F-Zero GX analysis: Multiplayer game mechanics

f-zero mute city

F-zero GX was one of my personal favorites during the Gamecube generation. Its sense of speed, focus on skill and ability to (almost) always create thrilling races made me and my brother play it for years on end. It perhaps isn’t strange it was a major inspiration for the Wheelchair Racer project during my second year at university.

This article is a translation of an analysis written during my third year at university. As part of a “game theory and play mechanics” course, the analysis mostly covered the mechanics of the game. Also, it was written in Swedish. As such, I couldn’t use it as-was on this blog.

About F-zero GX

F-Zero GX (Amusement Vision, 2003) is a digital racing game developed by Sega’s Amusement Vision and published by Nintendo. Its skill-based game mechanics imply that the audience will view gaming as competition, which goes hand in hand with the game’s name – a reference to Formula 1. The goal for the player is to make her way around the course for a number of laps and pass the finish line before the other players do. The game have several game modes, such as tournaments of five levels against 29 AI-controlled competitors, a single player campaign, a practice move, time trails, a store, build-your-own-vehicle feature and a multiplayer mode for two up to four players.

F-zero GX knot level

Limits of the analysis

This analysis is limited to the split-screen multiplayer mode (online multiplayer was very rare on Gamecube games), using the developer-made vehicles. This allows me to focus on balance, general mechanics without going into the combinatorial explosions the create-your-own-vehicle feature brings (it’s interesting enough for a separate analysis, though).

Although I would like to complete this analysis at a later date with an analysis of the single player mechanics and the presentation layer, as well as a usability study, I won’t in this post as it would become much too long.


This segment mention the terminology that will be used for the rest of the article, with explanation and references to the creation in which they are mentioned. Clearing all of these out beforehand makes the analysis easier, as terms will not need to interrupt the pacing of the text.

Operational and Constitutional rules are phrases from Rules of Play (Salen and Zimmerman, 2004), which together with Implicit rules makes up their framework for various kinds of rules. Operational rules is how you play the game and what constitutes the game, constitutional rules are the math behind the rules and implicit rules are, well, implicit, like acting with sportsmanship. There are other models to group rules, such as Parlett’s rules set, but I will use this one as it separates underlying game logic from the possible player actions in a practical way without being overly complex.

Positive and negative feedback are phrases introduced by Marc LeBlanc at GDC-99 (Salen and Zimmarman, 2004). Positive feedback means feedback rewarding the winner, thus destabilizing the game and will hasten the end of the game, while negative feedback rewards the loser, stabilizing play and extend the game. Well worth noting is they are not synonymous to punishment and reward.

From The Art of Game Design (Schell, 2008) comes the phrases interest curve and indirect control. An interest curve is a graph showing interest over time. Ideally, it makes a roller coaster-shape, with a spike early on, a fall after that which comes close but not quite reaches zero interest (because, if it does, the player/viewer leaves), and then slowly climb with peaks of interest along the way until it reaches a climax at the end. Indirect control is a way to set up the player’s mental “box” by cultural conventions and expectations, making the player perform subconscious decisions while not even having experienced a choice consciously.

Emergence is described by Salen and Zimmarman (2004) as when “complex possibilities are the result of a simple set of rules”. Schell (2008) describe it as when few verbs result in many actions. In both cases, it’s when the game become more than the sum of its parts.

David Sirlin once wrote about the term Yomi, a phrase for or “mind reading”, or predicting the opponent’s action before she makes them. Sirlin divides Yomi into layers numbered zero to three, where layer zero is to predict an action but not act on the prediction, layer one is to counter the action, layer two counters the counter, and layer three to counter to counter’s counter. In essence, it’s a four-layer rock, paper, scissors where one player can pick “rock” and “scissors” and the other “rock” and “whatever number four is”.

In the same article, he coins the phrase Lame-duck situation, which is when one player has lost but needs to wait it out for the winner to achieve the victory condition.

Jonathan Degann described in his blog the phrase Story Arc as when “the quality of having the situations and decisions metamorphose during the course of a single game so that the player has the experience of participating in a story with a wide sweep”.

In another article, he coins the phrase Agonizing Decision, which is self-explanatory. This may feel terrible for a player, but it is a quality of the game, as it needs the player to think through her priorities.

In regards to decisions, James Portnow’s article distinguishing problems and choices comes to mind (as can be seen as a video in this early Extra Credits episode). Long story short – choice is a matter of preference, while problems boil down to a measurement of value.


This segment will start by describing the operational and constitutional rules, whereby an analysis of the operational and constitutional rules will follow to understand how they create the playing experience.

Describing operational rules


A gamecube gamepad and the names for each component

  1. Before a race, the racers select the course and their vehicles. Every vehicle has values for weight (kg), body, boost and grip (ranking from A, which is the best, to E). [Translation note – these are really constitutional rules, as they affect how each vehicle handles, but due to the disposition of the original work, I will leave it here and perhaps deal with it in a future analysis]
  2. Before the race, the racers select a value in the range 0 to 1, where 0 represents maximum acceleration and 1 represent maximum high speed.
    1. They can also toggle the direction of the screen split, the number of AI opponents, if “handicap” should be on or off, if re-spawns should be on or off.
    2. Every player controls the game with a Nintendo Gamecube gamepad (Figure 1).
      1. Hold down A to gas,
      2. Hold down B to slow down.
      3. Press Y to boost.
      4. Hold down L or R to glide sideways to the left or right, respectively.
      5. Press Z for a spin attack,
      6. Move the left thumb stick left or right to turn.
      7. Move the left thumb stick up or down to tilt the vehicle up or down.
      8. Press X in addition to left or right on the left thumb stick to tackle the the left or right,
      9. Hold down L and R in addition to left or right on the left thumb stick for a drift (sharp turn) to the left or right.
    3. The race takes place on one of the game’s 26 tracks.
    4. After the race is over, you can view the racer’s finish times, scores are distributed among the players (10 for the winner, 5 for the second player, 3? Then 1?).
    5. Lastly, a menu allows players to choose a new track to play, change vehicles on the same track, change settings on the same vehicles or quit multiplayer game.

Describing constitutional rules

f-zero-gx headache level

Picture unrelated to the text, but it looks cool. You actually race inside that tube before you reach this view!

  1. The value on the range between acceleration (0) and top speed (1) result in the interpolation of two formulas for the vehicle handling.
    1. The closer to 0, the more speed is kept during drifts, the more speed is kept on slowing terrain, and the easier it turns.
    2. The closer to 1, the higher the top speed is, the more speed it gains from boost pads, and the better the grip is. The better the grip, the more friction with the ground (ignore for a while the fact that they hover).
  2. While a player gases, the speed moves up toward the vehicle’s top speed as a function of the acceleration.
  3. When the player breaks, the speed moves toward 0.
  4. When the player boost, it gains a speed increase for a time (if the player gases during the boost). After the time has ended, speed nears the top speed.
    1. This effect also occurs if the player drives into an “explosive barrel”. This will make the barrel to disappear and drains on the vehicle’s energy.
    2. This effect also occurs if the player drives onto a “boost pad”. If the player drives on the pad’s edge will cause a “double boost”, increasing the effect.
    3. The effect can be created manually by boosting (Y button), which costs energy. This can only be done after entering the second lap.
    4. Energy can be regained while driving on certain terrain.
  5. A spin attack damages all enemies within a certain range of player vehicle at the time of action, and lasts a short period of time. The attack reduces the player’s speed. A vehicle must gas to spin dash.
  6. A tackle moves the vehicle in a direction, damaging the first opponent it hits, pushes the hit vehicle with a force in the same direction as the tackle, and drains some of its energy. A vehicle must gas to tackle.
  7. A vehicle needs to move to slide or drift.
  8. If the player vehicle is on a dirt area, the speed is reduced and vulnerable to damage.
  9. If the player vehicle is on an ice area, it loses grip, as ice greatly reduces friction to anything on it.
  10. If a player vehicle’s energy reaches 0, it is out of the race.
    1. If re-spawn is set to active, it will re-spawn after five seconds. By default, re-spawn is set to active.
  11. If handicap is enabled, players away from the lead gain higher top speed than the player in pole position. By default, handicap is set to active.
  12. The players drive a set number of laps around the track. Default setting is 3 laps.
  13. The first player to cross the finish line wins.
  14. After the player next to last has crossed the finish line, the last player has seven second to cross the finish line before the race is aborted. This rule does not apply to AI racers.

Analyzing the constitutional rules

  1. The balance between acceleration and top speed is most likely intended as a strategic problem to solve for each track, rather than an agonizing decision. It mostly effects how the vehicle is handled, but could be considered a problem rather than a choice.
  2. The rule about acceleration has an obvious flaw: It makes it advantageous to sometimes avoid accelerating, something we’ll see more of in the section on advanced techniques. However, re-writing the rule to disable a cap would result in ever increasing speeds until players lose control. Which could be fun, but it would give the game a very different flavor.
  3. The rule about breaking doesn’t make much sense, given all the options to turn and use the speed gained. The only reason I can see is that it could strengthen the theme of the game (just imagine a car without breaks!)
  4. The rule about boosting has two main reasons: It makes the game feels faster, as the mind doesn’t get time to adapt to sudden speed bursts, and it offers a reward (or bait) for players.
    1. The explosive barrels are advantageous during the first lap, and become more tempting the further away from the lead a player is. Because the barrel disappears, it gives a clear advantage on the players who come after the player or a player who miss it. It is a form of reward that diminishes over time, but can spread players out in the early phases of the race.
    2. The boost pads offer a form of reward by incentivizing players to follow a certain path on the course. It could also offer time-limited problems by placing it as an excluding choice (which, really, is a problem) alongside a pit area – “will I go faster with more energy or by a free boost?”
    3. The manual boost may be the best rule of the game, and perhaps unsurprisingly a risk-and-reward rule. Energy is mainly used to gain speed, but it’s also your life reserve if some other racer tackles you. The more you boost, the more vulnerable you become to opponents tackling you.
      The manual to the game explains the unlocking at entering the second lap with “the slight margin between participants at the start of the race”, and there is a point to that. Vehicles are very close to each other at the beginning of the race, so claiming and losing positions is much more sensitive to using the regular move set. However, I suspect it has more to do with the interest curve of a race, which I rather crudely made a diagram for below. The start is fairly exciting in itself, but the interest falls as the lap progresses. Enabling boosters at entering lap two causes a spike of interest, which changes the perceived distances of the course, the game becomes faster and less predictable. Entering the third lap isn’t all that exciting, but it climbs as the finish line comes closer, and usually climbs to a climax at the final stretch before the finish line.

      F-zero GX diagram 1

      A subjective interest curve on how I experience an F-zero race.

      To test the theory, I try playing the game’s time trail mode, which unlocks boost powers from the get-go. The result is a race where the first two laps feel overly long and less exciting, although I should note this is a matter of personal opinion.

      F-zero GX diagram 2

      Another subjective interest curve, this time comparing the difference allowing manual boosts to be available from the second lap (gray curve) compared to available the entire race (red cure).

      This same rule also gives all players to learn the tracks before the game begins ”for real”, it creates variety within a race and, to some degree, a Story Arc. During the first lap, players are drawn to good areas or other advantages on the tracks (explosive barrels, boost pads). The second lap start off in an explosive fashion, where players have the full energy bar to deplete before the first pit area. At the last lap, the players can again deplete their energy bar, this time by the end of the lap.

  • The spin attack results in a one-to-many player interaction. This allows players to clear some space around them if they are tightly surrounded, but that is a rare occasion because the racers move so quickly. In combination with losing speed, it is rarely used.
  • Tackles result in a one-to-one player interaction. The target must be within the player’s vicinity, and thereby equally close to the finish line. This is also one of the few times where Yomi layers surface, as the tackle situations gain some attributes of fighting games. One could tackle the other, but it causes a short loss of control and speed. If the other manages to avoid the tackle (by boosting or intentionally falling behind a little), that player clearly takes the lead. That other player could also ignore doing anything, but it leaves the risk of the other player tackling.
  • The rule about needing speed to glide and drift means you don’t need to gas. This loop hole has resulted in some of the advanced techniques mentioned later.
  • The function of the dirt areas is to discourage players from going certain players. This can make straight roads to become short “puzzles” just by dirt areas.
  • The function of the ice areas is similar to the dirt areas, but they fill it in a different way. As it reduces grip, it can be used to make short cuts more challenging or risky.
  • Schell writes that risk taken need to result in some kind of punishment on failure for a risk to actually be a risk. In the case of depleting the vehicle’s energy, the player will lose, or at least lose several seconds. This is also positive feedback, which is too often so strong that it often ends the game prematurely. In a game often won by tenths of a second, five full seconds is a lot.
  • As the game lacks the negative feedback power-ups serve in Mario Kart, the handicap system is the only negative feedback to stabilize a game. It seems to mostly increase the top speed of a racer falling behind. This means the player falling behind has a chance to come back, but needs to perform to actually get there.
  • By default, a race is three laps, but why? Mostly for fun, I and my brother tested a 20-lap race on one of the game’s longer courses. The interest is usually the same for the first two laps, but slowly declines for the coming ten or so laps until you’re merely passing the laps. However, this means investment is in for laps 16 or so, when interest starts to climb again to be very, very intense by the last lap. It could be thanks to a few very certain settings, but it balances itself out in a way that makes the final stretch in these races among the most intense few seconds of gaming I’ve found (it’s pretty much the reason I’ve played the game for so long)
  • The rule of “first to cross the finish line wins” usually defines these kinds of games, but there are some who have experimented with this goal. In Excitetruck, crossing the line first is just one factor of many leading to victory – making big jumps, drifting, passing time limits (in other words, just generally do cool stuff) all gather points, and the player with most points at the end is the winner. It should be noted, though, that changing the victory condition can change the game substantially.
  • That the next last player triggers a timer for the last player is probably two-fold: Partly to avoid lame duck-situation as well as prevent players stalling a race to ruin the play for the others. Of course, the latter reason falls apart if two players stall, but in that situation it’s well worth asking if the stallers really are a nuisance or if it’s actually the other players who ruins their explorative play.

Advanced Techniques

F-zero in-a-cylinder level

This picture doesn’t explain any advanced techniques. However, it does display the kind of twisting level designs featured in the game. This level, for instance, is the inside of a cylinder.

This section covers “advanced techniques”, gathered from F-zero central. These techniques are not explicit in the rules, but have emerged as a result of how the rules are written. Academics would call this “emergence”; other people would probably call it “loop holes”. I prefer the former, and will use it for this article.


”Snaking” comes from using drifts back and forth in a, well, snake-like movement. By combining gliding in one direction, and drifting the other, you can utilize and repeat a small boost which adds up and can give considerable extra speed. Going for the maximum acceleration on the setting between acceleration and top speed seems to use the effect the best.

Snaking is probably not an intended strategy, as selecting heavy vehicles with good grip becomes a dominant strategy on anything but really narrow roads.

Momentum Techniques

Due to how the rule of acceleration is designed, there are situations when it is better to not gas – anytime when the speed is beyond the set top speed and accelerating isn’t a demand for boosting. This forms the basis for a group of techniques called Momentum Techniques, or MT for short.

Boost Delay Techniques

Some actions in the game require the game to gas to have any effect – boosting, for instance. However, you can wait for a short while with a boost before you need to, and still have the effect. This works on boost pads, manual boosts, and also when you’ve successfully managed to gain speed through diving at one of the game’s many falls.

Shift Boost

When a player falls of the road, it gains a short-lived speed boost. If a player saves herself at the very last second, she can use this speed boost and combine with momentum techniques to save a lot of time.

Suicide finish

I forgot to mention a few things about the “0 energy loses the race” rule. You see, when the vehicle is out of the race, the vehicle is thrown forward, and there is a slight lag between reaching 0 energy and triggering the loss/re-spawn. This was probably intended to give you time to observe the change, but it opens a small window where a player can, just before they finish a race, intentionally die to use that extra momentum to cut a few last second off the time.

Side Attacks

Because side attacks move the player ship, it can be used to move the player vehicles in tricky turns, and to combine diving with snaking and a good complement to one of the MTs on certain courses or with certain vehicles.

How advanced techniques change the game

These techniques change the game on a fundamental level

  • When the player should gas and not becomes a problem, rather than a non-issue.
  • The width of the road becomes more relevant, as broader roads are more snaking-friendly.
  • The edges of the course without rails become a risk/reward resource instead of a clear danger.

To visualize, here’s a completely new player (first embedded video) compared to a self-styled world record (second embedded video).

Strengths and Weaknesses

The game rules encourage gameplay focused on speed and competition. There are no luck-based mechanics (on the surface, at least), but there are some clearly intended for quick problem solving or to enhance the sense of speed. This is both an advantage and disadvantage – skilled players will enjoy being master of her own finishing time, but new players may perceive a very steep learning curve to become competitive. The sense of speed can also be a barrier of entry for new players (at least ten years ago when it launched), but is also the main hook (and selling point) of the game.

Getting players past these thresholds was very likely an important goal with the single player part of the game, which will be a clear focus in the single player analysis (if or when I write it).

The negative feedback handicap together with the tackle system result in rising tension in a player the longer other players are close. Racing games often fall victim to being “parallel single player games”, often a result of next to no player interaction, and this yomi-layered design avoids that potential weakness.

Sources (apart from the embedded links)

Schell, J (2008) The Art of Game Design: A Book of Lenses, Morgan Kaufmann Publishers, Burlington

Salen, K, Zimmerman, E (2004) Rules of Play: Game Design Fundamentals, The MIT Press, Cambridge, Massachusetts & London, England


Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s

%d bloggers like this: