Final Boss Design Discussion

[XIIIthHarbinger]

@bgillisp

True,

In FFVII the various weapon battles served as fairly decent trial runs for the end fights. In fact some of them were more powerful than Sephiroth.

Which is one of the reasons I think that people were willing to overlook just walking in & hitting him with Knights of the Round. Because you had earned those "God Killers" through some damn tough battles.

 

[bgillisp]

@XIIIthHarbinger

Not quite. You got Knights of the round by making a Gold Chocobo, then heading to a cave. Most I knew breeded and raced until they got it instead of getting it from the weapons, as in my experience you needed it to have any shot of winning the later weapon battles.

 

[XIIIthHarbinger]

@bgillisp
I honestly don't remember using it on Sephiroth, Ultimate, Emerald, or Bahamut either for that matter. I might have used it on Ruby, but I am not sure.

The wind up was nice, but is was just too "easy button" I suppose. Once you picked up the ultimate weapons for each character, got each of them to level cap, & used materia to boost your parameters to cap, all you really needed were limit breaks for just about everything.

If I remember correctly, I mostly spent a great deal of time grinding in a sunken ship or submarine. (shrugs) It's been a long time.

 

[D.L. Yomegami]

So...how do you make it feel like a final test, but one the player can fairly study and prepare for? And how do you make it interesting without making it feel like you are contradicting the idea of being able to study and prepare for it?

I think the best way of going about this would be making every enemy have lessons built into them. Have regular enemies be lessons for normal bosses, but have normal bosses be lessons for the final boss. By the time the player reaches the final boss, they've seen all the tricks enemies might throw at them and how to deal with them; the final boss is simply the best at using them. In that manner, the final boss feels like the culmination of everything the game's been building up to in gameplay.

Just to make doubly sure, the enemies of the final dungeon should serve as a review of sorts, just to make sure the player hasn't forgotten anything important before they face the final boss.

What I said earlier about designing the final boss first could also help here. On top of being the benchmark for building player skills around, the final boss could also give you an idea of what the other enemies should be capable of, which in turn should make them feel like they're lessons.

When it comes to optional stuff: On one hand, I don't think optional stuff should really be considered when it comes to designing the final boss. If the player really wants to do the super-hard sidequest for the best equipment in the game and beat the level 999 superboss before they face off with the final boss, then by all means they probably deserve to be able to kill the final boss in a single blow. Furthermore, there are players who put off doing optional stuff until after they've beaten the final boss for the first time anyway.

On the other hand, I think the final boss should be a reward in and of itself, speaking as someone who thinks fighting final bosses is one of the top reasons why he'd play games. If the player's stuck with the game for that long, they deserve a boss that'll show them a good time. For that reason, I often like to make the final boss and the game's strongest boss one and the same. If the player's done literally everything else the game has to offer, the final boss gets a stat increase and new skills to throw around. Might as well give the player a suitable challenge to use all the ultimate equipment on. Maybe even give them a new ending for their troubles.

Sidenote: FFVII's final boss (talking about the last boss in a chain of three bosses) actually does scale its strength based on the player's; its stats can increase based on things such as how many characters are at level 99, whether or not you used the ultimate summon on the first of the three bosses, etc. Granted, it never gets as powerful as the bonus bosses (its max HP is 400,000, compared to the strongest boss's 1,000,000), and it really doesn't make that big a difference if the player has all the ultimate equipment and whatnot, but it's there.

 

[Canini]

Although she is a boss from a platformer instead of a rpg, I feel that Gruntilda from Banjo-Kazooie is the gold standard when it comes to fun and challenging end bosses.
It is a sequential bosses where all the moves of the game are used. Although it is always really obvious which moves you should use actually executing it can be tricky. A good example is when she flies out of bounds and you have to use eggs to hit her. The method is obvious, but actually jumping up the spires of the tower and dodging her attacks is tricky. The battle is really challenging but also strictly timed and plays out more or less the same every time, making the player feel rewarding for getting further and further into the battle. The boss banter is also changed each time so replaying parts of the battle you have already won still feels fresh.

Translating this into a rpgmaker turnbased boss would involve a small amount of puzzle solving, while still being being very clear. Somebody mentioned barrier spells with a rock-paper-scissor mechanic to counter them earlier in the thread. What if a boss used combo-barriers and the player had to figure out what spells counter both barriers at once. To prepare the player for this, an earlier boss use single barriers instead of two. Switch the mechanics up during the battle but always be clear what the player has to do and you will get a great boss battle!

 

[bgillisp]

@D.L. Yomegami According to the guide, the final boss only scales at level 99, and no other time (though I did see the guide a long time ago, it might have been level 98 and 99. Either way, it only scales at a couple of places, and it is really high). So if you are say level 97 you get no scaling whatsoever. And honestly, I think scaling up a boss because you leveled up is a bad idea, as it penalizes them for leveling up if you scale up too much. Plus, what do you do if the player finds the scaled version too powerful to beat? It's not like the player can level down to go back to the other version. Nope, they are now stuck with the too powerful scaled version.

That in fact was the problem FF8 had with it's level scaling. I remember getting stuck on the Adel fight on disc 4 as I had gotten that far with my GF's, only to suddenly learn I can't use them (no warning, nothing, just I can't use them or I'll kill Rinoa and get a game over). But my physical attacks were pathetic due to the scaling, and had been getting weaker all game, not stronger. I still somehow got through that battle, but it was a grueling 45 minute battle. But, the point is, if I hadn't been able to get through it, I had no way to counter, as I can't level down to counter the level scaling, and the level scaling was giving the bosses more stat gains than I got at each level up!

 

[Adventurer_inc.]

@bgillisp
I personally hate those kinds of scaling. From my experience, scaling enemies is basically punishing players who enjoy your systems. "Systems" as in any systems that give numerical experience points that affect the scale: combat, crafting, exploring, etc. Punishing players who enjoy your game should clearly go against purpose, in my opinion. We're not all sadist. At least, I hope so.

Scaling brings to mind Final Fantasy Tactics. In FFT, random encounters scales but story encounters are never scaled. In this case, scaling is always optional because random encounters are, believe it or not, optional. Players could only ever be rewarded for enjoying the systems. Faster levels mean faster story progression. This even opens a lean way for optional scaling dungeons at the end because the monsters in the dungeon are considered random encounters.

Scaling also brings up another game that chose a unique way to do it. In this specific game, there are modes such as normal and hard. The difference between normal and hard is your old fashion AI tweeks: For example, the pools for normal are [Attack 30%, Wait 60%, Guard 10%] and pools for hard are [Attack 50%, Wait 25%, Guard 25%]. (I actually don't know exact pool but this was how it described it). The main difference was that hard mode, compared to normal mode, gave fewer experience points but more money. At first, this made no sense to me, however, I noticed that the game's "crafting system" was heavily money dependent. IE most of the material was store bought as well as monster drop. Basically, this implies that the main difference between easy mode and hard mode was how much of the systems did you want to master. If you chose easy, you could win the game through stats like any traditional RPG. However, if you chose hard, you will always be under-leveled but will have a pool of resources to explore the game's mechanic. This is unique in the way that instead of scaling monsters, the player is scaled. (The option to change modes mid-game exist also for those who realized mastering mechanics = difficult).

 

[D.L. Yomegami]

With the way I'd do it, the final boss would scale based on how much optional content the player has done, rather than solely their levels; they'd have to go far out of their way to get the final boss to its strongest form. If this involves fighting a superboss that's stronger than the final boss's initial form, then I figure anyone who can beat the superboss can probably handle the stronger final boss. I'd also couple this with making sure that the final boss (or any boss, really) isn't a "You do this one strategy or you DIE" kind of boss, and with a battle system that puts more emphasis on strategy rather than pure stats.

I might make the final boss scale based on the player's levels as well, but it certainly wouldn't be anything like FF8 (why people dislike that system is very obvious to me). The way I'd handle enemy scaling is by making enemies adjust their levels based on the player's upon initially reaching an area, but never have them level beyond that. This'd be true for the final boss, as well.

Even in the case where the final boss would always scale itself based on the player's level, it'd be more in the sense of "this boss will always kill you in five hits and die from twenty of yours, regardless of whether you're level 40 or 60." Though at that point I'd be more likely to just throw levels out entirely and make stat progression based on equipment and/or manual stat upgrades ala Dark Souls instead.

Enemy scaling is kind of getting off topic here, though, so I'll leave it at that.

To be perfectly honest, I'm the sort of person who likes difficult bosses for the most part. I greatly dislike it when bosses can be effortlessly breezed past without much work on the player's behalf, especially not when it's a boss as important as the final boss. So naturally, any boss I design isn't going to be effortlessly breezed past unless the player is really clever in how they fight the boss, or put a lot of extra work into it.

 

[mauvebutterfly]

To be perfectly honest, I'm the sort of person who likes difficult bosses for the most part. I greatly dislike it when bosses can be effortlessly breezed past without much work on the player's behalf, especially not when it's a boss as important as the final boss. So naturally, any boss I design isn't going to be effortlessly breezed past unless the player is really clever in how they fight the boss, or put a lot of extra work into it.

I guess this is just a matter of perspective. To me, doing a lot of grinding and over-levelling your party qualifies as "a lot of extra work" and I have no problem with players stomping on the final boss as a result.

If players are deliberately doing a ton of grinding in order to make the game easier for themselves later on, they know what they are doing, they know the consequences of it (easier or trivial boss battles), and who am I to then tell them no? I can't imagine that the pool of players who engages in grinding while wanting a challenging boss fight afterwards is particularly large.

 

[AMGLime]

I'm a fan of Endurance fights, Multi-phase fights, but also fights that kind of make sure you've learned stuff in the game itself. I'm actually going to post might Final Boss Draft as a spoiler at the end of this, to see what people think, and kind of just show the stuff I like. I don't think bosses need to be a damage sponge, they can just have mechanics that make them kind of control the rate of the battle. Turns where they have like a "guarded stance" where they take less damage, or things like Bahumats Megaflare that forces you to play more defensive for a brief time are great ways to do that. Status effects are another way to control the battle, not Bad Breath type status effects like got used at the end of FF9, but a super ice attack freezing someone solid for a turn, or just general stuff that can disrupt the rate of combat like a delay buster type attack can all be good for it. Sure it can be annoying, but its the final boss, I feel like you want to be kind of grinding your teeth a bit, y'know? Throw the tricks you may have used before, back at you as a way to catch you off guard, but at the same time, something that won't completely stop you in your tracks. I think FF7 Final Boss himself was a bad design, Safer Sephiroth anyways, offered no real risk. Supernova itself was just an enhanced Gravity and couldn't actually kill you, and it's not like he hit particularly hard either. The actual difficult boss at the finale was Bizzaro Sephiroth, who used status effects, had multiple targets with individual actions, and also, made you use multiple parties (if you wanted). I actually struggle to think of a final boss that I can look back at and be like, "Man, that was a great final boss fight", with the exception of Breath of Fire 3. There's always things I like about final boss battles, but as a whole they never really felt rewarding after beating Breath of Fire 3 lol.

Spoiler

I actually finished designing the draft for my final boss earlier tonight. It's a 3 Phase fight, with 1 and 2 being back to back, and then a save point and point of no return popping up, and then the final boss himself. Phase 1 is a repeat fight from earlier in the game, so the party knows the tricks to it, he just hits harder. Phase 2 is an "extreme" mode of that fight, with the damage being amplified to the point that you need someone who can deal with it, I have 7 Playable Characters, with 2 of them being designed as tanks first, and 3 of them being designed as Healers. The way the damage is paced out for the second fight you need one of those 5 characters to perform the duties they were designed for, or you're going to fall behind fast. At the same time though, the fight itself is scripted in that he attacks for 2 turns (Might increase to 3 after actually tested), but the 3rd turn he enters a 'recharge" phase, where he takes less damage, but gives you some needed breathing room. The fight itself isn't difficult as long as you follow the general, tanks and healers are strong, mentality I've been pushing in the game up to that point. Cause they've been mentioned as being really strong, but are rarely full blown needed as you can typically outgear and overpower most things up to that point but I'm putting his damage to the point that if you dont have a tank controlling the fight or reducing damage, or a healer who can save the day, unprepared people are going to get squashed.

After that it's the save point, and point of no return after before the third and final form. Like someone else mentioned earlier in the thread, I went the route of 1 boss, with multiple parts to act like fighting multiple things. The boss itself, 2 Arms and an Eye. The boss takes reduced damage for each other segment alive, but when the segments are dead, they end up returning to the battle eventually, not like 2-3 turns later, but they do come back. The Body itself does the least amount of damage, the Arms specialize in one being physical, and one being magical. The physical one hits harder, but the magical one has chances to inflict status ailments and debuffs with the spells it casts. They also follow a pattern, Turn 1 they both single target attack, turn 2 one aoes, and the other single targets, turn 3 they both single target, and turn 4 is a repeat of turn 2 but the arms are swapped, Turn 5 is controlled by the Eye. The Eye only acts every 4 turns, and focuses its gaze on a single party member, the following turn both Arms focus that party member with powerful attacks, unless one of them is provoked, or the party member is Covered. When a segment of the boss is defeated, not only does it take more damage, but it also gains a new ability while that segment is out of the battle. Defeat the Physical Arm, and the body itself will start doing physical damage, defeat the Magical Arm, and the body will start using negative status skills, defeat the Eye, and the body will use debuffing skills. The Arms themselves are a copy of a previous boss fight, as is the Eye, but now they're fought together, with the actual target being the body. As the final boss of the game I've designed it as an Endurance fight, but also forcing the party to be able to push out damage, otherwise Segments revive making the fight drive on longer. Again its still just the rough phase, and requires testing, but thats the goal.

 

[Titanhex]

Meant to chime in on this earlier. But here. Maybe this is a fresh perspective.

I'll disregard your first 3 questions for now to bring up this vital point.

A final boss doesn't need to be the hardest boss in the game or a final test of the player's skill.
In many games, it's not.

The Final Boss is just that. The most climactic fight in the game. It needs to feel like the final culmination of the story. The logical conclusion it was heading towards. The zenith of every point and thread left by the story. The end of the path.
It should, as the names suggests, be the climax to the narrative.
In many games, the gameplay and the narrative emphasize combat. The result should be a boss fight that provides a challenge to the player. When you've reached this conclusion, yes, those 3 questions you asked are appropriate.

But, @bgillisp , this is only half of the question, and thus, you are only getting half an answer. Infact, you are only getting 33% of an answer. Because we must delve deeper into the heart of the following 3 questions:

1. What is the best way to make the combat with the final boss feel climactic?
2. What is the best way to handle the narrative of the final boss to make it feel climactic?
3. How can I ensure both the combat and the narrative are in harmony for the final battle, resulting in a climactic end to both the players accumulated skill and their investment in the narrative?

And that is no small discussion. But one worth having. I'd gladly invite you to chat with me any time about it, but I'm afraid writing it all out would be a monster task.

I will answer your 3 questions now, through this lense.

1: How do you make your final boss feel tough, without making it a 100000000000 HP sponge that takes hours to defeat?
Make it relevant to the final level. The final level should contain difficult enemies, and the strategies those enemies use should be present in the final boss battle. It's really no different from any other stage. The Final Boss can also use strategies from past levels that add depth to the repertoire it's choosing from this level.
Infact, kudos if it uses something from the beginning of the game. A useful trope is to make the ending feel like it's coming full circle.
Example. In Undertale's true ending, you fight previous boss battles alongside the boss battle you're currently fighting.

2: How do you make your final boss feel fair without the RNG messing someone over an hour into their attempt at defeating it?
Simple. Cut RNG.
Instead, rely on patterns. Make the patterns hard to predict, but still predictable. Or reduce the window of success to safely navigate the pattern.
In a bullet hell, this would be reducing the space between each bullet. In an RPG, it could be reducing the parties health to 1, then following up with an AoE attack on the next turn.
Or, killing off a single party member.
You want the battle to have tension at the proper points. RNG does not do that, and should not be part of a climactic battle.

3: How do you make the battle feel fitting of a final boss without 10+ different forms? Or do you think it cannot be done?
The "10+ forms" are meant to represent rising and falling tension in the final battle.
I suggest having them. Even if it's not a physical change in form, it should exist spiritually. They provide tension and even a release from tension.
They're so tried and true, that they're a staple of the genre.
The problem is when people use them without realising why they're using them.

Hopefully this provides some illumination on what can conceivably be a very indepth topic.

 

[onipunk]

I loved the first two Golden Sun games, but if I remember right their final bosses (especially Golden Sun 2's Dullahan) were really frustrating to fight because they didn't use attack patterns. Dullahan could use his extremely annoying ultimate attack that practically disabled your whole party for 8+ turns at any time IIRC, which made it an extremely frustrating and tedious fight. Golden Sun was great, but for the love of all that is holy please don't design your final boss the way they did.

Funny, I actually came into the thread to suggest Dullahan, but not the version from The Lost Age. In Dark Dawn, as mediocre as it was as a whole, I think they did Dullahan a lot better by having him stick to a set pattern of eight moves that always proceeds in sequence, so once you figure out the pattern you know exactly what to do to counter each of those turns. Many of the moves were still the same, but the switch from him randomly using them to making sure you know what's coming next if you've been paying attention makes all the difference and turns the battle into a game of memorization, pattern recognition, planning and strategy rather than just simply scrambling to react to whatever move he just hit you with on his previous turn.

This one is less a mechanical point but more of a quality of life-type thing, because sure everyone loves impressive spell effects and summon sequences but if you're going to have a lengthy animation, either make sure it's skippable or it's used so rarely that the player won't get annoyed by seeing it. No-one likes sitting through Sephiroth using Supernova several times in the final battle. It's impressive at first, but sitting through its two-minute animation sequence several times makes it lose its lustre fairly quickly.

 

[Harosata]

I recall a RPG Maker 2000 game, one with relationships and HYPER SEA SICKNESS. You can only fight the final boss with your character and one other, plus the boss could only be defeated with an certain element (which are gained in that only battle). Yet despite the fact that made most of the skills useless, that was an entertaining finale.

The point is, final bosses don't necessarily have to be dedicated to the boss design of the game. For example, Kirby has been entrusted with the Dream Rod, Ribbon and the Crystal, and even Meta-Knight's ship, removing his previous abilities and power of walking. But like Miss Explosionman's tribute to Punch Out, changes in the boss structure can be entertaining, but it can also diminish the finale for some people. I think the reasons for such changes shows how strong the boss becomes and the hero matching him, which is why Yoshi's Island was rocking despite the aim and why I felt that Kirby's Epic Yarn and Rainbow Curse weren't as grand when rehashing the Tank and Rocket.

---

On a separate note, we can agree that Fable 2's final boss is the worst. A guy who ruins the whole world and gets what he wants, only to fall dead by one shot, and sometimes by one of the guys he captured. But since we required a music box to remove his barrier, I guess there was a reason to go there in the first place, but still...

 

[ScientistWD]

See... the thing is...
The best way to make a battle artificially feel like it was difficult is to make it long. It sucks to be killed by the RNGGoddess, but that's because in a lot of ways, she's the only thing that kills you in an otherwise 15 hour fight. Once the player has things under control, it's just a waiting game for the 100000000 HP final boss.
But on the subject of "tests" and "exams" or whatever, I think that the "multiple form" thing is a result of that. If a game has a lot of different sorts of mechanics, then the last boss has to test them all, right? That might take multiple "forms". It's the best justification I can think of, anyway. As far as "studying" for it goes...
Boss Rush, anyone? Fighting all the old bosses in the game should, in theory, be a great "study session" for players needing to review all the game's mechanics ahead of time.

For me, though, I have two final bosses planned. Let me quickly explain why.
The "Narrative Final Boss" takes place at the end of a weird and difficult dungeon. By the time you reach him, one of your party members is missing, another is forced to wear certain equipment, and a third has leveled up a suspicious number of times while you weren't looking, and looks exactly like the boss you're fighting. It's a plot-heavy moment meant to feel draining, tiresome, and "I just want this to be over". That's just how I want the story to end, so that's what I want the boss to feel like.
Then I have a "Mechanics Final Boss". The "Superboss", if you will. She's available from the start of the game, but naturally really strong. In her first form, she has a few tricks that other foes never had before, so for the most part the battle is about outwitting her and staying alive. In her second form, though, all party members get a super-form and have to beat her in three and a half minutes. Which, I guess, is meant to keep players on their feet and provide a controlled environment for me to test them. Also, it's super cool.
What was I saying... oh yeah. I have two bosses because I want to separate the "Final Exam" from "The End of the Story". Depending on how the game is, this might be a better approach. Hopefully it works for me.

 

[Ras101]

I have some pretty simple ways to make my final boss (and other bosses) feel hard but not make the battle too tedious.

My favorite way is to simply add "enrage" mechanics. For example, at the end of every turn, I have the boss cast a spell that hits everyone, with the damage increasing as the turns go by. That way, players deal with ticklish damage for the first few turns, but the pressure steadily builds, pushing players to focus on bringing the boss down before they have to deal with damage they cannot heal through anymore. This way, it forces players who play too defensively with turtling strategies to focus on the offense and they are punished for taking too long.

Essentially: add mechanics that push players to end the battle quickly rather than just bloat the boss' HP to make battles seem harder.

Of course, I balance the boss' HP accordingly. For example, if I make the enrage spell to start becoming dangerous after turn 5 and becoming impossible to heal through after turn 10, I make sure to set the HP so that most players at that point of the game should beat it within 5-10 turns. Players who make all their characters tanks might prolong that by a couple of turns, but it would be useless in the end if they can't kill the boss in time.

Some other ways I do bosses:
1. Having multiple enemies where if one goes down, it increases the strength of the others, forcing players to carefully get them all down to low health before finishing them off.
2. A boss that casts a massive AOE spell at certain intervals, forcing players to go on the defensive and guard or use other defensive measures to even survive it. It helps break the monotony of having characters spam the same skill over and over again.
3. Have additional enemies that cast buffs and heals on the main boss who otherwise would not deal that much damage. Forces players to focus on one minion at a time instead of just burning down the boss.

And of course, variations of the above strategies! It's quite easy to be creative with it.

 

[Ed19]

Multiple bosses-phase(s) might be a good idea, but I think it could be an optional thing to do.

You could make a side-boss before the big boss, defeating the side-bosses before you confront the big boss will make the big bad weaker.
Remember, Xenogears, everyone? That makes the players will try a different combination and it could make some replay value.
Of course the side-bosses should not be included as a a marathon boss fight.

In my opinion, I like the multi-phases boss, and even a 2-3hours boss fight, it's really epic and that's what the real fight should be, because you get the reward : "the ending of your game".

 

[cddeat03]

1: How do you make your final boss feel tough, without making it a 100000000000 HP sponge that takes hours to defeat?
The best way to do this is to make the boss more mechanically or strategically complex than other enemies in the game. The final boss should certainly be more resilient, but doesn't need to be a damage sponge to create greater difficulty. The boss could shift its attack patterns ever x turns or when it has lost x% of its HP. In a turn based game this forces the player to continually re-evaluate their battle strategy, rather than finding the winning combo and repeating it over and over until the boss dies.

2: How do you make your final boss feel fair without the RNG messing someone over an hour into their attempt at defeating it?
I think the answer to this one is to not have any RNG elements determining a boss' skill use. Each turn's skill use should be crafted in advance by the game developer. That way the player can't possibly feel arbitrarily punished by the RNG. Also, I would never give any enemy the ability to kill the player with one or even two hits when the player is at max capacity; at least not in a turn based system (which is what I'm using in my game). If a player manages their HP and MP resources effectively, they should be able to stay at least one turn ahead of death even when faced with the most devastating attacks. However, if they don't manage those resources well then the boss can punish them for it.

3: How do you make the battle feel fitting of a final boss without 10+ different forms? Or do you think it cannot be done?
Honestly, I think giving bosses is a way of developers trying to prolong the final fight and ratchet up the dramatic tension while also trying to prevent the boss from feeling too much like a damage sponge. If, however, you build you boss fight well you certainly won't need to employ multiple forms. The final boss of one of my favorite rpg's, Dragon Quest 3, doesn't have multiple forms and that fight is plenty epic. (I know technically he does, but it is functionally the same fight just with different stats.) Ultimately, I think the final boss feels like the final boss only in part because of the challenge of the fight. The other part is the narrative/atmospheric build up to that fight throughout the rest of the game.

 

[Alastor01]

I am gonna talk regarding general game boss concepts, not specifically in turn-based RPG terms.

I liked Chrono Trigger final boss, with multiple parts and transformation. AOE attacks, elements, status buffs / debuffs.
The HP pool was manageable, not too little not too much. The attacks were deadly if you didn’t heal enough.

Didn’t like FF10 dark aeons... Oh God the massive HP pool and bs attacks... You have to grind like mad!

I love Dark Souls bosses. Well most of them. They punish you for your own mistakes. You do have some time to learn attack patterns, before they wreck you.
So basically if you heal at the wrong moment you can get killed... Or if you get too aggressive.
DS1 final boss is great. It’s just like high powered you, with a similar move set.
In DS3 bosses would often transform or entertain high power mode. That makes you come up with new strategy.
DS2 relied on cheap boss numbers... (I hate you, Aava, and you siblings!!!) But that may work for final boss / es. For example boss guardians then boss. Or 2 final bosses which combine. Or an extra final boss if certain events have been done

Onimusha and Devil May Cry games use transformations very well.

 

[loki2007]

My final boss design is a multi part fight, but not the way you mention... The big bad isn't the ultimate threat, but it's existence will attract said threat.

Ultimately it's a possessing "spirit" that is the last remnant of a paradox. It didn't start out evil persay... It just craves existence and is fighting tooth and nail for it in any body it can obtain.

As such, as you hunt it.. it adapts, tries new things in each encounter in the game until you finally back it into a corner (final boss fight). When you put it on the ropes... the true threat emerges. You essentially try to survive what follows as the 2nd (action sequence) phase kicks in.This phase will be hard, draining, but offers a full heal and save when completed. I plan to finish up with a rather disturbingly easy 3rd phase where you have x amount of rounds to kill the final boss or make the decision to spare it. Failure to do either leads to a rather bad ending.

 

[Internetomancer]

@XIIIthHarbinger

Not quite. You got Knights of the round by making a Gold Chocobo, then heading to a cave. Most I knew breeded and raced until they got it instead of getting it from the weapons, as in my experience you needed it to have any shot of winning the later weapon battles.

I found FF7 Sephiroth a huge yawner, in part because I spammed Knights which took like a full minute watch, but also in part because it was just too easy at that point. I remember enjoying FF6's marathon boss battle, but I don't think it's necessary.

I guess it depends on what you're going for. If story is the main purpose of your game then the boss is just like opening the last door. It doesn't have to be challenging at all, and should be short. If combat is the main purpose of your game then the boss battle has to be interesting and difficult, and you DO want it to last a while-- maybe 15 minutes, you need time to appreciate the struggle, maybe almost die, and to try out a few different strategies.