Hmm. I would class different enemies as being a part of the same lazy design quirk you suggest. It is easy to throw in grunts with green hats that are immune to shotguns, but at the same time it’s quite easy in design terms to pass from grunts to ubergrunts who have retractable arms allowing them to reach out to you from further away. Enemies are just stat blocks ultimately, even if some take more effort to draw than others.

By contrast it’s quite difficult to keep coming up with tactical challenges without resorting to the same thing only dealing more damage. That’s a game AI problem.

for example in the D&D game I was running, it struck me as easier to go from 1 HD monsters to 2 HD monsters to 2 HD monsters that are immune to sleep. That’s a very quantitative approach to design. I found it more rewarding and tougher to think more about how the monsters would deploy and how they’d act. If you do that, in principle you can use the same monster over and over again. You’re just posing different challenges. The struggle is to keep coming up with new challenges.

In the case of GTA the “you run in, kill a load of guys, do something, run out and lose the cops” is a tried and tested challenge. Keeping things interesting without lapsing into that model is a new design challenge every time a mission is conceived. This is one reason why I think the design of GTA games have improved : each time they return to the games they have a wider idea of what works and what doesn’t. It’s just frustrating that the designers aren’t more ambitious in trying new stuff to keep that elastic band stretched.

To parse that slightly, many FPSs suffer from a phenomenon where to prolong gameplay big bads are introduced, or new monsters with immunity to your best attacks, or some variant on that theme. There tends to come a point where the game stops using existing foes in new and clever ways, and just throws harder foes at you.

That’s because, IMO, it’s much easier. To have a decent length game with a consistency of threat which varies because it comes at you in different ways is creatively hard, to prolong game length by bunging in a new and tougher threat is actually much easier – no longer a creative challenge at all, merely an easily solved technical one.

Equally, it’s creatively hard to come up with ways to make small number of opponents challenging for a decent length of game, but throwing in waves of them is easy. Once you’ve programmed for one, replicating them isn’t that big a deal. This leads to absurdities such as in the sometimes excellent game Mafia, where you wade through literally tens of people to your goal, along the way sacrificing all sense of verisimilitude (or even, there, fun). It’s easier to throw in waves of bad guys than to think creatively of how to use two bad guys well.

Few games get around this, because the challenge is a creative one, not a technical one. Off the top of my head Operation Flashpoint is the only one that comes to mind.

Good piece by the way.

I think you may be right that ‘fun’ may be an active ingredient in the band snapping back but I think it’s a good deal more subtle.

a) I think that all kinds of games (even table-top RPGs) struggle with increasing power levels and the desire to make each adventure better and bigger than the last. So I think that the longer any game lasts, the more likely they are to start including gonzo elements simply because they’ve run out of mundane stuff and need to go ‘one better’.

b) I think that there’s quite a narrow definition of ‘fun’ involved here. Yes driving a tank around the place is fun. More importantly, everyone knows it to be fun and it’s a proven value of the word fun. the GTA franchise is built upon rampages, even though the taxi cabs in GTA IV cut down on those kinds of distractions (whilst also brilliantly conveying what it’s like to live in a big city. I live in London and I know that Camden is a tube ride away or a cab ride away but I don’t know exactly how far it is or how I’d drive there. In previous GTA games you basically drove everywhere and so engaged with the cities in a way that only real world cab drivers do).

So I think it’s partly problem of game design inherent in continuing narratives and partly a question of innovation vs. conservatism. I also think it’s simply a question of people running out of ideas. As with GTA : SA, the innovation lasts a certain amount of time but then the tried and tested formula reasserts itself; simply put the designers run dry and still have game to fill.

In your analysis, it seems like the mundanity of GTA’s prologue is the achievement, and the regression to formula (mayhem!) is the failure, but their relationship must be more complex. GTA is not just a really good implementation of The Sims that happens to run off the rails into mass violence; the insane mass violence is there for a reason, has been there since the inception of the modern GTA formula (GTA III in 2001) for a reason.