With a Kickstarter page that name drops Super Mario Bros, I Wanna Be The Guy and Super Meat Boy, it's no surprise what kind of game Cloudberry Kingdom was designed to be-- a crazy-difficult side-scrolling platformer that really, really wants you dead. But what sets this game apart is the way the levels are made: procedurally, using a ridiculous algorithm (detailed here in this helpful Gamasutra article). We've seen a lot of games use similar algorithms (Minecraft, Terraria, etc.), and they tend to do pretty well. As that article says, though, no one's really done the same for platformers, and why shouldn't they?
This same Gamasutra article gives me a pretty good starting point to look at this because it outlines the ways the developers wanted the levels to turn out: beatable, challenging, and interesting. So let's go through each goal and talk about how well it does this.
Are the levels... beatable?
This is the easiest to talk about because the game doesn't let you play a level unless it can prove that you can beat it. There are a series of checks in place (go read that article!) to make sure, and if you don't believe it, you can spend some of the collectibles to watch the computer do it.
Perhaps more awesomely, the game also creates levels with the limitations of each hero in mind. Heroes range from a one-jump regular to, say, having a jetpack, or being strapped to a wheel. It's actually pretty neat to go into one of the modes, choose a type of hero, and see the game figure out what it needs to do.
But the thing about how the game makes its beatability checks is that it only makes sure that it can be beaten only one way, so each level, even the craziest looking, boils down to a specific, and usually pretty basic, path that you have to stick to, or you'll have a much more difficult (almost impossible) time getting through. This mostly means that even a level as crazy as this:
...can be reduced to running to the right and jumping over obstacles. There are levels that are more complex, but usually this just means you have to stop for a second (literally) and then retry that exact same method. So while they're certainly beatable, the game only gives you one way to finish every level, limiting the amount of fun you can have experimenting and trying to find the best way for it to work for you, instead just figuring out the only way the game allows.
But are the levels... challenging?
So even though each level has a best fit line and a "don't stop for nothing" mentality to beating it, that isn't to say it's not challenging. It is, and it wasn't long before the achievement for dying 1,337 times popped up. You will die. A lot. As another aid, you can actually buy a visualization of the path that sort of acts like a racing line from Forza, showing you where you need to be and even giving you a bubble to stay in that acts as a guide to when you need to speed up and slow down.
But the problem is these paths are ridiculously tight. If you're a little bit too fast or a little bit too slow, it's unbeatable, and even being a millimeter off from the guide bubble will result in quick death. This would be be okay if the controls were tighter, but I never felt great about the physics, especially since they're always changing throughout the levels. Even if you follow the path it sets out for you, following as close as possible to the dot that guides you, you'll still die all the time because the correct path is as big as your hit box. Vary a little, and you'll never make it.
And that's actually really annoying. You don't get the same feeling of overcoming adversity that you do in a game like Super Meat Boy, where there are so many paths to victory, and missing the originally planned path still gives you the ability to recover. Instead you'll have to try the same path so many times to the point of boredom. You stop wanting to play the level and quit out instead, so that the game generates an entirely different level.
So the game tends to fall between so easy it's boring, and so difficult it's boring. It seems to misunderstand what makes all of our favorite challenging platformers so fun. Even something as sinfully difficult as I Wanna Be The Guy rewards action/reaction, adaptability, and your own clever thinking to get past challenges, and doesn't just accept that there's only one way to beat every level. Doing so leaves you, the player, feeling less engaged, and coupled with having to restart levels so constantly, it leaves you really bored even in the face of a level so jam-packed with hazards.
So is it interesting?
In games, we deal with the uncanny valley a lot. The way that a human character looks super real, but there's just... SOMETHING missing that keeps it from actually seeming real. In a way, Cloudberry Kingdom creates the same sensation for the platformer genre. It's got all of the superficial things, but it lacks that spark inside of it that really brings those games to life. Though the level creator algorithm is no doubt fascinating and a cool idea, it doesn't ever create anything that actually makes you want to come back and play it again.
I know a lot of people probably read the line above about how each level can really just be beaten by running and jumping at the right time, and thought to yourself "isn't that most platformers?" But if you actually stop and think about what made those platformers great, you know that just getting from point A to point B isn't the only thing that makes a game interesting.
Do you remember how you felt the first time you found a secret block in Super Mario Bros.? Your first pipe that allowed you to go to a substage? We take these things for granted by now, but when you discovered that, it was amazing. You found something that just seemed to blow the game wide open and make it about more than just reaching that flagpole. Each level became intricately full of secrets to discover, and each one you found made it a more fulfilling experience than it otherwise would have been. It's something platformers have done for years, even up to recent ones like Super Meat Boy and Rayman Origins-- those secrets, that sense of discovery, the fact that there's more than one way to get through a level and the feeling there's still something to find is a great feeling that they do well to cultivate.
Beyond that, even the simple assertion of "all platformers can be beaten that way" falls immediately flat. This isn't even true for Sonic games, a series known as "hold right and win"; just look at this video of Marble Hill Zone and you'll see that, in less than a minute, the game's already switching things up. All of your favorite games in the genre do things to change that core mechanic because, let's face it: if you're just doing the same thing, over and over, you're going to get bored of the game.
Now imagine you're doing that same thing for infinite levels. Infinite levels of the exact same type of gameplay that just gets more difficult as you do them.
Even more than that is a complete misunderstanding of what makes challenge so fun. IWBTG or Super Meat Boy are difficult and full of all kinds of obstacles, true, but they give you more than just that one path to beat them. It's that action-reaction. That feeling that the game is testing your understanding of its physics and your abilities through a series of challenges. That's the kind of thing you only get when you have an actual person planning out a level. The time put into the design of a level really shines through, making the game more fun for the player. Sadly, this is completely lost in Cloudberry Kingdom.
Seeing as the game can literally make any kind of level, it does eventually spit out a couple of cool ones that test platforming skills instead of patience. Unfortunately, unless you make the game generate those types of levels, you'll almost never see them, and if you make the game give you just those kinds of levels, they're not varied enough to keep you coming back.
On top of all of this, the game has almost no personality. There are some cutscenes in story mode (starring one Kevin Sorbo as your main character, Bob), and while those have a cool look, the actual animation and writing are terrible. Perhaps it's no surprise that a company called Pwnee Studios filled their story with a bunch of bad puns, but it does the game no favors when it comes to actually being charming.
The look of the levels also doesn't really do much to invest you, and have the problem where all of the obstacles run over each other, making everything look like a confusing, obfuscate mess. Expect to die hundreds of times from traps you couldn't see or a projectile coming from the top of the screen that you'd never have been able to react to.
This isn't to say that procedurally generated platforming levels can't be fun, but you still have to have some better planning. A good example is the Rayman Legends Challenge Mode (available for free on the WiiU), where the game randomly arranges pre-made blocks to give you ever changing, and infinite, levels. But that still has the added benefit of having someone plan each individual block before the game mixes them up, and giving new goals and play types every day gives that game a lot more replayability.
While I certainly applaud the work done on Cloudberry Kingdom's engine, it has one big flaw: computers don't understand what fun is. You can have a computer make a level, but it'll just be following parameters. To really achieve something entertaining, there has to be more of the creator in it, or more interesting gameplay on top of a neat algorithm. By divorcing themselves so much from the experience, Pwnee Studios allows their game to suffer from not being able to judge whether a level is actually fun, making it come off pointless, shallow and uninteresting.