Free-For-All Falls Flat

January 27, 2008 at 12:38 am (Game Balance, Game Design, Game Mechanics, PvP)

As a general rule, I don’t like free-for-all-style games (also known as “every-man-for-himself”).  In this category, I include games with more than 2 teams, even if the teams are larger than one player each.

More precisely, I don’t like the fact that they’re free-for-all; there are many individual free-for-all games that I rather enjoy, I just wish they were organized differently, and I tend to enjoy them less than other people seem to.

Like many of my gaming preferences, I imagine this is partly just due to my own individual personality, but I also think it’s partly that free-for-all games tend to suffer from several systemtic problems.

Allow me to break this into separate cases.  Broadly speaking, there are two major free-for-all structures: games where players can obstruct each other and interfere with one another’s chances of winning (such as Clue, StarCraft, Quake deathmatch, etc.), and games where players are competing to fulfill independent goals as quickly or as effectively as possible (e.g. bowling, golf, races (sometimes), high-score competitions, etc.).

I don’t have anything in particular against the second category, except that it’s really composed of single-player games masquerading as multi-player games.  If I’m just trying to get a higher score than everyone else, there’s no need for us to play at the same place or time.  The games are often fun, but if I’m actually going to go to the trouble of arranging to play a game with a group of people, I’d like to, you know, play with them, not play in separate games and then compare high scores.

Policy vs. Politics

So that leaves the first category, where the multiplayer is justified by the fact that one player’s actions can actually affect the gameplay of another (this is the more typical case).  My chief complaint with these is that there’s a tendency for them to trivialize any actual gameplay-driven strategy in favor of (for lack of a better word) politics.

Here’s how it tends to happen:  your opponents each have, at their disposal, one or more tools that they can use to interfere with your attempts to win.  Additionally, you have more opponents than you have allies.  This means it is extraordinarily unlikely that you can win through sheer skill if your opponents all decide to cooperate in stopping you–collectively, they just have a lot more resources than you do.  If you throw your obstacles in front of your best opponent, and a dozen opponents all throw their obstacles in front of you, your opponent is almost certainly going to win, even if you’re a significantly better player.

Thus, no matter what the actual rules of the game are, your primary goal is to convince your opponents to fight each other instead of you, because unless you do a pretty good job of that, it often doesn’t matter what else you do in the game.  The game ends up being decided based on who looks the least threatening, or who is most popular among the other players, or who happens to take the turn right after the last ready obstacle has been played.

Now, granted, manipulating other players into doing what you want instead of what’s in their best interests is a complicated and subtle art, and the precise methodology probably does vary a little depending upon the rules of the game.  But the gameplay is no longer the primary activity.

And there’s nothing inherently wrong with that, if that was the goal.  But it’s less interesting to me, personally, and I think the result really belongs in a different category from other games; the game here is just being used as the background for another activity.  If the gameplay was the goal, then the design has failed (this is actually a game balance problem, according to my proposed definition).

Of course, there are some exceptions to this rule–some games manage to use free-for-all gameplay without emphasizing the politics above all other aspects.  Here are some traits that I think help games keep politics under control:

Lots of Hidden Information.  A game doesn’t really enable politicking if you don’t know anything about the other players’ positions.  You can’t rally everyone to stop the imminent victory if you don’t know that someone’s about to win, and you can’t systematically undermine a single player if you don’t know what moves disadvantage him more than other players.  It’s also hard to give someone reasons to attack someone else instead of you if you have no way to establish that the other guy is a bigger threat.

Forced Opposition.  Some games (FPS deathmatch comes to mind) are arranged so that you are pretty much forced to fight whoever you run into (because that’s the way you get points, and the other guy can kill you quickly if you don’t defend yourself), and so that the people you run into are largely random.  Sometimes it’s hard even to tell who you’re fighting until one of you wins.  Since you can’t effectively target specific opponents, politics doesn’t play much of a role–on the other hand, the game also plays suspiciously like a bunch of short one-on-one matches rather than a coordinated free-for-all.

Restricted Cooperation.  Games where you can act to someone else’s benefit (rather than just to someone else’s detriment) make the problem worse; if the best that someone can hope for is that you’ll leave them alone (rather than actively helping them), players are less likely to form alliances.  However, this one is hard to manage to any great degree, because even if there’s no way to directly help someone, it’s often possible to aid someone indirectly by distracting or restraining a third party that would otherwise target them.

Fast Pace.  If players are required to act so quickly that they can’t make appeals to each other, this limits the amount of politcking.  However, if it’s easy to determine and target the current game leader, spontaneous cooperation is still a definite possibility, so I think this one reinforces the others more than it stands on its own.

Win Ratio

A lesser, independent complaint I have against free-for-all games is that I don’t win at them as often.

Of course, that sounds selfish, and of course, the game is supposed to be more important than who wins.  But, nontheless, people like to win–that’s not going to change any time soon.  And the more teams you divide people into, the fewer players get to win any given game.

It is also true that people feel more satisfaction for winning when they know that the odds were against them, and less frustration for losing when they don’t expect to win, but I don’t believe it balances out; I think people generally feel it more when they lose five games in a row than when they win one six-way free-for-all.

And it doesn’t take very many consecutive losses before people begin to say things like “I just can’t win today” or “why am I playing so badly all of a sudden?”  People have lousy intuitions for randomness, and won’t expect long streaks to be nearly as frequent as they actually are, so when they lose lots of games in a row, they won’t think “OK, that’s to be expected every now and then,” they’ll think “wow, I suck.”  And that’s not what we want the player to think.

However, this drawback can be mitigated by allowing multiple winners, or “runners-up.”  Some games, because of their structure, need to end as soon as someone wins–others can easily let the winner drop out while the remaining players compete to see who can win next.  On the other hand, this feeds into the third issue with free-for-alls…

Inconsistent End Times

Lots of times, players want to do something together when the game is over (sometimes play another game–sometimes something completely different).  If some players drop out of the game before others are finished, this can cause problems.

Many free-for-all games don’t suffer this problem at all.  Many that do have this trait are generally played in a setting where it doesn’t matter.  Nontheless, there is a tendency in some free-for-all games to have an “elimination” mechanic (where you win by being the “last man standing,” rather than the first to achieve some goal, or best at the end of a time limit).  This can result in some players being unable to participate for significant portions of the game.


Two-team games are nice.  I think that a lot of games could do well to introduce (or place further emphasis upon) modes with two competing teams, rather than one-on-one or free-for-all styles.

One-team games can also be very interesting–games where all the players are working together to “beat the game,” which behaves somehow algorithmically (e.g. a computer AI, random card draws and dice rolls, etc.).

On the other hand, some games just don’t scale up gracefully.  If the game really only works as intended in a one-on-one matchup, then you really can just leave it at that.

In fact, what worries me most are the games “for three or more players,” as those tend to have gameplay that obviously falls apart in the absence of politics (though there are some exceptions to this, as well).

My primary principle here is that if you want people to focus on the gameplay, then they should be able to win by playing the game well–not only by playing the other players.  If you enjoy politics-centered games, I wish you many happy hours playing them, but they are just not the same thing as strategy games.



  1. spellman23 said,

    Zombie Panic fixes the different end times by spawning you as part of the growing horde of zombies. w00t.

    But seriously, very nice. However, with the advent of avatars in FPS games, it is now very possible to form alliances in deathmach.

  2. Colleen said,

    I agree with your thoughts about FFA. Especially regarding FPS’s, the win ratio becomes a serious issue. Consider that a beginner player might become a target for other players (there rarely being any incentive not to make an easy kill), might have a difficult time overcoming the learning curve when it does not benefit any of the other players to help her, and therefore might never end up winning.

    Conversely, an expert player can win a disproportionate amount of the time, partly because people might avoid him in favor of combating weaker players, or because he can come up with one sound strategy and use it against player after player before any single player sees it enough times to devise an effective counter.

  3. Anonymous said,

    Interestingly enough, the phenomenon Colleen describes seems like it might be reversed for Super Smash Bros. Melee. From what I’ve observed, better players end up ganged up on and weaker players ignored more (4-player FFA still being the standard setup around here), giving them a chance to learn. At least, that was my experience when I was just getting hooked on SSBM, and it’s how I play now; I don’t know how well that generalizes to the experiences of others though.

  4. Iffrit said,

    Just like FPS offer survival and time style matches, Smash Bros. offers, at the most basic level, time or stock matches, where the free-for-all tactics change between them. In stock, you will generally see Anonymous’ phenomena, save for when people want instant gratification of smashing at least someone. In time battles, however, easy kills are easy kills, so the optimal will again be going for the most disadvantages character, and with known relative skill among the player and full knowledge among the avatars, it is very easy to determine who that is.

  5. Antistone said,

    If you have a player match-up where you think it’s desirable for three weaker players to gang up on one strong one, you can always play an actual 3-on-1 team battle. I’ve done that many times in Smash Brothers. There’s also a handicap system that, while not perfect, is quite serviceable.

    Neither the time nor stock match scoring systems used in Smash Brothers are particularly good for more than 2 teams; the one rewards predatory practices and (unless you turn up the penalty) deliberate suicides, and the other rewards staying out of the fight. I’d like to see a more complex scoring system that also takes damage (and not just kills) into account, but when I try to explain this to most Smash players, they tend to cut me off after about ten words and say it’s too complicated.

  6. Iffrit said,

    Team matches and handicap matches are a good way of moving towards a fair playing field. However, then your simply removing the free-for-all from the game. As for taking damage inflicted or some other scoring method for rewarding the overall effort of a player seems like a good idea, but just like any other scoring system, there could be ways of exploitation, like creating punching bags out of the least experienced gamers.

  7. Ian Schreiber said,

    Good analysis of FFA. I’d add that these games tend to be more vulnerable to the “kingmaker” problem: two players are neck-and-neck for the win, and a third (losing) player ends up making the key decision of which one of them wins.

    Personally, I like one-on-many games as an alternative structure (for example, Scotland Yard, or Knizia’s Lord of the Rings in the expansion where one player controls Sauron). I suppose you could effectively call this a “two-team” game — it’s just that one team has only a single player. An interesting case is Shadows Over Camelot, which might either be a one-team game or a one-on-many game, and part of the game is figuring out which one it is. Another case is Betrayal At House On The Hill, which starts out as a one-team game and turns into one-on-many part way through.

    I’m surprised you didn’t mention the predator-prey structure, where each player is trying to eliminate one other player and avoid being eliminated by someone else (the college game “Assassin” is an example, as is the Vampire:tES CCG). In some ways it just plays like a series of simultaneous one-on-one matches, except that you may have to balance your offense with your defense since your one-on-one “opponent” is split in two.

  8. Tommi said,

    Many strategy games can be hacked via suitable map creation; if there are, say, six players on a ring and the central territory is undesirable or impossible to pass, the likely dynamic is that of six one-versus-two games. Some hidden information and it could actually be interesting.

  9. haslo said,

    I do see your point, but I disagree that FFA makes game mechanics entirely useless. Game mechanics do influence how you can mislead your opponents, they can influence who’s most likely to help you out in fear of you fighting them and who’s most likely to exchange favours with you.

    You brought up Starcraft … if somebody is building up a huge force of fighters, and a terran player has all Goliaths plus the tech to make them very good against fighters, the fighter guy isn’t as likely to attack the anti-fighter guy and will consequently probably try to make his own survival beneficial for the anti-fighter guy. Or if a Zerg player has tons of ground units, he’ll want to keep the Protoss with his Reavers and Carriers away from his planets, and help him out.

    Or take Twilight Imperium. The Hacan have really powerful trade agreements, and those can’t be broken by other players unless through wars – while the Hacan can break other’s trade agreements. If he gets the right allies early on, it can carry him through the game. Or take the Letnev, with powerful military – others will try to get their goodwill early on, and they can call in favours later in the game. (Funny enough the political cards are among the least important things in the game when it comes to politics, in the end)

    A counter-example, Age of Renaissance. In this game, game mechanics do have a really inferior role, to the point where politics actually are game mechanics. It’s kinda contradicting my point really, but here politics were embraced with open arms and basically, the game is inter-personal politics with a bit of added combat mechanic thrown in. Nevertheless, game mechanics can be used to bluff, or to threaten, or to force players to do things they wouldn’t do otherwise – just think of the Black Death card.

    Admittedly, I do like politics in games, I do think that they add a lot to the possibilities for interaction between players (and just like you, I like those games where the players don’t interact that much less than other games, be they FFA or not). That of course means that I’m not bothered by the existence of politics, that I actually welcome them. The fact that players can and will gang up on the leader, the fact that there’s usually a king maker, they are true and remain. Politics do take the game away a bit from the game mechanics alone, and add an interpersonal level to it.

    But I really think that your post underestimates the influence that game mechanics can have on politics, and thus the extent to which it’s possible to influence the game’s outcome through game mechanics alone after all.

  10. haslo said,

    Read before you submit 😛 With “fighters” in the Starcraft example, I meant “flyers”. Silly me.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: