star guard is very nearly a perfect game. it has a small, carefully-chosen cast of characters; though a single monster will necessarily reappear many times, the situations in which that monster is used are diverse enough, and of increasing complexity, that the game never repeats itself and no character gets more time on the stage than she needs.
in an earlier level design lesson, i wrote, “concise design doesn’t introduce new game elements needlessly,” and star guard understands this so well that i’m going to use it as the basis for another level design lesson.
star guard’s protagonist is a green pixel warrior who battles through the stark black mazes of a wizard’s tower[1] armed with a pistol. the pistol fires horizontally; the warrior can jump to adjust her aim vertically. that’s it. the first opponents the player encounters can return her horizontal fire, but don’t really move to adjust their aim. the player has a clear advantage over them, and can easily jump to dodge their shots, align her gun with them, and fire.
shortly into the middle of the fourth stage, the player encounters the creatures in the above picture: one-eyed octopodes that float between the floor and ceiling, firing bullets left and right whenever they change direction. though the creatures don’t react to the player - they’re essentially on up-and-down tracks - they nevertheless move out of (and into) the player’s line of fire and adjust their own as a result of their motion. they also take three shots of the player’s pistol to kill, making it more likely that one will get off a shot at the player before it’s destroyed.
in this image, you can see the player’s first encounter with the octopodes (the image above) with lines traced over the areas that the octopodes’ bullets pass through. what becomes noticable is that there are parts of the map that bullets NEVER pass. that upraised platform on the left, aside from blocking bullets from reaching the player’s flag (where she reappears after being shot[2]), allows a player to observe - and to shoot - the monsters while staying outside their range of fire (this is the green[3] line).
the player’s never seen these creatures before, and this safe spot provides a valuable opportunity to watch and learn their behavior, and allows (but doesn’t demand!) the player to resolve her first fight with them from a safe location. the enemy at the far right of the hall can’t be seen from the safe spot, and the player will have to fight it from a place where it can reciprocate. using the safe spot to kill the leftmost creatures, though, decreases the amount of bullets that are travelling the hallway when the player sets foot in it, a simple lesson that the game will reiterate.
this vertical shaft is the player’s second encounter with the octopus creatures, a little further into the level. every step in the shaft is “unsafe” - the player can’t attack an octopus without putting herself within range of its bullets: but note that the player is only threatened by a single enemy at a time. the safest way to play through this area is to take it step by step, defeating each octopus in turn. the vertical orientation of the area encourages this: the game window is wider horizontally than vertically, and the player can only see as far as the next step at any given time.
level four finds its climax in this big area - there are no safe spots from which to shoot at enemies without being in range of their bullets (though the flag is in a safe spot, there’s no way for the player’s fire to reach any enemy from there), unless the player manages to get behind one. and that’s hard: there are lots of enemies, and some places are covered by more than one, meaning that there are lots of bullets in the air.
but those unsafe areas become safe - the red lines vanish - when the player destroys enemies. the player was taught that in the vertical shaft - where defeating each enemy made the player’s current step safe and unlocked the next, and where the player was forced to fight one enemy at a time (and even in the first hallway, where the player is allowed to pick off some enemies from safety before proceeding to fight the rest). the game is teaching strategy: defeating enemies opens safe avenues to defeat the next enemies and eventually reach the exit. in fact, this is the essential strategy of the game.
this is all taught, of course, through the situations the player finds herself in and through the order in which she encounters them. why should the author need to explain the game in text[4] when there are already rules, systems, and a player who’s engaged with them? this game, star guard, is a demonstration of good design.
[1] the juxtaposition of technology and sorcery perfectly recalls the tone of older arcade games like wizard of wor and joust, which were unsurprisingly interested in the romance between machines and magic.
[2] one of the things i appreciate about this game is that it never limits the amount of times you can die unless you specifically ask it to.
[3] the game’s simplicity and economy is reflected its bright-color imagery: friends are green, threats are red (with white for accents and effects), and the walls of the tower are yellow.
[4] i do like the game’s use of text to put the events of the game into context - in the background, where it doesn’t interfere with the player’s movement through the game. i think it says something about the deliberate pacing of the level design that episode’s of the game’s exposition can be attached to specific moments in the levels.
19 comments
Star Guard is an amazing ‘tribute’ type game, taking the elements and making a stunning new amalgam. It plays exactly like something I would have fed quarters into in my childhood. The support troops add a nice bit of dramatic depth to the proceedings.
Great Analysis Anna — thanks for this.
I enjoyed Star Guard, but I found the boss implementation to be quite flawed. The game gives you infinite lives without loss of progress up until the final encounter when you have to start the battle from scratch each time you die.
It’s possible for virtually anyone to play up to the boss because you can die constantly without consequences. This means that by the time a player encounters the boss, they might not have developed the skills to defeat the boss. The game doesn’t force the player to learn these skills until this encounter, which was pretty frustrating for me.
Eh, I dunno, it just seemed like a bit of a giant change in terms of difficulty. Or am I alone in thinking this?
First of all, many thanks for the recommendation - I enjoyed the game a lot!
Secondly, as an answer to Sergio: There aren’t any skills you need to pick up, the game mechanics are so bare that there isn’t many skills you need to pick up. And I don’t think you can get to the wizard without picking up some serious dodging skills (the last bit before the wizard).
Anyway, what I really disliked was the end, or should I say: lack of end? There was this great gloomy build up from all the plot messages, that kept getting gloomier and gloomier, and in the end you just kill the wizard and that’s it. Dunno, I expected something :)
Hey Anna, great piece on a great game.
I know that you’re interested in growing a workable vocabulary when it comes to games and level design, so I wonder if I might make a suggestion:
When talking about the vertical shaft you say that each step is ‘unsafe’. I suggest using the additional term ‘control’, which is from Chess. Ie. “Each enemy gunner and octopus ‘controls’ the horizontal space on either side of them until they are destroyed, making each surface of the vertical shaft ‘unsafe’ for the player.”
The idea being that it’s easier to directly link the source of the danger with the effect had on the environment.
Anyway, again, really awesome stuff and I’m so glad you’re continuing with the series!
Too repeat a (by now) well known opinion, great piece on Star Guard.
I picked it up the other day (from the Indie Games blog) and was thrilled by how simple yet fun it was. Glad to see someone else noticed as well.
Keep up the good work Anna, and I’ll keep reading!
Thanks for writing this piece, it’s a good read. Educate! Godspeed!
Please could you capitalize your writing properly. It would make me happy.
is your happiness really contingent on whether or not a writer on the internet uses uppercase letters?
the wizard is a harder fight than the rest of the game, yes, but not inappropriately, i think. i can accept that the player should be capable of demonstrating a certain amount of discipline to defeat the wizard (and the HARDEST part of that fight does allow the player to continue at death without resetting the battle).
> with white for accents and effects
the white bits are not just decoration, though, are they? i can’t have a look at all the enemies[0] in the game right now, so i’m not entirely sure … but IIRC, hitting red pixels with your weapon always hurts enemies[1], whereas hitting white pixels does nothing.
[0] by enemies i mean everything red and hostile, not just the creatures.
[1] not sure about the triangles that fall down, but when your gun can reach them you’re safe and they’ll vanish in half a second anyway.
you’re right, white pixels are often used to signify enemies that can’t be hurt, though i don’t think it’s completely consistent.
in the meantime, i’ve downloaded the newest version of Star Guard and had another look. unfortunately i didn’t get very far, because it still has this bug where after loading a new level often the graphics are completely missing :(
anyway, the word “signify” totally nails it: it’s not pixel-perfectly consistent, but it’s good enough to help the player quickly figure out what’s going on.
I despise that last section in the 9th level. It’s awful on every level. The best part is that the firing pattern is completely random, and will often-times screw you over completely, so that it loses all thought and becomes a game of chance.
DISREGARD THAT I SUCK COCKS
As it turns out there is some sort of odd bug where the turrets fire faster if the game is slowing down, and since my computer is shit, it makes the section nigh unplayable.
Yup, the last section (dodging bullets just before the Wizard) is just pure unfair to all those who play on lower computers, because absolutely NOTHING up to this point suggests that computer’s speed would be a problem!
I took me about half an hour to pass it (which means to have enough luck in the pattern) and makes the lovely hard mode just impossible.
Boo.
I lean toward Sergio’s opinion of the resetting during the first stage of the boss fight. Depending on a player’s style, this can either seem like a reasonable challenge or act like a brick wall. I’m inclined to make the first stage’s health not reset in future versions, and would be interested in any second opinions people might have about this.
I’d also like to apologize for the slowdown some people have encountered in later stages. The turret sequences really are quite unplayable at low framerates. While Flash is partly to blame, I think most of the fault lies with my own second rate programming ability. (I intend to make some optimizations in future releases.)
On the topic of the random firing sequences in the later stages, I’d be curious to hear what people’s reactions are. The random turrets appeal to me in some ways, but I feel that perhaps they should be deterministic because then they’d work better for speed runs.
Excellent article. Thank you. This is a great explanation of how games teach. If only “educational” games could do so well.
I love the minimalism I see described here, and now I want to play this game. Downloading now…
I enjoyed your treatment of Star Guard’s level design. I’d like to discuss some beautiful, subtle elements that improve the game both on the first play and for experienced players.
warm-up animations - Enemies that shoot at you have long, exaggerated, easily-identified animations preceding the creation of a projectile. This helps players to learn the enemies’ routines and allows more daring navigation of the level: getting up close to enemies isn’t a bad idea if you know you’ll have time to dodge a shot.
hit-box trickery - this is popular in bullet-hell type scrolling shoot-em-up games but I haven’t noticed it before in run-and-gun platformers. The easiest situation to see it in is when firing at small, pistol-wielding enemies standing on ledges above you. If your shot hits the corner of the ledge, the enemy will still die. The player’s bullets have a sprite one pixel high, but the hit-box extends above and below. The result is that enemies are easier to hit and the player has a better chance in tight situations. This system could be called dishonest. This choice not only allows the developers to be more forgiving without compromising the “lone soldier” aesthetic, but also reflects an appreciation for the saliency of the objects on-screen. The player needs less information about where his own projectiles are (he put them there) and extra information about hazards like enemy shots.
re-jump tolerance - if you press the jump key just before you land, you’ll begin a jump immediately. This is great in tight spots: having to react to the fact that you mis-timed a tight jump almost always means death, and the end of an interesting situation.
The small yellow bump that appears whenever the player jumps is another really nice touch. The player can see visually how near to the ledge he ran before jumping, and over time this feedback helps him jump more precisely.
I haven’t been this impressed with a 2-d platformer since N. The reaction to Star Guard in the scratchware and indy-game community has drawn me closer to you all.
Bad editing above. “This system could be called dishonest” is a remnant of a previous edit, and should be ignored. I was going to say something like “This system could be called dishonest but such a hard-line stance would have to ignore the many advantages it offers.”
Regarding the “lone soldier” aesthetic: it works well but it always has some blemishes in terms of narrative. For example, the enemies and bosses in the Megaman X games are large and cumbersome, with simple movement patterns and minimal intelligence at best. The player cast as the underdog, a little guy fighting against long odds, but in fact he has the most mobility and firepower, so the enemies can seem a little bit pathetic.
post a comment