Sunday, April 28, 2013

Why VYDE Needs Fixing

1) You Fail A Lot
You fail often, you fail quickly. This process is not at all rewarding because I can easily tell that the player has no  incentive to go on when she hands the game back to me after trying only twice. There is no flow to keep you going and the game doesn't do a good job at telling you how exactly you failed. Running into an enemy and getting spotted in the open are the only two ways you can fail, but it's not made apparent because there are no field of view cones and the game waits for exactly one second before it moves to the end game screen.

2) Zero Conveyance
The game has to resort to tutorial screens to explain the simplest of rules, including it's control scheme. There aren't enough intro stages to explain all the rules one by one. The rules :

1) A matching background color makes you invisible
2) A perfect match makes you completely invisible
3) Enemies will spot you if inside their field of view and visible (1,2)
4) Touching enemies is allowed iff you're completely invisible (2)

Rules 1,2 and 3 are conveyed via tutorial screens in one level, and the last rule number 4 is explained in the one after. Players don't read the short tutorial texts and the rules are never made clear.

The control scheme is new, players don't understand how to properly use it unless I demonstrate it for them. Once they do understand, they have no trouble with it, they actually admit that it's better than a traditional joystick for the freedom it provides by better understanding the player's gestures.

3) Just Puzzles
The levels are very small (6x4) and every level has hardly more than one solution to it. The extra pickup is a nice twist but they're mostly too hard to execute. In effect, players never beat a level taking a path I don't know already. Some levels are less brains and more hands which are even harder to complete for those who are uncomfortable with the controls.

4) A Flawed Concept
I'd mention in an elevator pitch that it's a stealth game about hiding in colors, and to keep with the simple minimal art style I had in mind, the player character is mostly hard to spot and sometimes even invisible. I expected players to understand that the character in control is the one that's not moving on it's own without player input (patrolling enemies) but it seems it's not that obvious to most. I could make Bob stand out in some way but then I'd have to sacrifice the aesthetic style of the game.

How to Fix It
Only so much can be fixed really. A total of eighty levels have been completed and tested so it's not feasible to change them all. Failing over and over is not fun and the players need to be encouraged to try again, the level now restarts itself automatically within a second of getting detected. The tutorial screens are useless because nobody has the patience to read through them, the player needs to be able to learn the rules by PLAYING the level instead of wading through a wall of text.

Also now on starting a level, Bob is revealed from under a cloud of smoke indicating to the player that there's something special over there, perhaps the character they're supposed to control. I don't want to put more junk on the screen like a visibility meter, an outline on the player character or even a permanent joystick. If the game world is not doing a good job in explaining itself then there's something wrong with it.

Now, how do you teach players a completely new way to interact with a game world behind a touchscreen?

Saturday, April 6, 2013

Last Day of Winter

I've been meaning to give a detailed report on this game with a complete dissection of the things done right and the many improvements over it's predecessor, but I'm afraid we'll have to settle with an extremely short and informal discussion for now. Spoilers below, it should take little over six minutes to complete the game, I highly recommend you play it to the end.

Event Structure
The game holds a set of events that are initiated each time the player makes it across one scene, the first half of the game where the king is collecting gifts is where this system is employed. This is by no means a mechanism that is trying to generate a procedural story, they're just events, for example, a pack of ghosts, village folk crowding around the king or a wizard with healing powers. The game selects an event form this list and randomly populates the scene with other simple game objects like trees, plants and grass. When we speak of stories in games we often find the player's actions working against narrative, they're always fighting each other and it's our job to have them both somehow have an equal say in how things progress. A story in the ultimate sandbox game can be a problem because one can always talk the nemesis into a ditch and drop lava on them.

Almost every player action has a couple of numbers attached to it (only one in our case), numbers that describe how certain metrics are changed. The system tries to complete the feedback loop between what events the game is producing and the player's actions, ensuring a smooth dramatic arc (see Robleto, Dramatic Arc) by trying to measure things like dramatic tension and danger. So if you've been collecting a lot of rings for a long time, the game will ensure you get some ghosts to deal with in the next scene. The granularity of how the information is collected will ensure the feedback is dealt with appropriately, so if the player had enough mana to get by the ghosts without any issues, the system will understand that she's still ready for more trouble. The game currently only demonstrates a very crude and dysfunctional version of this idea.

Non - Combat
There is no combat but you're always racing against time, so in effect we've just switched the HP bar with a timer and altered how the game affects that resource. Not much of an achievement but at least our game doesn't involve knives, guns and bazookas. The spells were put to constantly change the way the game is played, it can be argued that some of the spells are 'evil' so I leave it to the player to decide against using them. Instead for this game, I wanted players to engage with the objects in the world, every object has an almost equally important role to play. The apples give speed, the grass gives you mana, the plants help keep the flies off and the cows can be kicked. I tried my best to have the objects work as players would expect. More importantly, I made sure the 'first order actions' (like cows hitting villagers) are wired properly for players to fool around and find new ways to interact with the world, discovering 'second order behaviors' (like forcing the Wanderer to teleport onto a portal).

Controls and Progression
I've worked on many games and shown them to my friends, but only recently have I been putting some thought into mobile and web games. Truth is I'm frustrated, I'm frustrated with players holding my iPod with the most lost expression asking 'What do you want me to do?'. These games assume that players somehow understand the basics, which is not true, so I wanted a way to always keep players in context. The game borrows heavily from Super Wagon Adventures, that's another reason for the scene structure we discussed above. Mouse clicks are used to move the king around, interact with objects and cast spells. The spell casting system in the last game had zero conveyance, so I decided this time to pause the game and wait for players to cast targeted spells. Every scene change and interaction comes with a small line of text that snatches all control and halts the game for a few short seconds. This interruption covers the screen in black and puts that line of bold text in the center, holding the players by their hands only when necessary. Although it's very slow and frustrating to be skipping these messages every few seconds, I can be sure the player is always aware of what's going on.

Please leave feedback and suggestions in the comments below, or drop a mail. Feedback would be great.