By Daniel Slawson
Tuesday, May 27, 2014
Mass Effect 3’s difficulty system is an unfortunate cautionary tale about how not to handle difficulty settings. Though reasonably well balanced for its default level (“Normal” in singleplayer and “Bronze” in multiplayer), the design team decided to increase the hit points of enemies to an unrealistically punitive level for higher difficulties, which I found disappointing and frustrating. This struck me as a particularly amateur decision in an otherwise strong game, and I wanted to define why.
The problem in a nutshell is twofold: first, 90% of the weapons you enjoyed using in earlier difficulties suddenly become pea shooters. The other part of the problem is really just a natural extension of the first: you can unload a full submachine gun or assault rifle clip into an enemy at point blank range (which can take up to 3+ seconds, a long time in a shooter), and anything above the weakest enemy will still be at half health.
It’s worth noting that inflating enemy hit points is something many games do, the examples are myriad. Mass Effect 3 makes an interesting case study because this turns out to be extra jarring in a real-time shooter (when it isn’t busy being a roleplaying game). The fact that I’ve logged over 335 total hours playing it doesn’t hurt either.
To get a better sense of what the devs were doing, here are the actual mechanics and numbers involved. This data was gleaned from two very informative threads on the Bioware forums, with more detail than I included here. Simply put, there are things that change as the difficulty level increases (whether singleplayer or multiplayer), and things that don’t change.
The core problem here is that weapon damage doesn’t scale to the increasing hit points of enemies. The most obvious effect of this is the feeling of frustration and unfairness when enemies your brain is sure you should have killed twice-over are still standing (and doing significantly more damage than you when they fire back).
Some will make the argument that this is working as intended, because it increases the challenge. It demands more of players by rewarding precision (players to need to utilize the headshot damage multiplier as much as possible) and requiring careful optimization of your character’s build. This argument is absolutely, 100% sound.
…Except for just one thing. It overlooks a mission-critical distinction: the subtle difference between challenge and frustration. Put another way, there’s something else fundamental that doesn’t change with difficulty level: players always like to be rewarded for their actions. If I get the drop on an enemy, I expect to have a meaningful advantage, regardless of my weapon. If I don’t enjoy one, the game feels unrealistic and unfair.
Players will always be rewarded for precision and careful build optimization, regardless of difficulty level; but having to “grind” through an enemy’s health isn’t fun, and fun is the ingredient that matters. Challenge is difficulty mixed with fun (or the possibility of fun); frustration is difficulty with just negative emotions.
This is blatantly apparent when fighting “boss” enemies, who have about ten times the health of average troops. These enemies act as durable force multipliers, and they’re mostly very predictable. This means that most of the time, your team is going to be taking these out last, resulting in a tedious, slow “dance” as you exploit their simple pattern. A full clip may not even deplete even one “bar” of their health (of which there are ten), so firing bolts of hot plasma actually feels like watching paint dry. With certain characters, taking out a boss solo can actually take 2 full minutes. Hard to believe someone played this and thought it worked just fine.
The next consequence is unfortunate on a different front: higher difficulties cause the majority of available weapons to no longer be viable.
Above normal difficulty, only a few weapons have the sheer stopping power to make an appreciable dent in enemy HP; typically slower-firing weapons. This forces players toward a limited selection of weapons, which is problematic for a few reasons: firstly, because both of those weapon types tend to be heavier than many classes can comfortably use (because heavier weapons slow down the recharge rate of your powers). They also strongly encourage (if not outright force) a specific playstyle to be effective. Sniper rifles, for example, have a cumbersome, overly high magnification zoom, and the ones powerful enough to warrant using are usually single-shot with long reloads (like the classic Mantis rifle). This can be a fun style for some classes, but ill-suited to most — and the long reload animation definitely gets old fast, especially if you aren’t a crack shot.
Shotguns can be a better alternative for the classes that can survive at very close ranges, but they tend to be on the heavy side, and only a few of the more “elite” unlockable ones really deal the damage you expect on higher difficulties.
Despite the appearance of a wide selection, players are incentivized to use the few specific weapons that are optimal for their class.
Another side effect is a major degradation of the game’s sense of realism. All game worlds, no matter how fanciful, have some measure of internal logic: a yardstick for your suspension of disbelief. Mass Effect’s science fiction setting features deadly high tech weaponry: a weapon’s design, recoil, sound effects, and firepower all enhance the immersive experience by grounding the player in the world. It feels fake when something that looks and feels so powerful does a pathetic amount of damage; it draws the player out of the immersive, moment-to-moment experience with a bit of incredulity (or worse, frustration).
Okay, here’s where we switch gears. I’ve gone to great lengths to identify the problems, now the task is to locate solutions.
The challenge for the designers of ME3 was to offer increased difficulty levels without compromising fun. But if increasing hit points wasn’t the answer, what is?
There’s no universally appropriate answer, which is part of the charm of game design. It’s possible to come to different conclusions depending on your design sensibilities, existing constraints, and what kind of experience you want: each game calls for solutions that fit its context. While some solutions will be objectively better than others, there’s usually quite a bit of room for variance.
The important thing is to identify your variables, and then carefully adjust them to maximize enjoyable gameplay. The Mass Effect 3 designers obviously recognized and altered many variables (damage, health, speed, AI characteristics, number of enemies, etc.), they just altered them (very) suboptimally.
So what’s a good framework for doing all this? Here’s my suggestion:
In my experience, each variable in a game has a “sweet spot”, a point at which altering it too much actually introduces problems (whether frustration, imbalance, or something else) that can start to detract more than add. When that happens, a sensible designer should consider backing off and adjusting another variable instead.
Put concisely: if anything you do causes the game to be less fun, stop and try something else.
Using that framework (and without further ado), here’s my own personal sense of where the sweet spot is:
Keep everything the same as the default difficulty level (normal/bronze), except for the following:
The reasoning is straightforward: the standard difficulty level (that is, normal or bronze in multiplayer) is really the ideal place to start from a fun perspective: a “known good state” where the game offers a good experience. Weapons and powers kill enemies at the rate you’d intuitively expect them to, and (for most classes) enemy damage is neither pathetic nor too overwhelming. Players enjoy the freedom of choosing any weapon (as they’re all viable), and still enjoy the rewards of careful valuation when selecting the right one for their character.
As the difficulty increases, upping the number of enemies makes for a more frenetic experience: while each individual enemy isn’t overpowered and can be dealt with quickly, larger groups will cause the AI to advance more readily, increasing the pressure for teams trying to hold position. Cover is also increasingly vital: taking damage from too many enemies at once will still be very dangerous. The action is tense, and the level feels less scripted and more like a battlefield.
This being said, pushing any variable too far clearly has the potential to unhinge the game: increasing enemy damage too greatly, for instance, will cause very similar symptoms to increasing HP (frustration, feeling the game is too punitive). In my solution, increasing the number of enemies too much could make area-of-effect skills way too powerful, making certain classes and builds too dominant compared to others. Perhaps too many enemies in the level will be too demanding on hardware, or won’t leave anywhere for players to run to, or players might die too quickly without cover. It’s all about brainstorming, playtesting, and being willing to find the sweetest sweet spot.
In the case of Mass Effect 3, I can’t help thinking that the higher difficulties were designed with the “feature checklist” paradigm rather than the “sweet spot” paradigm, perhaps because of budget or time constraints. Only minor, selective balance adjustments were made after release, so the team clearly thought the fundamental mechanics and balance were good enough (or worse, were good). Whatever the case, it strikes me as very wasteful: if you’ve run the marathon of making the majority of the game, it’s a shame to fall down in the last mile, especially when fans were already skittish about the inclusion of multiplayer (and still others were excited by its potential). A little care and attention to detail could have gone such a long way here.
It’s an understatement to say the game did a lot right, but that doesn’t mean it did everything right. Make sure to take a close look at the mechanics of your game, and don’t let the good be the enemy of the excellent.