on Nov 7, 2013
The Freddie Wong video we mentioned is here: Splinter Cell: Lightbulb Assassin. This ties in nicely with the conversation we had about stealth in Metro 2033 and Metro: Last Light. And really, it also relates to stealth in just about every stealth-optional shooter. The stealth is often silly, absurd nonsense.
I think the reason for this is that it’s prohibitively difficult to make stealth mechanics that are both sensible and fun. Stealth is complex and involves a lot of fussing with AI. If I was actually guarding a single dark room all by myself for hours, I would very likely be extremely sensitive to even the slightest sounds. I’d notice if a door was open, a chair was moved, a light was off, a shadow moved, or a friend was missing. I’d hear someone (particularly a grown man with a lot of gear) hitting the ground in the next room. And if anything spooked me I’d likely keep my back to the wall and turn on every light at my disposal. If not out of a sense of duty, then out of a sense of self-preservation and a desire to fill the time. I mean, I literally have nothing else to do. And even if I didn’t find any enemies, I’d be paranoid for the rest of my shift.
And of course, knocking out a human being in a single non-lethal blow is incredibly difficult. It’s very unlikely anyone could do it reliably and silly to imagine they could do it silently. There is no way a full-grown man in combat gear could slink around silently in an unfamiliar building while carrying three full-sized firearms, particularly when the inhabitants are bored, jumpy, and intimately familiar with the space. And if the place is made of crumbling concrete and creaking wood? And everyone spends their days worrying about monsters and ghosts? No way.
But none of this matters. Sneaking around in the dark and knocking people out makes for really fun gameplay. It adds tension. It has a lot of interesting mechanical trade-offs between safety and expediency. It usually adds another whole dimension to what would otherwise be a monotonous shooter. There are already well-established rules for how these systems work and most players have kind of made peace with these contrivances. It’s understood and expected that guards will walk predictable routes, that they will talk to themselves to communicate their current mental state to the player, and that they will drop back into patrol mode after only a minute or two. None of this is realistic, but it’s unrealistic in a familiar way and so it doesn’t shatter immersion the same way that unrealistic but unexpected behavior will.
The point is: It would be incredibly difficult to make stealth realistic, and if you somehow succeeded then it would probably just make stealth gameplay impossible or boring.
So any would-be game designer has to design their game knowing that somewhere, SOMETHING is going to not make any dang sense. Some part of stealth has got to be a silly contrivance.