• Welcome to Final Fantasy Hacktics. Please login or sign up.
 

Battle Design and Concepts

Started by CONMAN, August 16, 2021, 05:40:57 pm

CONMAN

I've found that one of the most enjoyable parts of making a tactics mod is creating interesting battle scenarios.  Managing to create an interesting and reasonably difficult battle event can feel extremely rewarding.  Crafting these battles really allow for a lot of creativity but it can also be rough hitting a wall after creating a few dozen story events and a hundred randoms. 

I looked into D&D for a while for inspiration and actually ran a fair number of sessions as a dm. That definitely helped expand my view of what can be done and what should be be avoided.

Anyway, I took a few guiding principals to heart and concepts that I try to include where I can.

Action Economy- This isn't quite a one to one with Tactics from d&d, but still (partially because d&d characters get bonus actions).  Essentially the number of turns and damage potential on each side should be fairly balanced.  Comparatively a normal 5v5 regular enemy battle and 5 vs one boss should require the boss to have 5x hp and deal 5x the damage in a round.  That's like asking the boss to one shot per attack or do something like 1/3 damage to the whole party....it's off.  Having a big bad and a couple minions (or lair actions in d&d) is generally a better solution.  Maybe a really good gimmick.

Using Terrain/Mobility- Not really a principal, but this is a tactics game and making the environment part of the battle is huge.  Choke points, obstacles, height advantages make for interesting planning.  Making the player consider the environment or suffer gives an extra consideration in battle. Balancing a boss with extra movement/flight or speed make for a fun otherwise under-powered unit to kite around and annoy the party.  The opposite can be fun as well- crafting a brute force melee user that is somewhat escapable but punishing in close quarters.

Avoid Gimmick bosses- The type that are near unbeatable with surprising status effect damage and chumps when you come back with adjusted equipment.  Also, bosses for whom one strategy is about the only way you can win.

There is such a thing as too many hit points- At a certain point the battle will just drag. 

I had a couple more but I'm running out of time on my end.  Any concepts/strategies/battle ideas, I would love to hear them.
  • Modding version: PSX

Jumza

(CONMAN join the Discord server if you aren't on it, people talk design stuff all the time!)

One type of encounter that I've only ever experienced once was actually in FFTA, in the fight against the second Totema. The first time I ever played through the game, I didn't really understand that I could protect myself from the charm ability the stationary crystals had, so what ended up happening was a pattern of fighting off a group of two or three of their attacks at a time until I had destroyed them, and then falling back to a safe corner where the rest couldn't attack me to regroup (heal up, raise dead allies, cure status effects). I think that that came about because the fight was against totally stationary enemies that had wicked strong offense (very high chance for charm ouch). That specific fight can easily be overcome once you know what to expect, but you could craft a similar fight that would be harder to cheese to create that play pattern that is totally different from what we're used to in Tactics games (where you can't really escape enemies on the field). It could of course also be designed to allow for multiple play patterns, bum rushing, lots of status prevention / healing, etc.

Quote from: CONMAN on August 16, 2021, 05:40:57 pmAvoid Gimmick bosses- The type that are near unbeatable with surprising status effect damage and chumps when you come back with adjusted equipment.  Also, bosses for whom one strategy is about the only way you can win.

This reminded me of a design question that I've thought about quite a bit. There are some games out there that give the player unbeatable situations, encounters where you are supposed to lose to advance the narrative, would FFT be an appropriate place to do that? Is that even ever a fun thing for players to experience? My concern is that A what happens when someone thinks they've just triggered a regular game over and just resets continuously and becomes frustrated because they can't 'win', or B that players then fall into a pattern of just giving up when they run into difficult fights (also can be frustrating when you find out you were supposed to win all along).
  • Modding version: PSX
Nyzer: Alma teleports out of her own possessed body.
Raijinili: Remember that you're telling a modding community that the game they love could use some fixing.

nitwit

August 17, 2021, 11:20:59 pm #2 Last Edit: August 18, 2021, 01:33:23 am by nitwit
John Carmack said that "procedural generation is a ... form of compression", so you can use that to create variety around a small core of expected scenarios in an encounter.

You could for instance set a battle to have 2 bosses and 3 goons, and vary the goons between 2 classes and their secondary skillsets plus class unlocks.
Boss A: Position 1, always present.
Boss B: Position 2, always present.
Goon C: Position 3, Class X, randomly present, de/buffing and support.
Goon D: Position 3, Class Y, randomly present, damager.
Goon E: Position 4, Class X, randomly present, generalist.
Goon F: Position 4, Class Y, randomly present, stat/gear breaker and damager.
Goon G: Position 5, Class X, randomly present, healer and minor damager.
Goon H: position 5, Class Y, randomly present, random skillsets.

This creates a number of possible configurations and so long as the player isn't save scumming reduces the severity of the gimmick battle problem.

To achieve synergy between enemy classes, think about your most effective tactics.

I keep my melee characters all within 1 Move and SP of one another, so they can stay together and focus on one or two enemies at a time. So if I was designing a battle around this, I'd want 2 or 3 melee classes with roughly the same Move and SP, and my long range characters need skills with 2 to 3 range for every 1 move and 2 jump that their melee partners have over them. If the mages and ranged attackers are wasting turns trying to keep up with the melee attackers, then you're letting the enemy surround your melee attackers and take them out. By the time your ranged attackers are in range, the enemy is ready to close with them and take them out one by one.

That's not to say that the AI will focus on one threat at time. There are a few threads that pick apart the Ai, and I'm sure someone will come along with a detailed explanation, but I think they just choose what will deal the most damage in range and do some primitive prioritization for healers and a few edge cases.

If you want to avoid gimmick fights, give your enemies plenty of options, or give everyone few ways to negate their options... though that is a general balancing discussion, for which I have a link if you are interested.

Idea: disallow perfect element/debuff protection to improve element/debuff utility.
https://ffhacktics.com/smf/index.php?topic=12688.msg229811#msg229811

Quote from: CONMAN on August 16, 2021, 05:40:57 pmAction Economy- This isn't quite a one to one with Tactics from d&d, but still (partially because d&d characters get bonus actions).  Essentially the number of turns and damage potential on each side should be fairly balanced.  Comparatively a normal 5v5 regular enemy battle and 5 vs one boss should require the boss to have 5x hp and deal 5x the damage in a round.  That's like asking the boss to one shot per attack or do something like 1/3 damage to the whole party....it's off.  Having a big bad and a couple minions (or lair actions in d&d) is generally a better solution.  Maybe a really good gimmick.
Due to the 1 hit KO issue you mentioned, it may be best to have a 1x5 boss do something other than deal damage, and have ways to turn the players advantage in damage against them via something like a Reflect Damage reaction skill or passive healing methods.

A creature whose attacks all deal 2 or 3 different debuffs, have AoEs, occur instantly, and have no MP cost would be very hard to deal with.

Alternately, you can have 1 boss that is a sort of extended health bar for it's goons, as it can revive/Haste them by sacrificing some of it's HP, and it can regenerate HP by countering your attacks with a HP absorb skill. Your only option at that point is to kill them faster than it can revive them, and either pin it down out of healing range or prevent it from taking actions long enough for them to die. You could also zombie a character and have them attack it to prevent it's HP absorb healing counter.

If the goons can break or steal your CT (or otherwise prevent actions), you are in for a very long stun-lock duel. There's a lot you can do with the boss-as-extended-goon-hp-bar - I would create 2 unique goon sprites, and give each 3 different palettes so you can have 6 distinct goon classes for the cost of 2 sprites.

You could make it thematic too, by making the boss a like a Mako fountain that regenerates enemy HP or something, and absorbs your HP when you attack it. Or make it a cursed item that has brought demons/undead into the world, and which supports them.

Other themes are just as intriguing.

Boss as buff stick: periodically buffs it's allies and boosts their stats.

Boss as negative energy field: it damages the player party by some percentage of their HP each turn, while it's goons do their thing. Or they could be free agents, and the boss their artifact or prisoner.

Boss as full battlefield healing: it heals everyone on the battlefield every turn, meaning you must find some other means to defeat them or destroy it.

Multiple bosses-as-XYZ: have 1 on each corner of a battlefield that you must destroy to depower the goons.

Quote from: CONMAN on August 16, 2021, 05:40:57 pmUsing Terrain/Mobility- Not really a principal, but this is a tactics game and making the environment part of the battle is huge.  Choke points, obstacles, height advantages make for interesting planning.  Making the player consider the environment or suffer gives an extra consideration in battle. Balancing a boss with extra movement/flight or speed make for a fun otherwise under-powered unit to kite around and annoy the party.  The opposite can be fun as well- crafting a brute force melee user that is somewhat escapable but punishing in close quarters.
Bolded text is the 80s slasher monster boss. Friday the 13th Tactics.

You can make terrain more important by reducing mobility a bit through various means - minimum of 3/2 Move/Jump, max of 5/4 - or by editing the maps. Former is much simpler, but map editing is a treat.

My plan is to edit the terrain effects table so more terrain and weather effect penalties and bonuses exist, vary terrain types on each battlefield, and subtly tweak terrain features on each battlefield. Things like:
  • 1-way leaps across obstacles
  • Obstacle bypasses accessible only with a spear or direct ranged weapon (crossbow, gun)
  • Bidirectional long-leaps across battlefields for those with high jump
  • Things that complement stepping stone classes like Dragon and Behemoth
  • Multiple conditionally traversable terrain types like lava and water streams.
  • Jump paths through and around trees like those in Araguay woods, perhaps over a swamp with limited traversal options for the non-aquatic.

I guess the way to look at it is to break down every way that you can move across a battlefield into it's constituent parts, and then build those parts into bigger pieces. You can then use those pieces on existing battlefields, or create new battlefields from novel arrangements of the pieces.

If you have multiple deep dungeon style areas on the world map, one could be in Bervenia Volcano. Then you could have the original battlefield with some minor additions, then you could descend into caverns to experience a walkway and platform jump fight over a sea of lava, a hot springs (streams and lava flows terminating into a lake), and a switch battle where you can punch a few switches to trigger a lava or water flow event.