Wednesday, July 21, 2010

3 is the loneliest number

Today I was doing some testing of various 3 player formats for FMB.  Because I am designing a free flowing wargame it is obviously true that 3 players will be a challenging format as traditionally 2 people simply gangpile one guy and kill him off before going to work on each other.  FMB has some things that tend to weaken that strategy because all units respawn at their controller's base after dying so actually wiping someone's army out isn't possible; you can push him back and take his Mines but you can't actually destroy him so alliances between 2 players cannot be as solid as in many wargames.  The other main thing that discourages gangpiling is the victory condition; cooperating with an opponent is simply foolish once the third player is behind in score so there is a real in game advantage to backstabbing your 'partner' early on.

The simplest rules version I came up with simply has all 3 players accumulating points as fast as they can and the first player to 30 points wins.  (The 2 and 4 player versions require 45 points to win.)  It has the advantage of being very intuitive and the strategies are easy to understand even for a new player.  If one player is particularly bad it shouldn't be a big problem because the other two players will naturally end up fighting each other more as the bad player falls behind in points.  I came up with a much more devious alternative though which was based on around the idea of attacking left.  In big group games of Magic this was regularly used to encourage action - the idea is that you sit in a circle and win by eliminating the opponent on your left.  In FMB I set it up so that the game ends when anyone hits 30 points but that the winner is the person whose left hand opponent has the *smallest* score.  This changes things drastically of course, mainly in the following ways:

1.  Alliances are impossible.  You cannot team up with an opponent to kill the other opponent because either you are hurting yourself by not attacking lefty or your righty is an idiot for attacking someone whose score he wants to maximize.  Everyone is forced to be on their own team.

2.  The game feels much less intuitive.  It feels very natural to defend your territory and attack anyone who is overpowering you and much less so to throw yourself at the throat of a random opponent.  Defending your own territory and preserving your army is of secondary importance which surely has a bizarre feel.  I imagine for new players this might be really hard to get used to.

3.  One poor player can really mess things up.  If a player plays very passively (badly) and simply defends themselves then the player to their left is nearly guaranteed to win.  Being able to constantly attack and not defend is an incredible advantage.  This does require a pretty determinedly stupid player though, as even if they send one fast unit to lefty's home area to put pressure on the game is probably going to work fine.

The question I am mulling is whether or not a loss in intuitive, straightforward play is worth a gain in 'theoretical game perfection' whatever that means.  From far away I really like the idea of attacking left and it certainly feels unique but the motivation of the wizard/general we are roleplaying seems strange indeed.


  1. As a gut reaction the idea of attacking left makes more sense to me if the win condition is related to the difference in scores between me and the guy on my left. The idea that I can somehow win with 1 point because the scores are 30-1-0 seems really bizarre. Extreme for sure, but bizarre.

  2. I could definitely do that and it would certainly make it more intuitive in terms of defending your holdings, but it would also make figuring out who is winning at any given point much less clear.

  3. I should really play before I sound off... but....

    Does it make sense to include "attack left" as a variant for "advanced players"? That way when people are learning the game the styles of play and win conditions are consistent no matter how many ppl they play with but those who want to try a more rigorous version of the three player game have it included in the rules. I think that learning a new game that changed the win conditions a fair bit depending on the number of players would throw me at first.

    Makes sense?

    I'm REALLY excited to play! Can't wait to admire your handiwork.

  4. Matt: That seems workable. Having the rules change entirely for 3 player when advanced players are playing doesn't feel perfectly elegant, but it does let new players have a simple game and advanced players have the option for a better game. I already have a basic/advanced set of rules set up anyhow, though those are only for setup to make sure new players can jump right in.

  5. You could still have the game end at 30, you just end up comparing different numbers to see who wins. Under the proposed rule if A, B, and C are sitting clockwise and have scores X, Y, Z then...

    if X is smallest, C wins
    if Y is smallest, A wins
    if Z is smallest, B wins

    Whereas in my suggestion (which hasn't really been fleshed out)

    if X-Y is biggest, A wins
    if Y-Z is biggest, B wins
    if Z-X is biggest, C wins

    Often the same person will win in either situation (if X is really small then Z-X rates to be really big and C wins either way) but it gives players two ways to increase their score instead of just one. C can either work to increase Z or decrease X.

    I think this will be most important as the game progresses and units start dying. When a unit respawns in your fortress, what should you do? Under your system where really what I care about is pounding on lefty I need to find a way to get to lefty, but if we're successfully rotating left he's going to be far away from my spawn point. Probably what I want to do is head right and hope righty ignores my guy. My goal is to go defend righty while letting righty keep his stuff. I want righty to hit 30 points (this means lefty can't win) and all I care about is finishing a little higher than lefty.

    Oddly, what righty wants to do here is give lefty points. That's how he makes me lose. No one actually wants to score points themselves as a primary goal, and it can actually hurt to do so if you take them from the wrong person.

    I'm not sure my plan is any better, but by including more variables in your score you have more different things you can do to improve your score. When my dead guy respawns he can go fight over my starting nodes because increasing my own score is good. Here if someone gets most of the map they're going to win as opposed to just being one of two people who might win, but maybe this also makes people just attack wherever instead of attacking left.

  6. Consider what happens when A respawns and retakes his node from C, who took it when he attacked left. A will gain a point, C will lose one, so X = X+1, Z = Z-1. Under both systems:

    if X+1 is smallest, C wins
    if Y is smallest, A wins
    if Z-1 is smallest, B wins


    if X-Y+1 is biggest, A wins
    if Y-Z+1 is biggest, B wins
    if Z-X-2 is biggest, C wins

    So in the first case, when A takes a node back from C he makes it less likely C wins and more likely B wins. In the second he keeps even with B and makes it less likely C wins.

    I don't even know that that is better.

  7. Your suggestion certainly would make people more inclined to defend. Taking a Mine for yourself would be just as good as denying lefty a Mine so it would make sense to prioritize both things. Another way to implement the same thing which would be much easier on the 'who is winning right now?' front would be to have each player gain Gold on their turn as normal but then also take away that same amount of Gold from righty. This would mean that the scores would all stay very close to zero (and would regularly be negative) but whoever had the biggest number would be winning. Having the game end at a particular Gold total wouldn't work as well in this case since the players could sit at near zero forever, but some other game end condition is easy to incorporate.

    Hm. Another crazy option would be to do the above but instead of your turn being
    Gain X Gold
    Subtract X Gold from righty

    It could be
    Gain X Gold
    Subtract X-1 Gold from righty

    So if I control 3 Mines I get 3 Gold and my right hand opponent loses 2. This would mean that net Gold income per turn is 3 instead of the usual 6 so I would probably need to reduce the 'you win' Gold total to 17 or so. This is a strange enough situation that I would have to actually play it a few times to have any sense of how it works I think.

  8. Under that system taking a mine from lefty would gain you 2 (you don't lose 1 on his turn and you gain 1 on yours), would cost lefty 1 (he doesn't get 1 on his turn), and would cost righty 1 (you steal 1 on your turn.

    Taking a mine from righty would gain you 1, would gain lefty 1 (he doesn't lose 1 on righty's turn) and would cost righty 2 (the one you steal and the one he no longer gains).

    Clearly in this case you generally want to beat on lefty but still gain something by beating on righty. Seems ok on the surface. How would gold related artifacts interact with this?

    Net gold income will always be 3 unless someone loses all of their mines. Normal game the net gold income changes based on mine swaps, so you probably want to have the win target more than half. 20 maybe?

  9. What about just limiting which mines people can get gold from. Player A can get gold from the mines in front of himself, those in front of Player B and the one in the middle. Holding the mines in front of Player C does not help player A win at all. That creates an attack left scenario but still lets the game end when one player gets the most points.

  10. @Ziggyny
    Losing all your Mines does not mean that the net influx per turn changes. You gain 0, righty loses 0-1, hence he *gains* 1.

    I figure Artifacts would simply gain you gold, same with Destroying enemy units. They don't need to interact with this.

    That is a clever idea. It means you can punish righty if he is ahead but you really, really want to attack lefty. I like it because it is very simple and doesn't make scoring weird but really enforces going left.