Wednesday, July 13, 2011

Postmortem: 6 things I did wrong in my game, Jellia's Friends

Type: Online Flash game

Price: Free 

Target age group: unspecified

Publisher: Developed by Batty Media for JenniferAnn's Group (

Genre: tile-based adventure

The Bottom Line: The dangers of teen dating violence are real, but Jellia's Friends gives a fanciful take on the problem in a clunky and unattractive old-school style adventure game.

Since 2008, the organization Jennifer Ann's Group has sponsored an annual Flash game development contest in an effort to raise awareness about the dangers of teen dating violence. In 2010, intrigued by the challenge of making an educational game and happy to do something to support a cause like preventing dating violence among teens, I submitted an entry called Jellia's Friends. Now, over a year later, I wanted to revisit this project, not because it really deserves attention as an especially high quality educational game (it's not), but rather because I want to reflect on what I learned from my first attempt at making an educational game.

My goals in creating the game were two-fold.  First and foremost, I wanted to improve my Flash development skills and I wanted to see a game project through from conception to completion.  I had purchased Flash software years ago, and greatly enjoyed its potential and flexibility, but I never learned much about programming with ActionScript.  This was an opportunity for me to stretch myself in new ways and get considerably more game development experience under my belt.  Secondly, I had been reading Ian Bogost's Persuasive Games at the time, and I had been thinking a lot about how the mechanics of gameplay could and should relate to the educational message of the game. I wanted to try and develop a game in which the mechanics and goals of the game itself mirrored the message I was trying to send.  I was hoping that by submitting my game in a competition like this one, I would get some exposure and feedback so that I might make better educational designs in the future.

The Concept

A less-than-ideal relationship.
When I first tried to develop a design for this game, I decided very early on in the process that I did not want the game to play like an interactive storybook or a point-and-click adventure.  In my mind, these two genres had some glaring weaknesses, particularly when used specifically for educational purposes.  First, they typically had little to no re-play value.  Once you played out the story from start to finish, there was no further challenge and no incentive to play again.  Second, it was difficult to make such games feel more than superficially interactive.  That is, they often took the feel of a didactic afterschool special than the feel of the kind of "game" that someone would actually want to play for fun.  And while this is by no means a criticism of the genre as a whole, I did not personally have any great ideas for breaking the point-and-click mold, so it seemed best to stay away from these genres altogether.

I began looking to classic games for inspiration.  After scrapping several preliminary sketches, I decided a Zelda-style, top-down adventure game, which I thought would lend itself quite well to my ideas about how a person in the real world might go about extricating themselves from a potentially dangerous relationship.  The main features that I wanted to include in the game were:
1. The story would center around the character who was being subjected to abuse, so that players would see how a relationship might devolve from happy to destructive.  The abused character would also be the character that players would control, to make the point that people in abusive relationships do have the power to help themselves.
2. Just as it would not be risky in real life for someone being abused to confront their own abuser alone, the player would be discouraged from directly confrontation the abuser.
3. The player would gain in-game powers from their friend as they played the game, just as gathering together real-life friends would empower someone against his or her abuser.  The main point would be to  illustrate through the mechanics of the game how an abused person might take control of the situation without endangering him- or herself.

The premise of the game and the gameplay itself actually ended up being relatively simple.  The opening scenes introduce the player to the main character, Princess Jellia of the Kingdom of Sentient Jellybeans.  As usually happens in fairy tales, the Princess is highly sought after, and the game opens with Jellia and her father scrutinizing the latest round of suitors.  But Jellia isn't just a generic princess, she's also a somewhat willful teenager, and like many teenagers, she is simply not impressed by the suitors her father endorses. Instead falls for the charming prince with the dashing smile, Prince Licorice.  As she soon discovers, however, the Prince has a mean streak. When he ruthlessly kicks her beloved pet Galen, Jellia finally realizes just how unsavory the Prince really is.  Not entirely sure of how to extricate herself from her situation with the Prince, she decides to seek out the help of her friends.

Players learn about the warning signs of an abusive
relationship with each friend they find in the forest.
This is where the introduction ends and the game begins.  In order to talk to her friends, Jellia has to sneak out of the castle and seek them out in the woods.  She must avoid the controlling Prince at all costs while also evading the forest's wild animals, who just happen to enjoy snacking on jelly people.  The player controls Jellia with the arrow keys, and Jellia has to avoid the hazards of the forest while she explores, trying to find her friends.  With each friend she meets, she learns about the warning signs of an abusive relationship and slowly realizes just how dangerous her relationship with the Prince actually is.  Some friends also have special items that will help Jellia navigate the forest.  However, Jellia's moves remain purely defensive.  She can only try to avoid enemies, but does not attack them herself.  If Jellia manages to gather all 5 of her friends, they persuade her to return to the castle and to tell a respected adult, her father, about her problems with the Prince.  The player then needs to navigate Jellia back to the castle, where the Prince confronts her, but her friends escort Jellia to safety, and with the help of her friends and her father, Jellia resolves to make a long-term Safety Plan to avoid the Prince.

All in all, it was a very ambitious project, and after several months of painstaking labor and many sleepless nights, I felt very good about the finished product.  Now, over a year since the game was completed, its much easier for me to see the game for what it was and to speak about the game with slightly more objectivity.  So today, I take a critical look at what went right and what went wrong in Jellia's Friends.

1. The game was ugly
Jelly friends or jelly zombies? With this art, who can tell?
I did my best to get past the glaring deficiencies in my artistic abilities by employing the most simplistic of characters, jellybean people.  But while I think this resulted in a passable splash screen, it was not nearly enough to redeem the game as a whole.  At the time, the graphics were a deficiency I recognized, but I didn't consider them important enough to correct.  In retrospect, I absolutely should have spent more time with this.  The 8-bit days are long since gone, and while this doesn't mean that every game needs 3D-rendering, textures, and particle effects, it does mean that player expect a certain amount of polish and visual consistency in the games they play.  If the graphics are so poor that they break the immersive experience completely, then the game has failed, no matter what else it might have going for it.  This was a difficult lesson for me to learn, but ultimately, I think it has been one of the most valuable take-away points from developing this game.

2. I didn't do my marketing homework
In Jellia's Friends, there's a huge mismatch between the age-group targeted by the graphical design and the age-group targeted by the dialog and content.  The cutesy, simplistic jellybean characters really don't work if the game's primary audience is intended to be teenagers.  After all, when you were a teenager, would you want to be caught playing an educational game that looks like it was drawn for 5 years olds?  No, probably not.  On the other hand, the retro-styling of the dialog boxes and even of the basic gameplay, clearly appeals more to gamers my own age than to an audience of teens who never experienced the 8-bit era.

Both of these mistakes could have been avoided, I suspect, if I'd done a little more research into the aesthetics and visual styles of game popular among my target demographic, roughly 11-16. I knew that I really did not want realistic, or even human characters in the game.  If the characters were too realistic or too blatantly stereotypical, I thought that could easily turn an angst-y teen player off.  But I also didn't have a clue how to translate that idea into something that would actually appeal to teens. So while the dialog I believed ended up pitched at the right level, it simply didn't sit well with any of the other elements of the game.  By not creating a clear set of visual cues targeted to any one particular audience, I left Jellia's Friends looking like an incomplete mishmash of random parts instead of creating a visually coherent game world.

An uninspired and boring forest.
3. Map design was bland, uninteresting, and downright poor
I did think a great deal about the numerous maps I made for the game.  I tried to make it non-obvious where the player should go next so that he/she would be compelled to "explore" the world. I also tried to create constraints and obstacles in the landscape that were most appropriately suited to the particular enemies on that screen--sometimes to make a screen particularly difficult (like the plant-laden screen with the little grey metroid-like creatures) or sometimes to make it easy (like the screen pictured to the left).  But despite my good intentions, I certainly felt like many of the screens were not that well thought-out.  Some screens could be brutal if players didn't have the time-stopping power-ups like the flute or the smokescreen. Others just had a lot of open space that served very little purpose at all.  Possibly, I made the screens bigger than they needed to be for what I had to put on them.  I was so taken with the possibilities open to me once I learned how to use arrays to store maps for my game, that I wanted to make my maps as big as possible.  But in the end, I don't think the gameplay fully warranted such expansive maps.

4. Building the game engine myself made it buggy
Developing the game was a learning process for me, and an invaluable one at that.  It only made sense to me when I was starting this project to build every part of it from scratch.  This meant that even though my ActionScript skills were minimal, I decided to build the entire tile-based game engine myself.  While I relied very heavily on the excellent tutorials by Tony Pa, it still meant that well over half of the development time spent on this went into re-inventing the wheel and making a game engine that has probably been made a thousand times over by much more skilled programmers than I.  If I were making a game for commercial release, not only would this have been a pretty foolish waste of time, but the residual bugs that I had neither the time nor the skills to correct would have quickly spelled the game's death.

I wouldn't say that I'd necessarily do things differently if I had to do it again. Some of the bugs, like the fact that some of the more erratically-moving enemies could occasionally get stuck on bits of the landscape, I knew about and decided were actually good bugs that I wanted players to be able to exploit to their advantage.  And of course, building this game from the ground up really let me cut my teeth on some down and dirty programming, making it was a very useful learning experience for me. Moreover, I still get a ridiculous amount of satisfaction out of knowing that every line of code in the program was there because I deliberately put it there.  But the other lesson I learned was to recognize and respect my limitations.  If I had used an engine that had already been built, I would have had a lot more time to do things like dream up better level designs, improve the graphics, and think about how to make the game more fun overall.

5. There should have been a way to skip the intro scene
I knew better than to force players who died and restarted the game to have to watch the lengthy opening sequence every time they played the game.  Had I not done this, I can't imagine anyone would have played through more than once.  But as a game with a message, I wanted to force players to watch the opening scene at least once, which meant no skipping it the first time you start up the game.  While this decision did make some sense given the context, it was nevertheless a decision I regretted almost immediately after submitting the game.  If someone actually wanted to play the game again, I shouldn't be discouraging them by forcing them to watch a poorly animated and lengthy opening scene again and again.

6. The GUI was not obvious enough
Although the game did have meters for gauging how much health the player had, and had many more jelly friends she/he needed to find, the display at the bottom of the screen was just a little too unobtrusive, and players probably found it easy to overlook.  There should also have been some icons that appeared when players acquire certain items, like the flute or smokebomb.

It is harder for me to speak objectively about the things that Jellia's Friends did well. I know what I thought I was putting into the game and what I think managed to survive even after it was butchered by my programming and graphic skills.  But a true postmortem would ideally have the perspective of player opinions to take into consideration, and that is one bit of feedback that I simply never received. My dogged belief in any virtues of the game must therefore be taken with a grain of salt.  Still, I do think that the game hit on some important concepts in educational game design that may be worth repeating in a more sophisticated and polished product.

Scary spiders are scary!!
Even though it was designed to inform, Jellia's Friends was designed to be a game
As stated above, I did not want Jellia's Friends to be little more than an interactive storybook, to be played once and never played again because nothing would change from playthrough to playthrough.  I therefore designed the game to be challenging and to have certain semi-randomized elements that would differ on each playthrough.
There are at least 5 different types of enemy with 5 different AI routines, and while they're not all genius, it does at least make the player have to learn those patterns and think about how to get past them.  Jellia's friends are also placed on the map according to 4 different possible arrangements selected randomly at the start of each game.  This means that power-ups like the extra heart, flute, and smokebombs are not in the same place every time.  This does considerably vary the level of challenge of each playthrough. And finally, the appearance of the most difficult baddie in the game, Prince Licorice, is also randomized, meaning that even the same arrangements will differ in difficulty depending on when and where the Prince shows up.

The educational message is reinforced by the gameplay
The primary message of Jellia's Friends is that a person in an abusive relationship can help him/herself escape it by reaching out to a network of supportive friends and responsible adults.  This message is largely supported by the goals of the primary gameplay.  Jellia does not have any offensive powers, and thus cannot engage in combat with either forest creatures or her abuser.  Her strength lies exclusively in her social network.  Not only do players only succeed by collecting Jellia's friends, but friends also possess unique traits that can help the player succeed as he/she progresses through the woods.  And successful escape from the abusive Prince can only be achieved through gathering all of the Princess's friends.

The game also discourages direct confrontation with the abuser, which would be quite risky in real life.  This is done by making confrontation with the abusive Prince also highly risky in the game.  If the player is not able to avoid the Prince, one unit of health is taken away, a dialog cut-scene is played through which the dangerousness of such a situation is made evident, and the player is relocated to the starting point.  Also, slightly more tense and sinister music is played when the Prince appears, to give the player a sense of heightened anxiety. All of this was done with the intention of invoked in the player the same fears and anxieties that he/she would and should feel if they were trying to escape from an abusive partner by him or herself.

Dialog attempted to imitate how teens might speak to each other
One of the most difficult things about writing any game dialog is trying to avoid sounding overly trite or cliched.  And one of the most difficult things about trying to write educational dialog is trying to avoid sounding either pompously didactic, or simplistic and patronizing.  It was quite a challenge to try and convey a lot of information about teen dating violence into a short game, and the dialog was definitely not as concise and snappy as might have been ideal.  Still, while at times the length of the exchange may have tried the player's patience, ultimately I think the dialog succeeded as being informative without being condescending or overly stilted.  This particular structure for conveying the bulk of the educational information in the game would, I hope, have made this material seem more convincing to a teenagers, who might be more likely to trust the advise of his or her friends than to listen to a lecture delivered by an adult.

Final Thoughts
As a finished product, Jellia's Friends was weak in countless ways.  It tried to do too many things, and as a consequence, it did not do any one of them particularly well.  However, I like to think that if you consider Jellia's Friends as a prototype for a game concept, then perhaps many of its flaws are more forgivable.  I think with better polish (cleaned up graphics, more interesting level design, more balanced difficulty, and more map screens to explore), the ideas that motivated the creation of the game might shine through a little better.

That being said, I learned a heck of a lot about programming, design, and project management in creating this game.  For all of its many flaws, I am still proud of having taken the game from just a few rough ideas in my head and turning it into a playable educational game.  Is it a good game?  No.  Was it a damn ambitious project for a non-programmer to pull together, solo, in a few months using outdated software and some online tutorials? Yes.  And for that, even if Jellia's Friends did not ultimately turn out to be the fun and revolutionary game that I wanted it to be, I still have no regrets about making it.

A picture-perfect ending to a less than picturesque game.

No comments:

Post a Comment