Sunday, December 18, 2011

L.A. Noire - Final Thoughts

Finally finished L.A. Noire, and I have to say that I'm no longer as impressed with it as I was when I started. First of all, the "driving around L.A." parts of the game got to be so tedious that I started skipping them whenever possible. I was doing the chase scenes, but any time when I was just driving from point A to point B, I'd skip it. It's a very pretty re-creation of L.A., but I really don't want to invest hours of my life sitting and waiting for imaginary traffic signals.

Second, the interrogation scenes started to get pretty arbitrary. In the last third of the game, it seemed like there was no logical connection between what the characters were saying and what option I had to choose to "win" the interrogation. It turned into a lot of wild guessing for me, and as a result I was only getting about 1 out of 3 questions right ... with no penalty or negative effect. Sure, maybe I wasn't getting as much XP, but the game still let me advance the story even after fumbling an interrogation, and leveling up my character turned out to be mostly useless anyway.

Which leads to my third problem, which is that there really wasn't much skill involved in finishing the game. Didn't do well in an investigation? No problem, you'll always get at least the important clue you need to proceed. Even if you screw up a really critical interrogation, they'll let you do it again until you get it right. Can't handle an action sequence? After a few tries, the game lets you skip it. About the only part of the game you aren't allowed to screw up is the clue-finding part, which got really annoying at one part of the game, where all the hints were indicating I had to do one thing, whereas the actual clue was somewhere else, and I eventually had to give up and hit the web looking for someone to point me in the right direction.

So, whereas the story was nice, and the scenery was fantastic, there really wasn't a lot of challenge here, since there was never any real penalty for failure. So while I did finish the game, I don't really have the sense of accomplishment I would normally get when I've battled my way through a typical video game.

And I know I keep bringing this up, but I think using a Grand Theft Auto kind of engine as the base for this game was completely the wrong way to go. This isn't a free-roaming, open-world game; you have to do missions in a pretty linear order. Sure, there are a few random "police band radio" missions you can pick up while driving the streets, but there's no real incentive to go around and explore, unless you absolutely think you have to find all the different hidden cars and landmarks. All the work they did to create a virtual 1940's L.A. was basically for no good purpose. There are thousands and thousands of fantastically realistic buildings and streets and such that you will glimpse only briefly as you pass by, or possibly never see at all. They could have easily just created maps for specific mission areas, modeling the entire city for this kind of game was a terrible waste of resources.

My next game is going to be Fallout: Las Vegas, which I expect to enjoy a great deal more. It's another game with a vast and detailed environment, but if it's anything like Fallout 3, it's going to be a challenge the whole way, and it's going to reward exploration of every last radiation-soaked square meter.

Sunday, November 20, 2011

L.A. Noire

I'm fairly deep into L.A. Noire for the PS3, and for the most part I'm enjoying it. I like the investigation and interview sections; you really feel like a detective at these times, finding clues, conducting interrogations, getting confessions. It's during these parts of the game that you get into the lives and the stories of the people whose crimes you're investigating.

The shooting and fistfight components are okay, though you do seem to rack up a heavy body count. Three-quarters of the random street crimes you encounter end in you putting a few holes in someone. I haven't had any luck trying to turn a gunfight into a fistfight; even if I creep right up on someone who has a gun out, I don't seem to get the opportunity to sock him one. The game seems to have pre-decided which fights are going to end in a fatality.

The platforming elements seem misplaced. Your character seems to be a regular acrobat, shimmying up walls and leaping across rooftops like a parkour master. Some points in the game take this to extremes, which I really dislike; if I wanted to play a platformer game, I would have bought one.

The part of the game I hate most is the driving parts. Yes, I know, this is a Rockstar game, and they're all about the realistic cities and realistic cars and driving all over the place. And they did make 1940's Los Angeles awfully pretty, and it is sometimes very immersive to be driving around at sunset, passing palm trees and billboards, listening to period music on the radio.

The problem is that they've made the driving bits too realistic. There's traffic on the roads, and traffic lights. If you wait for the traffic lights, it can take an awful long time for you to get from one place to another. This is anti-fun for me. I don't have a Playstation so that I can pretend to be sitting at a standstill in traffic. You can always run the red lights, but this sort of breaks the suspension of disbelief for me, and also winds up with me crashing into a lot of cars and pedestrians.

You can run the siren, which helps to get folks out of your way, but every so often some other vehicle will drift into your path as you're blasting through an intersection. You can turn the driving over to your partner, but then you miss out on the random street crimes, and so you're skipping a lot of the game content.

When there's an actual car chase happening, then the "real driving" experience makes sense. It's part of the challenge then; a high-speed chase through crowded streets, trying to avoid injuring innocents while not letting the bad guy get away. But the rest of the time? It really gets tedious going back and forth, block after block, mission after mission. Not to mention that the camera actively attempts to screw you up by spinning around when you shift from forward to reverse; a slow spin that throws off your sense of direction and keeps you from seeing either straight forward or straight back for a few long critical moments.

I also have issues with a few of the story decisions that the main character makes, which I won't get into for spoiler reasons, but overall, it's an excellent game. I just wish they could have taken away more of the Grand Theft Auto elements and focused more on the noir/detective elements.

Tuesday, June 21, 2011

Welcome back, Java

My previous job was almost entirely C++; my new job is almost entirely Java. I think I answered an interview question recently by saying that I like both languages equally, and the one that I choose depends on the needs of the situation.

Of course, I said that having not actually used Java for quite a while. Getting back into Java again reminded me that I do really prefer it to C/C++; that when I first got to use Java, it was liberating, exhilarating; that I couldn't believe I'd dealt with the hassles of C/C++ for so long when Java had been out there, waiting for me.

Say what you want about the relative benefits of both languages; for me, I could make quite a long list of the things that make me prefer Java, but I'll just mention two: one annoying, one more critical:

1. Header Files. Having to maintain header files alongside your source code files in plain C is bad enough. Having to maintain a header file for every class file in C++ is just tedious. And then add how you have to put #ifdef blocks around the content of every header file because C++ can't figure out how to resolve the situation where the same header file is included multiple times. Listen, if I have to add the exact same #ifdef structure to every header file to cope with a reasonably common situation, this ought to be solved by the compiler somehow so I don't have to type the same hacky solution over and over again. Sure, a modern tricked-out IDE will autogenerate all of this code for me, but this is just the IDE glossing over inadequacies of the language.

2. Memory Leaks. As a C/C++ programmer, you cannot ever focus entirely on the desired function of the code, because at least part of your attention always has to be on memory management. Sure, it's possible to write applications that never ever need to explicitly delete or free, but in my experience this is more of a theoretical paradise than a real-world result. I always seem to wind up having to get my hands dirty with pointers and with memory that's my responsibility. It's a profound relief to be able to say "new" in Java without having to immediately make sure that there's a matching "delete" somewhere that will absolutely positively execute.

So goodbye stdio.h, goodbye cout, goodbye char*. You'll always have a special place in my heart.

Monday, May 30, 2011

Why Vampire Hunters Aren't Cool Anymore

Well, not the modern stereotypical vampire hunters, in any case. I didn't even realize there was a stereotype until I saw it on television this morning. We were watching the beginning of some vampire movie (I don't recall the title) and there was a scene of the vampire hunters discussing some upcoming raid, and I realized that I'd seen all this before. Here's the checklist:

  • A bunch of edgy, misfit, gung-ho loners, dressed in t-shirts and tank tops
  • A clandestine meeting location, cluttered with the tools of the trade
  • The "tools of the trade" mostly consist of a mixed collection of firearms and medieval weaponry
  • A prevailing attitude of "the cops/public can't deal with this, so we have to"
  • Bonus points if there's a misunderstood vampire on the team, helping the good guys
And it struck me: why should this be the winning mix of ingredients? If there are enough vampires hidden around the world to give a small pack of vampire hunters a constant stream of work, is the "garage band" model really a model for a successful team?

First problem: if there are that many vampires, and it's possible for a pack of underemployed twenty-somethings to find them on a regular basis, the vampires really can't be that well hidden. Which means that the proper authorities would have found them by now, and that teams with proper government funding would be working the problem. Unless, of course, the government is "in on it" or something.

Second problem: what these DIY vampire hunting teams seem to be involved in is a continuous series of special forces raids against a foe that's highly resistant to conventional weapons. It seems to me that this team would require serious combat training (as opposed to on-the-job training, which would be suicidal). You need to be able to handle weapons and maintain discipline and coordination under the worst possible conditions. You also need to be able to give and receive bloody injuries on a regular basis without losing your sanity, which requires serious mental training. Yes, history has shown that rag-tag teams of rebels can make an effective fighting force. But what kind of losses do these real-life backyard commandos experience? Out of a team of a half-dozen or less individuals who gave up their barista jobs to fight vampires, how many of them would take themselves out of action from fumbling their weapons in the first few months? And how many lost to friendly fire ("Wow, sorry, Joe")?

It seems to me that a more effective team would look like a team of special forces operatives, with iron-clad discipline and chain-of-command, with standardized and well-maintained modern weapons and by-the-book tactics, supported by a substantial team of surveillance agents, doctors, intelligence experts, and so forth.

What I'd really like to see, though, is a return to the Peter Cushing style of vampire hunter. The intellectual hunter with a deep and well-rounded education, who went into a fight with nothing more than a few doctorates, a sharp stick, and impeccable manners. I'd like to see the kind of guy who has the fight won through strategy and preparation before he even steps into the vampire den, and who doesn't need five minutes of slo-mo gun-fu martial-arts action to defeat his foe. Vampires are typically represented as stronger, faster, and more durable than human beings. It seems to me that you want a hero who can out-think the nosferatu. That's the movie I'd like to see.

Sunday, May 29, 2011

Playing With a Full Deck

Recently I've become addicted to an online card game called Tyrant. I've been playing it on Kongregate, though I assume all the cool kids are playing it on Facebook instead. In any case, it's a collectible card game (CCG) similar to Magic: the Gathering (MTG), but streamlined for more casual play.

I'm interested by some of the design decisions they've made with Tyrant. First of all, they've removed a lot of the choices a player of MTG might have to make, leaving the player with only the decision of which card to play from his or her hand at the start of the turn. All other game rules are deterministic or randomized. So the strategy is simplified, which I expect makes it easier to write an algorithm to allow for computer-controlled opponents. The algorithm just needs to pick one card out of three to play, on every turn.

Also, they've made some cards measurably more powerful than others. In theory, this is balanced sometimes by increasing the activation time of the more powerful cards (in other words, it takes longer for these cards to become active after they're played). But even when comparing cards with the same activation time, some cards are definitely more powerful than others, which leads to the player eventually retiring some cards as "obsolete" in favor of newer, stronger cards.

I don't know how prevalent this is in MTG, but I would think that this would be undesirable from the point of view of a traditional CCG that's used for competition. After all, the point of a CCG is that the strategy is in deck-building and intelligent play. All cards are supposed to be good in certain circumstances, and some combinations of cards can create a synergy if played at the correct time. So, ideally, each card when taken in isolation should be just as powerful as any other, so that just having card X does not automatically give you an advantage over players who don't have it. Having intrinsically overpowered cards makes the game less about strategy and more about acquiring cards.

This aspect of the game actually works more or less fine in Tyrant, since it's not a tournament game. The idea is to complete single-player missions in order to acquire more powerful cards, and thus improve your deck so that you can beat harder missions. There is a player-vs-player aspect, but in situations where one player has collected more of the advanced cards, that player will have a strong advantage.

You can pay real money to buy cards in the game, but you can also just get new cards by playing the game and beating missions, so you can accomplish quite a lot for free. As a result, I've been on this game every day since I found it, opening booster packs and assembling decks and such. I do have a weak spot for this kind of game; it's a good thing I never really got into Magic, or I'd probably have a much lighter bank account by now.