Category Archives: Game Design

Dreamblazers Devlog: Week of June 2, 2014

Last week’s achievements

* Finished creating moves used by Miharu overworld enemies: Energy Siphon, Pounce, Pummel, Radiant Flare, Rush, Seize, Silk Wrap, Siren Song, Slash, Smash Takedown, Spiky Vines, Super Hammer, Super Tackle, Swiften, Tackle, Tempest, Thunderbolt, Trap Wrap, Vital Fang, Whiptail, Wild Charge, Wild Thrash, Zap
* Assigned Miharu overworld enemies to learn their moves
* Created and assigned battle AI to all Miharu overworld enemies
* Put together all enemy groups for the Miharu overworld

Current focus

Balancing and testing enemies and enemy groups.

Sample stuff

AI for Kobold Chief (ability names blacked out)

AI for Griffinaire (ability names blacked out)

Weekly goals

* Send in art feedback for everything that comes in
* Finish creating early-game moves used by the original player party
* Test a range of player party attacks against various early-game enemy groups and vice versa

Comments

I achieved everything I expected to this past week and a bit more; instead of only covering 25% of overworld enemy AI, I did all of it, which is great. I owe a lot to ORK Framework for making battle AI sensible to a designer. =) Just need minor polish for the early party, then it’s time to begin testing!

I’ll end with a few words (or a few hundred) about battle patterns…

Even Year 1999 Me (or so) understood in some primitive sense the value of challenge and battle AI, but today I’d say I was a bit heavy on the challenge at the expense of the AI.

For example, in a document that looks like it’s from 1999 or 2000, I have a paragraph about a battle with a single powerful boss, two archers who use a lock-on ability and then get guaranteed criticals on their target for the rest of the battle, two mages who boost enemy party stats twice each per turn, and a warrior who can attack twice per turn. The idea was that since the main boss has too much health to take it down first, the player has to chip away at the weaker foes like the mages and the archers while the main boss runs wild.

I can tell exactly which games my younger self borrowed most ideas from, but this particular battle seems incredibly forward-thinking. It’s a lot like a Bravely Default or Etrian Odyssey boss fight, but BD and EO didn’t exist at that time and I can’t name any RPGs of that era that could have modeled a battle of that complexity for me. If anything, I probably ported player party dynamics over to enemies.

However intricate it might be, though, I believe the battle is flawed: just like certain EO bosses who use their ultimate attack every fifth turn, it’s too consistent. Five of the six enemy party members do only one thing and four of those six only do that one thing to one specific character; the archers pick a target and stick with it while the mages will only target unboosted allies.

With no random chances modifying the sequence of events, winning the battle boils down to identifying a pattern and then going through the motions. For all its moving parts and hard-hitting enemies, it’s not especially more exciting than Final Fantasy VI‘s Whelk, the tutorial boss. Remember not to attack when it’s in its shell. Remember to target the mages and archers first.

To my credit, at the time Star Ocean: The Second Story was my newest Top 5 Game and the idea would fit much better in real-time action battles, which emphasize player execution instead of player comprehension. Any turn-based PvE (Player vs. Environment) battle system, however—even halfway real-time Final Fantasy or Chrono Trigger ATB systems—requires a degree of randomness to feel dynamic.

This boss fight will still exist in a new form, but winning won’t be as simple as figuring out the “trick.” The enemies might use the pattern above or they might do something different. Status effects will come into play. Critical hits will be likely instead of guaranteed, but they won’t be limited to one singled-out target. The enemy side will have healing. Compared to the original full aggro DPS idea, the battle will probably be less challenging, but definitely more unpredictable and engaging.

For me that comes a little later, though. I’ve got an overworld and normal encounters to take care of before I get to dungeons and boss battles!

Dreamblazers Devlog: Week of May 26, 2014

Last week’s achievements

* Minor menu fixes
* New “status effects”: Battle Cry, Expert Evasion, Focused Aim, Heart Rally, Hypercelerate, Light Poise, Reckless Abandon, a dozen variations of Speed modifiers, and ten more related to attacks/abilities/magic listed below with a (*)
* Minor fixes to traditional status effects
* Adjusted damage formulas
* New equipment for Tango, Evelyn, and Telia: Grand Guard, Charming Cap, Traveler’s Cloak, Black Pants (I’ll figure out a real name later), Traveler’s Boots, Brave Belt, Lilty Lute, Archer Half-Gloves, Three-Quarter T-Shirt, Sunshine Shorts, Bold Boots, Archer Core Plate, Brave Bandana, Battle Band, Lightweight Leotard, Telia’s Shirt (see Black Pants), Bruiser Gloves, Dragon Dares, Bruiser Boots
* Changed party to Tango, Telia, and Lucky
* New enemies (in terms of stats and names only): Impini, Archweaver, Gigarat, Greatwolf, Griffinaire, Harpilure, Impedi, Impega, Kobold Chief, Kobold Rogue, Kobold Warrior, Norian Raider, Ogre, Orc Mage, Orc Outcast, Slimary, Slugwall, Troll Warrior, Vivavines, Wyvern
* New attacks/abilities/magic (in terms of pure gameplay functionality only): Accelerate, Artful Combo, Battle Cry, Binding Vines(*), Body Slam, Captive Piledriver, Cling, Corkscrew Stab, Crushing Punch, Dark Force Fist, Diving Slash, Double Smash, Double Whip, Earthen Shield(*), Fireball, Fire Wall, Flame Spray, Flare Spark, Freezing Wave(*), Fury Gale(*), Giga Bite, Great Tremor, Half-Shield(*), Inferno, Javelin Throw, Light Flash(*), Mega Bite, Mega Fang, Mega Slash(*), Mighty Charge(*), Mirage Attack(*), Muck Shot, Needle Storm, Noble Hammer(*), Poison Drops

Current focus

Balancing and testing enemies and enemy groups.

Weekly goals

* Send in art feedback for everything that comes in
* Finish creating moves used by overworld enemies of Miharu (the first continent) and assign all the enemies to learn their appropriate moves
* Assign battle AI to at least 25% of all Miharu overworld enemies
* Put together enemy groups for the Miharu overworld
* Test a range of player party attacks against various early-game enemy groups and vice versa

Comments

It’s amazing how much more productive I get when I’m doing what I really care about: putting down all the groundwork to start watching numbers fly around in battles.

…or so I say, but honestly, the impressiveness of the above list compared to some of my previous weeks is mostly an illusion. It took ages to get my basic menu layout right too, but “made a menu layout” will never sound as cool as “made 35 attacks and 20 enemies” even if it took longer. Still, illusion or not, it feels so much better making gameplay than UI layouts. =)

More of the Miharu enemies’ moves are in place now than not. Once they’re all set up, the enemies just have to learn their moves, use them in balanced and appropriate patterns, and join forces to start battling my test party. Excitement ahead!

Dreamblazers Devlog: Week of May 19, 2014

Last week’s achievements

* Sent art feedback for Lash and Kelly round 1
* Minor menu fixes

Current focus

Balancing enemies and enemy groups.

Weekly goals

* Send in art feedback for everything
* Begin implementing a variety of (normal) enemy stats and enemy groups for the first continent
* Implement initial equipment for early-game characters and use those characters to test a range of attacks against various early-game groups

Comments

What time is it? Turn-spreadsheets-into-game time! I have enemy groups for the first continent drawn out on paper and about 70% of their stats penciled out, so I have enough material to enter into a test environment. Until now I’ve been testing with characters who are much stronger than the early game, though, so first up comes adding equipment for Tango, Telia, and Lucky.

Dreamblazers Devlog: Week of May 5, 2014

Last week’s achievements

* Various fixes to menus
* Various fixes to equipment
* Corrected Style stat formula
* I now know how to put in any of the following types of abilities: single-hit multi-target, multi-hit single-target, multi status effect, multi-hit plus single status effect, and multi-hit plus multi status effect

Current focus

Balancing enemies.

Weekly goals

* Send in art feedback for everything
* Begin implementing a variety of (normal) enemy stats and test a range of different attacks against them for balance

Comments

It didn’t occur to me until this week, but so far I’ve been conducting all my damage formula testing with playable characters and bosses. While boss fights will be Dreamblazers‘ centerpieces, I’m also aiming for interesting normal enemies. With only a few exceptions like Pokémon and Final Fantasy VI, I can almost always accurately judge how much I’ll enjoy an RPG by how much I enjoy its normal encounters.

So normal encounters are the focus this week! This is what I’ve been aiming for since the beginning, in fact. I underestimated how many things I’d need or want to do as foundations leading up to it, but now we’re here and a fun week lies ahead.

Dreamblazers Devlog: Week of April 21, 2014

Last week’s achievements

* Primarily damage formula testing
* Simplified energy attacks concept

Current focus

Battles!

Sample stuff

Pretty glorious.

Weekly goals

* Send in art feedback for everything
* Continue ironing out damage formula balance

Comments

Thought of the week:

I’d extend it to not only projects, but components of projects. As I implemented moves in battles, one of the first steps was setting up damage formulas. No problem, thought I; I already know what my damage formulas are, so I only need to add them in.

Little fine-tuning here and there, but things were going great! I picked eleven reasonably diverse characters and matched them against their own stats to test. All the calculations seemed fine, so hey, cool! Just as a final check and to have a reference lying around in a spreadsheet, I’d match them against each other—

Uh-oh. This quickly revealed that energy attacks were too powerful. They’re meant to be powerful because they bypass defenses, not because they’re inherently four times better than physical or magical attacks. I had modeled my damage formulas after the Pokémon series because there’s no RPG I’m more familiar with, but I had gone overboard with my changes for energy attacks.

Days of testing later, you see the spreadsheet above! Even now I’m not confident that I’ve settled on a final formula, but it’s infinitely better off than before. Continuing to plug away!