Monday, October 24, 2011

Reaganomics


Having received my review copy of Warhammer 40,000: Space Marine several days before its release date, I expected to have a harder time connecting to full multiplayer matches. As it turns out, there were ample opponents online, mostly located, according to the in-game leaderboard, in France and Germany. At the time, I assumed Space Marine must have released early in Europe: nearly all of these players had reached double-digit levels in Space Marine's XP-unlock system. Not only did they zip through the maps to secure choke points with the speed and familiar ease of old pros, but they were also kitted out with powerful weapons and perks. It was clear they'd had time to practice. 

I got slaughtered, of course.

(Turns out the game's official European release date was later than the North American date; I'm told some European retailers regularly break release date to sell extra copies. Knew those guys couldn't all be reviewers.)

Despite the relatively limited modes—team deathmatch and king of the hill varieties—I'd been looking forward to jumping into Space Marine's multiplayer. That's a rarity for me; I generally dislike competitive multiplayer shooting on the console, since I lack the appropriate twitchiness to properly acquit myself. But, as anyone who follows my lunatic Twitter stream knows, my 40K fandom knows few bounds. Besides, except for a few desultory forays into Call of Duty: Black Ops and Battlefield: Bad Company 2 matches, I hadn't had much experience with the now-standard persistent unlock model. 

Of course, I was skeptical; intuitively, this kind of system, which awards players more powerful gear, stat increases, and/or customization bonuses the more XP is accrued, seemed unbalanced at best, obnoxious at worst. Still, I was curious to see if the RPG-lite hook of XP and weapon unlocks might engage me when applied to a fictional world I was already into.


 The verdict? Sadly, no.

I can see why the persistent unlock model is so attractive to both designers and players. From a business perspective, it encourages a "long tail"—by lengthening the time players are likely to spend in multiplayer modes, you're keeping the product off used game store shelves at least for a little while longer. In the process you build word-of-mouth, since the player's online friends can see what you're playing and will want to join in the fun. The model also allows designers to mete out a steady drip-feed of "progress" to the player, creating the illusion of achievement, if not an actual sense of accomplishment. Leveling up means the player doesn't have to start fresh every time he boots up the game, promoting a sense of connection between the player and the multiplayer character. The playing field is not level every match; star players have paid their dues, and earned their perks. That sense of superiority is, I'd imagine, pretty important for people who invest a lot of time into competitive multiplayer. And the allure of new gear—well, that's been borne out by many a game.

But the persistent unlock model also comes with its share of downsides. The most convincing argument against this type of system is that bolting on a leveling up mechanic obscures the notion of skill. In a recent blog post, 100 Rogues designer Keith Burgun lays out a compelling case:

So, I hope that it's clear how in a skill-based game, having leveling-up is really out of place, and thus it makes your game much harder to balance. Assuming your leveling-up system is at all non-linear (which it really should be), it becomes impossible to track where your player will be in terms of his core-stats, and therefore impossible to balance the game appropriately. Combine with this the fact that most digital games are probably too inherently complex to begin with and we have a recipe for the current status of digital games: perpetual imbalance. 

Our games should not be fighting against themselves like this. When designing a game, we must decide what it's about, and stick to it. Is it a game about timing, accuracy, and twitch skill? If so, how does a leveling up system compliment that? I am not saying, by the way, that an action game cannot introduce new mechanisms to the player as he goes, in the form of new enemies, new weapons, new environments. These are all great ways to increase the difficulty that force the player to increase his own skill. 

There are a few interesting points here. The first is embedded in the first sentence and reinforced in the second paragraph: what kind of game is this, really? I'd argue that multiplayer shooters are primarily about skill. Success in a deathmatch-style game is mostly dependent on, as Keith says, timing, accuracy, and twitch skill. Knowing the maps helps a great deal, as does familiarity with weapon styles and use of bonuses or items. But without those qualities, upgrades will only take you so far. You need to have skill to win consistently.

With an unlock system in place, though, the game becomes more about addressing imbalances (i.e., exploiting your own upgrades or overcoming others' upgrades with your own skill) than about demonstrating mastery. The playing field is uneven. The player's objective thus becomes gaming the system to balance it in his favor—not about whatever he's actually supposed to do in the match.


For unskilled players, an unlock system creates the illusion of parity. By simply investing enough time into multiplayer—most unlock systems will award you at least a little XP for each round you play—you can advance up the unlock ladder, making up for your lack of skill with powerful weapons and perks. In theory, this will help you be more competitive. 

In practice, it's more like our tax code: the rich get richer. In a graduated income tax system, people are taxed at a higher rate the more income they earn. This structure is intended to ensure that those who have benefited most from society contribute more back to it. As we're all aware, it doesn't work like that in reality. Decades of supply-side theory, and I'm sure more than a little bit of lobbying, have convinced decision-makers that 99% of us can rely on the largesse of the wealthy for our own financial security. Because they create jobs, see.

Multiplayer unlock systems don't operate on the same principle—there's no "trickling down" of skill—but they have a similar result. Skilled players will advance up the XP ladder at a much faster rate, compounding their advantage over others. Unbalanced matchmaking ensures you'll be placed in lobbies with players at wildly ranging levels, pitting your level 3 character against at least a few level 30s. (I've no idea how matchmaking works on the back end, but I assume grouping players of similar skill levels is a challenge.)

And this is the irony of the unlock model: it de-incentivizes its own incentives. Lower-level players are often killed within seconds of spawning. Even familiarizing themselves with the maps becomes difficult, since it's tough to explore with a bunch of elites hunting you. It becomes enormously tedious to advance up the ladder. After several rounds with a pathetic kill/death ratio, earning that level 10 Lascannon feels a lot less enticing. Space Marine combats this obstacle by allowing you to copy the loadout of the last opponent who killed you for one life; but even this creative fix is really just a Band-Aid. Being able to preview a high-level weapon or perk doesn't erase the tedium of grinding levels to earn it.

The last time I invested a genuine effort in multiplayer was probably in college, when my dorm floor played regular Quake II matches. The Quake games are notoriously fast-paced affairs, heavily reliant on twitch skill. But they also feature finely-tuned maps that allow for creative tactics. One thing I loved about Quake II was that I could practice with bots before taking on my friends. There was a way to refine my skill in a customizable environment, to familiarize myself with maps, items, and weapons without getting my head blown off every ten seconds. Everything was accessible from the start; the game felt like a shooter, not a JRPG. The playing field was even.

Here's the best part: Translating that experience with bots to matches against my friends was incredibly satisfying, even if I still got slaughtered. I had learned something by practicing and applied it in a live context. If I got destroyed, it was because my opponent was more skilled, not because he'd unlocked some perk or weapon. If "fun" in games is about learning, as Raph Koster says, we need more opportunities to learn, not to earn.