2005 F.E.A.R. AI and physics, almost ten years without advancement
2005 F.E.A.R. AI and physics, almost ten years without advancement
Initially, I'm sorry for the long block of text. I'll try adding some visual cues to improve readability and highlight the points that matter. Playing Shadow Of Mordor and Alien: Isolation sparked my thoughts. I was hopeful about gaming's future after enjoying FEAR, but it ended up being an outlier rather than the standard. I wish it had become more widespread, yet games released since haven't matched its AI sophistication. Usually, increasing difficulty doesn't mean smarter AI—it's either a bunch of simple bots or a slight boost in health stats to make opponents slightly tougher. More often than not, this leads to frustration instead of challenge, forcing players to exploit AI flaws. The result is rarely satisfying; dying because you couldn't hit the button five times in a row or breezing through a tougher game isn't ideal. FEAR wasn't the first to offer enemies with good field of view, but it handled that well and added many other strengths. It didn’t rely on clumsy bots, unlike some titles that used severe visual distortions or late-stage vision loss. For me, FEAR marked a step forward in AI depth. If you face a squad of six and quickly eliminate five, the last one might quit. When they engage, they’ll use the surroundings—crouching, hiding, and using objects for cover. With two or more enemies, they might fire suppressive rounds to help flank others. Killing one can make the next hesitant to follow. The tracking felt realistic; enemies used flashlights effectively, unlike many games where NPCs ignore your moves. Even in HL2, which had better AI than most of its era, it fell short compared to other aspects like dialogue. The script was trimmed by 40% due to technical limits. Here’s a snapshot showing just how extensive the work was: NPCs spoke a lot, taking damage wouldn’t be noticeable, and paper could be torn apart if you wanted to. Back to AI—Monolith used STRIPS from MIT. As a casual player and fan of a nearly forgotten game, I’m puzzled why it wasn’t adopted more. Check out the detailed document here: [link to PDF].
I understand, right? It seems like developers might prefer making enemies easy targets for bullets and the AI’s weak sight is a hassle in the metro games. Sure, Darkness is helpful when you’re hiding, but it shouldn’t make you completely invisible if you’re just a bit to the left of someone...
The latest Alien AI game features an extremely "unpredictable" AI, which suggests they're heading in the right direction.
I’ll just wait for a sign and hope it shows up in more games. (I’m sure I’ll be the one getting hit for that comment.) After all, I’ve never really enjoyed those alien films. *Gasps of shock horror from the crowd and lightning flashes behind me*
Did you realize Half-Life 1 featured cutting-edge AI for its era? It included Bull Squids displaying defensive behavior, only engaging when threatened, and soldiers coordinating tactics while you battled them. They also adapted their strategies after taking down some enemies.
I enjoyed the AI in that game. It kept me entertained even though I struggled to complete it, and I expressed my frustration a few times.
Are you from Canada or a Canadian? Keep it up to retain your forum privileges.
Fear generated intense particles but the science fell short. Objects would fly off when stepped on and it felt like a whirlwind. Everything seemed to lack sufficient weight. I believe Half-Life 2 sets the benchmark for realistic physics. Garry's Mod likely wouldn't match that quality on any other platform, especially the classic version where you constructed rather than dodged in an office setting 