6 ms·

F.E.A.R is definitely one of the top 10 best PC games I've ever played. There are not many games that blend Action and Horror so delightfully (other than Half-L

F.E.A.R is definitely one of the top 10 best PC games I've ever played. There are not many games that blend Action and Horror so delightfully (other than Half-Life). That being said, I haven't been around in the gaming scene for a while and playing catch up after recently building a gaming rig. I do agree with the author about the lack of progress in FPS enemy AI. The recent releases from the Call of Duty franchise are nice games, but the AI enemy combatants look dumb next to their FEAR counterparts. Why has not FEAR's AI strategy been adopted by other games/studios? The article does not mention anything about this.

NikolaeVarius 10y ago
Because FEAR's AI was the product of an extremely focused game with very little "unpredictable" elements. It's easy to seem awesome when you have just do extremely little well. I would argue STALKER has the best AI in any FPS.
- boyter 10y ago
  What I loved about STALKER is that it felt like a living world. I remember watching a group of fellow stalkers being beset by bandits. During the whole firefight I was just watching it from a safe position. The fight ended and the stalkers continued to walk in the same direction they were following before. I did nothing but observe the action which was unscripted. I really got the impression that a million little actions were happening around me even if I was not there to observe them.
  - NTripleOne 10y ago
    Have you played any mods that expand the ALIFE system? There are some that run the entire ALIFE simulation in every map all the time, groups of stalkers can be having gunfights, trading, looting corpses, putting things in containers, etc etc - so every time you revisit an area after not being there for a while, people have moved, inventories have changed, piles of mutant and stalker corpses make appearances, all sorts. It really keeps the game fresh and exciting.
    - lovehashbrowns 10y ago
      Great, thanks. You just made me want to play STALKER again. :\
      - ionised 10y ago
        I play through all three games maybe once a year. They are easily my favourite gaming experiences.
  - Pica_soO 10y ago
    They had a AI-driven ecosystem in mind, but scraped it. The agents, just snowballed the whole Eco-system.. meaning.. a mutated dog chases a rat, both run into a mutated boar, thus the chase cascades through the level.. until some incredibly strong monster hunts a tsunami of mutated boards towards the player. Which is cool.. but not fun.
    - bryondowd 10y ago
      Reminds me of Everquest, circa 2000, when a player got in over their head and started a 'train to zone'. Enemies would chase you to the end of the map/zone, so when a player realized they couldn't win a fight, they'd beeline for the boundary, and often picked up dozens of other enemies along the way. When they crossed the zone boundary, all those enemies would suddenly have no target, and fixate on any poor soul who happened to be nearby, or along their route back to their spawn area. Haven't thought about that in a long time.
- JustForFun 10y ago
  Well, actually, STALKER used GOAP as well: https://aigamedev.com/open/interviews/stalker-alife/ https://aigamedev.com/open/interviews/stalker-alife/ Enjoy. -- JustForFun.
  - dkersten 10y ago
    Oh, I had forgotten about this! Thanks for posting the link (I actually read this article back in the day). The STALKER series was my all time favourite shooter series, and I'm sad that STALKER 2 never saw the light of day :( PS, a friendly note in case you didn't know: people typically don't sign comments on HN (as the guidelines say, your comment is already "signed" with your username)
    - ionised 10y ago
      There is a slim chance STALKER 2 will happen, as Sergey Grigorovich reopened GSC Gameworld in 2014, making Cossacks 3. When asked about STALKER he said that it is a possibility provided GSC can make a successful (and profitiable) go of it again, giving them the resources to make a true STALKER sequel. On the plus side Grigorovich resisted all attempts to take over the STALKER IP, I'm convinced he wants to do it right, if it is going to be done at all. On the other negative side, most of the original STALKER devs have joined other companies in the time between GSC first shutting down and reopening.
      - dkersten 10y ago
        Thanks for that information! Hopefully something will come of it, although as you say, the original revs won't be working on it... Still, I'm hopeful.
sigmar 10y ago
>I do agree with the author about the lack of progress in FPS enemy AI. I believe online-play preempted advancements in game AI. People will play competitively online if they want a challenge. Single player mode in FPS games these days seem less geared towards being a challenge and more geared towards being cinematic.
joe_the_user 10y ago
I do agree with the author about the lack of progress in FPS enemy AI. I follow AI much more than I following gaming. I definitely find it curious that enemy AI has hasn't made a lot of progress when game-playing in a different form is such a focus of deep learning.
- azernik 10y ago
  Game AI tends not to use machine learning approaches - it's all hand-coded strategies.
  - e12e 10y ago
    I suspect machine learning (and other, older ai techniques) might work for games - but one would probably have to train them for the right goal: fun. I suppose for some players and some games, it makes sense to train for "good" - but even with chess - is it really fun to play an ai you cannot realistically beat - ever?
- stevenwoo 10y ago
  I haven't worked in video games since the PS3/XBox360 but everything was bespoke on every platform for every game except for maybe low level audio in my ten years on PC/Mac/PS3/XBox360 and I had to review/examine code for games other than the titles I worked upon. Even games that used game engines/physics engines have heavy customization. I only worked at smaller studios but the practice of starting from almost scratch every time can make it hard to have a standard toolkit of AI, the reuse incentive was low.
- Radim 10y ago
  1. HW resources: games run on extremely tight HW budgets. It's probably the most competitive, cut-through industry in terms of what you can afford to do with a given CPU/memory/bandwidth per frame. While DL is still in the "let's throw a thousand GPUs at it, wait for a week, see what happens". 2. Quality control: DL is a blackbox. Bad for the playing experience for obvious reasons. And even if just using pre-trained models, the lack of control goes directly against what both the designers and players typically want (predictable fun). 3. Finally, DL hasn't made THAT much progress here, actually. There have been serious efforts to apply DL to modern games (e.g. DeepMind with Starcraft), but what you've probably seen is "just" DL applied to some finite-state, 100% deterministic, complete-information, zero-sum games. A far cry from complex dynamic environments. Which is not to say it cannot be done. Just that the incentives (both business and technological) are aligned differently -- too little bang for a lot of buck.

NikolaeVarius 10y ago

Because FEAR's AI was the product of an extremely focused game with very little "unpredictable" elements. It's easy to seem awesome when you have just do extremely little well. I would argue STALKER has the best AI in any FPS.

boyter 10y ago
What I loved about STALKER is that it felt like a living world. I remember watching a group of fellow stalkers being beset by bandits. During the whole firefight I was just watching it from a safe position. The fight ended and the stalkers continued to walk in the same direction they were following before. I did nothing but observe the action which was unscripted. I really got the impression that a million little actions were happening around me even if I was not there to observe them.
- NTripleOne 10y ago
  Have you played any mods that expand the ALIFE system? There are some that run the entire ALIFE simulation in every map all the time, groups of stalkers can be having gunfights, trading, looting corpses, putting things in containers, etc etc - so every time you revisit an area after not being there for a while, people have moved, inventories have changed, piles of mutant and stalker corpses make appearances, all sorts. It really keeps the game fresh and exciting.
  - lovehashbrowns 10y ago
    Great, thanks. You just made me want to play STALKER again. :\
    - ionised 10y ago
      I play through all three games maybe once a year. They are easily my favourite gaming experiences.
- Pica_soO 10y ago
  They had a AI-driven ecosystem in mind, but scraped it. The agents, just snowballed the whole Eco-system.. meaning.. a mutated dog chases a rat, both run into a mutated boar, thus the chase cascades through the level.. until some incredibly strong monster hunts a tsunami of mutated boards towards the player. Which is cool.. but not fun.
  - bryondowd 10y ago
    Reminds me of Everquest, circa 2000, when a player got in over their head and started a 'train to zone'. Enemies would chase you to the end of the map/zone, so when a player realized they couldn't win a fight, they'd beeline for the boundary, and often picked up dozens of other enemies along the way. When they crossed the zone boundary, all those enemies would suddenly have no target, and fixate on any poor soul who happened to be nearby, or along their route back to their spawn area. Haven't thought about that in a long time.
JustForFun 10y ago
Well, actually, STALKER used GOAP as well: https://aigamedev.com/open/interviews/stalker-alife/ https://aigamedev.com/open/interviews/stalker-alife/ Enjoy. -- JustForFun.
- dkersten 10y ago
  Oh, I had forgotten about this! Thanks for posting the link (I actually read this article back in the day). The STALKER series was my all time favourite shooter series, and I'm sad that STALKER 2 never saw the light of day :( PS, a friendly note in case you didn't know: people typically don't sign comments on HN (as the guidelines say, your comment is already "signed" with your username)
  - ionised 10y ago
    There is a slim chance STALKER 2 will happen, as Sergey Grigorovich reopened GSC Gameworld in 2014, making Cossacks 3. When asked about STALKER he said that it is a possibility provided GSC can make a successful (and profitiable) go of it again, giving them the resources to make a true STALKER sequel. On the plus side Grigorovich resisted all attempts to take over the STALKER IP, I'm convinced he wants to do it right, if it is going to be done at all. On the other negative side, most of the original STALKER devs have joined other companies in the time between GSC first shutting down and reopening.
    - dkersten 10y ago
      Thanks for that information! Hopefully something will come of it, although as you say, the original revs won't be working on it... Still, I'm hopeful.

sigmar 10y ago

>I do agree with the author about the lack of progress in FPS enemy AI. I believe online-play preempted advancements in game AI. People will play competitively online if they want a challenge. Single player mode in FPS games these days seem less geared towards being a challenge and more geared towards being cinematic.

joe_the_user 10y ago

I do agree with the author about the lack of progress in FPS enemy AI. I follow AI much more than I following gaming. I definitely find it curious that enemy AI has hasn't made a lot of progress when game-playing in a different form is such a focus of deep learning.

azernik 10y ago
Game AI tends not to use machine learning approaches - it's all hand-coded strategies.
- e12e 10y ago
  I suspect machine learning (and other, older ai techniques) might work for games - but one would probably have to train them for the right goal: fun. I suppose for some players and some games, it makes sense to train for "good" - but even with chess - is it really fun to play an ai you cannot realistically beat - ever?
stevenwoo 10y ago
I haven't worked in video games since the PS3/XBox360 but everything was bespoke on every platform for every game except for maybe low level audio in my ten years on PC/Mac/PS3/XBox360 and I had to review/examine code for games other than the titles I worked upon. Even games that used game engines/physics engines have heavy customization. I only worked at smaller studios but the practice of starting from almost scratch every time can make it hard to have a standard toolkit of AI, the reuse incentive was low.
Radim 10y ago
1. HW resources: games run on extremely tight HW budgets. It's probably the most competitive, cut-through industry in terms of what you can afford to do with a given CPU/memory/bandwidth per frame. While DL is still in the "let's throw a thousand GPUs at it, wait for a week, see what happens". 2. Quality control: DL is a blackbox. Bad for the playing experience for obvious reasons. And even if just using pre-trained models, the lack of control goes directly against what both the designers and players typically want (predictable fun). 3. Finally, DL hasn't made THAT much progress here, actually. There have been serious efforts to apply DL to modern games (e.g. DeepMind with Starcraft), but what you've probably seen is "just" DL applied to some finite-state, 100% deterministic, complete-information, zero-sum games. A far cry from complex dynamic environments. Which is not to say it cannot be done. Just that the incentives (both business and technological) are aligned differently -- too little bang for a lot of buck.