 3 solid choices that need investigation :)