Recommend
15 
 Thumb up
 Hide
7 Posts

Clans» Forums » Strategy

Subject: Understanding the endgame rss

Your Tags: Add tags
Popular Tags: [View All]
Rolf
United Kingdom
Unspecified
flag msg tools
Avatar
mbmbmb
Clans - The Endgame

Shapes and tactics

Every Clans player should study the basic shapes which occur at the end of the game. I am only describing stable shapes, i.e. a region of connected areas (from now on 'region') where no village can be formed immediately. Unstable shapes will occur as moves in stable shapes and will thus be covered automatically. Critical shapes are those shapes where any move will lead to an unstable shape. I will refer to 'move parity' throughout: this is the concept, which every starting Clans-player has encountered (either to his joy or frustration), where only stable, critical shapes are on the board. The moving player has to destabilize a shape which the opponent can then form into a village, again leaving only stable critical shapes. This way the opponent can pick up a lot of bonus points for forming villages and can typically decide where these villages are formed.
However the moving player can decide which villages are formed in which order. The opponent would have to give up at least one bonus chip (a two-point swing) by playing a 'tesuji', i.e. instead of forming a village from an unstable shape, destabilizing another stable shape so that two unstable shapes are on the board. The moving player can then form a (or multiple) village(s) from one of the unstable shapes, the opponent can then form a (or multiple) village(s) from the other unstable shape. Note that no matter what, the moving player will determine which village is formed last, i.e. gets the +5 bonus!
Note that I am assuming that all areas are 'sparsely populated', i.e. that no area ever has too many huts to be moved. This is obviously not necessarily the case - adapt the below in these situations.

The standard 3 (critical): Three areas each touching the other two

A1 | A2
-------
A3

(e.g. wheat, grass and mountain in the top left corner of the board).
There is not much interesting about these. If the huts in this region will be scored, they will all form one village. The move parity is preserved by scoring this group.

The standard 4 (non-critical): Four areas areas like so

A1
-------
A2 | A3
-------
A4

(e.g. wheat, grass, mountains and mountains in the top left corner of the board).
This is far more interesting and imho the most important shape to understand. Firstly, the player on the move may transform this into a standard 3 and changing move parity. However, there is a more interesting option: move A2 to one of the other areas so that a linear 3 is on the board; now the opponent may either divide this into two villages which preserves the move parity or can move one of the extremal areas (A1,A4) into A3 leaving you to form one village and changing the move parity (this is better for the moving player than forming a standard 3, since he gains one bonus chip and can usually decide on which of two terrains the village will be formed). In effect one is offering the opponent a two-point advantage and no change in move parity with the alternative of a one-chip disadvantage and a change in move parity.
Furthermore, depending on the colour distribution this can be used for defensive play by the moving player: suppose the area contains one of every colour and some colour(s) more than once; the moving player has only one hut in the region. If he transformed to a standard 3, this hut would be removed during the clan war and he wouldn't score at all. However, if he transforms to a linear 3 and the opponent forms two villages, it is very likely that the moving player will score in one of the villages - maybe even arranging a nice fat +3/+4 bonus. If the opponent does not form two villages, the original moving player then gains a bonus chip and can (usually) decide the terrain, either choosing it as barren, or at least prevent any bonuses.
A similar tactics can be done offensively, provided the moving player has huts in at least two different areas of the region one of which must be A1 or A4. Create a linear 3 with own huts in A1 and A4, and either score in both villages or be able to choose the terrain (especially good if a clan war results) and get a bonus chip. This is especially useful if only one village remains to be formed and there are no convenient standard 3s with sufficiently many own huts around: the moving player ensures that he will participate in the +5 final village bonus and only gives up one bonus chip!

The circular 4 (critical): Four areas arranged in a circular manner like so

A1
----------
A2 | | A3
----------
A4

(e.g. wheat, mountain, forrest and mountain in the upper left hand corner).
This is less flexible for the moving player, since he is forced to create a linear three with the same results as described above, i.e. either no change of move parity and two-point disadvantage for the moving player or a change in move parity and a one-point advantage for the moving player (however, the non-moving player gets to decide, whereas in the standard 4 the moving player could force a change in move order by transforming to a stanard 3). However all the offensive/defensive possibilities for the moving player still exist.

No other stable 4-areas-shapes are possible on a Clans board. Notice that there is no group of four areas such that each touches all others - if it existed it would be non-critical and the only option would be to transform it into a standard 3 and change the move parity.

There are a lot of stable 5-areas-shapes and only dedicated Clans players need to remember them and their possibilities. I will not go into tactical finesses in colour distribution etc.:
The standard 5 (non-critical): Five areas centered around one of them like so

A1
------------
A2 | A3 | A4
------------
A5

(e.g. the wheat, mountain, grass, mountain, forrest region in the upper left corner).
This can be transformed either into the circular 4, changing the move parity, or into the standard 4 also changing the move parity. What is advantageous depends on what other shapes are available: e.g. if only standard 3s are available, then transforming it into the circular 4 is (almost) a given since then the moving player will be able to earn all bonus chips (however, be aware that the opponent can then choose which huts will form villages - unless you give up some of these boni by playing a tesuji!).

The central 5 (non-critical): Five areas, one of which touches all others like so

A1 | A2 | A3 | A4
-----------------
A5

(e.g. grass, wheat, forrest, wheat on the left side of the board and a mountain to their right which is 'central').
The moving player can transform this into a linear 4 (move A5 to one of the other areas), a standard 4 (move A1 or A4 to one of the other areas) or a triangle with a stick. We've covered the standard 4; the linear 4 can be either made into a linear 3 (and then see the analysis underneath the standard 4) or the opponent can form a village and leave the moving player with a village forming opportunity giving both one bonus chip and changing the move-parity. The triangle with a stick can either be transformed into a standard 3, a linear 3 or the opponent can form a village and leave the moving player with a village forming opportunity as above.

The circular 5 (critical): Five areas arranged in a circle lake like so

A1 | A2
-----------
A3 | | A4
-----------
A5

(e.g. the areas around the lake in the top lake row on the left).
The only critical five area shape but quite boring. The only option is to transform to the linear 4.


Let me summarize the above with a couple of formulae:
The standard 3 (S3): 0 -> 1c
This represents the following: the moving player makes a move which forms no villages (0), the opponent then forms one village having the choice of terrain [or plays a 'tesuji'].
The standard 4 (S4): (0 -> (2 | 0 -> 1c)) | (0 -> S3)
Where S3 represents the string from the standard 3
The circular 4 (C4): 0 -> (2 | 0 -> 1c)
The linear 4 (L4): (0 -> (2 | 0 -> 1c)) | (1c -> 1c)
The triangle with a stick (T4): (1 -> 1c) | (0 -> S3)
The standard 5: (0 -> C4) | (0 -> S4)
The central 5: (0 -> L4) | (0 -> T4) | (0 -> S4)
The cup 5: (0 -> L4) | (0 -> T4) | (0 -> C4)
The circular 5: 0 -> L4

There are a huge amount of non-stable regions of size 5 which I won't cover. Knowing and understanding the above should prevent most tactical surprises and can be used against weaker opponents to salvage a technically lost position. Against more advanced opponents, you need a strategy.


Endgame strategy:

At the end of the game, there will typically be huts which are not part of any village. These are wasted huts! It is better to have a pair of huts (or even a triple) in a village without clan war than having the wated huts left over. However, for reasons of flexibility you need to waste some huts. This is because the last village scores a lot of points: at least 6, but usually 8+. With winning scores being anywhere between typically 20 and 35 points this is a third or a quater of the points you need. Thus you simply cannot afford being left out of the last village - unless you ensure that your opponent doesn't score either or you score at least one village with a 4-point bonus which your opponent doesn't score.
So, when the endgame approaches (5 or less villages to score), study the board carefully to see which regions you would like to score last and which regions you cannot afford to score last. There are two ways to deal with the latter: either you make them score early giving bonus chips and points to the opponent, or you try to form villages so quickly (again giving bonus chips to the opponent) that they will be wasted. In both cases timing might be essential (i.e. wait as long as possible with giving your opponent the chance to form villages and hence bonus points) as he will try to do the same and maybe you can get away with the quick end without giving lots of bonus points to the opponent.
Secondly, check the move-parity and note that you would like to form as many villages as possible and decide where they are formed. But keep in mind that you need to change it sufficiently early (possibly via a tesuji) to be able to determine the last village. Especially the formation of linear 3s can change the situation suddenly. Of course, if you're sure to score in all remaining villages you can concentrate on move-parity but then your opponent made a grave mistake in the middle-game.
The alternative is to make sure that you form all of the last five villages (5 bonus points) and give up being present in the last village. If you can get some extra points from terrain in the course of this, you might be able to offset the disadvantage of not being present in the last village. Especially if it is not plain which colour you are, this might become a very good strategy as it might force your opponent to take a 50-50 guess.
To prevent you from having to take 50-50 guesses, the easiest way to exclude all but your huts from a village is to create a standard region with a total of 6 huts, two of which are yours. I would recommend going for a standard 4 region since standard 3 regions are easily turned into villages (your opponent destabilizes and then forms the village on his next move as opposed as S4 to S3 to linear 2 to village, or S4 to L3 and then you at least get two tokens and prevent the opponent from getting two tokens. But this is getting into the middle game already...
8 
 Thumb up
1.00
 tip
 Hide
  • [+] Dice rolls
Michael Snedeker

Sparks
Nevada
msg tools
Ok, Rolf, you have way too much time on your hands!

That's quite a breakdown of the end game for Clans. How do you feel about this overall end game versus Colovini's adavanced sister game to Clans, "The Bridges of Shangri La"?

Well done. I just need to lay down now and take an aspirin!

Mike Snedeker, pres.
BGoR
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Rolf
United Kingdom
Unspecified
flag msg tools
Avatar
mbmbmb
board_games_reno wrote:
Ok, Rolf, you have way too much time on your hands!

Hmm, sometimes you just need to get things out of your system

board_games_reno wrote:

That's quite a breakdown of the end game for Clans. How do you feel about this overall end game versus Colovini's adavanced sister game to Clans, "The Bridges of Shangri La"?

I must confess I haven't played yet. Although I really enjoy Clans and the minimalist approach to design that Colvini exhibits in general, I have not managed to pass on my enthusiasm

Best,
Rolf
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Mitch Willis
United States
Kathleen
Georgia
flag msg tools
badge
Avatar
mbmbmbmbmb
nessi wrote:
Although I really enjoy Clans and the minimalist approach to design that Colvini exhibits in general, I have not managed to pass on my enthusiasm


I really enjoy Clans as well; it's my favorite abstract. However, I have trouble getting other folks to play...
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
sunday silence
United States
Maryland
flag msg tools
THe end game in this game is quite deep and I have not even explored it. ONe guy on that breitspeil site game me some pointers and I wrote them down somewhere.

I know there is another strategy where you try to isolate a person in a corner, assuming you know his color. And you stick him where he does not have a path of regions that connect to a an area with more huts. So he is stuck in an area where he cannot score many pts. or maybe he has two huts but the region he is in only has four colors so he cant get the bonus.

There were at least four other strategic issues at the end of this.I will try to find my notes..
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Eric Walkingshaw
United States
Corvallis
Oregon
flag msg tools
badge
Avatar
mbmbmbmbmb
This is an absolutely fantastic article.

I haven't played Clans in several years, but something in my research recently brought it to mind and I started thinking about how to apply my research and some basic game theory to it. After re-reading the rules and thinking for a bit, I stumbled across this excellent post.

I am not sure yet if I will actually follow through on this little tangent, but if I do, this provides an excellent foundation and terminology to work with. Thank you!

Edit: In fact, I worry you did too good a job, leaving too little that is interesting for me!
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nicholas Hjelmberg
Sweden
Saltsjö-Boo
flag msg tools
designer
Avatar
mbmbmbmbmb
Both the game and the article are old but reading this, I can't help wondering why Clans rate so low. It's a pity that the game is difficult to get hold of as I really would like to play it.

(Partly off topic, I designed a similar game before I learnt about this where the objective is to group tokens rather than isolating them. If anyone would like to offer feedback, you are welcome to http://boardgamegeek.com/boardgame/170540/iconoclasm).
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.