Overview

The team match will start on 2011-06-10 as described in the rules with the following teams:

white: Carroll, Thierry Pertuy, maraca

black: upendra, David J Bush, Klaus Hußmanns

Tools: Jtwixt

Jump to move (the reason why the game starts in a new thread):

1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40

Move 1

(Somehow the anchor was removed, I hope an alternative works on move 2)

I think Alan's first move chart on http://twixt.wetpaint.com/ is very good. If we examine first league games, then it seems to be somewhere in the middle of the opinions of the players what a good starting move is:

I would say there are too categories of starting moves:

1) those along the more horizontal diagonal and

2) those in the 'opponent's territory'.

2) is not often played and more difficult to handle. While a common answer to 1) is J10 (or a mirrored position), there seems to be no standard answer for category 2). You can look at the moves of category 1) like shifting the diagonals, that's the reason I think why D3 is a little bit too strong, it shifts 2 diagonals and not only 1, when we go closer to the middle the moves have more influence towards the center, but maybe don't help that much in the corner.

I propose the following moves for category 1): J4, I5, K6

and if we want to go for category 2): C10

What do you think?

Sorry G4 not J4

Hi,

I don't understand the 1-24 row in this Alan's chart. It seems nosense to me playing there, i.e. above or below our own line.

I think I would have prefered a very classical D3 first move for educational reason.

But it makes sense playing the most neutral first peg. So I am ok with G4.

As for playing a 1) or a 2) category move, I'd say that I would play a 1) category against a player with a lower rating, and a 2) one against a player with a better rating. In such a case the goal would be to drive the adversary in some unusual game configuration, admiting that this could be an advantage (what I'm not sure of...)

(Carroll told me he won't be here this week and we can decide which move to play)

The 1-24 row is there for completeness. I think it's pretty safe to say all moves smaller or equal to 45 are losing moves and all moves bigger or equal than 55 are winning moves, so you can ignore them.

My thoughts about when to play a move of which category are about the same as yours, however my experience with that strategy wasn't too good :) I would go for a category 1) move here too.

If you prefer to play a 'standard' game, then like you say G4 would probably be the best choice (the modern D3 ;-). This is actually still my favourite first move. I suggested I5 in case you think G4 is too strong. K6 or L6 would already be slightly experimental moves, because moves that close to the middle are also rarely played.

So we go with G4?

(@ the other team: I'll post the Move 2 header when we're done and a Jtwixt file if I find a place to host it)

I have no experience with G4, and as you may already know I'm not very experimented in Twixt as I really play it only for 5 months. So I'm not able to say if it is too strong or not.

So Ok. Let us go with G4 ! And will see if the other team feels it too strong or not...

Ok then we go with G4!

I don't have that much time right now, but I think all of us understand how the current position looks like without an image ;-)

(Btw. you can also give comments on our turn, it's not forbidden ;-)

Move 2

First we should decide whether to swap or not. I would prefer to swap, but whatever the majority decides is fine with me. Speaking of which, are we going with majority rule here?

I'm sure Thierry already know this, but since his remark could be misinterpreted, it bears explanation.

Although playing behind your own line on Move 1 is surely bad, it's not useless in the sense that passing would be if legal. For instance, a latter move could simultaneously connect to both that starting peg and to a peg closer to the middle of the board. It could do that in a situation where there was no other way to connect the edge to that more central peg.

@David J Bush: I thought you wanna swap :) The majority rule only kicks in if you couldn't agree on a move, here it might be difficult to convince the others why you wanna swap, but on later positions you can certainly bring some arguments why to play a certain move.

Would you also agree with the chart that D3 is stronger than G4?

@Dvd: that's right. And I'm 100% sure that playing behind your own line is a losing move. It is already on 6x6 and the border between losing moves and winning moves has the tendency to go downwards and to the right ( solved boards).

I believe D3 is stronger than G4, but of course I don't know.

We could use an empty board to examine variations for this game.

I prefer to swap too! G4 is too near to the center for leaving it to our opponents.

Hi everyone - sorry to be late to the party, didn't realize it started. :)

(And let me issue this disclaimer once, and it applies to the rest of my posts - I just started playing in the past couple months, so I have no idea what I'm talking about. :)

I would say swap:

1st mover advantage and having an extra peg on the board is of course extremely powerful *if* the first peg comes into play; so another way of asking the question is what is the chances that G4 will come into play later in the game. Either directly, or as a deterrent than prevents certain lines of play.

I would think that G4 would mean that we likely would win any corner battle on the upper left - it pushes the Achilles "heel area" south by 2 pegs, I think.

Of course if the probabilities of Alan's first move chart applies to this game, and the winning odds of G4 is 46% less than half, we shouldn't swap. But given the environment of this game, with the open thread, this isn't a typical game. :)

Anyway, I say swap!

Downloading Jtwixt for the first time; David, Klaus - are you mapping the game in JTwixt? Any way to share your files and notes and annotations?

@upendra, you can share them via save-files, but I would prefer if you use the empty board like David suggested. I think I'm only gonna use the commentator, my eyes are used to that layout.

Btw. the first move often already can be used when the 3rd peg is played. For example: 1. G4, 2. J10, 3. N11. Without the peg in the top left corner N11 would only be semi-connected, but with G4 it is more or less connected to the top.

The "empty board" link above is to Alan's servegame site. It is not possible to make a swap move on that board. I propose we swap sides, so now Klaus, upendra and myself are now white, and maraca Carroll and Thierry Pertuy are black.

Is that okay?

@Dvd: Of course you are right. I'm just surprised that some people did try to start there and that some statistics can be done.

@Klaus: I was sure you will swap :-)

@upendra: As maraca said, one peg can have an influence very far away on the board because of ladders.

Ok. So this our turn again...

What about the classic j10 ? (classic answer to d3 - no idea at what could be the classic answer to g4 if any). j10 cannot be attacked with a 2-0 move because we can reach the diagonals going to the top or the bottom. I haven't experimented a lot on the second peg (and when I did, it wasn't very successful...)

Now, is it too right ? Well I feel (more than I think...) it isn't. Is it too left ? Well I don't know.

I must admit that the game is so open that I cannot really answer and that any real prospection is too wide to me...

Move 3

Move 2 was swap, color update:

white: upendra, David J Bush, Klaus Hußmanns

black: Carroll, Thierry Pertuy, maraca

Image Hosted by ImageShack.us

First a rule of thumb for beginners: Always play the 2nd move inside the inner octagon and on the same half of the board as the first peg.

We shouldn't play too fast now because of Carroll. I will post my thoughts later.

@Thierry: I would say J10 is the classic answer to all the category 1) moves.

For the 2nd peg I think there are again 2 categories:

A) connected (to the border) or

B) semi-connected

J10 falls in category A), it connects to the left and it is also the position which is most towards the center in this category.

The same is true for J15, but J15 is only good with G4 as first move or a similar one, NOT with D3. You have to be a little bit careful with J15 that you don't end up on top right with a ladder going to the left, but it also prevents white from making use of the first peg already.

K17 is somewhere between those two categories, it is connected to the right but very weak (also not good with D3).

In category B) are all moves more towards the center, they need an extra peg to connect to the border. I played against those moves, but never used them myself, maybe David can tell us something about them.

So I would go for category A) with J10 or J15. They both also have the advantage of being 1.5 above the diagonal, so in case there will be a ladder, an Achilles have to be started very early to be succesful. In my games I more or less alternate between those two answers.

@maraca: Let me write it my way in order to check if I get what you mean. When you say that J15 is not a good answer to D3, I suppose that this is because D3 makes it not connected in case of a ladder starting with H15. Right ?

I don't undertand what you are talking about with "but it also prevents white from making use of the first peg already".

I'm not sure that I get you remark about the Achilles.

Regarding category A), J10 looks to me more agressive, and J15 looks to me having more potential because it lets wide open the diagonal to the top right and the horizontal to the right. White are very likely to answer in the horizontal right area, and black along the diagonal.

Regarding category B), I never used it neither myself. So, as often in Twixt, we should answer a way white will hesitate playing left or right. I'd say we shouldn't play more right than L.

My personal preference would go finally to J15.

@Thierry: Yes J15 is not a good answer to D3 because of the ladder. The distance is very long, it would be wrong to answer immediately with H15 or F15 then some kind of mousetrap can be played, but the threat is there and later it can get dangerous.

I also agree on category B) with not playing more right than L, this corresponds to the rule of thumb below the image.

Look here for the Achilles Defense. The last section shows how an Achilles can be answered with another Achilles and so on, so if the others want to be succesful with that kind of pattern they have to attack a ladder to the right very early.

The part with the first peg: Like already seen, a common opening is 1. G4, 2. J10, 3. N11 with connecting to the top because of G4. But if we play 2. J15 and white responds with 3. N14, this N14 is only semi-connected to the bottom.

Nice opening!

I can not write much. I agree with J15 but out of curiosity, wouldn't K16 give us more influence on right side or is it wrong regarding some sort of parity with the Achilles?

@Carroll

:) Thanks, we haven't done much yet.

Ok lets first look only at the power to the right side: I would say if the distance is that big, that one step closer to the border is less important than how much you are above the 'horizontal' diagonal starting/ending at W23. So I would conclude that the connectivity to the right is as follows: K16 < J15 < K15.

If we now also look at the connectivity to the left, then K16 and K15 aren't connected to the left, they probably need some kind of Achilles pattern that makes a new chain threatening to connect to the old one. I would say that full connectivity to the left is worth more than that little bit of extra connectivity to the right side.

Ok looks like we agreed on J15

(you can still ask questions about this move when the next one started)

Move 4

@maraa: thx for the answer, it seems convincing. As K15 is at distance 8 1/2 from the border is there the equivalent of Hex templates known? or how far would be the first white stone that would threaten this firm connexion?

Possible moves are Q14, P14 and O15, maybe M14 too: M16->Q15 and M18->Q17 look good, but O15 is difficult to play: N16- or L16- are bad, and 6.P13 is very complex...

I have difficulty following your moves, Klaus. I'm not sure what your -> symbol indicates. Please play out variations here. Registration is free. Click on the board to make moves and generate a move list. You can copy and paste the move list into a comment. Be sure to include the | symbol at the beginning when you copy the moves. When you post your comment, the moves will be click-able. You can click on the moves inside a post to see the position.

The -> symbol should be a "followed by". I wrote down my comment on Twixt Commentator too now.

The best move in a certain position has to

1. prevent, that the opponent can get a chain from his last set peg to one of his two borders of the game

2. most effectiv for the rest of the game (which means most often near the center)

In our position there is no chance to prevent that J15 is connected to the left border. Therefor we have to block a possible connection to the right border as L15, M15, N15, O15, P15, P14 or Q14 and maybe some more. Now we have to decide, which of these moves has the best influence for the future; O15 and Q14 save us a connection to the lower border, and P14 has no direct access to the lower border because of P16, but lies more central.

@Klaus: (3.m14) 4.o15 seems a strange answer to me. Why are (3.m14 4.o15) 5.n16 and 5.l16 bad ? Because of 6.m12 ?

I'd say that to be more effective as possible a move has to :

a. In the first half of the game, set up some influence or potential path

or

b. Set up a double threat

or

c. Prevent the opponent to do the same thing

...or, preferabily two of the three.

@Carroll: No there are no templates. To connect safely (can change in the game later) to the left, if there are no obstacles, you can be max 1.5 right of the diagonal because of ladders, so the most right pegs that connect to the left (on the lower half of the board) are: I13, I14, J15, J16 the moves on 17th row and below to be really sure are something like J17, I18, H19, G20, F21 but it all changes if it is not only a single peg, but a link and if there are some other pegs near it, so templates wouldn't help that much. It's better to try to understand the concept behind a move than to learn the exact pattern by heart. For example like I already said about the Achilles Defense: make a new chain that is connected to the border and threatens to connect the other chain.

What's going on team white? Klaus made his suggestions but David you didn't say which moves you wanna play or that you just agree with Klaus.

I'm ok w/ 4. q14...

I posted a response on servegame. I am also willing to play Q14. I just hope that is what Klaus meant.

Yes, I agree: Q14

Ok then it's our turn now. 4. Q14

Move 5

I suggest we use a new board for variations, it's easier to keep track then. LINK

At the moment I can only see 3 candidates for the 5th move:

Q12, Q11, P11

I will comment more on them later.

And maybe, O8.

The idea is that we have to make a short distance block or a block far away, if we choose medium range like P10, then white can play N11 and we are in trouble because of G4.

Ok, I posted some variations and explanations to the 3 moves mentioned above here. I think my current favourite is Q12, but I'm not so sure yet.

Hi Maraca,

I'm copy-pasting the comments I wrote in Twixt-commentator.

I agree with your consideration about a medium distance block : 4.p10 5.n11

As for white answers to 4.q12, what about

5.m12 6.n15 7.o13

5.m12 6.n15 7.n14 8.o17 9.m16

5.m12 6.o13 7.m16

...and I feel it as being too "local" at this stage of the game.

As for white answers to 4.q11, what about

4.q11 5.l12

I cannot see a good answer to 4.p11.

As an alternative, I'd propose 4.o9 more than 4.o8

But all in one I think that I'd prefer 4.p11

Thanks for your analysis Maraca, I would prefer a more centered move but that is because I don't play well! I think O8 is a bit to "defensive".

@Carroll: Yes I also don't like O8 that much. Normally we would try to play more centered, but white's move forces us to play there.

@Thierry: L12 is indeed an interesting answer to Q11, it could work. But M12 as answer to Q12 isn't good imho, I'll post the lines on the commentator.

4. O9 is dangerous, because of Q9 and R9.

Q11 doesn't form a nice pattern with J15, nobody seems to really like that move and Thierry proposed L12 as possible answer. So the question is now Q12 or P11. They span both a similar pattern with J15, P11 has M12/M14 and Q12 has M12/N15. Those points are farer away from each other for Q12, which is usually better. We also have to take care if we play P11, because all moves near the diagonal from W2 to C12 can be dangerous for our potential Achilles at R5, that's why I'm looking a little bit sceptical at P11. I also agree that Q12 looks somehow weak, strange and 'local' at first sight, but that could also be the case because we choose the opposing answer J15 to G4 and not J10.

Can you guys make your mind up - some of us are trying to learn from you experts! ;-p

:D Okay. Thierry is away for about 2.5 weeks now. So the decision is up to Carroll and me. Thierry wanted to play P11. I think Q12 is a little bit stronger, I also made some comments on the commentator. So Carroll you can decide which move to play, do we go with P11 or Q12? Or you can also ask for specific information about the moves first.