You know that feeling of being in a completely new place but feeling like you've already been there before? Perhaps hundreds of times? The French call this deja vu, and it's something I experienced every time I attempted to conquer Spelunky on the PC. That was years ago when I wrote my review of the original Spelunky (referred to as Seplunky PC). Now a XBLA HD spiritual reboot is here. Though the walls haven't completely shifted with the core gameplay design of this new version, it does feature new graphics, controls, and gameplay modes. The original game is free and filled with game design gold. The new Spelunky costs a bit of cash, so I wonder what additional game design gold can be mined from it.
Co-op is perhaps the biggest new addition to Spelunky, and it's the reason why I picked up the game for XBLA even when I didn't own an Xbox360. After extolling the wonders of roguelikes like Spelunky PC orShiren The Wanderer to my friends, and them never giving the genre a serious attempt, I knew that co-op would be the only way to share the highs and lows of the roguelike experience. Fortunately, the co-op design of Spelunky is great.
In Co-op players have access to new co-op mechanics boosts, which is level 5 co-op design. By picking up other allies, players can hitch a ride, reach high places by jumping out of holder's grip in mid air, or be violently tossed across gaps. One particularly good combination is when one player has the jet pack, the other has the shotgun, and they pair up to create a flying shmup like killing machine. This combination is very effective because, nearly everything players do in the game via their mechanics affects and likely injures other players.
"With multiple players interacting, co-op becomes a much different game. It's easy to help or hurt each other, so socialization is important!" ~Derek Yu
Tossing a rope hurts allies as well as throwing a bomb, rock, gift package, or just about anything else in the game. Likewise, every weapon in the game has friendly fire. Because all co-op players must play together on the same screen, the potential for danger and disaster dramatically (and hilariously) increases with every player added to the adventure. This design also means there's a lot of potential for co-unter-op opportunities, where angry players can work against the good of the group. This basic design is very similar to New Super Mario Bros. Wii where co-op characters become significant and dynamic platforming obstacles because they're solid bodies. For Seplunky, the increased danger potential also opens the door for potential cooperation, stressing individual player skills, team skills, and callouts. As far as mechanics design goes, Spelunky's co-op mechanics and design are of level 5 and 6 co-op meaning there is a lot of dynamic and interesting gameplay potential purely through player cooperation and coordination.
Beyond the mechanics level, there's are two major co-op design issues to consider; difficulty and participation. The whole point of playing co-op is to play together. The experience isn't the same if one player is removed from participation while the other player carries on through the rest of the game. New Super Mario Bros Wii let's players return in a bubble at the cost of one of their lives. Halo lets players respawn near a teammate as long as there's no danger around. And in Portal 2 dead players automatically respawn at the start of the level without disrupting the progress of the other player. All of these co-op design choices work great for their genres, but would inherently work against one of the core unique features of roguelikes; permadeath and the tension it adds to the experience.
Spelunky solves the co-op participation problem in a way that preserves the roguelike gameplay. When co-op players die, they cannot return to the game until any remaining players break open their coffin on a later level. So the remaining players not only have to survive to the next level, but they have to find and navigate to the coffin to bring back the dead player. Depending on what items and upgrades the living players have, this can be very risky and very costly. But overall, it's not hard to revive fallen players, who respawn with 4 health and no equipment.
This respawn design shapes co-op adventures differently when there are more players in the game. The more players, the more dangerous it is for everyone to survive. And because there's only one coffin per floor, when multiple players die, they have to wait even longer to be revived. With this design, keeping four players alive is a distinctly different challenge than doing so with 2 or 3 player co-op. You may find yourself sitting out of the game for a while. I appreciate these moments because they break up the survivors into different group combinations, while the dead put their hope in their friends.
Fortunately, when you die you never actually leave the gameplay experience. Rather than sit out patiently, you control the ghost of your character who can float uninhibited through the level, blow objects around, and disarm some traps. These ghosts can be very helpful, and interestingly, very harmful to the living players. One time, my friend blew me and my brother into a small pond with a single piranha in it. And to our horror, that one fish locked both of us in an infinite combo until we died. I had 6 health! For such a tragedy, it was pretty funny. In Spelunky, the co-unter-op design is maintained even beyond the grave.
Another design element I like that's particular to the co-op design is equipment drops. When players die, they drop their equipment and items, which include bombs, ropes, capes, rocks, shotguns, and other items that cannot be manually removed when alive. This design makes the location where a player dies a potential hotzone of risk-reward. Venturing into danger to recover dropped equipment has costs me several promising runs.
The biggest problem I have with the co-op design of Spelunky is the camera. Only one player controls the camera in co-op, and this player is marked with a white flag. When other players move outside the screen view, a countdown starts. When it reaches zero.... well, I'm sure you can guess what happens. It's clear that the camera was designed around the single player gameplay. But because of the camera design, successfully progressing as a group often slows the gameplay down as the camera player not only has to scout the areas for him or herself, but for the rest of the group as well.
Though I love the clarity of the single screen presentation, I wonder what other design options could help in this area. Perhaps a dynamically splitting camera system like Bionic Commando Rearmed that splits the screen vertically or horizontally when players move too far apart from each other. In general, giving players separate screens to do their own thing tends to compartmentalize and separate their gameplay experiences, which works against what co-op gameplay is. Perhaps it would be better if the little magnifying circle that indicates the relative position of the off screen character was a little bigger so that players could see enough of the level around them to make it back on screen.
There are a few design issues that I want to point out in the core Spelunky design. As I've described in detail when I talked about the raw design of Spelunky's Deathmatch mode, there are many design feature that have that indie feel to them. I've been reading Derek Yu's thoughts and comments on Spelunky since 2009, and one thing that he's been consistent about is that he wanted Spelunky to be a fast paced game. In many ways Yu succeed because the character movement in Spelunky is really fast. But I argue that it's too fast. It pushes the limits of the game's ability to provide adequate feedback to the player so they can make skillful, informed decisions. In other words, when your character moves too fast and erratically to actively see and comprehend its positoin, then the best you can do is make guesses.
Derek Yu and the Mossmouth team did a great job polishing Spelunky by making all kinds of tweaks and adding details to everything. However, just like the original Spelunky, there are still many questionable interactions that I consider to be bugs. For example, I've cleanly jumped on a frog only to have it hurt me and then die. Trading hits like this is cool in fighting games, but not in platformers. For another example, I was crushed by a bolder that wasn't moving. I'm pretty sure I pushed my character up against the rock in a way that made the game think I was pinched between a rock and a hard place. This is just like the time when I was "crushed" while resting on top of Olmec's head when I died on my best run playing Spelunky PC so many years ago.
I do not like some of the hitboxes tuning in Spelunky. There are times when I think I dodged or hit a target only for it to hit me anyway. Programming wise, I know that the faster characters and targets move, the harder it is for a game to calculate hit detection. And I think the high character movement speed may be a reason why the collision interactions are occasionally janky. Another reason may be that with the upgrade to HD graphics, the enemies and objects are no longer designed in a blocky, quantified degree. In Super Mario Bros. almost all of the enemies are designed around the brick unit. They're either one brick in size, or their size can be approximated by stacking these brick units. With the the increased fidelity of HD graphics, there's a possibility that the hitboxes in Spelunky are drawn outside of any kind of quantified measuring system. So when you see a frog that looks to be about the size of a brick, it may actually be slightly larger. When you see a bat that's smaller than a brick unit, it may actually be larger.
Just like Super Meat Boy, I find that Spelunky is oddly lacking in sound design, which is another indie feel example. Spelunky enemies don't make much sound at all. Bats make a soft flapping sound when they drop from their roosting position, destroying an enemy with a jump makes a pallid pop, ropes make a whoosh sound, and bombs are bombastic. But nearly everything else lacks sound. The bees don't make a buzzing sound. Getting hurt doesn't make a sound. Jumping off an an invulnerable enemy hardly makes a sound at all. If the most important actions are combat and platforming actions, these sounds need to be distinct in the soundscape so players can better time and understand their actions as well as hear incoming hazards.
It seems like a mistake not to have a sound effect for jumping or landing or grabbing a ledge. My theory is that many indie developers create soundscapes that are similar to the experience they have playing games while listening to other soundtracks. This means that the music is foregrounded while the gameplay sound effects are subdued and minimized. Designers who put gameplay first must not create soundscapes like this.
And then there's the HUD. In a game with looming permadeath where the levels are randomly generated, and hazards can attack from any direction, it's essential to inform players. With the somewhat lacking sound design, more stress is put on the visual feedback in Spelunky. Unfortunately, the HUD elements often obscure critical visual information, just like the HUD in Super Smash Brothers Melee. So, taking an innovation from Smash Brawl, it would be better in Spelunky if the HUD went semi transparent when it obscures an enemy or player. Better yet, the HUD should go semi transparent whenever the player is moving. It's not like the HUD presents information that is constantly important to the player. Especially with 4 player co-op, I'd rather have a bigger screen view than a persistent HUD.
In part 2 I'll wrap up by addressing the topics of randomness, control, skill, and teaching.