Thursday, January 29, 2009

Reward Schedules

One of the fundamental things I expect from a game is that it rewards me in proportion to the amount of time I spend playing it. This is why I can't stand games that are too long, because the rewards tend to come too far apart. But there are other ways that game designers miss the point and fail to reward in proportion with play time.

Wipeout Pulse is a pretty standard racing game. You do events, collect medals (ideally gold); collect enough medals and unlock more events. One of these events is a Tournament, which is a set of several races back to back. The problem is, you only get one medal for the tournament. There isn't any more or less or different skill than a single race, but I'm only rewarded 1/3 or 1/4 as much per unit of time played. This is really annoying, and I've now cleared the first nine (out of 12 and 4 bonus) race grids having played exactly one tournament. I understand the need for tournaments in an online game, to test player performance on a variety of tracks - but in the single player game, you've already done that by making me play all of them against the same AI controlled cars I'll play against in the tournament. [I wrote this paragraph before Wipeout HD came out. It has the same problem.]

There's a fine line between rewarding players for time invested and for skilled play, and I also get no satisfaction from rewards simply for time invested. David Sirlin has already written everything I want to say on this subject:
Here is what World of Warcraft teaches:

1. Investing a lot of time in something is worth more than actual skill.

If you invest more time than someone else, you "deserve" rewards. People who invest less time "do not deserve" rewards. This is an absurd lesson that has no connection to anything I do in the real world.
The problem isn't specific to WoW, though. Too many games, MMO and not, reward time investment over skill improvement. Skill improvement is often the result of time investment, so it's hard to distinguish sometimes. Still, there are clear cases of games that have unmotivating rewards, not because the rewards aren't interesting, but because they're handed out in the wrong ways.

My desire, which I don't think I've seen a designer formulate explicitly before, is that I want to be rewarded a number of times proportional to how long I play, but with a quality of reward proportional to how well I play; additionally I want my reward magnitude to diminish as I play things far below my demonstrated skill level. This is an oversimplification in some ways - varying the reward frequency randomly is much more effective than a reward every hour, for example - but is what I basically want. If I play a level for an hour and get 100 points, playing for two hours should get me about 200 points. If I play a level further in the game, I should get at least as many points, and also a special hat; if I play that level twice, I should get two hats.

To bring this back to Wipeout: Tournaments should reward me proportional to the number of races in them. A gold medal in a tournament should be worth 12 points instead of 3. The game already does a good job of doling out rewards with magnitude proportional to my skill level, by unlocking new harder, faster tracks when I beat the old ones. To balance that, playing levels from grids before my current one cannot unlock future ones, only raise my raw score.

Although the application for Wipeout and other arcade-style games is obvious, the general principle works when analyzing other games. In Sim/Tycoon games for example, it's easy to get a stable supply of money and rake in profits slowly forever. But you can keep building more and more complex parks, taking on more risk to prove your skill at the game and get a bigger reward - manifesting as more money faster, and a chance at an even more complex environment. Design elements like tech trees also make sure that to get the coolest things in the game, you need to task those risks and not just amass money.

In RPGs, I can grind for XP and level up my characters. But superlinear level progression means this stops being effective unless I push myself, and I'll never see new content - plot, environments, items, and so on - unless I move past the area I'm in.

Poor reward handling has hurt a lot of games for me recently - Assassin's Creed was enjoyable except the game's main reward mechanism is making you better in combat, when a key element of the game is avoiding combat. Fallout 3 stopped handing out primary character advancement 20 hours into an 80 hour game (by contrast, my Oblivion character has a dozen stats I could increase after 120 hours, but it stopped giving me exciting equipment, which Fallout managed up to hour 50).

Saturday, January 24, 2009

They Only Met Once

What we found out is that each of us is a bro, and an athlete, and a princess... and an onnagirai, stripper, crossdresser, and... some sort of bear thing? Plus the dude who seems to be an empty shell. I guess that doesn't really answer anything. Sincerely yours,
There's two things I want to point out about Persona 4.

It has the emotional maturity of a John Hughes film. It's pretty obvious in this case because of the similar setting and themes, but we're staring at a bog-standard high school coming of age story; not just standard in its themes but also its structure and presentation. Game designers should be pushing more comfort zone boundaries and offering a deeper analysis, especially in a game like this where the game is going to sell to the same group of people regardless of its content.

However, it has the emotional maturity of a John Hughes film. Standing next to the Porky's standard of the video game world - whether that's the incessant whining of a 15 year old Final Fantasy protagonist, or the precocious and unrealistic maturity of a 16 year old Final Fantasy protagonist - it's a huge leap forward in the same way Hughes's films were. For as little as it pushes, it's much further and harder than its contemporaries. Especially contemporaries in the same genre.

Monday, January 19, 2009


I realize I'm a bit late to the new year's party, but trying to summarize an entire year is hard enough. If I tried to do it before the year actually ended, I'd be doing a disservice to DJ Max Portable: Black Square, which came in the mail December 30th. While one of my favorite games of the year, it's not going to make my "game of the year cut", for reasons I'm not entirely sure of myself. When talking about the best games of the year, how do I measure that? Disregarding that game preference is subjective, what priorities do I give to innovation (successful or otherwise), graphical polish, longevity? At least for me, there's a tendency to consider the ratio of my enjoyment to my expectations, which is why Black Square and others, like Metal Gear Solid 4, don't make the cut. Not unlike information theory, the games that meant the most to me this year are those that surprised me. Even if Black Square is what I'm still playing at the end of 2009, it'll have blurred together into itself and other DJ Max iterations. In reality the games of 2008 are not going to be decided here and now, but sometime in 2011 or 2012 when I say "holy crap, wasn't it awesome when I shot that barrel?" (I hope that never happens.)

There is another tendency which people have noticed, and this list is guilty of. It contains more games released near the end of the year than games from early in it. Pessimists claim it's because these are the games played more recently. I'd like to posit that, just maybe, we're getting better at designing games over time. The alternative is incredibly depressing.

No More Heroes
I cannot praise this game enough. While it's come under fire from many reviewers for its supposedly shallow open world and linear level design, those are exactly what makes it a perfect example of a game's narrative themes being reinforced by its game mechanics (and vice versa). Travis's world is stark and existential. Literally the only things to do in the game are kill, sleep, eat, consume, and (the ultimate goal) fuck. The existential crisis of a Final Fantasy character is shallow, because they will invariably face down their destiny - their purpose - sometime around hour 60. The existential crisis of Travis is very real, and the player faces the same one. Have you found an exit?

And aside from that, or maybe because of it, the game dodges the pretentiousness of MGS2 or Braid and is fun on a visceral level. Whether it's beating down a hundred goons or avoiding a crotch laser, you're enjoying yourself the entire time playing it.

Mega Man 9
For all the handwaving about "nonlinear gameplay", "player-driven narrative", and "immersion", here is a game that does them all with aplomb and no one even realizes it. The choices you make in Mega Man are those that really matter - the ones that affect the game mechanics, not the ones that change some text in a dialogue tree or chance the appearance of an otherwise irrelevant NPC. The Mega Man universe is immediately familiar to anyone over the age of 20 (and many of those under it) and the game leverages that to draw the player in with as much meaning (i.e. none) and far more emotional investment (how many broken contollers?) than most other games this year. As for immersion - do you seriously think about something other than this game while playing it? There's no downtime to rest and no mercy for the distracted.

The marriage of new design ideas to tight and polished old games that we saw in Bionic Commando: Rearmed and Pac-Man: Championship Edition is here as well. Downloadable expansions, time trials and leaderboards, and achievements - many of which acknowledge the game's fundamentally old design (can you imagine Halo 3 having a "beat this game five times in a day" achievement?) - all drive home the fact that we really did have some good ideas back in the 1980s.

Fallout 3
I wrote of the ratio of expectations to reality earlier; my expectations for this game were insanely high. Bethesda overdelivered. The game is deeply flawed, from technical and design views. But there's an equally deep satisfaction to be had crossing the D.C. ruins, seeing familiar landmarks, and exploring areas that feel genuinely undiscovered. In a year where level design took a backseat to flashy graphics and writing in so many RPGs, Fallout 3 shows just how much good level design can carry a game just as last year Mass Effect showed how poor level design can destroy one.

Soul Bubbles
A victim of the recent retailer-exclusivity trend, Soul Bubbles is a puzzle-"platformer" with a focus on a slow pace and accessibility. It should be studied by anyone who wants to know how to make a game with a shallow learning curve appease nearly everyone regardless of gaming preferences.

Honorable Mentions
I'm reserving final judgement on Mirror's Edge until the DLC comes out. I think that's the game this really wanted to be, even if the developers didn't know it until after they wrote such a horrible frame story.

Rock Band 2 per se is only an iterative improvement to last year's version. The key this year was that Harmonix showed they knew how to manage a community that now dwarfs most other online games in players and dollars.

LittleBigPlanet is the tip of an enormous iceberg. I think by the end of 2009 I'll know whether it should've been here or not.

Persona 4, Valkyria Chronicles, and World of Goo are simply games I have not played enough yet. They probably belong here.