For the Challenge, Part II

Difficulty can, of course, be affected through programming as well as or alongside design.  Mario was programmed with hard limits on his jump height and the levels were designed accordingly.  There are some games where programming and design are complementary, and the result for the player is beautiful, lively physics.  The design and the programming tend to be set in stone for games – especially the latter – and that’s a good thing!  You need consistent ground rules in order for your game to be considered as, well, a “game.”  But when some elements are static, like the programming, other elements are often demanded to be dynamic, like the difficulty.

The most classic example of what I mean by “dynamic difficulty” can be found in the presence of multiple difficulty settings/modes in a game.  Easy, medium, hard – or if you’re feeling creative, “Kid Mode,” “Impossible,” “I am Death Incarnate!”, and so on.  They’ve existed for a long time, often as adaptive measures to player testing of a “baseline” difficulty.  The idea is that difficulty can be more flexible by tweaking certain parts of the game one way or the other depending on this setting.  This is more prevalent in games designed around complex stats and numbers – less so in platformers like Mario where player “stats” are as simple as failure upon taking one or two hits.

The problem is these tweaks are usually just that: tweaks.  Developers can easily program things like player/enemy stats to scale with the difficulty setting, but when they take this approach, every other aspect of the design is still operating under “normal mode rules.”  If you design a final dungeon with an arduous gauntlet of tough enemies, then double their stats as a hard mode condition, it’s unlikely that the difficulty of the same dungeon will remain well-balanced and tested.  Instead, it’ll simply take the player at least twice as long to complete a segment that was already balanced for a completely different pace – or else it’ll be unintentionally and incredibly difficult relative to the player character’s capabilities.  I mean, I’m convinced that oftentimes developers don’t even bother to test alternate difficulty modes at all, but that’s a little harder to examine and prove.

All joking aside, sometimes it’s painfully obvious that a game’s multiple difficulty levels are implemented more as an algorithm than a series of design changes or even tweaks.  Look at Fallout 3 or Skyrim.  When you slide the difficulty up or down, the only alteration you see is how much damage you deal and take.  Enemy configurations and dungeon puzzles remain exactly the same throughout.  The same as what they were when balanced for a “medium” difficulty (Fallout: New Vegas at least gets credit for introducing a “hardcore” mode that actually adds extra features and overhauls the mechanics).

I find this to be a slightly lazy solution to the difficulty “problem.”  What it really does at its ugly core is change the pace of a game, rather than the difficulty; you take twice as many swings at enemies, have to spend twice as long dodging them, and probably die a lot more.  Essentially, longer and more numerous iterations.  Like other fake difficulty, what this comes down to is a waste of the player’s time rather than a test of their skill or tactics.  I could seriously come up with some kind of fancy axiom… “the amount of enjoyment derived from a video game is inversely proportionate to the amount of control the player has over its pacing.”

A smarter way of adapting difficulty higher or lower might be changing enemy AI to not fall for the same tricks: a prime example would be the fighting game genre, where CPU opponents of higher levels will imitate correspondingly more complex tactics and reflexes.  Take level 9 CPUs in Smash Bros. games, which can destroy beginner players in much the same way a more experienced human opponent would: through better coordination of offense and defense.  Sure, you do often spend more time playing against a high-level AI than a low-level one, but the pacing is ultimately dictated by the skill of the player rather than the amount of bullets the enemy can absorb before going down.  If you practice a little more at Smash Bros., you’ll eventually be able to conquer an advanced AI (or human player) faster and faster as you adapt to the level of play.  If you practice a little more at Fallout 3, that irradiated mutant animal is going to sponge the same damage at the same rate no matter how much you wish it would die quickly.

But even stat changes can make for a more interesting challenge if balanced thoughtfully.  Look at Critical Mode in Kingdom Hearts II: Final Mix.  Sure, enemies are programmed to deal out more damage, and the player’s max HP and MP gains are cut in half, but at the same time the player does more base damage and starts out with more abilities and ability points.  This results in a real dynamic difficulty where ultimately the player can take their time with cautious play to offset increased risk, but also apply skill and knowledge of the mechanics to actually complete the game faster.  In other words, the game rewards a higher difficulty with a potentially more streamlined pace.

But you’d better believe it took more thought to implement Critical Mode in Final Mix than “Very Hard” mode in Fallout 3.  There simply isn’t a shortcut to well-designed difficulty.  It bears repeating: simply making gameplay take longer does not make it more (legitimately) difficulty.  The whole experience has to be fun for the player in exchange for (and in proportion to) the extra effort.  Even if that requires a little extra effort.

By the Numbers

Let’s talk about the black sheep of the Kingdom Hearts franchise: Chain of Memories.  This one still sticks out like a sore thumb next to the other installments: it had sprite-based graphics, strange (but oddly charming) writing, and especially card combat.  The card combat part has proven quite polarizing among the fanbase, which is kind of understandable.  The first game, which was the only other one out when CoM was released, had quickly and easily accessible gameplay that didn’t run roughshod too much all over the wonderful aesthetics and sweeping crossover ambitions.

CoM is almost a direct antithesis, recycling the same worlds wholesale and slapping on excuse plots for those (leaving the freshest material for the non-Disney segments) while its gameplay mechanics demanded the player’s full attention.  Every action you take is determined by playing a card, and almost every card exists as part of a numbers-oriented system of “greater than, less than.”  If you play a 4, an enemy can play a 6 and “break” your card, for example (“0” cards have the distinction of breaking any card but also being broken by any card depending on when they’re played).  Your ability from the first game to Guard attacks directly is gone, and Dodge Roll is much less useful, so card breaks are an essential mechanic to self-defense.  For added fun, you can combine up to three cards (sacrificing the first card chosen for the rest of the battle) in a “sleight” for a special combo attack – as well as higher numeric strength, given that the three card values are added together for the attack.

To the player who enjoys more “pure” action games this probably sounds like torture, but it actually works very well, and here’s why in a word: visibility.  Most JRPGs, for all their reliance on numbers, tend to make those numbers highly opaque.  If you want an example, just look at the EV mechanic in Pokemon or the way character stats affect their offense/defense in Final Fantasy.  A complicated series of behind-the-scenes equations drives most of the character progression and leaves the player with just a vague sense that a higher STR stat means harder hits.

With CoM, all cards are on the table.  So to speak.  If you play a 7 and your enemy plays a 9, your card gets cancelled and you immediately know why.  If you want to access a certain sleight, the game tells you exactly which cards of what numeric values to combine to use it.  If you want to engineer your entire deck around a set of boss-annihilating DPS sleights, or only load cards with higher numbers than what you know enemies can play, that’s entirely doable.  To wit, the mechanics are conveyed well enough that you can master them early on by using strategy and basic arithmetic, rather than simply follow them with varying degrees of awareness throughout the game.

And the rules of the game are consistent for enemies too.  Mobs have access only to pre-determined card values, thankfully never 0s, that are also shown on screen – such that you can learn them after a couple battles and revise your deck to counter if you want.  Bosses have the same options as you in terms of playing 0s and sleights, but also the same weaknesses.  Their 0s and sleights can be broken, they can be attacked while reloading, and they still have finite cards in their deck to play before doing so.  If you take the time to optimize your deck, you can trivialize battles.  And no enemy in the game is exempt from all these rules.  It’s all-in-all more fun to learn the combat system here because it’s always expressed clearly.

A “what you see is what you get” approach serves RPGs better than you’d know from examining a lot of entries in the genre, since the defining trait of their gameplay is that it’s based on strategy rather than timing or execution.  How they tend to play out, though, is that having stats is the key to victory, not employing them.  The reason many of them can’t have clearer expressions of how their mechanics work is that it would reveal how little the battles are influenced by the player compared to Random Number Generators and the raw stats themselves.  The near-automatic battles in Final Fantasy XIII get a lot of flak for their nature and exclusion of the player from the whole process, but it’s really just more honest about the player’s level of intellectual involvement than other installments in the series.  These strange and large numbers floating above the characters’ heads don’t have to get in the way of the graphics and the narrative that way.  And that’s fine, if that’s the kind of game you want to make.

Just, if you want to make one where you claim player involvement is supposed to be central, then make sure you communicate the rules of your game fairly enough that they can be involved and have it matter.