Acerca de
About Crea
Crea is a 2D RPG Sandbox in the same vein of design as Terraria.
Crea drew my attention with strong modding support early in it's development, and I modded it as a hobby for several years. This continued up until I was hired as an intern to work on Crea after which I eventually became a part time employee.
When working on Crea, my primary jobs were content design, gameplay programming, and bug fixing. The gameplay code is written in Python. The engine code is written in C++. I regularly worked with both.
The Anomaly System
The Anomaly System was a system I proposed, designed, and implemented into Crea.
The Anomaly System was designed to fix 3 problems we had identified in the core gameplay:
-
Players often felt that the world felt static and dead.
-
Players felt that they lacked challenges outside of boss fights.
-
Players rarely explored more than a small portion of the world.
The solution that I proposed was to create a system which randomly created time-limited events in the world, which provided challenging encounters with significant rewards should the player choose to travel across the world to the location of the event. I implemented 6 events, many of which had additional variability. More events have been implemented since I stopped working on the game.
-
Minibosses: Called "Apex" monsters. This event utilized a previous system that could create randomly empowered monsters in order to create a set of fixed empowerments that were designed to be stronger and more mechanically interesting than standard random empowerments. By utilizing that system, it became easy to expand the list of possible sets of empowerments as the empowerment system was expanded. Since each monster is able to become a miniboss, further variability is available. These minibosses drop unique, rare equipment.
-
Darkness: The game is locked into nighttime for a time and monsters are tougher. This event has low player interaction, but received feedback that it makes the world feel more alive.
-
Swarms: This event created a swarm of monsters that roam the world. Upon defeating them all, the player is rewarded with a rare item. This event provides a form of combat not otherwise available in the game, as monsters typically don't occur in large quantities.
-
Blessing: Players find an item that teleports them to an area, wherein they are able to make a choice between a prolonged buff to their character, or a large amount of rare resources.
-
Defend the Objective: This event creates a pillar that if activated spawns monsters in waves to attack it. If players successfully protect it they are rewarded with a rare item.
-
Fast Travel Infection: This event infects the fast travel network, shutting down a single known node and requiring the player to clear it to regain use of it.
The Anomaly System received positive feedback, and was later expanded upon with more events. In particular, the Apex monsters were later integrated into the dungeon system as particularly challenging dungeon bosses.
The Stalker
The Stalker was a monster created as an experiment based on feedback on the game. Initially during the creation of the game, about 4 monsters were created for each biome. However, as the game neared the end of Early Access, it was becoming clear that more monsters were needed for variety within the game. There was a notable problem though, our single artist on the team was backlogged and unable to provide monster art. As such, I proposed creating monsters utilizing existing art, and was given the go-ahead to try and create one.
This resulted in some specific limitations:
-
The new monster could use no new art (other than minor recolors).
-
The new monster had to be easily distinguishable from other monsters, despite using a recolored sprite.
-
The new monster had to provide a significantly different gameplay experience, it couldn't merely have one new ability.
The issue of using no new art, I chose to resolve by using an existing attack of the monster in a very different way. Normally, the Stegara (the monster that the Stalker shares art with) operates as a durable brawling monster, hitting players with multiple hard-hitting melee attacks. I chose to alter the attack pattern to a low-health, hit and run attack pattern. This kept the same attacks, but dramatically altered the combat experience. This also solved the problem of providing a significantly different gameplay experience.
The issue of having the monster be easily distinguishable was solved by turning the monster mostly transparent. While the silhouette was the same, after some testing I found that leaving the monster mostly invisible caused players to immediately identify it as a completely different monster.
This resulted in a "gotcha" style monster, in the same vein of design as Mimic monsters exist in. In order to ensure that the monster was fun to play against, I did some research on other "gotcha" monsters, and found a couple of important aspects that were present in those that were well received:
-
Players felt like they had a fair chance to catch the monster before it was able to attack them.
-
Catching the monster before it attacked resulted in the player immediately winning the encounter, or making the encounter almost trivial.
-
Failing to catch the monster resulted in the player taking significant damage, but not outright killed.
-
"Gotcha" monsters only appeared after several hours of gameplay.
This caused me to alter the design of the Stalker further with the following attributes:
-
The monster has very low health, but very high damage
-
The monster is mostly transparent when not in combat (and re-enters stealth after about a minute out of combat).
-
When the monster notices a player, it walks towards the player very slowly until it gets within attack range or takes damage.
-
If the monster gets within attack range, it becomes visible then attacks a single time for a large amount of damage before running away at a fast rate.
-
If the monster takes damage before getting in attack range, it becomes visible and runs away very fast.
-
The monster can never kill a player who's above 80% health with a single hit.
-
The monster only spawns at higher levels of play.
This monster was well received by the player-base. During testing players would rarely express frustration at the monster, but would often express extreme pleasure at killing it. Furthermore, the experiment of "monster variants" proved to be successful, and many more monsters which used art from previous monsters (with minimal changes) were added to the game.
The Winged Bow Rework
The Winged Bow was a weapon (and classification of weapons) that was notorious throughout Crea's Early Access for being difficult to balance.
The original design of the Winged Bow was that it would shoot 3 arrows in a cone (instead of the normal one). The problem was that if the arrows did full damage, then they did a significantly higher amount of damage than their equivalents. If the arrows did 1/3 damage (or maybe some middle point such as 1/2 damage) then the bow felt useless unless you it was used in melee range (and there was already a series of bows with the design space of a melee range bow that was more interesting). As such after several failed balance attempts, I proposed reworking the weapon and was allowed to take the task.
The design space I decided on for the weapon was the following:
-
Maintain the original intended design space:
-
Good at hitting multiple monsters as an AOE weapons.
-
Poor accuracy.
-
-
Should have above average power, but require more skill to achieve that power than other bows.
The solution I decided upon was to make the arrows split at the zenith of the arrow's arc rather than immediately. Each arrow retained its original strength and dealt full damage. This resulted in nullifying its benefits completely at close range, limiting its benefits against single targets without clever usage, and amplifying its benefits against multiple targets compared to the original.
This rework proved to be successful. The weapon archetype remained popular, but was within an acceptable level of balance compared to other bow weapon archetypes.