How to (not) design (real and fake) unwinnable boss fights

Joyfully Rewarding Players' Guts: Winning or losing - are you really the boss in this fight? How to (not) design (real and fake) unwinnable boss fights.

There are a few JRPG tropes I really don't like. Jail time, for example: The party is almost always forced into imprisonment even though the player gets the impression he could have avoided this predicament through (further) fighting. Most times, the player gets to keep his equipment, which makes no sense at all story-wise, but (s)he wouldn't be able to escape without his/her equipment. Long story short: Almost all games don't get prison scenarios right, so I would prefer it if they didn't use them at all.

Can't win, don't like it

This article is about another of my least favorite tropes: unwinnable boss fights (real and fake). What's the difference? Real unwinnable boss fights are really unwinnable (who would have thought of that?), because the developer designed them properly. Fake unwinnable boss fights can be won or survived by grinding beforehand, and the developer rewards the player for his efforts (there are also hybrid forms, for example when the player can win the fight because of a bug, but the game still pretends that the player lost). Why do I dislike them? They don't integrate well into most stories. Losing a boss battle (or any battle) usually means game over. Thus, the integration of an unwinnable boss fight means the developer has to bend story and logic in order for the party to survive. Best-case scenario: The villain can't finish off the party because his final attack blasted all of its members out of reach. Worst-case scenario: An arrogant villain spares the party, because its members don't deserve death by his hands - or even his attention. Not only does the player feel bad about his powerlessness in this situation, the story is also spoiled, similar to a young boy who knows he could only win a game between siblings because his big brother let him win.

If a developer really insists on implementing an unwinnable boss fight, (s)he should at least implement them properly both story- and mechanics-wise. Let us explore how to (not) do this in various constellations.

Unwinnable - for real

One could be tempted into thinking that real unwinnable boss fights are easy to design. I mean: How difficult can it be to design an insurmountable obstacle that utterly destroys the player's party? This depends on whether you consider your approach (absolute or relative) for both offense and defense. Relative approaches are hallmarked by simple augmentations of a boss's level and stats, which means they're more likely to not fulfill their purpose when the balancing is off or the developer underestimates the eagerness of usually over-leveled players (like me). Two (negative) examples:

The unpatched version of Galer: A Plague of Heroes included a boss fight that the player is supposed to lose in order to advance. I - over-leveled as usual - won the fight during my first try, but couldn't proceed afterwards, since I was returned to before the boss battle started and could only trigger it again and again in an endless loop. Of course, I noticed that this boss fight was rather on the hard side, but only after a few tries I realized that I might be supposed to lose in order to continue. This example shows how spectacularly the relative approach for both offense and defense can fail. Many games, especially having been made with or having been published in the dominant era of RPG Maker 2000, simply pretend that the player still lost the unwinnable boss fight, even when (s)he didn't (which means this particular fight is kind of a hybrid between real and fake unwinnable). While this solution is preferable to the assumption of having encountered a game-breaking bug, it's still far from ideal. On the other hand, it's even worse if a game is so boring and grinding would take forever at this point that the player doesn't even try to win a potentially fake unwinnable boss fight (example: The Ino Chronicles - Ascension). Some games even punish the player with a game over, because the developer wrongly assumes the player must have cheated in order to win the boss fight in question, instead of blaming himself for having used a relative approach, which makes cheating possible and alluring in the first place (even when it isn't necessary to cheat, because grinding can suffice to emerge victorious).

Breath of Fire III features a pair of undefeatable bosses (Balio and Sunder) early in the game. Unfortunately, the game's official strategy guide mistakenly claimed they could be beaten. This boss battle features a relative approach only for offense (the bosses even run out of AP at some point, but still waste their turns by trying to cast magic), so as soon as curious players managed to permanently survive during this battle, they tried to dish out enough damage. Some fought for hours and even days to no avail, while others even powered up their party using a Game Shark, but finally they all arrived at the same conclusion: The bosses are invincible (absolute approach for defense), though the battle can be dragged on forever (because of the relative approach for offense).

Ergo: If you want to play it safe, use the absolute approach for offense and defense. Absolute approach for defense means that the boss can't be defeated no matter what (see the Breath of Fire III example above), either because it has infinite HP or because all attacks launched by the player always cause 0 damage. In order to prevent the fight from lasting forever like in the Breath of Fire III example, an absolute approach for offense is needed. This means that a certain skill of the boss in question (used at the beginning or after a few turns) either damages all party members for the maximum of HP that can technically be reached (fixed amount independent of other factors) or fulfills a special victory condition like causing (unavoidable) instant death for all party members. Either way, while the absolute approach for offense is technically the safe way, the story will suffer from using this approach. A gap of power this large can - story-wise - rarely be closed in a convincing manner, which is one more reason to avoid the implementation of (real) unwinnable boss fights. By the way: If you really want to piss off players, create an unwinnable boss fight whose outcome also leads to the player losing all of his/her items (which can be experienced in Astoria: The Holders of Power Saga).

Not really unwinnable

Being by far the more preferable one of the two, fake unwinnable boss fights only pretend to be unwinnable. While the player is supposed to lose the fight and is therefore allowed to lose without suffering any disadvantages, (s)he is also allowed and able to win the fight, which grants him/her a special reward and a slightly different story continuation. It goes without saying (I'm saying it nonetheless) that fake unwinnable boss fights require relative approaches for offense and defense, since an absolute approach for defense (the boss is invincible) or offense (the player's party is inevitably obliterated) means that the boss fight is really unwinnable. By implementing fake unwinnable boss fights, developers can demonstrate they really know what they're doing. The reward for winning a fake unwinnable boss fight must be valuable (but preferably not unique, since unique items shouldn't be permanently missable, but that's just my opinion). Once again two (this time positive) examples:

Estpolis Denki II (better known as Lufia II: Rise of the Sinistrals in the USA or just Lufia in Europe, because Lufia & The Fortress of Doom wasn't released in Europe) features one of the first and most prominent fake unwinnable boss fights. Gades is one of the four Sinistrals (i.e. one of the four big bosses) and the first one to challenge the player when he's still unprepared. His attacks are so overwhelming even against highly-trained parties that a player could believe Gades was really unbeatable, yet it's still obvious for experienced players that Gades's approach for offense is only relative (nowadays; the game was released 23 years ago during a time when experienced players had "only" completed the first six Final Fantasy games and the first five Dragon Warrior games). It comes as no surprise that many players that grinded for hours and/or found the optimal strategy were finally able to defeat Gades over the years. Their reward: the Gades Blade and its infamous octo-strike special attack (something Octopath Travelers can only dream of). What's so special about this reward (apart from slashing an enemy eight times in a row)? You may have heard of the Ancient Cave, a (within certain parameters) randomly generated dungeon in this game that consists of 100 floors. Entering the Ancient Cave means that the levels of all party members are reset to 1; they also can't use any previously acquired equipment in the Ancient Cave - except for equipment found inside blue treasure chests that are rarely generated inside the Ancient Cave. It usually takes the player several "dives" into the Ancient Cave, every time a bit deeper and as deep as possible, but the most important thing is to make it out alive with any blue treasure chest items found (the player can't save inside the Ancient Cave and loses all acquired items upon defeat). Securing these blue treasure chest items which can be used freely inside the Ancient Cave means that each subsequent "dive" gets a bit easier (especially on previously reached floors). Guess what can also be found in blue treasure chests inside the Ancient Cave? The Gades Blade. And when is Gades encountered? Before the Ancient Cave can be explored for the first time. That means: By defeating Gades, the player is allowed to wield a powerful weapon inside the Ancient Cave from the beginning, enabling him to conquer this beast of a dungeon significantly faster.

While not being as rewarding as in the aforementioned example, Forever Home still offers a decent fake unwinnable boss fight. The first boss fight in the game is fought in the very first village (the protagonists' home village) against a future companion (mistaken for an enemy). This boss fight ends automatically after a few turns (assuming or feigning the player's defeat), and while this means there's an absolute approach for offense in play, it also means the incorporated time limit gives the player a fighting chance (combined with an indeed relative approach for defense). Grinding for an hour outside the village enables the player to win this fight just in time. The reward: a whole lot of PP (equivalent to at least twenty successfully defeated random encounters) this early in the game. PP are used to level up shards (which essentially means Materia just like in Final Fantasy VII). (Leveled-up) shards grant (higher) stats boots. In addition to the aforementioned grinding, the player can gain a significant edge when it comes to the game's early parts by winning this fake unwinnable boss fight. As a bonus, both outcomes (winning or losing this fight) appear to be reasonable story-wise.

Fake unwinnable boss fights like the two I mentioned are prime examples of the principle "winning isn't everything, but winning is always better than losing".

Win them or lose them

Hopefully, I was able to demonstrate that unwinnable boss fights only compliment a game when done right both story-wise and mechanics-wise. By achieving that, it's up to the player to recognize if and how an unwinnable boss fight can be won and if it could be worth the effort. As long as many developers still don't realize the scope of problems inherent to unwinnable boss fights, I still prefer JRPGs in which every battle can and should be won, but I give every developer the opportunity to surprise me nonetheless - hopefully in a positive manner. Time will tell if implementing unwinnable boss fights is a fake or real unwinnable boss fight in itself.

What's your stance on unwinnable boss fights? Please tell us in the comments.