« July 2009 | Main | September 2009 »

August 08, 2009

First Impressions

One of the things I find important to any first person shooter, and perhaps any linear game¹, is the "introduction" -- how you are introduced to a new enemy, new weapon, new environment, anything. I came by this importance working with Nathan Martz, now the lead programmer at Double Fine, on Star Wars Republic Commando. As our enemy AI programmer, he really drove forward the idea that each introduction of a new enemy should be interesting and should say something about the character in question -- its tactics or its personality. In SWRC, that became important to how we introduced everything, from your squadmates to your weapons to your enemies, and it has become one of the lenses I examine games through.

So, recently I played Gears of War and in this area there were a few things I felt they might have worked better. The game is very successful in the running, gunning, and cover areas, but less consistent in how it introduces its enemies. Here are a few rules that came to mind while playing.

  • Personality is key. The Berserker was a good example of this, a blind, raging character who bursts through walls. Unfortunately, the introduction of the Boomer was far less successful -- two Boomers walking down a corridor who... step on a rat. It just doesn't fit with the character -- their in-game behavior is far better. While I understand that this was meant as an attempt at levity, it was completely out of place with the tone of the rest of the game, involving slapstick rather than the over-the-top macho gallows humor prevalent throughout.
  • The player has to see it. The first time I ran across the Corpser, one of my squadmates pointed him out and made some remark. Problem was, I was looking the other way, and as of this writing I still don't know what I missed. This can be a real danger in in-game introductions, where you have no control over the player's current attention. In Republic Commando, all of our introductions were in-game, so we did our best to put them in spots where it would be hard to miss them. An ameliorating factor in the case of the Corpser was that there were several introductions, building tension, which gave me multiple opportunities to see it. Many of the introductions in Gears are done through cutscenes, which avoids this issue entirely, but takes you out of the first-person perspective.
  • Give clues as to gameplay. The Gears "Berserker" is a good example of this; we're told that we can destroy him if we can get him into the outdoors. However, we're in a sealed-off room. Lo and behold, this critter is superstrong and will run at any noise...
  • Introduce the enemy in an appropriate location. If you're always going to fight a critter in a tight hallway, don't introduce him in a wide-open cavern. The Wretches were a good example of this done well -- they spew out of a hole in the ceiling in a tight hallway, which is a space which works well for them.
  • The first impression can't be far from the payoff. With the Corpser, there were several opportunities to see what was coming, and a lot of tension built up... that was squandered as the critter disappeared until the very end of the next act, several hours of gameplay away, which for me was as much as a few days. As a counterexample, the Reavers are shown in an introductory video and immediately pay off by attacking the train you're riding in.
  • Make the meaning clear. It must be clear to the player what they are seeing. In the case of General Raam, I had no idea who this guy was, and he ended up being the final boss. He appeared briefly and shot some random soldier, and then disappeared until the end of the next act, depriving that final encounter of the emotional impact it might have held.

Hmmm... I seem to be getting a bit Bullet-Pointish these days, I'll have to watch out for that. No idea what I'll be talking about next, but you can meet me back here in a week or so.





¹This is not to say that I don't believe this can be managed to some degree in open-world games as well, though it's a more significant techological and artistic challenge. (back)

Posted by Brett Douville at 10:31 AM | Comments (0)

August 03, 2009

RSS feed

I've updated the RSS feed link to point at the atom feed, which works much better with Google Reader and I suspect others. For full text and better formatting, I suggest you unsubscribe from the feed and resubscribe.

Hey, while you're at it, you can follow me on Twitter.

Posted by Brett Douville at 09:01 PM | Comments (8)

The Perfectly Executed Mini-Mechanic

As you know, I've been playing lots of games again, which is nice. My travels have included the planet Sera and two trips to the Middle East, albeit a few centuries apart, and lately I've been relaxing at the mall. In those travels I've noted down a few impressive "mini-mechanics" -- little bits of games that aren't the center of it but do a good job of reinforcing the main game.

The mini-mechanics that work best for me tend to follow a few guidelines:

  • They reinforce the main character. An old game development hand once said to me that characters in games had nothing to do with their personality, and everything to do with what they could do. In other words, while Mario's name was changed from Jumpman, that name almost exactly defines Mario as a character in his first game (and might therefore have been too limiting). Indiana Jones, in the context of a videogame, is his whip. Buffy Summers is a stake. Riddick is his vision. You get the idea. A perfectly executed mini-mechanic reinforces the gameplay traits of a character.
  • They are brief. Hence "mini". They should fit right in with the rest of gameplay.
  • They are infrequent, but hardly unique. If you're doing this every few seconds, it's not a mini-mechanic, it's probably the mechanic. Similarly, if you're only doing them once, in a boss battle, it's not a mechanic.
  • They aren't required. These are purely optional; more accurately, it might be better to say that any individual instance is optional.
  • They should be in situ, not apart from the bulk of gameplay. If you come out to another interface screen, as in many mini-games, you're really doing some other task, and not using a small mechanic to reinforce the main game.
  • It's not meant to challenge the player. You aren't doing it enough, probably, to be able to do it perfectly every time. So, it can't be a high skill proposition; the risk-reward ratio can't be too high. It should never be a game-ender.

The first of these that really struck me was the "leap of faith", from Assassin's Creed. In each of the game's 11 locations¹, Altair may ascend a set of towers and other high points to get the lay of the land below him. Now, having gone to all the trouble to make Altair able to climb up or down every vertical surface, I would have forgiven the designers for making me control Altair back down to a reasonable location. Instead, they turned a very real gameplay problem into an opportunity for a beautiful little mechanic. Generally speaking, the place I need to go next is way down there amongst the ants, so adding a way to get me down there quickly was an excellent choice. It's flawlessly executed -- the animation is fluid and frankly gorgeous, and got me every time. Finally, it reinforces the character -- this is a man who is able to do thrilling things with his body through years of training.

The next two worth mentioning from my recent games both come from Gears of War, though I think one of them was better executed than the other.

In a few years, when I look back at this first year of HD gaming, I'm going to remember the "roadie run" from Gears of War. For those who haven't played... oh, who am I kidding, I'm last to the party on this for sure. In any case, the opportunities to use the roadie run are somewhat rare -- in any large spaces where it would be most appropriate, you're often too busy with enemies to be able to use it for pure navigation, and in smaller spaces, it's less effective. But as a means of moving up in a sort of semi-cover (crouching to make for a small target), it's terrific. It accentuates the gameplay -- Gears is almost entirely a game about cover in combat -- and it furthermore reinforces the physicality and bulk of the character, as Marcus Fenix rarely looks as built as he does when he's running in a crouch carrying all his heavy equipment. And it uses up a resource, though a hidden one; there's some sort of fatigue going on, since Marcus can't use it indefinitely.

The other is the "active reload". I was less enamored of this mechanic, primarily because in order to get the feel for it, I was constantly drawing my eyes away from the action. The active reload mechanic involved exactly timing a button press to an on-screen indicator, but I found that you could train up on it. Early on, splitting my attention between the action in the center of the screen and the reload bar in the upper right was more than I could manage, particularly with a variety of weapons, and I think this is largely the reason why this mechanic didn't quite feel right to me. Later in the game, once I had taken a look at the achievements list, I spent a minute of relative calm just doing it a dozen times in a row to get the achievement, and after that it felt good and reinforced the character. It felt like the sort of thing that some Army grunt might know -- oh, if you bang the butt of the rifle at the right time, you can kick the first round into the chamber quicker, but watch out, the gun can jam if you misstep.

That all said, this was a mechanic I appreciated rather than loved, unlike the other two. The leap of faith would have gotten old if I used it to jump off every building. I would have wished for better running and gunning had the roadie run been constant. With the active reload, I would have been more impressed had they showed a little restraint and made it only available on the one gun -- the gun with which a Gear would theoretically be most familiar. At the beginning of the game, where I was switching amongst several weapons to find the one that fit me best, the reload was a hindrance more than a help. By the end of the game, I had given up trying to time any weapons other than the main assault rifle that you start with; it was simply too distracting and not worth the time to invest to learn each of them.

Finally, it's not necessary to always introduce a mechanic, or to make a mechanic more complicated than it need to be. I appreciated that in Call of Duty: Modern Warfare, there was no special trick to place a charge or tank-buster or what-have-you on an armored vehicle or other target. In these cases, it was simply a "get near and press A" proposition -- an absolutely minimal implementation. Pressing the button on the controller was like opening a door or doing something else similarly trivial. More than that would have likely led to player frustration, which should be avoided like the plague in mass-market titles.



Well, that's about all I've got for now. Join me in the next week or so again for some discussion about "introductions".


¹Sure, it's an open world game, but there are three cities, each with three districts, your home town, and the "Kingdom". For purposes of this article, the modern office building to which Desmond Miles has been kidnaped can be dismissed, as it contains no actual gameplay. (back)

Posted by Brett Douville at 08:35 AM | Comments (0)