Your opinions on hard and mandatory bosses?

[Soryuju]

I definitely appreciate a spectrum of difficulty in a game’s boss fights, assuming the difficulty comes from actual mechanics and not just inflated stats. Failure isn’t pleasant in the moment, but it presents a chance to understand the game more deeply as you evaluate what’s gone wrong in each previous attempt. There’s huge satisfaction in seeing yourself get closer with each rematch and then eventually triumphing. Hard boss fights are some of the classical moments which allow gamers to not simply play a game, but to master it.

That said, it’s nice when the harder bosses in a game are the ones which are actually important to the story, and not the random big monsters which happen to crop up in between plot points. I’ve played some games where certain major villain fights are designed around a single unusual gimmick, and the boss barely puts up a fight if the player is prepared for it or figures it out quickly. But then later the player may need multiple attempts to take down the less important villains or the generic mech/golem/whatever because the developer built them like actual boss fights instead of trying to be cute with off-the-wall abilities.

You shouldn’t have to guess whether players will take your big bads seriously and feel the drama of those encounters, so if you’ve got a particularly experimental mechanic intended to raise a boss’s difficulty, maybe resist the urge to put it in a boss fight which holds a lot of narrative weight.

 

[Grunwave]

My Gf and I have been playing through Stick of Truth together. Her only exposure to RPGs is Pokemon.

We encountered a boss fight last night where you simply drink a potion and auto attack each round, and your victory is in inevitable. When the boss was at 50 percent health, she said "This is a really long fight." The boss had a huge health pool.

I immediately thought of this thread.

So maybe a good rule: if your player feels like they are wasting their time, your boss sucks.

 

[trouble time]

I like tough boss fights. Especially it its got them good MECHANICS and a slick theme song. My favorite bosses are always the most memorable to me

 

[punchybot]

I always wonder how to make a good boss and maintain the eyes of the player. But perhaps that's for another thread.

I love challenging boss fights. Love them! I definitely think a game should have them, but not every boss needs to be difficult.

 

[bgillisp]

Just be careful if you decide to reference SMT though. Persona 3 also is known for having the cheapest boss fight in history (The floor 110 Tarturus Boss specifically), and the final fight was clocked by someone to take 3 hours to even attempt it once if you fight it at the expected level.

And I'm still not a fan of how Persona 5 handled boss battles. You *don't* introduce new status ailments the player has no idea what they do and cannot cure but they will kill you in 3 turns if you get afflicted with it (I'm looking at you, stupid Sphinx). Or make a boss that if they use their ultimate it still KO's you on normal difficulty even if you guard and have full HP, but you're higher level than all the monsters in the dungeon (actually happened to me with the boss of whatever that Burger place was).

So yes, they got some good boss designs, but when they put a bad boss in, it's REALLY bad.

 

[Wavelength]

I find myself in complete agreement with the majority of posts here, with little extra to offer. Hard boss fights are great, and even taking a few game overs to bosses once in a while is great, as long as I - as the player - feel like the key to the cell is in my own hands. I need to feel like I'm in control of my success or failure. The best game overs to take are the ones where you can say "well, if I did this and that better, I'd have won".

A few pitfalls to avoid when designing hard boss fights:

• Don't make the difficulty spike too hard from everything that's led up to it. A slight unexpected spike is fine; a big unexpected spike is bad.
• Don't ever make the only viable solution to losing "grind more". Allow the player to win by becoming better at combat.
  • This also follows the very important design rule that the more skill expression there is in combat (the more ability for the player to "get good" at combat regardless of stats), the more satisfying your combat will be in general.
• Gimmicks are fine, but don't make the battle impossible if the player doesn't figure the gimmick out. Make it tougher, but winnable.
• Instant Death skills and Skills that can one-shot party members from 100% HP are no bueno. They're actually even worse if they have a chance to fail, because then you as the designer can't design the pace of battle around them.

I personally don't have a big problem with bosses that are "too easy" for a lot of players - they generally still allow you to have fun - but if you can get the difficulty right, to the point where it's challenging or even outright difficult, while giving the player what they need to beat it on a good run - it adds an extra dimension of enjoyment and accomplishment to your game experience.

 

[Aesica]

Instant Death skills and Skills that can one-shot party members from 100% HP are no bueno. They're actually even worse if they have a chance to fail, because then you as the designer can't design the pace of battle around them.

This is one that can actually be acceptable depending on how widely available and accessible resurrection abilities are. In my project, this would actually be a good mechanic to use on a few fights for the following reasons:

• Raise is pretty inexpensive, both in terms of items and MP, plus several characters have access to it.
  
• There are (or rather, will be) both buffs and gear with immunity to instant death as well as taunt mechanics. A tank could easily take a hit that might kill a squishy mage from sheer damage, and can be outfitted with instant death immunity to trivialize such attacks by drawing them to him/herself and negating them via immunity.

However, if the player's toolkit lacks these kinds of counters, then yeah, I agree that it's a terrible way to go. And things like aoes with a random chance to kill each party member that the boss uses randomly are cancer incarnate.

 

[JazzGotBlues]

Bosses should also be one of the last steps of your game. (Not the actual characters, but the battle)
You just playtest your game twice, once doing every battle there is available up to that point and one where you only fight when needed. Compare the 2 levels you have and choose a minimum requirement to be able to defeat the boss.

I see nobody talking what kind of mechanics they exactly want for a good bossfight.
But for me it requires:
-Endurance: Small fighst before the main boss to weaken you and maybe require you to use some heal spells/potions
-Tactical Items: Some of wich might need you to search for them In caves to give a reward for grinding other then higher levels.
-Punishment: Ever played a pokemon game with as less fights as possible? You will not last it out there at a certain point. Be sure they have to explore your world, get a higher level and defeat a boss. Not just at the end, also in the middle and maybe even at the beginning, if you want to show what type of game the person is playing.
-Optional powers/team members: Only having that 1 team of main characters really gets boring. Make sure they either have powers they can learn optionally or you can choose the members each with different skills. Changing up people's skillset based on different events in the game also works quite nice.

I just think of pokemon most of the time. The game was as simple as grind them sweet xp, but allowed for the use of items to boost that grinding. Also the use of items in battle. Your rival was a good way of seeing if you were on a high enough level and I believe he's based on the average level you can reach upon that point (or a bit lower). At least everyone enjoys pokemon, and that game really wasn't too complex

 

[Milennin]

As long as there's a solution to a boss fight that goes beyond grinding for more levels or buying more potions, it should be fine...

There are a few important things to follow when making a 'difficult' boss that doesn't feel like a cheat:
-Make whatever the boss is doing clear. Give its skills that when you see them used give players an idea of what it does. If it uses status effects, make sure the player gets to know what the status effect does, and how to counter it. If it uses powerful attacks, maybe have it telegraph them a turn in advance (and provide tools for characters to defend themself in such a scenario). If the boss has a certain weakness, make clear what it is, so players aren't guessing for turns on end what they need to do.
-Don't make the difficulty revolve solely around overcoming big numbers. Big numbers are a huge part of RPGs, but if it's the sole reason a boss is difficult, it's bad. It needs to do more than just hit very hard, so players are encouraged to think of strategies to use against it, rather than go grind more levels.
-Don't put RNG mechanics in there that can wildly swing the outcome of a battle in either direction. RNG is fine as long as it's manageable under most circumstances. Losing to RNG rolls is the worst...
-Don't assume your boss is more difficult just because you put it in front of a long cutscene that players can't skip or save the game in between. Wasting player time doesn't equal higher difficulty.

 

[Accendor]

No, not everybody should be able to finish the game. Making a boss hard is totally fine, making him unfair or frustrating is not.
E.g. the last boss fight from FF13 is a good example here. That fight was not only hard but also frustrating because of the following mechanic:

- If the main character of your party dies, it is game over (every other character is not a problem, they can be resurrected)
- The boss has a instant-kill-spell
- You can stack resistance for that spell type but never become completely immune

Take into consideration that the fight lasts ~40 minutes and the only starts using that spell at the ~20 minutes mark and it is instantly game over if he ever gets through to your main character you can see how this gets insanely frustrating very fast if you are unlucky.

-Punishment: Ever played a pokemon game with as less fights as possible? You will not last it out there at a certain point. Be sure they have to explore your world, get a higher level and defeat a boss. Not just at the end, also in the middle and maybe even at the beginning, if you want to show what type of game the person is playing.

Again Final Fantasy 13 has a good example here. The game has a "auto battle" feature and you can through it very only using that and only do the mandatory number of fights. However at some point you fight the first Eidolon, Odin. That fight is so insanely more difficult than everything before it, that you not only can not use auto fight but you probably also need to grind up. After that fight however you can get through the entire main game with auto fight again, except for the last dungeon and the final boss fight, which is both again a really, really big jump in difficulty. That felt just weird.

Another example from loooong ago was Terranigma for the SNES. Basically you could go through the whole game with doing minimum fights until you reached a certain boss (4 Dolls / Bloody Mary). If you are under level 25 now your attacks only do 1-2 damage when they usually should do A LOT more than that. As soon as you hit 25 it normalizes and you never have this "feature" in the game anymore, but it is still not a good design imo.

P.S. I freaking love Terranigma.

 

[Wavelength]

[One-Shot KO's] is one that can actually be acceptable depending on how widely available and accessible resurrection abilities are. In my project, this would actually be a good mechanic to use on a few fights for the following reasons:

• Raise is pretty inexpensive, both in terms of items and MP, plus several characters have access to it.
  
• There are (or rather, will be) both buffs and gear with immunity to instant death as well as taunt mechanics. A tank could easily take a hit that might kill a squishy mage from sheer damage, and can be outfitted with instant death immunity to trivialize such attacks by drawing them to him/herself and negating them via immunity.

While I'm sure it's possible to pull off, I don't think I can name a single good battle system where resurrection came cheaply and reliably (and was necessary to use). Generally in RPG combat having characters "die" is the worst thing that can happen; the one thing you're trying to avoid. Making death so cheap that you can make it go away with no real penalty (besides a loss of turn - which is something you'd have to invest anyway to heal the character to prevent death, or to remove the other status effects that death removes for you), in my experience, kind of turns the battle into a weird kill-revive cycle where you are either:

• Resurrecting a character each turn while chipping away at the boss' HP (similar to the attack-heal-attack cycle in some games but decidedly less satisfying), or
• Having more than one character KO'ed on some turns, and spending your time hoping that the boss doesn't use their multi-KO attack (and spending multiple turns gradually cleaning up the damage when they do)

Such systems might achieve what feels like a good level of difficulty if the balance is done well and the RNG cooperates, but I think it will still be less fun gameplay than it would be without the frequent Instant Kills and/or One-Shots. This is true in both turn-based and action-based battle systems.

Your taunt-and-tank dynamic, in combination with immunities, is interesting, but I feel like preparation is usually a less interesting dynamic than intra-battle decision making (so, for example, you could give the tank skills that negate status effects including instant death for a few turns rather than a piece of equipment that negates instant death), and since equipment choices should always involve tradeoffs, the immunity to instant death should leave the tank with less ability to tank damage (if they can still do this very effectively after making tradeoffs that minimize their DEF, then isn't your combat system too easy in the first place?).

I agree that there are better ways and worse ways to do the One-Shot, and that you've correctly identified which ways are the better ones; I just feel that using any point on that spectrum is probably a handicap when you're trying to design something that's truly engaging.

 

[Aoi Ninami]

While I'm sure it's possible to pull off, I don't think I can name a single good battle system where resurrection came cheaply and reliably (and was necessary to use).

One that comes to mind is Legends of an Otherworld. By the point in the game where instant-KO is a fairly frequent tool in bosses' arsenals, you have access to a character who can put ATK, MAG and AGI puffs on the entire party for free. Therefore, in any serious encounter, you want to get these buffs on as quickly as possible. So while revival is fairly cheap in terms of item / MP cost, it still has the severe cost of up to six turns necessary to reapply buffs to the revived character. This means that death hurts enough that you want to avoid it -- leading to interesting puzzles of coming up with strategies for each boss -- but if death happens because you make a careless slip in a long battle, it doesn't mean it's all over.

 

[TheoAllen]

Everyone else already threw their cents on how to make a proper hard boss battle. I'm gonna assume the boss is properly designed, so I'm not going to add more into it. What's the merit?

Having relatively relaxed boss battles could make the game being taken for granted. I mean, those bosses probably quite a trivia or not even memorable. Having hard bosses could lead into your player base could build a community and talk about how to beat the boss. And some people probably gonna talk about "This boss is so hard, I failed like x times just to realized I need to this and that to beat it". In relatively easy boss, such talk might not happen much because people are not gonna find it memorable or "important" to talk about.

So yes, make a hard difficulty boss has its own merit. If you want to make your boss battle memorable, make it difficult.

 

[MushroomCake28]

@TheoAllen I suppose it all depends on the type of game and the type of community you want.

 

[Wavelength]

One that comes to mind is Legends of an Otherworld. By the point in the game where instant-KO is a fairly frequent tool in bosses' arsenals, you have access to a character who can put ATK, MAG and AGI puffs on the entire party for free. Therefore, in any serious encounter, you want to get these buffs on as quickly as possible. So while revival is fairly cheap in terms of item / MP cost, it still has the severe cost of up to six turns necessary to reapply buffs to the revived character. This means that death hurts enough that you want to avoid it -- leading to interesting puzzles of coming up with strategies for each boss -- but if death happens because you make a careless slip in a long battle, it doesn't mean it's all over.

You make a very good point that in games where stacking buffs is a usual thing to do (in boss fights), losing those buffs upon KO can be a reasonable and engaging punishment to compensate for the fact that the actual Revive is so easy and cheap.

If designed well, losing a character to KO here and there will still allow the player to remain "above board" while the character's allies get them back up to speed, whereas losing too many characters in a short time could be a backbreaker that allows the boss to get so far ahead in pace that he can eventually wipe your party.

In Legends of an Otherworld, do you feel like the player has a good amount of agency to avoid (or minimize the effects of) the Instant KO's coming their way, or does it feel more like force majeure that the boss happened to choose an Instant KO and all you can do is just hope he doesn't choose it again soon?

 

[Soryuju]

While I'm sure it's possible to pull off, I don't think I can name a single good battle system where resurrection came cheaply and reliably (and was necessary to use). Generally in RPG combat having characters "die" is the worst thing that can happen; the one thing you're trying to avoid. Making death so cheap that you can make it go away with no real penalty (besides a loss of turn - which is something you'd have to invest anyway to heal the character to prevent death, or to remove the other status effects that death removes for you), in my experience, kind of turns the battle into a weird kill-revive cycle where you are either:

• Resurrecting a character each turn while chipping away at the boss' HP (similar to the attack-heal-attack cycle in some games but decidedly less satisfying), or
• Having more than one character KO'ed on some turns, and spending your time hoping that the boss doesn't use their multi-KO attack (and spending multiple turns gradually cleaning up the damage when they do)

Such systems might achieve what feels like a good level of difficulty if the balance is done well and the RNG cooperates, but I think it will still be less fun gameplay than it would be without the frequent Instant Kills and/or One-Shots. This is true in both turn-based and action-based battle systems.

Your taunt-and-tank dynamic, in combination with immunities, is interesting, but I feel like preparation is usually a less interesting dynamic than intra-battle decision making (so, for example, you could give the tank skills that negate status effects including instant death for a few turns rather than a piece of equipment that negates instant death), and since equipment choices should always involve tradeoffs, the immunity to instant death should leave the tank with less ability to tank damage (if they can still do this very effectively after making tradeoffs that minimize their DEF, then isn't your combat system too easy in the first place?).

I agree that there are better ways and worse ways to do the One-Shot, and that you've correctly identified which ways are the better ones; I just feel that using any point on that spectrum is probably a handicap when you're trying to design something that's truly engaging.

I think Bravely Default is another pretty good example of a battle system where it works, especially since the relative cost of revive skills gets cheaper the later you get in the game. Similar to the example @Aoi Ninami gave, Bravely Default has a fairly buff-heavy combat system. Also, every single action costs at least one “BP” to use. You can opt to skip a character’s turn to store BP and then use the “Brave” command to take up to 4 actions in a single character’s turn. Under this system, action economy is everything, and even though KO’s can be reversed cheaply and reliably, they can wreck your team’s efficiency and send you spiraling if you don’t respond effectively. While only a handful of bosses have actual instant KO/guaranteed 9999 damage moves, most bosses do very hit hard, are capable of acting multiple times in one turn, and only need a few good hits to kill unprotected characters. You can typically stop them in their tracks with a good defense, but again, it all comes down to managing those buffs.

To go into a little more detail, a KO in BD causes the victim to lose all of their buffs and stored BP, and they won’t regain BP while KO’d. So right away, a single KO can easily cost you 3-7 actions just to reset back to where you were. Then there’s the fact that characters regularly act multiple times per turn in BD, so the cost of not reviving a character immediately is often higher than in a game where everyone only gets a single turn each round. If it was a healer or a support who got KO’d, it’s especially painful. And finally, there are some expensive AoE defensive buffs which are very popular to apply to the party in lategame fights (e.g. AoE status immunity, AoE elemental damage immunity, etc.). Characters lose these buffs as well when KO’d, and if you don’t invest the significant amount of resources needed to reapply them, bosses can exploit those vulnerabilities to repeatedly KO or otherwise incapacitate your revived ally (thus draining your resources anyways).

So with all that said, when a character in BD is KO’d, it presents a unique challenge to the player - they must revive the character quickly (especially non-attacking roles) to minimize the loss of action economy, but they also need to create safe conditions for the revive which won’t allow the buff-less target to be KO’d again immediately. If the conditions aren’t safe, it’s often better to accept the situation and wait. Stretching your resources too thin will eventually cause your defenses to break and allow other allies to be KO’d. Every KO becomes a test of the player’s preparation and judgment, and each failure drives them further down the spiral of defeat. Just having cheap revives available on every character isn’t enough to erase the risks or the challenge of that situation.

Now in the sequel, Bravely Second, you can pull some pretty simple shenanigans to revive your entire party automatically at the end of every turn, and that destroys the whole balance of the dynamic from BD. So there are definitely limits to the ideas above, but as long as the revives available aren’t too crazily efficient, it can be a fun and dynamic type of system.

 

[Wavelength]

@Soryuju Great example - yes, the Bravely Default system is another mechanic that allows cheap revives without necessarily making a KO cheap nor inconsequential. Particularly when a character loses multiple actions, it can be a big blow (without necessarily crippling the player entirely), giving a lot of incentives to keep your party alive.

This is probably a good approach for designers to take in general - revives during battle causing a loss of multiple actions could allow for engaging combat loops even outside the Bravely Default system, as long as the actors aren't being one-shot out of the wild blue. Perhaps something along the lines of when a character is Revived, they can't move for a couple of turns, but also can't be targeted/damaged, or similar? In BD you're likely losing actions (as well as buffs) up-front, so this wouldn't be necessary.

Looking back, I was casting too wide a net when I said that battle systems barely ever work when Revives are cheap and reliable - I really should have said that they don't work when the consequences of getting KO'ed in battle are negligible (which requires cheap revives plus other factors). I do think that when KO/death in battle is trivial, you inevitably get caught in the weird, non-interactive combat loops that I mentioned several post above.

 

[Accendor]

Such systems might achieve what feels like a good level of difficulty if the balance is done well and the RNG cooperates, but I think it will still be less fun gameplay than it would be without the frequent Instant Kills and/or One-Shots. This is true in both turn-based and action-based battle systems.

I mean, there are ways to influence the range here. E.g. make sure that the boss does not sure his AoE kill spell in the next 3-4 turns after he used it the first time.
However, I still dislike the concept of cheap shots / instant kill mechanics in general.

 

[Aesica]

While I'm sure it's possible to pull off, I don't think I can name a single good battle system where resurrection came cheaply and reliably (and was necessary to use).

Final Fantasy Brave Exvius. It may be a mobile game, but it's actually a pretty good one, and even though resurrection skills are dirt cheap to cast (and auto-life is also pretty cheap and plentiful) death is still something you really want to avoid because of how important buffs are. Since a raised character comes back without any, it's very easy to get stuck in the KO/Raise/KO/Raise loop because of how important everybody's turn is. Tanks will have to reuse provoke and/or cover next turn and during that time, enemies can easily wipe out a few other characters.

Final Fantasy 4 also had Raise costing a mere 8 MP and the game (the proper version) still manages to be somewhat difficult in spite of that. And while dated, the battle system is still better than most of the RM games I've played.

Both of these games focus more on a quick defeat in battle if an important character dies and you're unable to get them back up efficiently. If you survive the battle in spite of that, it's easier to continue on at full strength.

On the other hand, what I see more often in RPGs (and actually kind of dislike) is the slow defeat you suffer when a party member dies and the way to get them back up is either difficult or nonexistent. So somebody dies and you're stuck enduring the rest of the fight (or maybe even dungeon) from that point on with 1 fewer member, then possibly another, then possibly another.

That said, I'm not against bumping the cost or even adding a cooldown if the need presents itself. I noticed that while FF4 had Raise only cost 8 MP, it went up to about 30ish MP in the next 3 games before dropping down to 8 again in FF9, then up to 18 in FF10. I'll just have to see how it works out.

Generally in RPG combat having characters "die" is the worst thing that can happen; the one thing you're trying to avoid. Making death so cheap that you can make it go away with no real penalty (besides a loss of turn - which is something you'd have to invest anyway to heal the character to prevent death, or to remove the other status effects that death removes for you), in my experience, kind of turns the battle into a weird kill-revive cycle

My thought here is that, if everybody has an important role and set of things to do on a given turn besides just "hit the boss as hard as you can until it falls over," losing a party member is still painful even if you can get them back on their feet during the next few turns. If your healer is the one who dies, then nobody is getting healing. Likewise, (if you're using any of these roles) losing a tank, buffer/debuffer, etc means that role isn't being performed so you could be facing more incoming damage than your healer can keep up with.

 

[Lornsteyn]

Hard bosses are a must, I hate easy bosses.
I would say even some normal enemies should be challenging.
But it must be fair, not instakill attacks out of nowhere.