Check In: Did tech-support today, which may require me to keep short my blogging. My brain has a hard time shifting gears between composing copy and troubleshooting technical failures.
Introspective: The Thief 2014 thing from yesterday is still bothering me. More examples occurred to me of game elements that suggest the developers really didn’t want to make a Thief game or a game about stealing in favor of one that was either less niche or appealed to a wider audience…but I’m too angry about it today to turn my growing conspiracy hypothesis into readable prose.
Yesterday I discussed how the guards in Thief 2014 were particularly accurate with their stone-throwing. I was really angry. Validly so, I think. But more at the stupid corridor than the stupid rocks. But as a matter of fair assessment, overly accurate rock-throwing is neither a unique problem to Thief 2014, nor a new problem. In fact, it’s one of the most common problems in 3D shooter games, right up there with death by ladder.
In STALKER: Shadow of Chernobyl (2007) I was pinned down by a distant sniper. I was both distressed and exhilarated because I was determined to acquire what was assuredly going to be my first sniper rifle in the game. Much zigging and maneuvering and distracting and sneaking later (not to mention a few game-reloads) I finally got close enough to him to shoot him with my one handgun, after which he dropped… a poorly maintained handgun. His weapon was worse than mine. STALKER saves some cycles by not taking into account the inaccuracy of handguns versus rifles when in the hands of enemies. Enemy weapons also perform as new no matter how bad their condition. I’d eventually come across some sweet rifles and sweet scopes to mount on them, and found I could usually take out enemies before they even knew I was around, but my most pronounced STALKER moment remains the pistol-sniper.
Even in Far Cry 2 (2008) there are instances where one enemy’s knowledge of your location translates into all the enemies pinpointing you, including the guy on another island with a light mortar. …who’s accuracy is uncanny. He’s close, man. He’s real close. (Link is NSFW) The smoke marker landing right next to me is usually the first indicator my day has just become super complicated.
Some games that focus on tactical detail take into account weapon accuracy and gunman skill*, but as we’re seeing, this is still not the norm. Fortunately, it’s a common exception. But additional factors take additional time. Precious, precious time. The computer is trying to project and render a three-dimensional world full of objects and make the objects respond according to consistent mechanics and even give some of those objects agency and intelligence enough to behave like a living thing. A long term gripe of mine has been that AAA games often focus on better, shinier graphics when I’d personally rather see humongous levels, smarter baddies, consistenter physics, or casts of thousands (of extra zombies). The AAA industry is pressured not by their audience, but by platform clients to utilize graphics hardware as best as they can. Hence we’ve seen shiny graphics take priority even at the expense of level size and smart foes**.
It was notable that the levels of Thief, the Dark Project and Thief II, the Metal Age were significantly larger than those in Thief: Deadly Shadows which were accounting for better graphics and console memory limitations. Thief 2014 levels are also smaller, and while its detail may seem more atmospheric, lost is the ability to pick up any object, and even play basketball.
More generally, we’ve seen in AAA gaming a development focus on bold, brilliant details that draw attention, rather than small details that contribute to the game’s internal consistency. I’ve talked about this before. And for now I’ll just have to keep gushing about lovingly detailed games with swoonable moments to encourage developers, AAA, indie or otherwise, to continue making the sorts of games I like.
* Actually, it’s the reverse. The game takes into account gunman and weapon inaccuracy. When a bot gunman wants to shoot a thing, the game computes a direct hit from the target back to the weapon to create a perfect muzzle vector. Then it adds deviations, first based on skill (what he’s actually shooting at) and then based on firearm spread (where he’s finally pointing) which creates a final not-so-deadeye muzzle-exit vector. Games with fancier mechanics will take into account parabolic arc (bullet drop), wind and intervening mediums when shooting through light cover. The upshot of this, then is that you have much less to fear from a guy shooting you with a pistol from across the field. Math!
** Smart foes in this case refers to baddies who can navigate the environment without getting stuck on flaws in the map, or at least can get themselves unstuck. There are some indie projects that look to develop sophisticated squad tactics (suppress and flank, bang and clear, flush and sweep, etc.). But for now most efforts in game intelligence are towards getting units to consistently get from (any given) point A to (any other) point B without fail. Being able to utilize cover in the meantime is gravy.
Smart foes in stealth games refers to bad guys who respond appropriately to the limited knowledge they have. This is tricky, since computers don’t generally make intuitive or logical leaps once perceptive contact has been lost, hence games often have to fake it with a certain amount of blue sense (i.e. cheating). Computers are pretty terrible at hide-and-seek…so far. Also communication paths are unclear even when when accounting for radios to call in artillery.