Monday, November 26, 2007

Game Difficulty: Front loading the pain

In recent months, my work focus has been tuning player weapons and abilities, enemy AI, and game systems. I spend a lot of my time thinking about the "feel" of how each of these systems interact and how they affect the play experience. As a developer, systems are tough to balance because everyone has a different play style and preferences. After a fair amount of internal and external playtesting, two conflicting opinions come up regularly - some want more challenge while others feel like they aren't powerful enough. While these two opinions seem at odds with each other, they reference different areas to improve. I address the first issue by crunching numbers on permutations of different combat outcomes and poking around for where the numbers break down. To address the second, I look for ways to improve presentation, because the perception of power is equally important to balance when crafting player weapons and abilities. Because it relates to my current work, I feel extra sensitive to difficulty curves in games I play these days. Three memorable examples come to mind.

First off is Blue Dragon, the RPG brainchild of former Final Fantasy series director Hironobu Sakaguchi and developer Artoon. RPGs are an interesting difficulty case study because by and large, they give players the ability to determine their personal difficulty curve. Your characters become stronger gradually via combat and equipment upgrades, and you can tweak the difficulty down by spending more time fighting enemies to collect experience and money. Blue Dragon presented an interesting conundrum because I found myself dying a total of 5 times in the first 2 hours of the game. The first time this happened was when I stumbled upon the first boss of the game. While I should have been wary once I had received a 'checkpoint' notification, I had gotten lost when wandering around the map. A short cutscene ensued and I found myself in the midst of combat. The first boss requires a relatively simple but specific strategy to defeat. Unfortunately, that strategy rendered one of my three characters almost totally useless and forced another one to heal every turn. If the character that could effectively damage to the enemy was knocked out, I fell into a vicious cycle of reviving that character, then healing them, only to have a different character die. After about 3 tries I had the strategy down and won the fight.

Shortly after this encounter, I walked into the nearby wilderness. Having just defeated a powerful enemy, I felt confident in my abilities. Unfortunately for me, this victory was a ruse to boost my ego. In a battle with one of the 'normal' enemies, I was on the receiving end of a spell that did enough damage to kill my entire party in one round even though they were all nearly at full health. Lesson learned - don't mess with that monster. Then I found a path that looked like it might be the next area I should explore. That path was blocked by a sleeping bear. Once again I decided to tempt fate and bypass this bear. He proceeded to annihilate my party with only a few swings.

After this death, I became a bit dismayed. I didn't know where to go and it seemed like every time I tried exploring somewhere new, death was the only one there to greet me. The solution seemed simple - kill lots of easy, well understood enemies, right? As a general rule, I tend to over train my characters when playing RPGs anyway. The problem lies in the fact that the location of these enemies relative to early parts of the game clearly didn't indicate to me what was dangerous and how strong I needed to be. In addition, the game punished me for one of my favorite aspects of playing RPGs - exploring the world. Also, the game hasn't taught me much about how I can fare better in those unexpected circumstances, other than grinding enemies over and over. All of this added up to a poor early impression. I was also left to wonder if the developers spent much time observing people play through the early game or thought about how powerful a player could be at the time they encountered these early enemies.

The second example I'd like to call out is Stuntman:Ignition for the Xbox 360. As a caveat, I must admit that I have never played Stuntman games before so I had no idea what to expect. The result was ultimately an exercise in frustration. I failed the first level within 30 seconds of playing the game. I proceeded to fall in holes, instant death lava, and missed the timing on scripted sequences about 10-15 times before I completed that first level. One failure lead to another. In addition, there was a timer which served as another failure mechanism. This creates an additional pressure that pushes you forward while you attempt to avoid the numerous hazards. Each of the hazards can instantly set you back to the start of the level. In addition, the stunt sequences are fixed, but you encounter several in rapid succession, and you have to be familiar with the geography of the level and timing of the events to complete them. You also have to get within close range of each objective to complete it, with very little wiggle room and a ticking clock staring at you.

On average, this game probably reinforces failure more regularly than any recent game in memory. In addition, once you succeed, you rarely get a perfect score, and several of the achievements are tied to perfecting each level. After playing the game for a while longer, I made a couple of adjustments to my strategy, one of them being to set the difficulty to easy. This mode was still quite challenging and as a long time gamer, I feel aversion to feeling forced to play on a mode called "easy" to make progress. Gears of Wars' naming convention of 'Casual' difficulty has a much nicer ring to it, but I digress. While switching difficulty temporarily stopped the frustration, the challenge naturally ramps up over time and the levels become longer, and at one point a 'shelf moment' (not my term, but a good one for declaring when the game has an event that makes a player turn it off and leave it to sit on a shelf for the remainder of its life) was triggered. I stopped playing without any intention to return.

For my last example, I'd like to offer a counterpoint to these games. Guitar Hero 2 has one of the smoothest difficulty ramps I have experienced in any game. Easy mode felt like a difficulty that was fairly unchallenging for me until I shifted my focus to perfect each song before moving onto Normal. The tail end of easy mode flows nicely into the early parts of Normal. Attempting to perfect each song on Normal segues perfectly into Hard, and Hard difficulty introduces another chord which makes you feel like you need to expand your toolset even more. Finally, expert feels as it should, an exercise meant only for those that seriously practice the game in order to develop the muscle memory needed to reach the peak of their skills. Along the difficulty progression, Guitar Hero 2 made me feel as if I had genuinely improved my ability, and that I could consistently perform at that ability, even after putting the game away for some time.

My appreciation for Guitar Hero led me to the GDC talk Harmonix gave at GDC last year. It was enlightening to hear the evolution the company went through in making abstract games such as Amplitude as well as mass market titles such as Karaoke Revolution. To them, each game felt like a success in some ways and a failure in others, until they arrived at Guitar Hero. The lessons they shared about making games deep for core gamers and accessible to casual ones were well stated and resonated in my personal experience playing Guitar Hero. Harmonix managed to create a game that hooks you early, teases you with challenges that are ever so slightly out of your reach, and generously rewards you along the way with digital gifts such as new songs as well as an intangible sense of satisfaction from reaching the next plateau in your personal ability. They also created a game that truly follows the "simple to learn, difficult to master" mantra. It's refreshing to have games that balance their challenge so well when there are other well produced but poorly paced games that also cross my path. As I stare at the combat numbers and playtest results at work, I think of the Harmonix design principles and aim for the goal of achieving the finely tuned balance present in their latest creations.