Browsed by
Month: October 2017

Weapon upgrade system designs and limitations

Weapon upgrade system designs and limitations

I’ve been contemplating on design of the weapon system and upgrades for a lot of time. I wanted the game to be based on skill but have a variety which would add to the replayability of the game at the same time. Here are the basic definitions i have decided on:

  • Game will contain about 30 weapons
  • Every weapon can be upgraded 4 times (levels 1-5)
  • Ship has two weapon slots available, you can freely switch between weapons in game
  • You can’t have same two weapons equipped
  • You eject the currently active weapon by picking up a new one that’s different

For example, you have Pulse Gun Level 1 equipped as active weapon and Biter Level 1 equipped as inactive. After blasting the enemy transport you come across Ripper and pick it up. Since you don’t have it equipped in any of slots, it will replace the active weapon and eject Pulse Gun Level 1. If you wanted to replace Biter, you could simply switch weapons to make Biter active and replace it by picking up Ripper. This will be a common occurence for adapting to the enemy types because of their vulnerability or resistance to certain type of damage (ballistic/explosive/energy/special against normal/armored/shielded types of enemies).

I could make things simpler in design and coding by simply omitting the part where the replaced weapon is ejected since there’s a small chance of picking it up by accident since it involves pressing a key while you hover over the weapon. However, two player mode requires that feature for the weapons to be interchangeable between players and that is a great way to improve cooperation, gameplay and combined firepower.

Due to some design limitations i had to make a hard choice that can affect the future gameplay on upgrading the equipped weapon and few solutions came to my mind.

1. You can only upgrade the weapon if you pick up the exact same weapon. That way, either equipped or not, the weapon in players possession is upgraded to the next level without any ejection which only happens when you are picking up a weapon you don’t have equipped on any of the slots. While challenging with high long-term impact on decision-making, you only have 6% chance of getting the same weapon from the transport which is slim to none and could severely hamper the player experience. If weapons had only one level the approach would be viable, but with total of 150 weapon levels it would only be frustrating.

2. Whenever you equip a weapon that is not equipped it is always at level 1, but when you upgrade any of the weapons on ship to level 2, the weapon you replace the level 2 weapon with will also be level 2. Basically, if we modify the first example a bit so the active weapon (Pulse Gun) is level 2 and inactive weapon (Biter) is level 1, when we pickup a Ripper instead of Pulse Gun it will automatically be upgraded to level 2. Opposite to first approach, it is less challenging and encourages experimentation, but it comes with a design problem which i’ll explain thoroughly.

When we picked up Biter it is upgraded to level 2 on the ship and Pulse Gun level 2 is ejected. This enables us to switch Biter to active weapon, pickup the level 2 Pulse Gun, eject the Biter, and then pick up the Biter again which will automatically be upgraded to level 2 now. While requiring a bit more speed to do it in a chaotic environment i would considering it cheating since you’re upgrading both weapons that way and that is certainly not something i plan to implement. As i noted in the introduction, i could simply disable the weapon previously equipped to be ejected, but that beats the idea of switching weapons between players which i find to be a great gameplay feature of a co-op mode.

Maybe i’ll disable the feature of ejecting only for single player mode for now.

3. Make weapons upgradeable only by picking up the same weapon as equipped, but increase the chance of spawning a weapon you already have

The maths on this one are simple, though a bit tedious to code. You have 25% of transport spawning active weapon, 25% of spawning inactive weapon and 50% chance of spawning a new weapon. This comes with a different kind of trade-of. Though 25% is a lot it may happen that you rarely run into a weapon you want to upgrade. On the other hand, you may always run into a weapon you already maxed out. This discourages experimentation since you will always want to hold on to your maxed out weapons, no matter how good or bad they are. There are no bad weapons per se, but holding on to weapons of the same type greatly decreases success.

4. Weapon upgrade pickups

Though not originally meant to be implemented, this could pose a good solution combined with approach 1 or 2. It is simply a kind of a joker card which levels up your active weapon without worrying if it’s the same one. If you pick it up, the active weapon gets upgraded and you just keep on blasting.

Which solution would YOU like to see implemented?

Spawning System Overhaul and overcoming the obstacle of enemy pattern making

Spawning System Overhaul and overcoming the obstacle of enemy pattern making

The Grid

Wow, it’s been a while, but things are moving forward slowly but surely!

I’ve been busy with lots of coding and programming enemies and i encountered some difficulties in proper positioning. On the image below you can find how the spawn points looked (5 spheres on each side) and how they look now (the red X signs)

Spawning Grid
Obviously, 12 times more spawning points offer much more flexibility

Obviously, it offers much more in terms of positioning. More than a year ago, when i first started working on a spawn system i wasn’t apt enough to make it the way i wanted to (grid system) so i had to be satisfied with only a few spawn points and additional repositioning upon instantiating. Needless to say, it adds much more work to simple spawning and positioning of those enemies.

By using this handy tool from the asset store (https://www.assetstore.unity3d.com/en/#!/content/20502) i created a grid made out of objects completely automatic. A fine tool indeed. After that, i simply added those to the hash table and now i can simply reference the object whose position i want the enemy to use as spawn point and voila. Besides using it to spawn enemies already in a pattern, i can use them to actually create random patterns on runtime by referencing a different object from the hash table upon predefined parameters to avoid completely random clutter. Not only that, a finer grid enables me to avoid spawning the enemies too close to each other or overlap. Since i’m using Core Game Kit for spawning, i’m waiting for the developers to implement the feature based on sphere raycasting, i.e. if there’s an object of certain tag or layer (enemy) in a defined range, the system won’t spawn any more to avoid the overlapping. It will work great with the system i made and described few devlogs earlies which is based on enemy pool values and enemy quantities.

Also, Easy Save 3 Beta will soon get a full release (i hope VERY SOON) which will enable IMPORTING variables from a .csv file. It will be of an immense help for tweaking the gameplay.

Enemy Pattern Making (Squadrons)

I must admit, though i am passionate about making a game, some things are quite tedious. I’m having problems with making enemy squadrons, and the way i make them is so boring and uninspiring it really halts my progress.

Before i was well into Unity engine limitations on nested prefabs (only one child per object, i.e. child cannot have it’s own child as a prefab, only when instantiated on runtime due to way serialization works) i thought it was going to be a breeze, i just drag and drop enemies in a formation, put them under a parent prefab and voila! Except it doesn’t work like that. All of my enemy prefabs typically have two children, Gunpoint and Thruster. Gunpoint hold the shooting logic and muzzle flash animation, while Thruster has the, well, thruster animation. It is on a separate object to avoid being colored with the rest of the enemy ship when it changes color on hit by a player weapon.

So i guess i’ll keep my work and make an empty squadron prefab which will spawn and then spawn the enemies in a desired pattern coded into it. That’s all nice and dandy until you actually start working that way. No more cosy drag and drop, just selecting what to spawn, input coordinates and hit play too see what you’ve done. If something’s not positioned correctly (it usually isn’t), reposition the enemy, copy the coordinates, stop, and paste them. Repeat 10 times for 10 enemy objects in a squadron, and i should have hundreds of them! Horrible!

Prefab and Runtime comparison
Prefab with multiple children with position setting on runtime

So i decided to change my ways. I need to make a reverse approach. Instead of creating an enemy prefab with all the children attached, i’ll attach the Gunpoint and Thruster on instantiation, which is only a two step process compared to setting the position of multiple enemies inside the squadron. This way, i have a clean enemy prefabs without children which i can joyfully drag and drop into positions i want and simply save them under a prefab which will be used for spawning.

Prefab and Runtime Setting 2
Prefab with multiple children instantiating on runtime

Though it is a bit more work initially, it provides an immense saving of time later and makes it more visual, fun and intuitive to work with.