First published in Develop, issues 7 & 8 (June & July 2001)
A number of conventions exist within the realm of video game design. Some conventions have arisen by a process of pseudo-evolution, with designers of early games being forced to think of new features for their games, their rivals recognising which features worked well or encouraged players to spend money, and those features being copied. Other conventions arose due to technological necessity - the level structure seen in most games exists due to a combination of design and technological requirements.
Early arcade games such as Space Invaders (Taito) and Pac Man (Namco) proceeded by facing the player with an identical level set-up, but increased the game's challenge with each subsequent level defeated. Games such as Phoenix (T.P.N. Corporation) introduced levels with different features; in Phoenix, Space Invaders style stages were followed by egg waves, and then the flying saucer stage. This last involved not many, but a single, large opponent for the player to blast. The saucer in Phoenix rained fire upon the player while its pilot hid beneath a destructible shield.
This varied level structure quickly became popular, then standard. Nowadays a great deal of variety is required in graphics and play to 'justify' a new level to the player. Once variety of play had become standard, the issue of game pacing arose. There's no point in slavishly designing variety into a game if it all feels the same to the player, despite the differences. The key to keeping a player on their toes is pacing the game experience. At its most basic level, pacing involves creating a feeling of climax at certain points of the game (after which the player will naturally feel relief).
The prime method introduced to vary the game pace was the boss. A game boss is any enemy that:
- requires defeating for the player to progress
- is significantly larger and more powerful in both offence and defence than the standard enemies
- acts as punctuation in game flow, usually being featured at the end of a level, as a climax
Bosses are also usually used as a graphical showcase, to impress the player.
(Later game development introduced the 'sub-boss', smaller bosses used to mark half-level points, or similar pacing milestones).
In a game like Konami's Nemesis, the role of the bosses as pacing aids is fairly explicit, with the bosses for all but the final two levels being identical (and being preceded by a distinct 'challenge' section which also provided game variety). Later games were forced to employ more imagination in boss variety. Modern day bosses are usually thematic; lobster boss, avian boss, fire boss. Their use has branched beyond the standard shooter genre, and they are used as punctuation in practically any game genre. Bosses usually employ predictable attack patterns which, once understood by the player, render the boss more or less vulnerable.
The point of this article is to state that bosses have become overused. Bosses have got out of hand. Their use as pacing aids has been almost forgotten, and bosses are now included in games because convention demands it. It would seem that players require their fix of screen-high poly-filled spikey death. But bosses cause all sorts of problems to game structure.
The biggest flaw in the boss philosophy would seem to be that bosses haven't kept pace with the requirements of the game design. Back in the arcade days, bosses were extremely valuable to a game; players would pay to play because they wanted to see the bosses, but would have to work hard (and pay more) to beat a boss to get to the next level. Nowadays, games for home consoles are designed to give the player an entertaining play experience which ends in the completion of the game, rather than being cash-generation engines. Though bosses may be useful from pacing and 'gosh-wow' points of view, their other uses no longer apply. Battling through a difficult boss level is rarely satisfying; rather than creating a feeling of triumph, a long, torturous boss battle leaves players with a feeling of relief - that they don't have to do it again.
Especially difficult end of game bosses are even more troublesome, often preventing players from seeing the game end. Why? A player who has spent thirty hours on a game surely deserves the closure of ending and credits sequences. Difficult bosses that cannot easily be replayed (because they are featured after a five minute unskippable cut-scene, for instance) become even more frustrating. It's harder to see your sequel selling if the player has bad memories of the prequel's end.
A similar crime is that of placing the save points after bosses. Save points shouldn't be designed such as to create challenge for a game, they should be placed as an aid to the player (though I accept the usefulness of save points as a pacing aid and don't necessarily insist on 'save anywhere' technology for the majority of games). There is nothing more irritating than playing an entertaining level of a game only to be quickly destroyed by a boss, and have to play the entire section again.
A special mention should be made of bosses with long, repetitive attack patterns which are nevertheless easy to avoid. After five minutes of concentration, players are likely to make a slip here or there, and if a boss battle takes fifteen minutes and ends in the careless death of the player, don't expect them to come back to your game.
Are bosses necessary? It's a fact that players expect them, but at one point players expected a high score table. Score mechanisms have evolved to keep pace with narrative-based progressive game styles, and so should pacing techniques. Bosses may still have a valid place in video games, and no technique should be fully ignored. But if video games are to make the most of the potential as story-telling devices, or even to capture a truly mass-market, games have to work with the player, not against. Challenge is a good thing, but pointless challenge less so. Players should only be restricted in interesting ways. Too many bosses simulate the real life experience of knocking through a wall with one's head. This is not fun.
It should be noted that many major game developers are already changing they way they use bosses. Super Mario 64 (NCL, N64) decentralised its boss structure in a manner that worked well, changing the emphasis from 'completing the level' to 'completing the task'. Many newer games decrease the difficulty of their bosses, using them more as plot points than game obstacles. Dino Crisis and Resident Evil 3 (both Capcom) explored the narrative based boss, using the boss to heighten tension, but always supplying enough firepower options for the boss to be defeated when necessary (by introducing ally characters, or allowing the player to solve a bonus puzzle to obtain superior ordnance). The feeling is that bosses need to change at the very least, and may change form so much that they become unrecognisable. I would applaud this move.
Do we need alternative pacing mechanisms? Several already exist. Metroid (NCL, NES) introduced the last minute 'race against time' climax, which also worked well for the original Descent (Interplay, PC, PSX, Sat). Games like Driver (Reflections, PSX, PC) eschew standard bosses in favour of missions of heightened difficulty at chapter's end. The original Resident Evil allowed the player to weaken a boss by solving an optional puzzle, a technique which might be extended to giving the player an option of whether to fight a boss or perform some alternative task instead. Of course, these techniques can fall prey to the same pitfalls as boss design itself. The solution may lie in plot and character based climaxes as opposed to challenge based pacing structures. Narrative based games are becoming very popular amongst players and developers, and the inclusion of a story automatically allows for previously unexplored pacing mechanisms to be used.
An interesting point concerning boss design is that the climax method of pacing works far less well within a non-linear game structure. The industry seems divided upon the usefulness of non-linear storytelling techniques in narrative based games, but it seems inevitable that someone will push forward and create a truly non-linear game at some point in the future. This style of game structure will require a far finer sense of pacing than a classic shooter did in its day, and it is likely that the boss will have no place within this style of game, in his current form. It is even possible that the convention of the boss is one of the factors retarding the willingness of the industry to explore non-linear narrative. If so, it is another fine reason to ban the boss.
There will always be room in the video games world for classic arcade style gameplay, and bosses will always have a place in the hearts of long time gamers. But as narrative gameplay becomes more popular, it seems that pacing must serve game progress, rather than restrict it. The goal of the game designer has moved away from creating 'one more go' titles, to that of creating 'just a little bit longer' style games. Good pacing will keep a player hooked, and will allow for the further development of narrative style games. If the boss has to finally die for good for this to occur, then so be it.
International Hobo Ltd