Wednesday, December 5, 2012

Matters of Scale

In your traditional roguelike, or basically any typical level-based RPG, the player and enemies will scale up on a scale. These scales usually aren't equal, but it is still noticeable. Game balance has a great deal to do with these scales: if the enemy scales up too fast, the player can quickly hit a brick wall and lose interest in the game. On the other hand, if the enemy scales up too slowly, the player will steamroll through the game. It is worth mentioning that the former situation tends to cause players to quit before they finish far more often than the latter. Spending an hour or two being an unstoppable super hero is fun sometimes - being pitted against a challenge that seems impossible for hours is generally considered substantially less fun.

I believe a well designed roguelike game falls somewhere in the middle. The brick wall should only become apparent if the player has poorly utilized their available resources, but the player should never really go full super hero mode either. I don't think brick walls should be arbitrary requirements that force a player to start over, though. One commonly cited example is System Shock - if you don't build your character in a certain way, you basically can't win. And System Shock wasn't exactly the type of game where you could go and right click some random NPC and reset your skill choices.

Another example, at least for me, is Deus Ex: Human Revolution. I played through most of the game on the hardest difficulty setting. It was pretty fun and engaging. I had to restart a level occasionally, but I never felt like I needed to tune the difficulty down. Than I reached a boss that electrocutes the room, which can only be easily resisted if you've taken the electricity resistance armor skill (which I had not). Supposedly this boss is easy enough on Easy or Medium without the skill, but I didn't want to tune it down. I felt like a brick wall came out of nowhere - the game didn't exactly scream at me and say, "Hey! Get the grounded armor perk!". It let me comfortably stride through the game and utilize a variety of solutions for every level and than shoved this random boss into my face. I put in a good two dozen tries on her and even looked up a few tips online, but I still couldn't get it down and I had no desire to reduce the difficulty. So I quit. Admittedly, I'd still like to  go back and finish the game someday, but the moral of the story is: don't invent completely arbitrary challenges and, if you do (which you shouldn't), don't opt to not tell the player about it beforehand.

Time to get this topic back on track and prove that this article isn't just a thinly veiled rant about Deus Ex: Human Revolution and is actually relevant to @Star Wars. Basically, the game currently attempts to emulate the typical RPG scale. You scale up, they scale up, we all scale up together. That's all fine and dandy, but I'm not sure if it lends itself to the Star Wars cinematic feel. I think Star Wars instead offers up the interesting idea of a "yard stick" enemy. Let's say there is one particular enemy in the game that exists throughout every level. Early on, this enemy could come alone or in a small group and be quite a challenge for the player. Later on, these enemies could come in larger groups but be more easily cut down by the player.

In A New Hope, Luke has to take on a few stormtroopers, despite being a fairly novice Jedi and fighter in general. Later in the series, Luke never really stops fighting stormtroopers, but he does get stronger and they turn into a kind of cannon fodder. I think there are interesting ways to build upon this kind of cannon fodder into @Star Wars. The idea isn't to eliminate scaling - the player will still meet new and increasingly challenging enemies. Beyond that, the cannon fodder could also diversify. Perhaps in the first few levels the player only engages against small squads that feature mostly marksmen, but later on they might have grenadiers, snipers, medics, etc...

Basically the idea is to introduce a different type of scaling that would work alongside traditional scaling. The player will have to fight these reappearing foes throughout the game, but they won't get statistically stronger - only with larger and more diverse groups. The real fun is when traditional scaling comes into play. Perhaps early on the player fights tiny groups of stormtroopers and local wildlife. Later on they might have to fight large groups of stormtroopers, AT-STs, dark troopers, and so on.

So, what would the point of this be? To increase the cinematic feel of the game and, subsequently, its Star Warsiness. I think if this idea was pulled off correctly, the game would feel similar in terms of balance, but also feel much more appropriate to the Star Wars universe and hero it is trying to represent. This is still an infancy-level brainstorm and I don't have any hard details ironed out nor do I know if I'll let actually go through with it, but I thought it was an interesting enough idea to throw out there.

Editor's Note: In retrospect, I realize that the connection between the first and second half of this article may not be crystal clear. Basically the goal with @Star Wars is to still hit that middle ground of game balance, but go about it a slightly different way.

P.S. Overall, I thought the rest of Deus Ex: Human Revolution (that I played through) was very, very good. But the boss fights were just an incredible let down. As far as I can tell, there are two ways to solve every boss fight. You can own the seemingly randomly selected skill that lets you cheese the fight or you can bash your face against them until they die. Keep in mind that DX:HR is a game that, besides the boss fights, lets the player pick from a pretty substantial number of choices to get through standard levels. You also couldn't avoid the boss fights. This seems like a major oversight in a game where you can successfully sneak through all of the levels and fight nobody.

P.P.S. This is the second longest blog post to my knowledge. It is beaten by the very first post - my essay on roguelikes.

No comments:

Post a Comment