Three Pillars of Combat Design: Characters


To get a full view of the systems design of Final Fantasy Tactics requires an examination of the many ways that characters, enemies, and maps vary. For this first article, let’s take a look at the composition of a character. We can start by taking a look at Ramza, the hero of Final Fantasy Tactics, at the very start of the game. Below is a screenshot of Ramza’s status screen in the first battle where the player has full control.

This screenshot shows most of the relevant attributes of a character–there are 34 in total spanning stats, equipment, and abilities. Analyzing these attributes, we can gain a handle on the design space of characters in the battles of Final Fantasy Tactics.

All battles have an objective of either killing a specific enemy unit, or killing all enemy units. Since reducing a character’s HP to zero is the most common way to kill it, HP earns a special place in the design space. All successful strategies have unit preservation (through minimizing HP damage) as a key consideration. Though the player can win battles through spending all of their effort dealing damage via attacks and abilities, many strategies also take advantage of powerful status-inflicting spells to temporarily hamper opponents, or take them out of the fight permanently (regardless of their HP total) by turning them into statues, frogs, or outright killing them by directly inflicting the “Dead” status.

I’ll cover the 33 different status effects in depth in a future article. For now let’s focus on what we can see on the status screen and understand the statistics that make up a character by taking a tour of combat.

The Basics of Combat

As a player of a turn-based game the first two questions that come to mind are “when do I take a turn?” and “what can I do on a turn?”

(From HCBailly’s Let’s Play. If the player were to select “Time Magic” or “Black Magic” here, a further submenu would open letting them choose the specific ability they’d like to use.) (From HCBailly’s Let’s Play. If the player were to select “Time Magic” or “Black Magic” here, a further submenu would open letting them choose the specific ability they’d like to use.)

Final Fantasy Tactics has an interleaved turn system where the player gets to take a turn for each character in an order determined by the characters’ statistics. Characters go individually when they’re ready.

A character’s Speed stat determines how often they get to take a turn. On a character’s turn (called AT by the game) the player can have the character MOVE, ACT, or both, or neither. Choosing not to MOVE or ACT makes the character’s next turn comes faster.

A characters can MOVE at most a number of tiles equal to their Move stat, and any vertical distance between two tiles along the path of movement can’t be higher than their Jump stat.

When the player chooses to ACT with a character, they get to choose to Attack with their weapon or use an ability. Some of these abilities are resolved immediately, while others may take some time during which other characters may take further turns.

Abilities used offensively, i.e. to deal damage or inflict status effects, can be magical (like the Wizard’s spells), physical (like the Monk’s martial arts), or neither (like the Mediator’s talk skill). There are a number of different formulas to determine hit chance and damage for different abilities, but understanding if the ability is physical or magical keys you into what stats the game takes into account when doing the calculations.

Many abilities have a base 100% success rate, like physical abilities, but others have a success rate based on some formula involving the relevant attack stat (Physical Attack, Magical Attack, Speed, and Weapon Power, typically), some constant, and Zodiac compatibility. (I’ll be covering Zodiac-related rules in a future article.) After the die roll for success rate succeeds, then the ability is subject to the defender’s evasion depending on what kind of ability is being used and some other conditions.

Think of evasion as layers of defense. The attacker has to pass each layer by rolling a 100-sided die and getting a result higher than the defender’s evasion percentage. If any die rolls end up lower, the attack misses.

If a magic attack is used, the defender uses their Shield’s (“S-Ev”) and Accessory’s (“A-Ev”) magic evasion percentage to try to cause the attack to miss.

If an attack is physical, the process is slightly more complicated. Which evasion percentages apply to a given physical attack is based on the relative position of attacker and defender. If…

  • the attacker is in front: all evasion stats apply;
  • the attacker is to the side: shield, accessory, and weapon evasion apply;
  • the attacker is to the rear: only accessory evasion applies.


If the attack passes all of these checks, it takes effect on the defender by dealing damage or applying a status or both. Damage formulas for the ATTACK action are based on the weapon involved and usually include some combination of Weapon Power (WP) and Magic Attack (MA), or Physical Attack (MA). Guns only rely on WP, squaring it to determine their damage, and there are a few weapons, as mentioned earlier, which use Brave as a part of their damage calculation.

Damage calculations for magical and other attacks typically depend on MA or PA and some constant factor. Magic attacks damage is additionally multiplied by the Faith of the caster and target as a percentage. I.e. Damage * (My faith / 100) * (Target Faith / 100).


Final Fantasy Tactics has a relatively complicated battle system, where characters are throwing damage and status effects around to try to gain the upper hand–but how do they get these abilities? They get them through the Jobs they’ve had.

Jobs are akin to classes in most RPGs, but act a little differently than you might expect if you’re familiar with how classes tend to work games that borrow from Dungeons & Dragons.

The player can select one Job for each character, and change each character’s Job between battles. This Job determines the character’s stat growth when they level up and provides percentage bonuses to HP, MP, Speed, Physical Attack, and Magical Attack. For example, a Knight has more HP and Physical Attack, less MP and Magical Attack; a Wizard has lots of MP and Magical Attack, but not much HP and Physical Attack; a Ninja has great Speed, good Physical Attack, but low MP and middling other stats.

Each job has a base Move stat and Jump stat, as well. Ninjas are much more mobile than Knights not only because their Speed is higher, but also because they can Jump and Move one tile more.

A character gets a group of abilities from their Job. Knights get “Battle Skill” abilities which let them break opponents’ equipment; Wizards get “Black Magic” abilities which let them deal elemental damage at range; Ninjas get “Throw” which lets them toss various items at enemies to damage them based on the item’s stats. Characters can use these abilities during the ACT part of their turn. But before a character can use such an ability, the player must unlock it for that character by paying its cost in Job Points (JP) between battles. Each character has a JP pool for each job. A character gains JP for a Job when they successfully ACT while they have that Job–abilities have to hit their target, friend or foe, for JP to be awarded.

The player can equip each character with an additional ability group from another Job, as well as the “automatic” ability group granted by the Job the character currently has. This secondary ability group can be used just the same as the ability group given by the character’s current Job.


Abilities are what really bring the characters to life in battle–the abilities the player equips to his characters, and the abilities that appear on the various AI-controlled enemies, are the main source of variety in the battle system.

In addition to the two job ability groups mentioned above, the player can equip each character with one Reaction Ability, one Support Ability. and one Movement Ability. Below is a screenshot of the Summoner’s Active Ability set, known as “Summon Magic”.

(The Summoner’s Active ability set. From SnapWave’s Let’s Play.) (The Summoner’s Active ability set. From SnapWave’s Let’s Play.)

Reaction abilities have a percentage chance of triggering their effect based on the character’s Brave stat when the character has certain kinds of abilities used on them. Reaction abilities give the character a chance to immediately react to being hit by, say, countering a spell with a spell of their own, or doing a physical attack in response to an enemy’s physical attack. Other reactions include increasing the character’s speed stat by 1, adding a positive status effect, and automatically using a potion to restore its health.

Support abilities are a motley assortment of passive abilities and a couple of active abilities. Examples: Increasing physical or magical attack power, granting the ability to throw curative items to nearby allies, and granting the ability to go into a defensive posture that doubles all evasion. Support abilities can also do much more unusual things, like unlocking more powerful abilities for nearby monsters, and “poaching” rare items off of monsters killed by this character.

Movement Abilities give the character some benefit each time they MOVE during their turn, or somehow enhance their MOVE action. Benefits can include finding hidden items, increasing the jump or move stat, ignoring height all together, or gaining HP, MP, JP, or XP after each MOVE action.


Below I’ve enumerated the stats and bonuses for the equipment on two late-game characters, so you can get a sense of what role equipment tends to play.

(From GetDaved!’s Let’s Play.) (From GetDaved!’s Let’s Play.)

Equipped Item Effect
Kikuichimoji 15 Weapon Power, 15 Weapon Evasion
Crystal Helmet +120 HP
Genji Armor +150 HP
Bracer +3 Physical Attack

This Ramza is configured to be able to absorb physical attacks. His high HP allows him to take the hits that his relatively high evasion doesn’t prevent outright. The Kukuichimoji sword allows Ramza to deal PA * Brave / 100 * WP damage, which in this case works out to 225 (This is lower than you would expect from the equation because the PSX doesn’t support floating-point math and will drop the remainder of PA * Brave / 100 before multiplying that result by WP.) Having the sword also allows Ramza to use a Job ability which deals MA * 16 = 144 damage to enemies in a line eight tiles long from either Ramza’s front, one of his sides, or rear.

The only way the player can increase the HP and MP of his characters, short of leveling them up or changing jobs to something tankier, is to equip them with different items. Typically body and head items provide the most HP and MP. Ramza, who is built as a melee combatant in the screenshot above, gets 270 of his 398 HP from his equipment. In this particular case, it’s clear how armor plays a large role in Ramza’s ability to absorb damage, since it accounts for two-thirds of his HP.

Ramza’s Job, Samurai, contributes 20% to his evasion rate (listed under “C-Ev”) while his katana contributes 15% (listed under “Weap.Power” after the slash), meaning an opponent physically attacking him from the front will only hit 68% of the time. Meanwhile, Ramza has no magical evasion whatsoever, so he’ll be quite vulnerable to magical attacks. He could put on an accessory, such as the Feather Mantle–the best such item in the game–which gives 30% magic evade and 40% physical evade, to improve his survivability against magical attacks. Remember that evasion granted by an accessory such as the Feather Mantle applies to attacks regardless of where they’re coming from–even if an enemy is attacking Ramza from behind he’ll get that 40% physical evasion!

(From HCBailly’s Let’s Play) (From HCBailly’s Let’s Play)

Equipped Item Effect
Blaze Gun 20 Weapon Power, 5 Weapon Evasion
Thief Hat +100 HP, +2 Speed, prevents Don’t Move and Don’t Act status effects.
Black Costume +100 HP, prevents Stop status effect.
Germinas Boots +1 Move, +1 Jump

Mustadio, like Ramza, relies on equipment for a significant chunk of his HP–in fact, even more than Ramza–but Mustadio’s build has more of a focus on staying away from enemies and hitting them with his gun. Guns are unique among weapons in that their damage is entirely based on the gun’s Weapon Power and has nothing to do with the wielder’s Physical or Magical Attack stats. Guns also can strike from the longest distance of any weapon attack in the game, with a range of 3 to 8 tiles. In contrast, crossbows tend to have a range of 3 to 4 and longbows have a range of 3 to 5. Mustadio’s gun, the Blaze Gun, is a magical gun, which acts more like a magical attack in that its damage is proportional to the Faith of the caster and target–its effectiveness still has nothing to do with PA and MA stats, though! Another important, unique detail about guns is that they ignore evasion entirely. They typically do less raw damage in exchange for hitting more often. It’s a great way to counter evasion-rich characters like the Ramza we just took a look at.

To stay away from enemy combatants who can take advantage of Mustadio’s woeful evasion of 5% from his weapon and 5% from his class, Mustadio relies on both his modest Move and good Jump, as well as his equipment’s capacity to prevent the most common status effects that would pin him down: Don’t Move would hold him in place and Stop would prevent him from taking turns all together.


This is Just an Introduction

It took many words just to give you the basics of the way characters are designed in Final Fantasy Tactics, and there is still so much to talk about. Over the next month expect  a number of articles breaking down in detail many of the design facets discussed here. Stay tuned!


Reblog: Camera Design in Wario Land 4 by Daniel Johnson


In this video extension of Daniel’s book, Game Design Companion: A Critical Analysis of Wario Land 4, Daniel breaks down the various types of cameras used in the game.

Richard says: I read that book! Daniel is a good friend of ours. We fought in the games criticism trenches with our blogs back in 2008. His book is comprehensive and organized, listing every mechanic, level element, enemy element, level challenge, boss, and secret in the entire game. Best of all, he frames it in a game design context. A great example of how to study and reverse engineer a video game.

Marcus says: Like Itay’s feature Scroll Back that we reblogged previously, Daniel takes a look at the specific 2D camera techniques used in Wario Land 4. Even though Daniel’s work focuses on one game, his terms and definitions for describing what actions the camera takes and how those actions affect gameplay falls in line with Itay’s more comprehensive look of camera across many games. The biggest difference in their presentations is their delivery of the material, namely Daniel’s video vs Itay’s animated gifs. As I’ve said before, I think the bite size repetition of an animated gif lends itself well for quickly communicating, sharing, and referencing visual information (e.g. camera techniques). So, I took an animated gif approach to Daniel’s video.

Position-Locking (Horizontal, Vertical) Edge Snap Region-Based-Anchors Position-Locking (Horizontal, Vertical) Edge Snap Region-Based-Anchors Position-Locking (Horizontal) Position-Locking (Horizontal) Region-Based-Anchor Forward Focus Region-Based-Anchor Forward Focus  Forward-Focus *Threshold Triggered

Now freed from the confines of youtube, these bite sized pieces of Daniel’s work can now move at the speed of social media. They’re so quick and effective that you might even find something like it in future Design Oriented tweets.

Terms used above are from Itay’s glossary:

  • Position Locking – Camera is locking to the player’s position.
  • Edge Snap    – Set hard edge for camera positioning
  • Region-Based-Anchors – Different regions (even within levels) set different anchors for position and focus
  • Forward-Focus – Player direction changes switch camera focus to enable wide forward view

Reblog: Perfect Dark Retrospective by Mark Johnson


[Perfect Dark Retrospective] is a post by Mark R. Johnson that addresses critical claims about Perfect Dark.  He takes a detailed look at several levels in Perfect Dark and concludes by considering the validity of the critics’ claims. Topics include level design, difficulty design, tutorial design, and modern FPS trends.

Richard says: Let’s go Perfect Dark! Still my favorite single-player FPS campaign. Mark R Johnson does a great job setting up the conversation with quotes from reviewers. Mark picks a few of the most interesting levels in the game due to their differences between difficulty and their use, and re-use, of space. His descriptions are accurate, though I imagine the screenshots are hard to follow if you’re not familiar with Perfect Dark.

Mark analyzes Perfect Dark levels for “completeness” to counter the claims made by many reviewers that the game is filled with dead ends. Mark illustrates how interconnected the areas, missions, and optional sub-missions are across the level and across the three different difficulty modes. Perfect Dark, after all, is a rare FPS that adds objectives, alters objectives, and changes the path of a level to create distinct experiences.

Mark addresses Perfect Dark’s unique “complete” level design to counter part of the critic’s claims, and then agrees with them regarding the game’s feedback issues. The environment design of the game makes it fairly easy for players to get lost at times. Coupled with the non-linear level design (players can take many paths and accomplish mission objectives in different orders) getting lost is more likely and more frustrating than in linear games. Mark explains the problems of Perfect Dark’s level design, and ends by arguing that “contemporary FPS games” have gone too far in the other design direction to remedy this problem by designing levels where the path to take is more obvious and the level design is more linear. 

Perfect Dark is a complex game to analyze in terms of how it teaches, guides, and challenges players. Mark has only scratched the surface by highlighting the coherent details with Perfect Dark’s level design.

Mike Says: It’s interesting how players, even those who have been playing games for more than a decade, have come to rely on the linearity of modern shooter level design on a subconscious level; how it’s far from certain–in fact it is becoming less common by the day–that a player will pick up on the more advanced or subtle level design concepts that could clarify the seemingly arbitrary locked door here or samey-looking room there. A game like Perfect Dark ends up suffering not because of some inherent objective weakness in its level design, but because the kind of headspace required to enjoy it as it was enjoyed shortly after release just isn’t practiced and available to players like it used to be, either because those players never had to practice it, or because they’ve long since been retrained by the zeitgeist of shooter level design.

Pac-Man Design: Deterministic and Random Ghosts


My recent deep dive into Pac-Man’s design and high-level gameplay got me thinking about randomness and deterministic design elements. I’m quite a fan of randomness in game design, though it has its pros and cons. Pros include variation and unpredictability that force player improvisation. Cons include a lowered skill ceiling and a lack of player control. 

The original Pac-Man arcade version strikes a delicate balance between randomization and determinism with its Ghost AI design. In Pac-Man the Ghosts change back and forth between chase and scatter AI modes. In both modes, Ghost movement is determined by a number of factors including the other Ghosts’ positions and Pac-Man’s position. Given the same positions, the Ghosts will act the same way every time.

When Pac-Man eats a power-pellet, the Ghosts run from Pac-Man in a frightened state. At every turn, they make a random decision which way to go. During this time, Pac-Man usually has zero threats on the field, so the random Ghost movement mostly affects the player’s ability to maximize score as opposed to limiting their ability to make informed decisions at times when their life is on the line. It’s a touch of randomness that that proves the skill of the designer and the potency of randomness as a design element.

On the highest difficulty levels, Ghosts don’t turn blue and don’t become vulnerable even when Pac-Man eats a Power-Pellet! Here, the exact positions of Ghosts is the result of a complex web of deterministic rules. If you play precisely the same way every time, the Ghosts will move predictably through the maze. Of course, it takes considerable planning and skill to work out a whole pathing “patterns” for these levels, but some elite players have done it and can pull them off reliably.

images from images from

If the perfectly planned pathing pattern is mistimed by even a frame or two, if Pac-Man’s movements are not in exactly the right place at the right time, the Ghost AI will react differently, starting a chain reaction of unplanned and less predictable Ghost movement. When that happens, even the best players have to improvise.

Low- and high-level players both experience the pros of randomness playing the Pac-Man arcade game. The random Ghost frighten movement creates unpredictable challenges without threatening the player in a way that creates unfair deaths. Both knowledge and adaptation (improvisation) are required to succeed. But on the final level of difficulty, expert players engage Pac-Man in a new way, using their skill to maximize score and minimize the unknown. To play at this level of skill, players need impeccable timing and incredible precision. With a microscopic misstep, even the most pre-planned gameplay turns back into the Pac-Man experience that everyone is familiar with.

Reblog: Scroll Back by Itay Keren


in Scroll Back Itay Keren goes on a journey through the history of 2D game cameras to help him understand and ultimately solve camera design problems he had developing his own game, Mushroom 11.

Marcus Says: As Itay notes in the introduction, it is surprising that the common and long standing topic of 2d camera design has so little literature. I thought I had a good grasp of 2D camera design. But reading through Itay’s post, I experienced numerous Oprah “aha” moments.  The camera types outlined in the glossary are simple, yet their application varies greatly from game to game especially when combined with with other camera techniques. And Itay backs everything up with examples, which include animated diagrams. The lines he overlaid on each gif to show tracking, position, panning, ect. is a simple and effective way to see the nuances of the camera design. It’s a great trick for his gifs and a handy tool for critics; all we have to do is hold our thumb up to a screen of game footage to get a sense of how the camera is designed. The whole post is slick. I’m convinced that animated loops might be the best way to show camera movement techniques, not just in games but movies and shows as well.

Reader Response: “On Perfect Imbalance”

In response to my article On “Perfect Imbalance”, Travis Clark wrote in with his insightful perspective on balance and memorization in Chess.


As for the state of balance in Chess, Travis makes a great point about the subtle nature of the imbalances present in what appears to be a symmetrical system:

“It is almost symmetrical, and visually tricks people into to thinking so with the optical-illusion of black and white squares and the pieces in front of you line up with pieces in-front of your opponent. Fold the board in half along its supposed line of symmetry and you will find every black square folds on top of a white square. The board itself isn’t meant to be symmetrical.  Its a very small difference but bishops are assigned to move along a specific color square so the board itself creates some very slight but uneven gameplay.  Now take a close look at the pieces. Things look a little different if you are playing black versus playing white.  If you are playing as white your king is on the fourth square from the right.  If you are playing black, you must look from black’s perspective, your king is on the fifth square from the right.  The situation for the queens are similar.  I believe it is this way because it is aesthetically pleasing.  However, every opening strategy is hinged on this very slightly imbalanced set-up.” ~Travis

In the original article, I had considered making a comment on the fact that Chess is not a symmetrical game, primarily because of White’s first-turn advantage, but Travis brings up a good further point about how the aesthetically-pleasing apparently-symmetrical setup is hiding some one-square positioning differences that rest at the core of the strategy-space of the game. He later also mentioned the importance of first-turn advantage as well:

“The most drastic way chess is imbalanced is the fact that white always goes first.  You may wonder how much of a difference white’s privilege actually makes. …Because of the small differences in starting positions and the advantage white has in time there is a Chess theory which states that if two perfect computers were to play each-other, with all the Chess databases known to man for reference material the best outcome black could hope for would be a draw in every game.” ~Travis

Travis pointed out a memorization-free style of learning chess that worked well for him:

“Josh Waitzkin is one of my chess heroes.  He wrote an excellent book called The Art of Learning that I highly recommend, and in partnership with Chessmaster(Ubisoft) created a fantastic chess course.  The reason why I was drawn in to chess so deeply and find it fascinating are his annotated games.  His style of play and teaching didn’t revolve around the learning of specific patterns or memorizing databases. In place of the rote portion he would explain the theory of why certain opening patterns developed and what the Grandmaster’s that used them were trying to accomplish what they gave up and got in return and how it suited their style of play.  I never once had been one to memorize openings.  Using his training I could look at a completely unfamiliar opening, that my opponent may have memorized, and look at its goal and strengths and weaknesses and choose a move based on that.” ~Travis


And he went on to talk about the mind-games that can be involved in Chess:

“But the heart of chess is this competitive spirit.  Connecting with someone through competition is the key.  Memorization is a piece of the puzzle but I argue that it is a much smaller piece than the current image of chess projects.   Raw calculation is indeed a  factor along with pattern recognition, logic, strategy, poise and other mental talents; however, I believe the biggest portion is the psychology of competition.  You have to be in tune with what you opponent is thinking and find a move that counters, disrupts, or solidifies your opponent’s thought process while fitting the current position on the board.  The correct move isn’t always the mathematical best move, rather it is the move that will beat the opponent in right in-front of you in that moment with their current state of mind.  As annotated by Josh, this psychological warfare happens on the grandmaster level, the beginner level and everything in-between.  That is why I argue that it is the heart of this slightly imbalanced non-symmetrical game.”  ~Travis


It’s a great point, and one that I think is too often thrown by the wayside when we try to talk about design in the abstract. People are playing these games, and seldom can they near the mathematical precision of a computer, in terms of memory and applying algorithms and valuation strategies effectively and consistently. Even though Chess is a perfectly deterministic game involving no hidden information that would affect objectively optimal play, a degree of psychologically-hidden information exists because of the limitations of the human mind’s ability to apprehend the many branches of the decisions tree. Even at the highest levels, says Travis, players benefit from exploiting the limitations of the wetware of their opponent while trying to avoid being exploited themselves. Memorization might be an “ideal” solution to chess, but as the metagame stands now, there’s plenty of room for human limitations to be exploited by using the mind and emotions of the opponent as a weapon.


Thanks for the great email, Travis!

Reblog: Assault Android Cactus – Official Trailer

Assault Android Cactus is an arcade style twin-stick shooter set in a vivid scifi universe. Currently on Steam Early Access, Assault Android Cactus will be released on PC, Playstation 4, Playstation Vita and WiiU later this year.

Richard says: I got a chance to play Assault Android Cactus about a year and a half ago at GDC2014. It was one of those experiences where I kinda got a feel for the basics and then struggled to stay alive as the game got progressively harder. I’m not sure if playing it co-op made the game easier. Enemy chase AI and aiming has subtle complexity when there are more players on the field. Instead of the Geometry Wars style cone of enemies following the player, new enemies patterns are created as they switch back and forth between co-op players. Just having more “present bodies” on the field is a big deal.

Mike says: I’m curious to see how this game will rate on the Vlambeer Scale. Looks like it could do pretty well. Lots of big shiny bullets and explosions with a relatively minimal but effective amount of screenshake. I’m concerned a little about feedback clarity, because it looks like there can be a lot going on at once. Perhaps 3D graphics add to the clutter here, as well, since they’re more complex to interpret than the 2D sprites typically used in bullet hell shooters. The bullet count in Assault Android Cactus looks like it’s much lower than your typical bullet hell shooter, so maybe there won’t be a such a feedback issue. I sure hope feedback works out well, because this looks like a game I’ll enjoy.

Guest Jonathan says: The dynamic shifting arenas and enemy variety make it stand out from other twin-stick shooters that I played from the PSN store. The glowing blue bullets are great and as long as enemies and particle effects don’t obscure that, the game should be good with visual clarity.

Look forward to our Design Oriented Chat with the developers of Assault Android Cactus next week. 

There Are No Nuclear Explosions in Baseball


In Pitching a Nuclear Baseball Marcus considers and answers two interesting questions concerning the game of baseball. First, “What would happen to a hypothetical game of baseball if the pitcher were to pitch a ball at 90% the speed of light?” And second, “What should happen in that hypothetical game according to the rules of baseball?”

In answering the first question, Marcus relies on science to conclude that the pitch would cause a nuclear explosion that would obliterate the baseball stadium and all of its inhabitants. Because I accept the judgments of science, I agree.

In answering the second question, Marcus relies on an interpretation of the rules of baseball to conclude that the batter – being hit by the pitched ball in the ensuing nuclear explosion – should advance to first base. Because I accept the judgments of common sense, I disagree. There are no nuclear explosions in baseball, because that would be absurd!

But Marcus has an argument:

As long as there is an umpire to interpret the rules, the game of baseball is able to be played in all sorts of abnormal conditions. This is not because the rules account for the unexpected but because they don’t! Apparently the rules check for a legal ball, a legal set up, and a legal pitch, so everything that happens in between the pitcher’s mound and the batter’s box is up for grabs. Broken sound barriers? Baseball doesn’t care. Rapidly expanding walls of plasma due to nuclear explosion? Baseball doesn’t care.

Marcus seems to be saying that if the rules of baseball don’t explicitly ban nuclear explosions, they must be implicitly supported. So his claim in this particular instance – that nuclear explosions are allowed in baseball – looks to be founded on a more general interpretive stance – that only the explicit text of a game’s rules need to be consulted in order to determine what those rules authorize.

I advocate a different interpretive stance for answering questions about what a game’s rules authorize. This interpretive stance goes beyond merely consulting the explicit text of a game’s rules to consider other factors related to a game’s design. The explicit text of a game’s rules should of course be your first, best guide. But one can’t exclude from consideration other factors of a game’s design space – game objectives, balance, and strategic variety (to name just a few).

Consider the central game objective in baseball. The players’ objective is not – nuclear baseballs notwithstanding – to physically annihilate their opponents. It is to score more runs. Since allowing for nuclear explosions in baseball would run contrary to the fulfillment of baseball’s central game objective, it’s implausible to think that the rules of baseball would allow this. Or consider baseball’s overall balance. One needn’t have a perfect understanding of all of the components of the game to understand that allowing for nuclear explosions would be a completely game-breaking strategic option akin to allowing players of Monopoly to legally and literally flip the board whenever they were at risk of losing the game.

The explicit text of a game is useful when confronting circumstances in which there is only one plausible interpretation, but when it comes to the gray areas of the implicit – or the mushroom clouds of nuclear explosions – we have to rely on more than this. We have to rely on ourselves and our own sense of a game’s design space.

So you can pitch the ball as hard as you like. Just be careful. If the speed of your ball approaches the speed of light, it’s game over.

Reblog: Rabbit Hole Game Design

In Rabbit Hole: Game Design Philipp Zupke comments on just how precarious game design can be as a discipline. It’s hard to know what players will enjoy. Trial-and-error methods when designing games are often all designers have to rely upon. It shows just how far we have to go as analysts and designers alike.

Mike Says: As I continue to participate in the design process of games–in my case it’s all turn-based strategy games–I can’t help but notice how little I know and how easily I am surprised by the results of playtests. I’m sure that with more practice my ability to predict player reactions will improve. I wonder if improved prediction abilities will only result in having improved intuition, or if it will also give me the ability to put my knowledge into words and help others start from a more advanced place and make fewer mistakes when designing their own games.

Marcus Says: I know Disney is often criticized for an overly positive outlook on any given situation. But when Disney makes a movie about how not thinking Disney-like will result in the destruction of the planet, I think it’s in our best interest to feed the right (positive) mindset (wolf). Worried about whether or not you’re qualified to charge for your games? Don’t. Let the individual decide whether or not they want to buy your game. Worried that your struggle in grasping the science of game design will always be a struggle? Don’t. If literature, radio, TV, and movies are any guide, we will never develop that science. It’s easy to let our fears get the best of us. There is no need to feed them a continual cycle of negative feedback from questions that can never be answered.

Richard says: Philipp Zupke has a pretty good attitude. Though I think he underestimates the potential of trial and error learning. You can use such a learning method to merely separate the right answer from the wrong, or you can use it as a necessary information gathering step before extrapolating rules, which can then be applied to other situations. Trial-and-error is just experimentation. You can do it efficiently and well, or waste your time. I can relate to the feeling that there isn’t a good way to understand game design or good guides to follow. The last time I had that feeling was about 8 years ago, just before I started the Critical-Gaming blog to teach myself the ins and outs of game design. Since then there have been many more books published, youtube videos made, GDC quality talks posted, and a stronger social network to leverage. So maybe the somewhat awestruck and gloomy attitude about “fun” and “game design” is just a necessary first step before one dives head first into the ocean that is this creative field. Philipp, you can always chat with us about it.

Reblog: The Problem with the Roguelike Metagame

In The Problem with the Roguelike Metagame Mark R. Johnson analyzes the impact unlocks have on roguelike players’ psychology. He makes the case that players see the unlocks as the reason why they can’t win, and become fixated on the fact that they don’t have the unlocks instead of focusing on learning the game and overcoming challenges by improving their skill. Mark holds that unlocks are therefore a detriment to the player experience if we design roguelikes with the intent that sufficiently skilled players should be able to win the game roughly 100% of the time.

Mike Says: I hadn’t given much thought to unlock design in roguelikes, but Mark makes a believable argument. In my experience if a roguelike doesn’t give me access to options unless I fulfill some arbitrary condition in a prior playthrough, I’m much more likely to get bored with the game before unlocking even half the content: I’ll probably lose a bunch of times without unlocking anything, get tired of the character I have to play, and at that point shelve the game.

Richard Says: Mark does a pretty solid job explaining his ideas. He starts by breaking things down. Then he presents a clear argument. Drawing from and listening to the community is a move straight out of my book! He’s non-aggressive and open to constructive debate. Someone give this man a gold star, or a ship, or power-up,  or whatever!

Mark’s model of the learning player is simplistic to a fault, and he puts too much genre focus on the conventions of the roguelike genre instead of the genre’s underlying design, which includes similarities to most skill-based games. Pick any genre or any hard game and you have a similar situation where players are forced to learn from their play attempts. If there’s any persistence, unlocks, or even suspended power-ups like in Super Mario Bros., you’ll find that most beginning players will lean on those systems, but not to the exclusion of understanding the consequences of their gameplay. It’s because they understand (or are beginning to) why they lost that they put together a plan of action that involves using an unlock, power-up, etc.

Part of Mark’s central argument builds on a weak correlation between unlocks and how players learn complex games from their attempts. Roguelikes are strategy games, which stress player knowledge skill most of all. There’s typically a lot of information to learn and no wrong way to learn it. Until the player understands a large amount of gameplay complexities (rules/data) they will make uninformed decisions and therefore their play experience won’t be one of interesting choices. So I don’t blame players who reach for helper items, unlocks, or other “cheating” upgrades. Being thrown head first into the deep end of a gameplay system like classic roguelikes and learning your way to victory is a pretty ridiculous task. This is probably the reason why roguelikes are so niche. The logic behind Mark’s argument reminds me of the theoretical player perspective.

Chris says: I liked this article a lot, though I think Richard is right that he focuses a bit too much on a niche genre to make his central point. Mark’s argument can be applied elsewhere. I do wish that Mark had expanded a bit more on what he considers to be the “undesirable mindset” generated by unlock systems in roguelikes. He says this mindset consists in thinking “that the game is to blame for their deaths.” But the game is always to blame for a player’s death. Clearly, Mark knows this and is trying to get at some unique sense of unfairness players feel when they die in roguelikes that feature an unlock system as opposed to roguelikes that force players to confront their own strategic failings. But he doesn’t fully elucidate where this unique sense of unfairness comes from.