Shadow of the Colossus is another attempt by Sony to produce a game with a unique artistic
vision. It does not appear to have been motivated by a desire for sales – this
could never be a mass market game – but rather as a gift to the Hardcore gamer, to generate goodwill towards Sony. In this regard, the game succeeds
admirably, delivering an established play pattern (bosses) in an
extremely polished package. Fumito Ueda and his team are to be commended for
their work on this game, but this is not to say that there is not also room for
The game is emotionally narrow, delivering chiefly awe (wonder and fear) and fiero (triumph over adversity), with the usual suspects of excitement and relief coming along for the ride. Indeed, delivery of fiero is the only reason to bother with a boss battle, which is a tired mechanic only surviving on the back of game-literate players having a great desire for fiero; witness Clive Thompson’s sermon extolling the virtue of the fiero of bosses (although he does not use the word himself).
This narrowness actually serves to heighten the experience of the game in many respects, and should not necessarily be interpreted negatively, but it comes at the cost of the player gradually becoming inured to the emotions being focussed upon.
The awe of the game is at its strongest
when one faces the first colossus, specifically when one uses the L1 button to
frame the camera such that the avatar’s and colossus’ heads are both in shot.
The impact of this moment is tangible. However, as one makes one’s way through
the sixteen colossi, the impact diminishes. Although there are several truly
exceptional beasts the player is forced to slay, there are also many which feel
about the same quality as the best bosses found in other games. Perhaps some
players were able to sustain the feeling of awe longer than others; I would be
interested to hear different viewpoints.
The diminishment of the awe of the game is all the greater because the player’s experiences in the world of the game are emotionally flat. Boredom, and occasionally frustration, are the key feelings many players will get riding around the somewhat repetitive landscape. The luckiest players are those who can find some beauty in the desolation of the surroundings or who, like me, are so in love with animals that simply riding a beautiful horse is in itself a satisfying experience. Nonetheless, even satisfaction and fleeting moments of beauty are not enough to break up the pacing of the game which is essentially back-to-back bosses.
These bosses are remarkable and, unlike
other games, each uses the same basic subsystems such that the player rapidly
improves in their capacity to tackle these foes. There is much fiero to be
gained in defeating them, heightened by a beautiful if slightly repetitive
orchestral score, and some players will also experience a certain sadness and
regret in killing such beautiful beasts. However, fiero is rarely delivered in
large doses without the player having to struggle through a considerable volume
of frustrations, and this game is no exception. I personally found this rapidly
killed any sympathy I might have for the colossi – I became resigned to killing
them, because it was the only choice afforded me and anyway, they annoyed me.
The game does afford more protection against frustration than most; the avatar is significantly more indestructible than any other I have been given to take into battle against arbitrary bosses. Indeed, one rapidly learns that this is a “rough and tumble” game – even more so than the recent GTA games. One can fall huge distances, survive being punched in the face by a hundred ton giant and always has the option to hide away and heal up to full health. This was perhaps the most enjoyable aspect of the game for me, as I relaxed into letting myself be thrown around like a rag-doll. It made me wish (as I often do) that more games afforded great latitude to the player in terms of their survival. It is sad and incongruous that the game still uses killzones to bound some of the play areas, especially as it is not always possible to understand why one great fall takes but a fraction of the player’s health while another is (defined as) fatal.
The main task of the game is to solve the
puzzle represented by each colossi, then climb and slay them. It becomes second
nature after a while. Placing puzzles directly on the spine of the game like
this narrows the appeal of the game, but given the game’s intended audience
this isn’t a problem. It’s a nice touch that a sequence of hints is provided to
bail the player out if they get stuck, although usually these verbal hints
merely tell players what they have already worked out. Still, even this is
helpful, as it reassures the player that they are on the right track. Anything
to alleviate the frustration of play should be seen as an asset.
Associated with this is locating the colossi, which is carried out via a somewhat clunky compass system, in which the player must hold their sword up to the light and move around until the beams focus on a particular point. This adds challenge where none is strictly necessary; if the game was interested in a wider audience, it would have made more sense to simply have the light from the sword point to the path forward when it is held up in this way. Since it is not, it is less of an issue, but it is quite telling to watch players in a shop try and work out what is going on when they activate the sword compass. Unmitigated confusion is the most common response. One can argue that since the audience for this game is going to be enjoying challenge, adding more challenge is an asset. To this I must ask: did anyone find the compass operation rewarding? I am open to the idea that they might.
Speaking for myself, I preferred to use the
sword compass as an orienteering guide. I would take two readings in separate
locations by conferring with the map, and use this to point out the location of
the next colossus (a skill I learned from CB radio “foxhunting” games in my
youth). I found this much easier than co-operating with the sword compass, but
it did mean I had to use the game map which is perhaps the worst game map I’ve
seen in recent years. The map is sufficiently unclear to begin with, without
the strange clouds which clear away only temporarily when the player explores.
I find it incomprehensible that the other half of the gameplay is “secret”. Indeed, I am grateful to Jack for telling me about the fruit and lizard tails (which serve as power ups, thus making the game easier) as I have no way of knowing if I would have discovered this feature otherwise! I consider it to be very bad form to hide (i.e. never tell the player, even in the manual) the mechanism by which the game can be made easier. All games are improved by granting the player the capacity to adjust the game difficulty to wherever they are most comfortable – whether at higher or lower difficulty. I can see no justification, beyond mean spirited denial, or an attempt to provide a small additional fiero payoff for players who “discover” this feature, for not stating this game element somewhere, at least in the manual.
Given how large and dull the game landscape
is (although I personally entertained myself for many hours just riding around
and mentally cataloguing the flora and fauna of this curiously spotty biome),
why hide from the player the only two activities that can be performed?
Shooting and collecting fruit, and hunting white-tailed lizards. (Not counting grabbing
hold of birds and fish for a free ride). My enjoyment of the game reached a
comfortable space only after being pointed towards these “hidden” tasks,
providing a logistical backbone to the play, and also allowing me to control
the game difficulty a little better.
The horse, Agro, while undoubtedly one of the better horses in games thus far, is a curate’s egg. On the one hand, it is beautiful to look at, the controller’s rumble feature is used to great effect, and the capacity to negotiate narrow trails by simply holding the X button and letting Agro guide himself through is enormously satisfying – one of the few places in the game design that the player has been afforded a certain luxury. On the other hand, insufficient motion capture animations were taken for mounting the horse, making this procedure rather annoying in its exactitude. I also was disappointed that while there are animations for hanging from the horse, one cannot grab hold of Agro as he runs past and then pull up into the saddle (or indeed mount neatly from a jump). These disappointments are minor, but they hurt what otherwise is one of the stronger features of the game.
Ultimately, Shadow of the Colossus attempts to deliver its artistic vision on a backbone of gameplay which is old and familiar. No matter how impressive the colossi may be, we are in little doubt that what we are experiencing is a supremely polished rendition of old familiar play, rather than anything new. Since what makes the game worthwhile is the unique experience of play, which is wonderfully realised, this need not be problematic. Still, the result is a game whose artistic vision can only be received by players who are already game-literate; already experienced gamers. I long for a game whose artistry is open to all comers; a game which can truly bury the ludicrous notion that games cannot be considered art once and for all. This is not that game.
The following is my play specification for
the game; apart from various complaints that I will discuss at length below,
this is a neat piece of game design: the gripping mechanics work superbly well
and mark out the climbing elements of the gameplay as exceptional, and for the
most part the toolkit the player is afforded feels surprisingly well rounded
despite remaining completely static throughout the game.
Shadow of the Colossus (Sony, 2005)
Specified by Chris
Move (walk, run, roll, swim - stick; swim underwater – R1)
Climb (climb up, climb across, hang on for dear life - R1 plus
stick for movement)
Jump (horizontal, vertical, hanging, directed - Triangle)
Ride (walk, trot, canter, gallop - X;
horse-jump – context sensitive;
stand on horse, hang from horse, jump from horse,
full stop, quick start – special)
Call (shout, whistle – X; used to call horse, or to lure colossus)
Attack (swing sword, stab sword when hanging - Square in
Aim and Shoot (Square in Bow context)
Look (at colossus ‘twin head view’ - L1; at horse – hold X;
look from side to side ‘vista view’ - L1 with no colossus;
zoom – L2)
Navigate (map reading - Start;
sword compass – Circle in Sword context)
Crouch (and restore health faster - R1)
Pick up (fruit or lizard tail – R1)
Forward roll (required just twice in the game – R1 + Triangle)
Comfort (Circle must be in Hand context and standing
in correct position, annoyingly)
Switch Context (D-pad although also context sensitive)
The following is my structural specification for the game:
(16 | Briefing. [Navigate World][Cross Area][Defeat Colossus].)
[Find Tree][Collect Fruit]
[Find Temple].[Hunt Lizard].
Before I go any further, I want to discuss what I feel is unnecessary complexity in the control scheme. Putting aside the issue that Ueda-san obviously wanted to make this game in a certain way, I nonetheless want to attempt to ‘fix’ what I see as a careless error in the interface design.
There is no value in there being three
selectable contexts – Sword, Bow and Hand. The game would be improved by
ditching the Hand context entirely. If this was done, the D-pad would simply
toggle between Sword and Bow, with the advantage that any player would know
what hitting the D-pad will do. (Although I’m sure some players quickly get to
grips with the three states, this certainly has not happened with any of the
players I have watched play the game, or with myself, even after some twenty
hours of play).
Why do we not need the Hand context? It does two things that I can ascertain. It allows one to comfort Agro using Circle (a hidden feature), and it is automatically selected when one picks up a fruit or lizard tail. Now since picking up items is context sensitive there is no reason for the Hand context to exist for this functionality. (In fact, I question the need for the pick up to be manually triggered at all – why not pick up fruit and lizard tails when you are in the correct position?)
As for comforting Agro, there is no reason
for the player to hit Agro with their sword, and indeed doing so has no effect
except to produce an incongruous tinny noise. Once again, there is no need for
a Hand context at all: the Square button could comfort Agro as a context
sensitive action whenever the player is standing nearby.
I therefore conclude that the Hand context is entirely unnecessary and could be removed from the interface design completely, improving the game a tiny amount at practically no development expense.
While we’re picking at the loose threads of
the interface design, I want to suggest that the Forward Roll verb is
superfluous, and actually faintly annoying. It is used twice in the game – once
in the tutorial segment, once in the final Colossus. However, I have frequently
seen players trigger the Forward Roll by mistake (and done so myself) because
Jump and Climb are used constantly throughout the game, and hence it is
easy to trigger the Forward Roll accidentally while trying to climb.
I strongly advocate avoiding overloading controls (multiple functions on a single control) except when context sensitive (and uniquely demarked) or as part of a game such as a fighting game in which arcane control systems are part of the challenge of play.
Lastly, I want to note that the swinging of
the sword (Square in Sword context) actually does nothing much at all in the
game (except being one way of killing lizards). You don’t use it to fight
colossi (you stab them). This actually means that it would be possible (but not
necessarily desirable) to do away with the contexts entirely. Square could be
sword, activating the sword compass if not hanging on, or stabbing otherwise;
Circle could be bow. I mention this as another example of how this interface
could be simplified; it is debatable whether this particular change would be an
improvement. Doubtless different players would have very different opinions,
according to how patient they were with the context system. I believe if there
were only two contexts, such that pressing the D-pad would have unambiguous
effect, this would actually be sufficient.
Shadow of the Colossus is a triumph of minimalist game design, stripping away much of the
conceits of a typical platform-action game and focussing on a handful of core
strengths. The game asks the player to walk a particular line; those that
manage to do so without experiencing excessive frustration are treated to a
unique and marvellous experience. Others are left feeling they would rather
watch the game than play it.
This is a master class in videogame bosses; I would love to believe that there would be no need to make any more bosses in the wake of this game, but of course, they are the well-charted fiero mines of the Hardcore gaming community, and will be with us for some time to come. At the very least, I hope that other game designers will note how focussing on just a few key subsystems and working upon those produces a more satisfying game (and a more satisfying boss fight) than throwing in every idea one can think of with no regard for how they will fit together.
Like any masterpiece, Shadow of the Colossus has flaws. But many of these flaws are chiefly problems endemic to the games industry as a whole, such as an obsession with both killing and fiero. Indeed, this is a game which could be seen as the pinnacle of pre-existing game conventions; a highly refined product which fails to deliver any significantly new play, but instead offers a polished gem of a more familiar kind. But as with any jewel, be careful not to cut yourself on the sharp edges.