In my previous post: “The Game Design Rabbit Hole” I have conducted my struggles with game design and it being not as a solid craft as I wish it was. This is yet another attempt to shed some light into the dark. And of course, this merely reflects my own opinions and learnings.
Raph Koster’s “Theory of Fun” is that the source of all fun is learning. Confronting the player with a challenge and giving her the capabilities to overcome it. It’s a task for the brain to figure out through an iterative process in which players refine their strategies up to the point where they are able to overcome the opposed obstacle. I want to dive into that a bit more and share my thought on how exactly this approach can help to take apart the process of designing games into more tangible pieces.
The Game -Loop
The player and the game are communicating with each other. The player speaks the “language” of input and the game answers with “feedback”. Players then form a mental model of the game system and the inner working of its parts. Through Inputs, they can modify parts of that system. The game will then alternate its current state and communicate that change to the player. These Inputs can be intentional, if the player has already learned what the resulting state change can be, or experimental if the player is still figuring out that part of the game. Either way, the feedback system will inform the player about the impact of these inputs and the player can validate it. Even more importantly, the players gain knowledge about how the game-system works and can in turn, update her understanding of the game-system. The intention is informed by goals which either emerge intrinsically in the players mind, or the game can explicitly give these goals extrinsically. More extrinsic goals are better for the early game when players still need assistance in learning the basics of the game. Extrinsic goals, if rewarded properly, will align with intrinsic goals eventually. “Reaching level 90” in an RPG is an extrinsic goal, while “I want to become more powerful” is an intrinsic goal. Ideally, they match, but first players have to learn that leveling up makes them more powerful. With that, the next loop of the feedback cycle starts. Every iteration of that cycle is an opportunity for players to learn. If the players knowledge of the games system has been established well enough (by running the feedback loop a sufficient number of times) the inputs players make can become intentional towards the desired game-state. The “desired game-state” can be an explicit goal stated within the game (e.g. finish a level) or a player-internal goal (I want to build a nice city). Throughout a game, the players understanding of the increases and a well-paced game will alternate between applying and learning strategies.
Strategies & Goals
I use the word strategy very broadly in this text and it means the following:
“A strategy is the intentional alternation of the game state through a sequence of inputs that lead to a desired state”.
The “desired game state” implies an intention on the player’s side. She wants the game to do something specific, so she uses her knowledge of the game to manipulate it. It is somewhat related to what Daniel Cook describes as “Skill Atom”. Although a “strategy” in this text can describe the usage of multiple “Skill Atoms” to achieve a particular result within the game world.
This definition can be applied to all levels of a game system. From a simple mechanics e.g.
Press the jump- button at the right velocity & position to perform a jump collect a pickup in Platformers
Placing units behind cover so they receive less damage in XCOM, spread units to mitigate Alien-AoE Damage
Throw a smoke grenade into a tunnel to deny an enemy rush in Counterstrike
Only shoot in small burst to maintain accuracy (FPS with recoil mechanic)
Place an observer at the enemy base to gain high ground vision ( Starcraft 2)
Find the biggest match in order to get the biggest possible cascading effect (Candy Crush)
Deploy skeletons first when attacking an enemy base (Clash of Clans)
Spend all of the mana in a turn in Hearthstone
Defending enemy attacks in a Beat’em’up
Execute headshots for increased damage
Stock up on pokeballs before leaving a town
To more complex ones– these strategies require a deep understanding of a games systems and require some iterations of the feedback cycle before they have been formed and mastered
Find efficient city layouts to maximize population happiness and thereby tax income
Min-Maxing Character stats in an MMO
Building a deck in Hearthstone around a new card
Using certain attacks to force the opponent into a bad position and then exploit that weakness with a finishing move/combo
Only pick up the most valuable goods in “This war of Mine’s” scavenging mode
The skill a player uses to execute a strategy varies greatly from genre to genre. Executing a headshot requires accurate mouse movement & reaction time while placing buildings efficiently requires strategic thinking. Both are strategies that work towards a goal in the context of their game. The applicability of this model to almost all genres makes it such a powerful. Learning strategies and combining them is what makes games entertaining and what keeps them challenging – given that the game provides more complex goals and deeper systems to interact with. A game becomes boring when there is nothing new to learn, or there is no incentive to combine learned strategies to reach more complex goals. This can be achieved by either increasing the depth or the complexity of the mechanics. Players attempt a solution of the goal with a certain strategy. The game will inform the player whether the attempt was successful or not. This iterative process how learning works and is engraved into the human brains by rewarding its success. When it comes to difficulty I consider creating challenges that gives players the opportunity to show how smart they are rather than showing them that they are inadequate. I think this is exactly where the border between challenging and punishing difficulty lies, but that is a totally different topic in itself.
Actionable: “Look at the mechanics of your game and list possible strategies. Map these strategies to possible explicit and implicit goals. Validate whether all strategies are tested by the goals and whether some your mechanics are underrepresented in your strategies.”
In the players Mind
The player’s ability to act inside of the game context hinges on 3 aspects:
Understanding the System: How much of the games system has been understood
Understanding the Inputs: To what extent is the player aware of all the possibilities she has to interact with game-systems (not the physical inputs)
Transmuting: Combining system-knowledge and input-knowledge to create a hypothesis about a new strategy: “I wonder if I can avoid taking damage by jumping over that enemy”, depending on the target audience this part is either left to the player through experimenting or has to be explained explicitly
Applying: Ability to use acquired strategies to solve the games challenge
This perspective allows to break down “player skill” into more tangible parts that can be addressed individually. Especially the first 2 are directly under the control of the designers through good tutorials and an elegant pacing of introducing new game elements and systems.
The latter two are not in the direct control of the designers are they happen mostly in the players mind, but with proper guidance (e.g. highlighting vulnerable parts in a boss-fight) we can assist players in finding solutions. Games test the “Transmute” and “Apply” categories with the goals they present to the players. When a player solves a challenge, she has proven that she has mastered the strategies associated with that goal.
Actionable 1: “Understand how your games performs in teaching the systems (through feedback) and the inputs (possibly through tutorials). Ensure that the challenge matches the player’s abilities to transmute & apply (Flow). “
Actionable 2: “What strategies do you teach explicitly and which ones do you leave for the player to discover? Discovering new strategies is one of the most rewarding moments in a game, but leaving the player too helpless can lead to frustration.”
The Hook: Player Agency – Rewarding Curiosity
Players invest upfront into a game by buying, downloading and launching it. So the game starts with a small debt towards the player. We know what the player want in general (“A nice experience”). So it should be the games’ first task to repay that debt and give players something in return. The more we know about the target audience for a certain game, the more accurate we can deliver that. Introducing the players to the core principles of the game and letting her execute that and yield first successes (as in ascertaining the games systems by experiencing the feedback-loop) will hook the players and motivate them to further explore the game. How do we do that? We show them that their inputs have an impact and lay out breadcrumbs for players to discover and figure out. These breadcrumbs can be a piece of narration about the world, a new ability, a new enemy. Everything that rewards curiosity or that introduces new elements keeps the players interested. This spiral goes on as long as the game actually makes the effort to encourage the players’ curiosity. At no point, the game becomes a “no brainer”. The game always needs to promote progression.
Empowering players through learning and applying the knowledge/skills, conveys a sense of accomplishment and mastery which ultimately is the core of a game experience and is what hooks a player.
Games provide an obstacle (be it a QT in a “walking simulator” or a boss fight in Dark Souls), but also the tools to solve them. Beating a challenge is already rewarding in itself intrinsically. If the player has reached the intended game state by using the abilities the game has taught her she has a feeling of agency and power within the game world.
Despite the resentments from core gamers and even some stubborn game designers, “Heavy Rain” just works like “Dark Souls” in terms of player agency, Players are in control of the games system an progress through the game-world on behalf of their intentions and the resulting actions. And that progress is communicated through whatever extrinsic rewards the particular genre has to offer. A cut-scene (narrative reward), a level-up (power reward), a new level (content reward) etc.
Actionable: “Is the feedback for the mechanics clear enough so that the player know what consequences their actions have? Carefully consider what events in the game world are tied to the players input and what are systematically generated (e.g. by randomness)”
Many things in this article were “stating the obvious” for most designers out there. But my approach to these articles is more of a “What would I have liked to have known when I started designing Games”? I hope that this approach of dissecting the individual parts of the game experience will help to single out certain design challenges that we face along the way of creating awesome games.
My belief is, that the better I understand how games work and what drives players, the better I can “design” systems that fulfill a certain purpose: Bringing fun to players
The process of designing “fun” has always been and is still is an undefined matter for me. All the books, all the articles, all the thoughts expressed on how we designers should approach this task just left me with even more questions than before. I became to realize that there hardly is a common understanding of our craft. Though I don’t believe that good games are just happening by accident or good processes (such as iterative design & prototyping). There seems to be some truth that every successful designer has his or her version of.
There are plenty of ideas and concepts on how to view the player, how to define the interactions between players and the game, what a game is, what a game needs, what makes a game fun. Checklists, lenses, models as plenty as there are designers. I believe, eventually a game designer will end up with its very own and unique variant on how to design games and how they work. But where would I start? How would I become one of those designers who seems to understand how all the elements come together, crafting an experience that is enjoyable for others.
Prototyping! That must be the way. And there is no doubt about prototyping teaching loads of valuable lessons about _finding_ fun. Cranking out prototypes, iterating their design and testing them is one of the best ways to unveil the inner workings of games. It will teach about what lets fun emerge, not so much about why it emerges though. The fact that we must rely on prototyping, a trial and error approach, tells me, that we still have a long road ahead of us to bring together those two domains. Sharpening the design skills will reduce the amount of iterations a designer needs to get from A to B and there will be less “failures”, although each failed iteration is the chance to learn something new. I believe that by learning and understanding more of our craft, we can speed up this process and make it more efficient – simply by avoiding mistakes.
Trying to Improve my Skills
I had my own experiences from playing games. At least I knew what fun is for me or could try to tell why certain things where not. In addition I could dwell on the shared thoughts of more experienced designers and people who have walked the same path that I am was about to pursue. I wanted to stand on the shoulders of giants.
Since I had already played a lot of games, and read a lot on the subject, my goal was to absorb all the different ideas expressed by other designers and researchers. Then I would reflect upon them by applying them on my own experiences when playing games. In the hope of eventually building up my own understanding of the elements and their arrangement to create “fun”- games (I would like to apply them also to other people’s experiences but that is rather hard unfortunately).
Underneath there is one terrifying fear of which which I don’t know whether other game designers share it: What justifies me earning money from something that I can barely grasp?
Daniel Cook has expressed an analogy that I appreciate very much – comparing today’s game designers to alchemist in the days before there was chemistry ( The Chemistry of Game Design). Experimenting with the known elements in order to create something new, making up their own version of the truth without knowing the underlying rules. Eventually alchemy developed into an actual science. I would not give any money to an alchemist for creating gold from common materials, would you? Game design as it appears, is not yet at the stage of science. However, it has already left the field of a mere art form. It is something in between. And that, to be honest, makes me uncomfortable by itself. I would like to figure it out. As I will probably not be the one to discover the truth™, I will have to just try my best to find at least some degree of certainty that allows me to ask for a salary. Oh yes, and if you want, you can read about my very own personal discoveries as I write about them.
The purpose of this blog is to document game systems and insights into the balancing of games as I encounter them in games. I hope that writing about these things will not only improve my own design skills, but also help others out there too. I want to learn from the mistakes of others and want to look behind the scenes why certain things work and how. Raph Koster has written an amazing article on his blog about that.
In this piece I will focus on small design decisions that have a great leverage on the dynamics and emotions of the game. Even as a designer the first perception of a game is the own experience. The challenge then is to identify the underlying layers of that experience. Ultimately we will reach the deepest layer of a game; its rules. The fascination of understanding the relations between the rules and the experience they create is what kept me going while analyzing Halfway.
Disclaimer: I am not associated with the developers of “Halfway”, everything stated here is purely based on my observation and publicity available data.
Actually there is no special reason. I bought the game, I played it. I noticed a few things, wrote them down in my notebook and suddenly realized that I had few interesting things on my hand which came quite close to my goal of unveiling how the mechanics inform the game experience.
Halfway is a turn based top down squad-fighting game.
It combines “XCom: Enemy Unkown”- Gameplay with a “Dead Space” Setting and pretty 2D pixel graphics. From their home base, players venture with their squad through decks of a spaceship whose crew has been infested and is battled in turn based fights. The goal is to unveil the events that led to the AI of the ship becoming hostile and the mutation of the crew.
“XCom: Enemy Unknown” as the best in its class will serve as a reference where needed. Not as a benchmark – it is not about what game is “better”.
This piece will be about how slight modifications of the rules lead to completely different game play (which is way more interesting than merely discussing what game is more “fun”).
Theme, not setting, not art-style or story. Theme as in Jesse Shells book “The art of game Design”:
“[..] but what is a theme? The theme is what your game is about. It is the idea that ties your entire game together — the idea that all the elements must support. If you don’t know what your theme is, it is very likely that your game is not engaging people as much as it could.” – The Art of Game Design p. 48.
Themes exist outside of genres, outside of graphics styles and outside of stories. From my point of view Half-ways Theme is “Surviving in an environment depleted of resources”. The survival based theme mostly emerges through scarce resources. Players constantly have to consider their ammo and health at all times. But creating a deep survival-game-experience is not simply achieved by limiting resources.
Many of “Halfways” mechanics are designed to reinforce the survival-theme of the game. Also the story of an “out of control-AI that is hostile towards the player” further reinforces the theme by adding a component of immediacy.
Progression & Game Flow
Players fight with their squad from level to level while meeting increasingly tough and more enemies along the way. Between levels the player is in a safe haven where she can restock on consumables such as ammo and medi-packs besides obtaining small chunks of the storyline through dialogues.
The player has control of a certain number of units. New units are added to the squad as the player progresses through the game and triggers certain story-points.
Health (when health drops to 0 in combat, the unit will become unconscious and can not fight in that encounter)
Agility (agility influences movement distance in combat as well as the ability to dodge enemy fire)
Aiming (influences hit chance of firing a raged weapon)
Items in Inventory
AP (Action points)
Every unit has 2 Abilities. A passive and an active one. Using an active ability in combat uses 1 AP and has a cooldown (turns). More details about the abilities will be covered later.
Players explore and navigate through the levels with an avatar that moves to specific location by clicking on that position. There are two types of movement-commands, left click will only command the selected squad member to move to the target position while a right click will move the whole squad (a very useful feature).
Levels have interactive objects such as doors or terminals that can trigger dialogues and/or encounters. When an encounter is triggered the Combat Phase starts. Most levels have multiple encounters. When a fight ends, the game goes back to the exploration-phase.
These 2 phases are the main tool for controlling the pacing of the game. The exploration phases are the intermissions for the stressful encounter phases. Unfortunately the games has no mechanic to control the pacing WITHIN one encounter. As soon as the combat phase starts, all enemies involved in the fight will spawn and from that point it is up to the player to meet the challenge. In XCom for example, the levels are also divided into encounters (single enemy groups), but those are much smaller and the player can concentrate on fighting 3 enemies at a time (usually). Whereas in Halfway the enemy groups are much bigger in each encounter (the reasons for that will be covered later). Depending on enemy selection and number of enemies the encounters can be tweaked to match the desired stress-level.
Levels also contain some loot crates where the player can find consumable items.
Resources and Items
Energy can be obtained by recycling unneeded items (a weird choice to name that resource energy imo). Energy can be used to purchase consumables in the save haven between levels.
Players can find ammo-clips in levels or purchase it. Each ammo clip fills the magazine of a weapon no matter how many bullets that weapon can hold.
In Combat; When all loaded bullets of a weapon have been fired, the player needs to reload. Reloading costs 1 AP and 1 Ammo clip.
The value of one ammo clip thereby varies on the weapons that the player chooses for his squadmembers. 1 Ammo clip can give 5 shots from a sniper rifle, or 30shots for a SMG. Ammo will not be restored after fights. Every bullet that is fired originates from an ammo clip that has been obtained at some point.
These are quite rare but deal massive damage in an Area of Effect (AoE). Buying them costs 225 Energy which is 9x more than a magazine.
Medkits & Shield Cells
Like their name says, these consumables can be used in combat to restore a certain amount of health or shields (there are small, medium, large version for each of these item types). Using a Medi or Shield Pack costs 1 AP.
Global Resource System
This mechanic alone, creates interesting choices that support the survival theme. In fights players have some control over their ammo usage, while using more ammo is usually the more safe decision within one encounter, but not the best other the long run because every bullet used, is one that the player might be missing in the next encounter – or level. To use ammo efficiently the player needs to carefully choose which weapon to bring to a fight. The weapon choices are influenced by the squad-members using them (they have some unique abilities & properties that determine their role). This is how the ammo system is entangled with all the other systems in the game to deliver the survival-theme. When creating levels and encounters, the designers always need to account for the impact on the global resource systems. These are some tough constraints for the encounter design which sadly led to the encounters being quite repetitive. The variety of the encounter difficulty cannot differ too far from the current “power” curve. Otherwise the impact on the resources might be too harsh. And the most limiting factor is that as the designer / balancer, you cannot be entirely sure what the resources the players have at a given time. A player can be in level 5 with 20 Clips and 10 Med Kits in her inventory, or she might have one 1 of each left. What do you balance the encounter for? The middle? That would be too easy for good player and way too hard for the less advanced. For the Minimum-case? Then there would be no challenge. You get the problem..
This is why shooter nowadays have auto health refill instead of health pickups. As a level/content designer, you can create more interesting fights when you know that the player enters a fight with 100% HP. You can add spots where the player can regen and then crank up the difficulty again. Knowing how many resources a player has at a given time, allows designer to create better tuned content.
The designer of XCom for example want the players to focus more on the individual fights without forcing the player the consider the non-fight-systems of the game too much. Ammo is not relevant at all. The only way XCom emulates an ammo system is by limiting the number of shots until the weapon needs to be reloaded (at the opportunity cost of shooting). But the player has unlimited reloads, hence unlimited shots. This liberates the designers to put more action into the single encounters as there are no consequences for the global game-economy. (While of course there are a couple of mechanics in xcom that make the players consider the global-consequences of their actions such as wounded-players, using explosives-destroys-weapon-fragments). Health supply is limited in-fight by the charges of the med kit (which can be increased by bringing a support class). There will be no situation in an “XCom” Encounter where the player has an unforeseeable amount of Health-Refills or Ammo.
“Halfway” gives players the ability to recycle equipment ( suits + weapons) which gives them a certain amount of energy. With energy, players can purchase consumables that can be used in combat for 1AP. As there is no random loot and a finite number of levels which can only be cleared once, the energy available in the game is capped. This means the designers have a control over how the income of consumables for every “progression”-stage. From that point onward though, players can spend the energy however they want, opening up the opportunity of an almost-ammo-deadlock (if players are out of ammo, it becomes insanely hard to obtain new ammo because players can only melee in fights).
In any games with a challenge component, the games difficulty increases over time in order to keep the game interesting. To also convey the feeling of achievement and mastery, the players “power” can also improve.
Many games utilize a leveling-system for that purpose.
In Halfway there is no XP-based-leveling system that increases a units stats or adds new abilities.
The only way the players power level is increased to match the increasing difficulty of levels, is by equipment (armor & weapons), and by
Stim packs can be found in levels and use one inventory slot. The player use the stimpack to increase the respective stat of a unit of her choice. Stim packs act as distributable stat-points (in RPG’s those are mainly given to the player when she levels and can be distributed freely in some games (e.g. Dark Souls) or increase automatically ( MMO’s ). Stim-packs are quite rare and given that the player has 5+ units every increase in a stat should have a significant impact. The frequency of stat increases determines how much impact that increase has. So there is a tradeoff between how often the player is rewarded with level-ups and how much more powerful the player becomes.
To allow for tactical choices during fights, games that incorporate group-combat mechanics rely on unit-roles. From Team Fortress to Final Fantasy all the genres use a system of weaknesses, strengths, armor types, abilities and usable equipment to create distinct roles that are played in different ways. Players can then strategize on how to set up their squads to solve the fight challenge.
The unit roles in “Halfway” are created by different Weapon Types whose efficiency depend on their accuracy at specific ranges (more details on weapons soon).
Every unit comes with 2 abilities which somewhat determine the weapon and thereby also the role of each unit. The connection between Abilities and Roles, hence weapons, is quite loose in most cases. The “Rifle Expert”-ability cleary specifies the role & weapon, while the “Medic”-role can be legitimately played with all weapon types, whereas some weapons make more sense than others.
Xcom handles this topic differently. Characters have a specific class that come with very unique abilities that cleary distinct the different roles.
Also there are explicit rules for what weapons can be carried by what unit. This means that the players has less choices when it comes to micromanaging the unit development. XCom limits these choices to those within one role while “Halfways” choices allow for some blur between roles. Giving players the option to do that also bears a higher risk of the player messing up.
Health & Shield System
In combat, units have Health and Shields. When taking damage, shields will be reduced first. When the shield-hitpoints are down to zero, incoming damage will reduce the health-points. Incoming health-damage is reduced by suit-armor. When the health-points reach 0, the unit becomes unconscious and is not able to continue fight/moving. When a fight ends, shields are restored and unconscious units awake with 1 point of health.
There are different suit types that have an important part of this system.
Scout Suits have a large amount of shield-hitpoints
Tank Suits have a high armor value that mitigates incoming damage
Assault Suits are the halfway blend of both
The damage a unit can take, depends son its hitpoints –
hitpoints = health+shields
Balancing-wise armor can be translated into hitpoints – a 50% reduction of incoming damage for example is basically doubling a players hitpoints.
hitpoints = health*armorMultiplier + shields
This means that armor scales with health: each health-point added to a unit, is more worth with a higher armor. The value of shields does not scale with hit-points. So at a certain health-threshold, a unit will benefit more from a tank suit, than from a shield suit of a specific type. At a given Equipment Tier, the suits are about equally strong in terms of total hitpoints (I estimated a certain health value on a unit for a certain Tier because hier tiers of equipment become available later in the game where units have more HP).
This has been calculated with the actual configured values from the game. Luckily the developers created the game to be modable and all the balancing files are disclosed ad JSON files. I collected some of the data for suits and weapons in this excel. Download Collected Data.
Shield suits come with an advantage though. Remember that shields are restored after a fight? Health points are not. The advantage of shield suits are free hit-points. Also the drain on the economy system is reduced because the player will need less medkits which are worth a certain amount of energy. This advantage does not come with another cost as far as I can tell. This would mean that shield-suits are simply better than the other types.
The choice for the player regarding which suit to assign to which units becomes unclear as the values between the different armor types are only slightly different. And the clear benefit what armor is better is not visible unless the player sits down and does the math. Although the balancing of the numbers makes sense as in “Cost vs. Benefits”, the distinction between the different suits is too blurry to have an interesting choice here. XCom is being slightly simpler in that matter too, armors merely increase the HP, but the different armor types come with special perks, such as flying-capabilities, or fire restistance.
This is where all the systems come together. The fight is executed in alternating player & AI turns. Each unit (enemy and players) has 2 Action Points (AP) per turn. Unlike XCom shooting does not end the turn. Since there are no explicit classes, there are no “special” rules regarding how different units work during a fight (e.g. XComs “Sniper move or shoot”-rule).
Actions During a turn
Moving (1AP per zone)
Each AP is worth a certain movement distance that varies by the agility stat (IMG movement radius).
A regular shot costs 1 AP, but the player can choose to do a shot for 2AP with an increased accuracy value (this will be covered later in more detail)
Use Item (1AP)
Players can use consumables: Use Shield-Packs, Medkits (on self or others),Throw Grenades, Reload.
Use Ability (1AP)
Every player comes with 1 active ability.
Retaliate (ALL AP)
If the unit is hit during the other sides turn, it will shoot back.
“Halfway” uses a half/full cover system that affect (incoming & outgoing) ranged-shots-accuracy similar to Xcom.
All these systems add a dimension to the “decision”-space of the combat:
The spacing and positioning of coverpoints in the level-design
Different ranges of weapons
Positioning of Enemies
Global Resource System
Different Enemy Behaviors
Consumables & Abilities
As mentioned before, weapons are the component that define the role of a unit, to be specific: the accuracy at different ranges does. An additional metric to consider when it comes to the choice of weapon is the magazine size. Ammunition is a scarce resource and therefor player want to maximize the damage per magazine.
Damage per magazine = BulletsInWeapon * Accuracy * DamagePerShot.
The amount of bullets per weapon and the damage are defined in the weapon and does not change. But the accuracy completely depends on the players choices (hence, the skill) during a fight.
Other properties such as a weapon-damage range and critical chance that adds damage allow for some RNG to create some uncertainty which keeps the fight interesting and enables potentially exciting or threatening.
Avg Damage Per Shot= (BaseDamage + CritChance*CritDamage) * Accuracy
To decide what unit will receive which weapon the player must first consider the desired role-mix. Then, within each weapon type, there are multiple variations regarding the ammo consumption, damage, and accuracy.
Each weapon type has a distinct accuracy-curve which ultimately defines their role. On problem with these accuracy curves is, that while in the balancing, there are quite distinct curves as you can see – ingame, nothing refers to them. It merely states the “best” range for a weapon, which does not imply a certain falloff curve. The game hides this mechanic behind the scenes.
Short-Range Weapons are only useful when players get quite close to enemies which makes them particular good for a unit that will be pushing forward to hinder enemies from moving towards the rest of the group – off course this dynamic can only emerge because the AI is programmed in a way that it will not just run into the player group attacking a random-player unit. Short Range Weapons also have a smaller magazine, leading to higher Ammo-Clip cost and less shots per action-point because of an increased reload-frequency. The steep accuracy falloff means that the 2-AP shot does not help the player to maximize ammo-efficiency. The accuracy is at 100% (which is the maximum) for 3 tiles and then goes very fast to 0. So any increase in these values will not make the weapon more accurate.
The limited range is not only compensated by the high ranged DPS but also by the melee Attack which has a 100% accuracy. Since users of short-range-weapons will be closer to enemies, they will use the melee attack more often – which will also deal more damage because short-range-weapons have the highest melee-damage.
Medium-Range-Weapons have the largest magazine which adds the option of firing 2 shots a valid choice to increase DPS at the cost of bullets. Shot Accuracy remains high until a range of 8 tiles and then starts to drop. The balanced accuracy and the best Damage per Magazine value make it the most versatile of all weapon types.
Long-Range-Weapons have the lowest accuracy at short and medium ranges but their high base damage make up for that to actually succeed medium-ranged weapons even at medium range in damage throughput. This comes at the cost of a very high Ammo consumption is about the same as for short ranged weapons. Due to the high bullet cost and lowest average-accuracy, this weapon benefits the most from the 2-AP shot.
When it comes to picking a shot players have 3 choices:
1AP for a single shot
2AP for two single shots
2AP for a single accurate shot
I mocked up some example balancing for these 3 shot options to show why the different are balanced the way they are in “Halfway”. The exact formula for accuracy calculations is hidden in the source code, but the general idea is this:
For adding 1 AP to the shooting action, players can increase the accuracy of that shot. Increased accuracy leads to a higher damage. In the first column I added a few example rows where I simulate different accuracy increases for the additional AP. In the game, the additional AP gives an accuracy increase of about 150%. If the additional would be twice as high, the damage would be higher than 2 single shots. Then there would be no choice because it not make sense to use 2 single shots as it uses 2 bullets for the same amount of damage. By picking a value lower than that, the designer actually created a choice where the player can control the ammo consumption vs. damage. The increase must be high enough to make a reasonable change though. The accuracy does not increase by a multiplier though, there is a second accuracy curve for each regular curve that is used when shooting with 2AP.
Halfways balancing & design adds up when looking under the hood. How these system are entangled with each other is not clear to the player though. The presented choices in the unit-progression (items & weapons) are not clear enough because the connection with the unit-roles only becomes obvious when fully understanding the combat-math. Interesting choices stem from a distinct functionality and or values with a clear trade-off. The differences should be reasonably high and must be communicated explicitly in order to explain the consequences of that choice. Only a clear choice can create player-agency.
The unit progression is quite slow due to the missing XP-leveling-system. This leads to a very slow progression which fits the theme. Also – each improvement is more rewarding because the impact of is more recognizable. This shows how much control the designers have when simply adjusting the power-curve.
The global resources systems leads to some combat-design limitations which make the fights slighty less interesting than they could.
Still, “Halfway” is a very carefully designed and tuned survival-game-experience that will attract some players and retain those who are already very familiar with the game-systems of the genre.
Thanks for creating this nice little game whose simplicity allowed me to learn a lot from its systems.
In my previous post about the progression mechanics in Sim City BuildIt I argued that shaping the players sessions is an important aspect of the games core mechanic, the crafting system. This time I want to scratch the surface of how certain game systems create and utitlize a certain player behaviour.
Session Pacing can be summarized like this:
How oftendoes a player interact with a game per day and how longdoes that engagement last?
So why is it so important?
Getting the session pacing parametres right, is important for creating a successful free to play mobile game because their monetisation model is based on games being a “service”. This means that the games systems need to engage players for an as long as possible timespan. To make this possible, the content of the game has to be packaged in small chunks to not burn out the players interest in the game. Each of these chunks should be a rewarding and concluded experience. Players want to feel, that at a certain point, it is okay to leave game . That they have accomplished something. They can put a “done” on their mental checklist and know that in a couple of hours there will be something waiting for them. This is what players of mobile games want and expect, and this also an advantage for the creators because we can build games to satisfy that desire.
In a session the players repeatedly engage with the core mechanics of a game up to a certain point. The choice regarding what game systems brings the player to that point of session-closure is depending on games the target audience. But a session must always end with setting an incentive to return, a promise, the prospect of an reward in future, putting a new “Todo” on that mental checklist. Just to make sure players don’t forget; also a local notification is queued at that point, ready to trigger when it’s time for the player to come back. The point at which a session ends is usually when a certain resources is exhausted (by the core mechanic).
The session frequencyis determined by speed at which this resource needs to refill and the session length is driven by the amount of resources consumed per round (and of course the length of one round).
This principle applies to all flavors of session pacing-systems.
In this easy to understand system, each gameplay a certain amount of a resource is consumed. This resource can be given out as reward and is usually restored over time. A strength of this system is its simplicity and the scalability of the cost and income of the energy-resource
This qualifies as a subgroup of the energy system. The difference is that one gameplay always 1 requires life. For a casual audience this is even easier to conceptualize but it comes with the disandvantage that lives can not be rewarded in “small” amounts. Limiting the designers possibilities to to tie lives into their reward systems.
Candy Crush Saga
Now it becomes a bit more abstract. The resource that is “spent” during a session and replenished over time, in these systems are “available-production-entities” . A player starts a session and harvests all the crops or collects all the produced resources. During the session the players use up their production slots by assigning a recipe/crop to it. The production then takes a certain time and the players can return to the game, when the production has ended, hence, when the production-slots are available again. Session pacing is controlled by scaling the amount of production slots and tuning the production timers & cost.
The trigger for the next session is stronger than in the Lives & Energy system because the player becomes emotionally invested when queuing up these crafting timers.
Sim City BuildIt
In these systems, players build up their resources to use in the core gameplay. As the player progresses she increases the possible amount of units. So it becomes more expensive to max these out. In clash of clans for example, players build units which they then can use to engange in PvE or PvP content. There is one aspect that makes this system quite interesting. Players can choose how much of their troops/resources they want to invest. It becomes a risk vs. reward choice. The players decides how many units they want to deploy and put at risk of loosing them. The more they can keep, the less time they need to spend waiting for their units to rebuild or heal.
This aspect of rewarding skill for a target audience that seeks challenge in a game. The better a player is, the more she can get out of every session which is a strong motivation for optimizing their strategy in the core gameplay. This “strategizing” that involves the players skill can come down to the actual fighting mechanics or merely to troop selection before the fight starts.The trigger to initiate the next session is strongest in this systems because the players had to use their hard earned currency to build the troops that they wanted.
Clash of Clans
Finding the system that triggers the right player emotions in your target audience is a top priority design decision when building a free to play mobile game. Some systems will work better with certain players for various reasons. All the systems presented here are built on the same principle which leads to the same goal: Motivating players to return the game. But every single implementation has been carefully crafted to meet player expectations & behaviors. To choose and build the right system to achieve that, we must first fully understand our players and their needs. This concerns not only the high level decision of frequency and session-duration but also what motivates players to come back to a game.
I took a close look at the balancing of the successful mobile free to play game “Sim City BuildIt”. I recorded many data points during my play-through in order to interpret the designers’ intention behind these numbers.
and look for yourself. It has more content than I will refer to in this article so taking a closer look might be useful.
I will start with a brief introduction regarding the focus of this article: Player Progression.
Games are about players pursuing a goal. In order to reach that goal players accomplish a subset of goals. “Progression” describes how far a player has come, reaching that goal. This often implies consuming a games content while moving on.
The content is consumed by engaging with the game mechanics. For example, in order to progress in an RPG (progress = Levels, Items, Story), the player needs to fight (core mechanic) enemies (obstacle).
What is progress in “Sim City BuildIt”? In Simulation games, or in social games in general, there often is no final goal which after reaching it, concludes the game. Progression itself is the goal in these games. Expanding, growing and exploring. Usually
economic challenges are imposed on the players as an obstacle to reaching their goal. Controlling the progression in games like that is crucial to their success in terms of longevity and monetization.
Goals of progression Control
Learning Curve & Ramping up new players
To not overwhelm players with too many Gameplay mechanics or systems in the beginning, the access to new features needs to spaced out. Giving players one bit of the game at the (right) time gives them the chance to learn the existing ones before presenting the next one. Although there is a fine line between spacing out unlocks of features for the sake of merely stretching the content (mostly for monetization reasons) and actually providing a smooth learning curve. Which leads to:
To overcome the progression blockers the designers have put into a game, the players are often asked to repeatedly execute the same loops. This increases the time spent with the same content. This only works if the core gameplay is highly replayable.
Having a lot of things happening and unlocking, keeps things more interesting. Bread crumbing the player through the early stages of a game to keep them engaged and helps getting them from install to showing the games full feature set. Once the player is hooked, things can slow down.
Long term goals
Having a clear measurable goal such as a level gives the player something to look forward to. Especially if they know what elements of the game are connected to reaching a level (e.g. unlocking new buildings, features or other content)
How it works
Unlocking new content or mechanics can be slower because players can spend increasingly more time engaging with the previously unlocked capabilities.
The number systems which the players can interact with or the complexity of the existing ones usually increases over time. This means, that the further the player has progressed, she can spend more time with different aspects of the game.
Example Lego, in the beginning you will need a lot of content (pieces) fast in order to build things. While too many pieces might overwhelm a child. Although obtaining more pieces (long term goal) is worthwhile (reward), the child can spend that time waiting for cChristmas with all the pieces it has collected up to that point.
How progression is controlled
Some examples for gameplay elements that have an impact on the player progression.
In games that use level-systems it is common to control to pace the progression by adjusting the XP curve. The XP needed to level up, increases with every level. Usually also the XP income increases (numbers always need to grow so players can see the progress). But at a lower rate than the thresholds. So overall the time to level up increases (Fig 1).
Adding Systems or Complexity
In MMOs sometimes non XP rewarding systems (such as crafting or PVP) are introduced which will slow down the player’s progression. During a round of Civilization turns get longer because the complexity is growing (number of cities, units, diplomatic options).
Difficulty / Randomness
Controlling the rate at which the player can consume content by letting her fail and repeat more often. Found in puzzle Games where some levels are simply harder (or more random) in order to require more rounds-played to finish.
Sim City Buildit
I will mostly cover the topic of the relation between the crafting system and the player progression. If you are looking for a more broader analysis of the game especially in regards to its free to play mechanics, I recommend reading the great deconstruction at the deconstructors of fun.
“Sim City BuildIt” boils down the concept of the original Sim City Series down to 2 systems: a Hay Day like crafting and old school city building mechanics as found in previous Sim City titles or the “Anno” series where the placement of support-buildings is crucial in order to grow a city. Notions of Sim City’s well known RCI Zones (residential, commercial and industrial) can still be identified but they merely continue to exist thematically.
Crafting is the pacemaker of the whole economy. Everything hinges on players capabilities to create products which she then can either use to upgrade residential buildings or convert them into soft currency through trading, the Cargo ship, disasters, or randomly appearing Deals.
Players expand their city by building residential buildings which will increase a city’s population when upgraded. Residential buildings require coverage of Utility and Service buildings such as power plants, hospitals etc.. Adding a residential building to the city will come at the cost of providing a sufficient service and utility supply. These supporting buildings can be built with soft currency. To maximize the effects of supporting buildings, players need to re-layout their cities from time to time. Meeting the citizen’s demands increases their happiness which will in return increase the tax income (SC).
Progression Sim City BuildIt
The player’s goal is to expand her city. Many of the systems to control progression are tied to the level of a player. Although they may appear simply as a burden when looked at from a systematical perspective, for players they are tasks and challenges that they want to overcome in order to reach their goal. Finding the sweet spot between challenging and driving players away is among the tough tasks when balancing the progression of a free to play game since this is where the monetization hinges.
The Soft Currency economy is as important for the progression as upgrading houses. By adding residential buildings the player expands the city and increases the population’s demands, and SC is used to meet these demands.
Players have two options to increase Population (Fig 2):
Build more houses (which will increase utility or service cost)
boost the population of the existing houses by building Parks or “Key” buildings
The latter does NOT add to the utility costs. And since there are no new buildings added, the service cost (buildings that provide happiness through area-coverage) also remain the same. But of course the designers have accounted for that strategy by adding a distinct cost to increasing population in the existing houses: players need to upgrades the roads for a large SC amount to keep the population in these houses happy.
SC is generated in 3 ways.
Sell products (either Trading or random-deals)
There are 3 Sinks for SC:
Utility costs (water, power, sewage) are increased by +1 for every additional house built.
Services (buildings that satisfy a certain demand by covering the area around it)
Parks & Departments boost the people living in a house
Random SC Rewards for upgrades
Coin rewards per upgraded are very fluctuating. They are always with a certain range. The average amount of earned SC increases by level but also the range (min max value to determine the amount) is increasing SC Ranges (Fig 3). By randomizing the rewards the game becomes less predictable the overall feeling for progression is smoother than with fixed rewards per level.
Happiness and population size is used to determine the SC production of the town hall. The happiness determines the tax bracket, and is then scaled by the population by a certain multiplier.
The income is not increasing linearly though. As you can see in Fig. 3 the tax income per citizen decreases with a growing population, meaning a diminishing return for each population increase. In other words: Increasing the population size becomes less lucrative as the city gets bigger. But the costs for each population added stays the same. Since the tax system does not have such a big impact on the progression of an engaged player this is only a minor progression dampening mechanic.
Demands & Utilities
In the early phase of the game, no services need to be provided to the population. Players gradually unlock them. If services are not provided, the happiness of the population is reduced (Fig. 4).
This introduces a new coin sink without any additional benefit for the player. The player needs to invest SC in order to maintain the current game state. So every time the player unlocks a new demand, or builds a new residential house, this creates a cost that needs to be accounted for with SC. In Fig. 6 you can see how the “cost” per additional building evolves throughout the game.
The cost per residential building is calculated by stacking the average cost of meeting the demands. The farther the player has come in the game, the more SC it will cost to get the max happiness out of a residential building. And because happiness is provided by buildings purchased with soft currency, the time needed to support a residential house increases as well.
Players go through a constant cycle of expanding and grinding making their progression plateau until enough SC has been earned & invested.
There is also another progression dampening side effect to this. The player needs to change and optimize her city layout when she unlocks a new demand. When plopping the new buildings they make every other building that supplies happiness to residential buildings less effective. Because the residential buildings are spread out wider to make room for the new utility/specialization building hence, the area that needs to be covered with services becomes larger by an amount of tiles that is larger than the newly added residential building.
The whole system of coin-sinks is a good example for a Red Queen Dilemma (the player needs to work more to maintain her current gamestate) or the good old carrot on a stick.
This way, every level up where the player unlocks a new Population-Demand becomes a potential monetization point. These kind of systems are the meat of Free to play economy balancing.
An issue I see with this system in “Sim City BuildIt” in particular, is that leveling up doesn’t really feel good and actually avoiding to level up is good strategy because the game is very tight on the Soft-Currency Economy. Leveling up is not rewarding and does thereby not qualify as a goal for the player.
XP & Levels
Upgrading Residential Areas
Upgrading buildings is one of the core activities and goals in “Sim City BuildIt”. It is the only way to earn XP and one of many ways to earn Soft Currency. A residential building can be upgraded 5 times, each time granting XP and coins and a population boost.
This means, by adding a new building, the players also create a source for coins and XP that can be exhausted. The total amount of XP that can be earned by player is thereby limited by the residential buildings that have been built.
In Fig.7 you can see the XP curve, showing the amount of XP that is required for a certain level. This curve doesn’t look too exiting at first glance. The interesting question for these type of curves is, at what rate does it increase?And usually the rate at which increases, is increasing as well, creating an exponential growth. So to get this number, we need to look at the XP-Requirement-Increase from level to level.
Now things get a bit more interesting! As you can see the XP-Requirement does not grow steadily. For each upgrade of a residential building, the players will receive XP. The amount of XP rewarded is a random value in a range which seems to be configured by level. The base-rate is increasing by level. This is expected since XP income usually grows. To convey a feeling of progression to the player, numbers always need to grow.
Since upgrades equals XP, the number of upgrades required by level can be simply calculated by looking up the XP income for an upgrade at a specific level and use that to divide the amount of XP required for that level (Fig. 9).
Fig10. shows how the number of required upgrades per level differ. This means that players will have faster and slower level ups, alternating by 1 upgrade more or less per level.
By having some fluctuation the progression is not too rigid and predictable. A progression with small bumps feels better than absolute linearity. Slower and faster levels can be created.
It is interesting that the required amount of upgrades sometimes decreases. But it makes sense – by that, the designers can have a strong “bumpiness” by maintaining an overall slow growth.
Crafting is the core of the gameplay and everything revolves around it. The whole session design and most of the progression is handled by this system. Through crafting players generate XP and SC, the two resources that players need to progress. The player is therefor under constant pressure to always keep the crafting queue running.
There are several implications in the crafting system when the players level up. Each of them eventually leading to a slowed down progression due to an increased complexity in the core loop.
Increasing Material Requirements per Upgrade
Material requirements for a residential-house-upgrade depend on 3 Factors (Fig. 11):
the number of materials slots
the material type
the amount of materials
Product Requirement Slots
To upgrade residential buildings in the first session, the player only needs to manage a very limited set of products. This leads to a very high upgrade-rate for the first sessions. There is almost always a match between what the player has in her inventory and that what a house requires to be upgraded. The core loop gets executed often and fast because timers are fast and requirements are low.
This is exactly how the early phase of a free to play simulation game should feel. And it is entirely controlled by the crafting system. Additional product-requirement slots are introduced over the course of the game (Fig 12).
Increasing required material naturally leads to the player spending more time on the core loop. In other terms, each upgrade will ask for more player engagement.
There is a side effect to this mechanic: The material types and counts are more or less generated randomly per slot (that will be covered in a later section). So by increasing the possible slots for an upgrade, the stress on inventory management is increased.
Unlocking Product Types
As the player progresses, she unlocks new products making the range of possible requirements larger. Gaining access to a new product also implies building the production-building for that type of product if it is the first of its kind. Building this production facility of course requires coins. As soon as the player unlocks a new product Type, it can show up in the upgrade requirements for residential buildings.
Gaining access to more crafting products will also force the players to care more and more about inventory management. With each new product unlocked it becomes more likely that players will hit the inventory limit. Which will then push them into engaging with the trading systems in order to obtain the materials needed to increase the inventory or to get rid of excess products.
Although new material types put the players into a disadvantageous position, it keeps the game interesting. Always producing the same items would be too boring after a while. With multiple possible products, players have to plan and organize their production queues which becomes one of the few challenges in the game. This is also what is limiting this system in terms of scalability. Adding too many crafting products with longer timers or more complex recipes doesn’t make the game more interesting after certain point.
As the game progresses, timers on products not only get longer, also the recipes become more complex. In the early game, the products needed to upgrade a building are often simply some raw products.
In the mid game, more products are unlocked that require raw materials as an ingredient. This significantly increases the production time for this product because this time is added to the actual production time.
And then there are even some products that need ingredients that need to be crafted from raw materials.
3 Dimensional Scaling
Having these 3 axes to balance the core loop gives the designers to capability to tune the system in different ways but with the same outcome, modifying the “Time to level”.
XP meets Crafting System
Now I will combine all these mechanics to show how the progression actually develops over time. Again, progression means the “moving towards the end goal of a game”, which in Sim City is, expanding and optimizing the city. To reach this goal, players need to complete the core loop of producing, collecting and investing products.
The time to level in Sim City can be calculated like this
Time to Level = NumberOfUpgrades * ProductsRequiredPerUpgrade * ProductProductionTime
All of these factors are increasing, all of the small increases, lead to huge increase of the Time-To-Levelup.
Increasing number of Upgrades per Level: The number of upgrades is growing because the XP-Threshold is increasing faster than the XP Income Per upgrade.
The Slots per Upgrade are increasing up to a maximum of 6 at level 15.
The amount of products per slot stay about the same so the main increase from products per Upgrade comes from the added Slot
Production time increases as more and more products with longer timers are thrown into the mix
There is one mechanic that compensates a bit for all these increasing requirements. Over time the players can build more factories with more production slots (Fig.13).
This compensates only the production time though as each product needs to be queued up and collected. So the amount of work (queuing up the products and collecting them) on the player side is still increasing proportional to the amount of products. Even if the growing production capabilities are set in relation to the increasing upgrade cost, the ratio increases over time, letting the players spend more time in the core loop until they can progress.
Distribution of Products
This section is about the probability for certain item types to get chosen as an upgrade requirement for residential buildings. The data I could collect to make this analysis is by far not enough to draw very accurate conclusions about the probabilities but there are at least some quite obvious tendencies.
The probability for each product is calculated by looking at how often that item was a requirement compared to all the recorded upgrades. To account for items being unlocked later in the game, the “RequiredCount” has been normalized (Check the spreadsheet for more insights into this). I just want you to be aware of that normalization when looking at the charts because some products just have a few data points but are treated equally to those with a lot of data points.
Ratio of “Raw Products” to “Manufactured Products” is 1:1 – Raw products are created in factories where all production slots are active in parallel. Manufactured Products are those built in “Stores”. Stores only have one 1 active production slot, but players can queue up products. Manufactured materials are always created from Raw Products.
Quick products are more frequent. At least for raw products it is clear that those that can be produced fast, are more often used to upgrade buildings (Fig. 15).
For Manufactured Products this no clearly supported by the data (Fig. 16)
The distribution of products is a system the designers can utilize to tune the session design.
A mix of different item types with varying production times is preferable. But shorter timers need to be more frequent in order to shape a “good” session:
Players are kept engaged during the session as long as they have a reason to produce quick materials until only those products with long timers are left, letting the players end their session after queuing them up. The goal of most social games is not to have the users playing as long as possible, but in small chunks distributed over the day, so the content doesn’t get exhausted to fast but the player gets used to engage with the game on a regular basis.
The means of controlling the progression in “Sim City BuildIt” are subtle but very effective. This stems from a lot of small progression-dampening systems working together on multiple levels (Crafting, Soft Currency Sources & Sinks, and City Lay outing). The designers have tuned these systems in a way that they are not too harsh to the player and ease them smoothly into the habit & hobby phase.
In a nutshell
SC is the gating resource
Gradually increasing SC sinks by unlocking newdemands
For each expansion players need to invest a growing amount of SC in order to maintain population happiness
SC income increases at a lower rate than the SC sinks
Each level requires more and more upgrades of residential buildings
Upgrade time increases by asking for more products and different kinds of products
Efficient session design through the crafting system
Necessary cookies are absolutely essential for the website to function properly. This category only includes cookies that ensures basic functionalities and security features of the website. These cookies do not store any personal information.
Any cookies that may not be particularly necessary for the website to function and is used specifically to collect user personal data via analytics, ads, other embedded contents are termed as non-necessary cookies. It is mandatory to procure user consent prior to running these cookies on your website.