Talk:Homeworlds strategy

From IcehouseOrg

Jump to: navigation, search

Contents

Mid-Game Strategy

I've been trying to learn this game on SDG.... and so far I'm not really "getting it". This past weekend I actually sat down with a buddy who I'd been trying to teach via SDG and we scrapped it out with me winning one game due to causing a catastrophe in his homeworld, and the next game, I caused a catastrophe in my homeworld (not thinking things through). The thing that we both struggled with was the Mid-game. I mean we got to a point where we knew our objective, we knew how the pieces operated, but didn't know how to get from creating ships and stars to destroying a homeworld. Any discussion on this fact would be greatly appreciated.

--Tuxhedoh 20:46, 10 Oct 2005 (GMT)

I'll give it a shot. I'm not really an expert on Homeworlds myself; I had played it several times before SDG, but I had missed a rule, which changes the strategy drastically. Since then, I've been revising my strategic thinking, and so while I have some idea of what's going on, it's still a little new to me. — Lambda 16:39, 11 Oct 2005 (GMT)

Just remember that this game is really about controlling the economy. Since all the players must play from the same stash, they can also control (to a limited degree) what ships are available and when. Since Binary Homeworlds gives it's players such a wide array of choices, the level of strategy is very, very deep (even more so than chess I would wager). The unfortunate thing about this is that it is highly difficult to predict what your opponent might do during the Mid and End games. I find that it's much easier to predict an opponent's moves during the beginning of the game. If your opponent has developed a strong defensive fleet at his homeworld, you can win pretty easily by creating an overpopulation at his homeworld and triggering a catastrophe. This will destroy at least one of the stars in his system, which will probably seriously cripple him. Just watch out for counter attacks since your opponent may now have direct access to your own homeworld. Also, remember that the object is to remove the opponent's ships from their homeworld, not neccessarily to destroy his/her homeworld. --GameBrain42 17:30, 28 Dec 2006 (GMT)

Questions/Clarifications

  • Can you have catastrophe conditions and not have a catastrophe, that is, if your opponent is feeling particularly generous or forgetfull and doesn't call a catastrophe, or does it happen automatically?
    • Catastrophes are always optional; however, any player can trigger a catastrophe in any system and the end of their turn. --Aaron 22:23, 17 Oct 2005 (GMT)
      • Are there ever any conditions in which it is not to the advantage of one player or the other to trigger a catastrophe? By this I mean, assume both players are vigilant... is there ever a time when both of them would see a catastrophe and decide not to trigger it? --Jabberwocky 13:00, 18 Oct 2010 (GMT)
  • What would it look like to sacrifice a yellow ship, does that allow any of your ships in any other system to move? Can it be multiple pieces?

--Tuxhedoh 16:04, 17 Oct 2005 (GMT)

    • If you sacrifice a size-3 yellow then yes, you can move any three of your pieces in any system one space. You could also (according to my understanding of the rules and the way it is implemented on SDG) move 1 piece 3 times or 1 piece once and a 2nd piece twice. --Aaron 22:23, 17 Oct 2005 (GMT)

Question: Abandoning the homeworld?

The "official" rules[1] seem to contradict themselves on whether a player is allowed to temporarily abandon his own homeworld:

  • (A) As previously noted, you must always own at least one ship in your Homeworld; if at some point you don't own any ships in your Homeworld, you are eliminated from the game.
  • (B) You may temporarily abandon your Homeworld during your turn, as long as you own at least one ship in it at the end of your turn.

Here's a minimal (and fairly silly) scenario: Alice has a y2 at her own homeworld, and two other ships elsewhere. She wants to sacrifice the y2 to make two moves: "advancing" one of those two ships and "retreating" the other one to her homeworld, so that her homeworld is temporarily abandoned but occupied again at the end of her turn.

If sacrificing the y2 would leave Alice's homeworld completely unoccupied, I believe the star would be immediately dismantled and returned to the stash before the rest of Alice's move, causing Alice to lose immediately. The "banker's strategy" depends on this interpretation for green moves: sacrificing the lone g1 at a star allows you to turn that star's piece into a ship. Therefore, Alice is never allowed to sacrifice the very last ship in her homeworld.

But! If Bob also has a ship at Alice's homeworld, then Alice's sacrifice does not immediately dismantle her homeworld, because it's not left completely unoccupied. In this case, we have to pick one of the "official" rules to follow. I think it feels better to follow (B) and ignore (A), but I'd like to know which rule the world champions have decided to follow. :) --Quuxplusone

Answering my own question: I am now confident that (B) is the correct interpretation. The interesting scenario is
   Alice (0, y1g3) y2-
   (b2) g1-
   (b1) g1-
   Bob (1, y2g3) -g1g3

In this scenario, Alice (player 0) has no winning move yet, because her y2 is the only ship at her homeworld. However, in this scenario...

   Alice (0, y1g3) y2-r1
   (b2) g1-
   (b1) g1-
   Bob (1, y2g3) -g1g3

Bob's small red will keep Alice's homeworld from vanishing in the middle of her turn, and she can catastrophe green at Bob for the win. --Quuxplusone 21:10, 15 February 2009 (UTC)

Not for nuthin'... but why not ask this at the Icehouse List and see what Andy thinks? Perhaps the official rules need an update to clarify? --David Artman 16:29, 18 February 2009 (UTC)

Puzzle: Zugzwang

Zugzwang (wikipedia:Zugzwang) is the term for a position in which the current player would prefer to pass because any legal move will weaken his position. Can a Homeworlds position be constructed in which both players have legal moves but (assuming they both play optimally from this point on) will prefer to pass, resulting in a draw?*

*  Note that an optimal Homeworlds player will always make the same (optimal) move from a given position. So if Alice and Bob are optimal players, and it is Alice's turn in a given position, and Alice chooses to pass, and then Bob also passes, it will be Alice's turn again in the same position — so Alice will make the same move again (pass), and so on forever; therefore, if both players pass in succession, the game may be considered drawn.
Therefore, if Alice ever passes, she is effectively offering Bob a draw. Therefore, an optimal player who has the upper hand will never pass.

I suspect that the answer is "no", but I haven't found any proof one way or the other. --Quuxplusone 21:10, 15 February 2009 (UTC)

The answer is "yes". Here is a trivial example:

   Alice (g1g2) g3-
   Bob (g1g2) -g3

Alice's only legal move is to construct another green ship at her homeworld. But if she does, Bob can then trigger a catastrophe in Alice's homeworld, winning the game. Therefore, Alice would prefer to pass. Then it would be Bob's turn, and since his position is identical to hers, the same argument applies for him. --Sapphire 00:06, 30 October 2011 (UTC)

Opening Combinations

I analyzed 1431 Homeworlds games played on SDG (games ending up to 11 Mar 2013, at least 10 moves long, only 2 players, no systems with two stars of the same size). The results I got are as follows:

  • Goldilocks vs. Banker: Dead even; Goldilocks won 312 times and Banker won 311 times.
  • Goldilocks vs. Fortress: Goldilocks fared slightly better here; 104 vs. 84 wins.
  • Banker vs. Fortress: Banker was more successful; 276 vs. 208 wins.

--AaronSDG 18:48, 20 March 2013 (UTC)

Personal tools