The Anatomy of Super Mario Gaiden III: The anatomy of bad games

[Blog post topic requested by Will Coon]

After the first write-up on World 1-1 of Super Mario Bros. (yes, I wrote three posts on that single stage, perhaps as a result of insanity), Fritz F. posed a question: Was Super Mario Bros. really that well-designed, or is everyone who dissects the game simply playing The Bible Code with its level layouts? This, I think, is a totally fair and reasonable question. However, it’s also one I feel perfectly comfortable in answering with a “yes.” That is, the game really was that well-designed.

Rather than simply leave it at that, though, I’d like to use one of my blog post topic request slots to provide a counterpoint. Let’s look at three side-scrolling Famicom platformers released within a year of Super Mario Bros., all of which demonstrate what mediocre design looks like. By comparing these games to Mario, we can get a better sense of Nintendo’s legitimate genius.

Challenger: Hudson, October 1985

Released only weeks after Super Mario Bros., Challenger is interesting in the way it demonstrates the parallel evolution of video games. If Nintendo hadn’t made Super Mario, platformers still would have happened — Mario had no time to influence this game. And, as we can see here, games probably wouldn’t have happened nearly as well without Mario.

I admire the ambition behind Challenger. It’s a much more complex game than Mario. This, however, is also its downfall. Challenger does so much for a game of its vintage, yet it does a terrible job of communicating its mechanics and rules to the player. It begins with something almost akin to an endless runner as you travel along the top side of a speeding bullet train, leaping birds and criminals, evading random lightning storms, and grappling with your greatest foe of all: The terrible controls and collision detection.

As you can see in this video, once you reach the end of the train, you duck down inside it and… eventually, the player becomes invincible. I don’t know why that happens; it’s never happened for me when I’ve played, at least not that I know of. There’s no indication of the change in the player’s state… it just kind of happens.

Once the train is conquered inside and out, the game undergoes a modal change, transforming into a top-down adventure. There’s even a life bar! Occasionally you duck into special chambers where you collect items, and you can acquire offensive weapons. But the top-down sections are grueling. Enemies appear constantly, homing in on you while moving with enough randomness to frustrate. The world sprawls enormously in a bad way, with no visual cues for where to go and tons of empty space to wander through. As you can see from the video, completing the game takes about six minutes if you know what to do; there’s very little substance here, and the unfairness of the design pads out the play time by forcing you to repeat the same clunky sequences over and over again until you finally nail down a path.

Amusingly, the finale of the game (which I’ve never seen in person) resembles a really terrible Donkey Kong or Donkey Kong Jr. clone. It’s set in a single screen, you climb to the top by evading monsters and objects circle a rock, and take out the bad guy. But it sucks, with little in the way of visual interest or appealing challenges.

Atlantis no Nazo: Sunsoft, April 1986

One of Japan’s famous kusoge, Atlantis no Nazo may be a bit like cheating in this topic. It’s aggressively unfair, deliberately unfriendly, methodically baffling. This action game spans 99 stages, but finding your way through them is an exercise in luck, blind luck, dumb luck, and impossible persistence. Also, buying guide books and begging friends for advice. The play controls are awful, and the hero’s only offensive skill consists of dropping a rolling bomb that can just as easily kill him as a foe — and often is necessary to reveal hidden secrets, at times in places where it’s nearly impossible to detonate the bomb correctly without effectively committing suicide.

Opaque goals, no guidance, invisible and untelegraphed items and passages necessary for progress, difficult platforming with poor controls, level design that goes from zero to insanely difficult almost immediately: These are not the ingredients that make a good game. The first stage consists of a flat plane with birds pooping at you, but subsequent stage designs are all over the place. Yes, with 99 levels, Atlantis no Nazo is three times as long as Super Mario, but it’s nowhere near as good (or fun) an experience.

Milon’s Secret Castle: Hudson, October 1986

Of the three games here, Milon’s Secret Castle is definitely the most polished. In the year between Challenger and Milon, Hudson really doubled down on making a more enjoyable game. The controls are more responsive, the design more consistent, the power-ups clearer. The latter change is probably the most effective: Milon upgrades his skills with permanent boosts purchased in shops. No confusion, no ambiguity, no mysteries.

Now, getting to the shops, that’s another matter altogether. Finding your way forward through this castle requires blasting pretty much every single tile of the game, both walls and empty spaces, in search of hidden objects. Like doors. This is the case from the very beginning of the adventure, the very first level. Milon’s Secret Castle mostly consists of shooting every inch of the world in search of goodies while trying to avoid foes.

Said foes, incidentally, spawn constantly all around the screen, move quickly, and follow erratic patterns. The further you get into the game, the thicker and more unpredictable they become. The game does follow a difficulty curve, but it leads to a point that feels insanely frustrating and can end you in a moment when you least expect.

I encourage everyone to try these games for themselves and compare their opening areas to that of Super Mario Bros. and see for themselves just how systematically Mario rolls out its design elements, one by one. Like these games, it contains hazards, but in Mario they always behave consistently and predictably. Mario contains plenty of hidden secrets, but unlike these games those invisible elements are strictly optional bonuses. Your path forward is never in doubt; your abilities and expectations alike are clearly communicated. You’ll rarely come across a moment in Super Mario where you don’t know what to do or die unfairly, and when you’re left to potentially learn through failure the resulting penalty is always mild and easily recovered from.

There’s a sometimes subtle difference between Super Mario and Milon’s Secret Castle, but that difference marks the divide between a decent game that was quickly forgotten and a timeless classic that inspired an entire industry.

Please feel free to discuss this further in the comments!

8 thoughts on “The Anatomy of Super Mario Gaiden III: The anatomy of bad games

  1. I could be missing some games, but personally at least on the NES I don’t think there were any well designed sidesrolling platformers approaching SMB standards until either Contra or Castlevania. While I like Ghosts and Goblins it was too punishing to be really considered well designed. And of course those other two games practically spawned whole new side scrolling genres.

    • I almost mentioned GnG, actually, but I don’t think it’s bad per se. It was designed with a different ethos, as an arcade quarter-gobbler.

  2. It is very true that there was no NES game before, and very few games after SMB that could hope to touch the level of polished controls and superior level design of this masterpiece, but there´s not saying SMB didn´t have its moderate share of opaque design… Namely the repeating maze castles in worlds 4, 7 and 8. Those are pretty confusing the first time around, and even when you know exactly what´s going on, the lack of a visual or aural cue of one being on the right path make them a bit daunting. That´s why in Super Mario All Stars for the SNES that flaw was addressed by adding a chime when you chose the right path and a buzzer when you went wrong. This feature coul´ve easily been included in the original nes version, so I guess it was a slight design oversight the first time around… Granted, those mazes don´t appear until the midpoint of the game, and by then a player has to be well versed enough in the nuances of the Mushroom Kingdom as to face this guessing challenge with aplomb. Unlike those other games you mention and dozens more where the very beginning is already baffling and discouraging. Just pointing out the only flaw I can remember in SMB.

    • True. I considered mentioning the mazes but its been so long since I’ve played them I don’t remember the details and didn’t want to misstate anything.

  3. There are numerous examples of less-than-remarkable platforming design – more than could be touched upon in this blog series – but another one came to mind as I was reading this. Mystery Quest (or Hao-kun no Fushigi na Tabi in Japan) is a 1987 FDS release that reminds me a lot of Milon’s Secret Castle, except more linear. There are signs in the castles that point you in the general direction you need to go, and the breakable blocks, while frequent, aren’t quite as maddening as in Milon. However, the difficulty curve is weirdly inverted, since the enemies don’t really get that much harder later on, and the inertia of your jumps is a little fussy. Not to mention that the US version was severely gimped, with large sections removed and two of the castles being made into clones of the other two castles. And you have to play the game four times in a row with minimal difficulty increases to see the real ending.

    It’s worth giving a three-minute test drive if you ever get a chance – it’s a little more decent than the visuals let on – but you’ve probably had your share of platforming garbage over the course of your lifetime.

  4. >> Of the three games here, Milon’s Secret Castle is definitely the most polished.

    Well, that’s a sentence I thought I’d never read.

    Fair enough, though :) I do still wonder how much of what makes SMB better than these other games was done *on purpose* (vs. Nintendo just getting lucky), but I guess it’s hard to make that many levels purely based on lucky.

    And I gotta admit I have a weird nostalgic fondness for utterly obtuse NES games like Milon’s Secret Castle, Legacy of the Wizard, etc. When I was younger, the fact that I didn’t understand two-thirds of what was going on made me feel like these games contained all the secrets of the universe, and I’d discover it all if I just explored enough.

    • Haha, no doubt.

      As someone mentioned before, Miyamoto has talked pretty extensively about the design of games like Mario and Donkey Kong and how much thought they put into simple things like the character’s starting position.

Comments are closed.