It looks like you're new here. If you want to get involved, click one of these buttons!
I could be wrong, but it seems like Mr. lizardbones just touched a quite important subject.Is it possible to build engaging and fun encounters with dumb AI? AI is another name for NPC brain. Dumb AI means –dumb NPC. Is it possible to build engaging and fun encounters with dumb NPC? Yes, yes and yes! Just make him stronger!
Generally we are looking not for some qualities in NPCs. We are looking for a challenge. NPS with advanced AI is a challenge. But strong dumb NPS is a challenge too. Just for very short period of time. Then those encounters will become very boring. They become repetitive. Because this is what dumb AI is. Dumb AI (by its definition) besides other limitations has very limited number of possible variation of behavior. So, dumb AI also influence game replay ability.
It is not a secret that progress in game AI development and utilizing was significantly deprioritized by development of PVP, which in some degree satisfied hunger for tactical challenge.
It is actually not about what exactly some NPCs would do in response to players actions, or in anticipation to potential players threat. Scenarios could be developed to your liking. It is about being challenging enough, but not overpowered. Or you will have situation Mr. MMODesigner has described.
Actually modern state of AI capabilities is quite high. Higher than capabilities of average persons. So, what the problem? When you fight another player, you expect him to be not very different from you in mental capacity. Meaning, you expect to win because of your skills.
Different players might have different skills (and not only skills in a game). Simple AI is not operating on skills level. It operates on a simple tactical level (by analyzing of outcome of multiple scenario variation (it is not absolutely accurate, but will do for this discussion)).
Whatever your attempts are to analyze what enough advanced AI will do in some situation, you will fail, because even dumb AI is enormously faster than you are, and capable of producing right (on its level of dumbness) decision in real time. It is very hard to balance that kind of AI with real people capabilities. Besides, real people make mistakes. We love them for those. We can defeat them because of those. AI may be dumb, it may be predictable, but it doesn’t make unintentional mistakes. And it is boring.
General (most common) direction in AI development (not in a game industry, but in all AI science) was to make AI as advance as possible, potentially better than human intelligence. Not to make AI to be precisely equal it. This is what we need to make fight challenging and interesting.
So, on my opinion, any hopes that improvement of AI will improve our game experience in nearest future, just hopes.
I mainly played defenders, and only rarely ever farmed missions or participated in "cheaty" stuff like power leveling a lowbie or whatever. As such, the "Burn nerf" never really affected me directly nor did it bother me. In fact none of the game balance stuff they ever did bothered me one way or the other. This might be because it was mostly beneficial to the defenders I guess. I will say that the "smarter AI" crowd isn't necessarily the PVPers, at least in CoH. There were people who had a fire tank just to generate swag then used a different toon for PVP. These people complained when the nerfs came down because it meant they couldn't as easily corrupt PVE to their personal swag engine anymore. These were the type of people who would complain "Aw, now you're making me go back and reroll a new guy because my old money/XP machine got dismantled, dangit, now I have to play MORE PVE that I hate just to generate enough swag to support my PVPing that I love. This sucks. " My response to those people at the time was to quote Judd Nelson from The Breakfast Club: "BEE OH OH, AITCH OH OH" Listening to the babies who don't want their "automatic win" button taken away is never the right answer. They'll complain when you nerf stuff, but that only proves they're invested. If it all becomes a cakewalk, they'll quietly stop playing altogether and go find another game. As for people not wanting a challenge, I would say that not EVERY mission or game experience needs to be totally challenging, but also, in the last few months of CoH, I regularly participated in repeated Triumph Server-wide attempts to get the "The REALLY Hard Way" badge. We were only successful once, but we had a great time doing it. Some stuff should be that challenging.
"Well sure, the FrinkiacVII looks impressive - DON'T TOUCH IT - but I predict that within 100 years computers will be TWICE as powerful, ten THOUSAND times larger, and so expensive that only the five richest kings of Europe will own them." -Prof. Frink
The issue with standard tanking is that it is pretty unrealistic for enemies that are not dumb. I like the concept for druaga tower that it is about the tank as a blockers. For example, narrow corridor or door tanking partially works that way.
An interesting AI would bring a less scripted boss encounters.
Boobs are LIFE, Boobs are LOVE, Boobs are JUSTICE, Boobs are mankind's HOPES and DREAMS. People who complain about boobs have lost their humanity.
Originally posted by Ozmodan Good AI takes a lot of coding and lots of memory. While today's computers are pretty robust, not sure they can yet handle good AI, besides that would take significant resources from the programming staff too.
This sounds like it would make a good game in itself - give players a little scripting language and a finite amount of memory/clock cycles, then have their AI-driven parties of characters duke it out in an arena.
We need smarter PvE opponents, hell most MMOs I play need smarter PvP opponents.
The Holy Trinity is the best PvE mechanic devised, but like any mechanic it forces players into a certain style of gameplay. Smarter mobs can exploit the HT to get an advantage, this can cause the HT gameplay real issues. So players want the mobs to be dumb. If there was a tactical response that the players could make to the moves of smart mobs I think they would be OK with it.
In the example given you had to have a Controller to be able to have a tactical response to the smart mobs moving out of the fire. That's a problem as not every team would have a Controller. This could have been solved by having other players "cover" the tank. The other players close of gaps in the flames around the tank simply by having LoS through a gap to the tank, they just move to get that LoS. Or you could have a group mechanic like a finishing move which held mobs already in dot zones.
We don't need the HT going ahead in MMOs, but if that's not used then we need a replacement. The blitzkrieg tactic of a player group with no structure is simply not as satisfying.
You received 25 Agrees. You're posting some good content. Great!
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Now Doesn't That Make You Feel All Warm And Fuzzy? :P
I am all pro Holy Trinity. But the division actualy is :
1. Crowd control - tank, debuff, curse etc ... whatever fits the enemy
2. Support - heals, buffs ... again whatever works in the given situation
Thus the Holy Trinity still exists, but the roles are filled by different classes agains different enemies. In your case, the crowd control and damage was the same character, so it was advantaged. Making the mobs smarter split the roles in 2 again (tank/cc could not do damage) thus you got a pushback :-)
I am all for good AI. Make the enemies less in number and more varied in group composition and skill/tactics used. I fail to see why mobs don't generaly call for help (invalidates pull tactics most of the time) or change targets (classic tank is out without other CC means) or change tactics. That is not hard to implement but lowers the fun of the more straight forward thinking players.
When I can find one such MMORPG that combines all the above, I'll buy a lifetime sub in an instant (I am not interested in KillX/FetchX quests, I can do with the grind if enemies are interesting and challenging).
A rather unimaginative piece to be honest.
Albeit with some grain of truth, if you believe the trinity is synonymous with stupid AI then it's indeed not the artificial kind of intelligence you need.
Just take a few minutes to imagine. You're an MMO designer. You've got classes that fit into the trinity. You're tasked with coming up with interesting, challenging and intelligent enemy tactics whilst keeping the trinity intact. What do you do?
If you really can't come up with anything at all then I'd say that your own lack of creativity, not any flaw of the trinity.
Do note, none of this implies that the trinity is the best combat system. It's simply a combat system among several others. Whether or not you have a trinity system does not determine how interesting, challenging and intelligent your combat is, it all comes down to what the developers do with that system. A good dev can make great trinity systems and great non-trinity systems. A bad dev will produce a boring combat system either way.
We are the bunny.Resistance is futile.''/\/\'''''/\/\''''''/\/\( o.o) ( o.o) ( o.o)(")("),,(")("),(")(")
In CoH it wasn't a simple matter of "Dumb AI = fights are winnable, but too easy" versus "Smart AI = fights are nearly impossible, or actually impossible, depending on your team". There were difficulty sliders you could set that would dictate the level and number of badguys that would spawn. So the nerfs to Burn and Taunt were mostly just to take away the ability of those tankers to just set the "mob size" control to "max" then aggro an entire map full of Freaks and immolate them all. When you get to the level of being able to "batch process" an entire map with one particular build, it's a problem. If anything I think the AI might have been handled best as a difficulty setting. If nothing else, the little kids and casual players could set it to an easier setting and the powergamers could set it to the max just for the challenge. I mean, it was certainly true that some people really made a point of maxxing out their build, whereas others didn't have the time or inclination to bother getting every Very Rare thing they might need, or even know what "the best" stuff even was. I certainly didn't, and by the end I just about had the Influence for it if I wanted to (CoH currency was Influence, not gold). I'd also like to point out that the Lord Recluse TF may have suffered from a too-smart AI, I wouldn't know because I never played it, because I never got my one villain to a high enough level to do so. I think people would still have done TFs even if the AIs were smarter than the average bear, but only if it were consistently like that for all such content, and if the rewards were proportionally higher to make up for the added headaches. As it was, it was hard enough to keep a TF together long enough to finish it for a lot of them (see "Dr. Quarterfield TF" in the Shadow Shard) so people tended to start more of the shorter ones (the "Quicky" being the Katie Hannon TF, for instance) because the rewards were the same anyway, regardless of the length or difficulty. They eventually changed that to some extent, but even the difficulty settings you got on the TFs didn't include smarter AI as far as I remember and they didn't give added rewards for anything you could set, I don't think.
GOD I miss that game.
I'm more in the middle.
I dislike farming and feel your first solution was correct.
But that being said I play games to have fun and I don't think continually having to chase down mobs after I engaged them is much fun either.
So maybe one or 2 of the group are dumb enough to keep attacking and the rest path as you described in your first solution.
Just My 2 Lunars
Originally posted by MMODesigner After this article hit, one of the CoH designers reminded me about the Lord Recluse Strike Force AI that was created; the Arachnos Lieutenants were efficient and brutal, targeting healers and avoiding taking damage. We thought it would be a nice challenge for players to figure out how to beat it. Instead they just never played it because it was felt to be unfair, the enemies didn't act like any other enemies in the game. We eventually relented and put the AI back to more traditional aggro tables and the players were happy again.
I was wondering, what incentives did you provide the players to tackle and master this more challenging scenario? Did beating them provide some special title, reward, item etc?
Most MMORPG players are strongly motivated by progression, and will generally gravitate to activities that provide the greatest reward for the least effort/risk.
They don't play these games for "fun" or challenging game play (OK some do) but in general you have to provide extraordinary reward for extra challenge/effort/risk.
If this wasn't part of the design, no wonder the players wanted it changed to be traditional, no reason for them to do this otherwise.
"I need to finish" - Christian Wolff: The Accountant
On hiatus from EVE Online since Dec 2016 - CCP continues to wander aimlessly
In my day MMORPG's were so hard we fought our way through dungeons in the snow, uphill both ways.
Don't just play games, inhabit virtual worlds™ "This is the most intelligent, well qualified and articulate response to a post I have ever seen on these forums. It's a shame most people here won't have the attention span to read past the second line." - Anon
Matt as much as I loved city of heroes and I did love it, five year subscriber here, there are some things that was messed up from the beginning about that game. This was one of those things. How in the hell that mechanic made it through internal testing is beyond me to comprehend but the better solution was to make mobs run out of stuff that would hurt them which if I remember correctly you guys did eventually do as well as limit how many targets a tank could hit. The real question I always wanted to ask is why you guys let Jack Emmert push Enhancement Diversification through right after a huge nerf to all powers. That nearly killed the game and in my personal opinion was the killing blow that caused the games demise although it did take a long time for the game to bleed out.
Anyways how come people cant like trinity yet like smarter mobs. It is not always as simple as just go after the healer. I go back to my pen and paper dnd games where the npc mobs was controlled by a human person. Our dm would always go after the healer even if we blocked off the healer and was beating on those mobs, to me that is just another form of stupid ai and I would tell him so. My whole argument was if I am sitting here beating this person or mobs face in he is not going to leave me alone to go after someone that is not even doing damage to him. This to me is where threat mechanics come in to make smarter AI. If a healer gets too much threat then the mob will target him but as long as he is seen as not a threat then the mob wont but the mob shouldn't and wont just sit there and die to fire just because it is supposed to be attacking that person. At that point unless someone is making the mob stay there through the use of hold or another control type power then that mob should make an effort to get out of the fire.
I played as a Fire/Ice Tanker, and the reason I disliked the burn-fear was that it was a rather narrowly selective application of the mob AI getting "smarter", so it came across as half ass and vindictive. If the AI was truly smart, they'd never hug up around and Invulnerability Tanker, nor keep trying to run into a Storm user's Hurricane, nor would they stand in any other location based AE that decreased their chances of winning.The fair thing barring a global AI upgrade of that kind would be to simply tweak the damage to make burn herding less OP instead of flat trying to invalidate it.
Edit: That being said, I did like how ruthless the Gw1 AI was in that it did a simple check of who was close, who was low armor, and who was low hp. So if you tried being a stereotypical frail healer xp leech you got gobsmacked. It was fucking glorious.
Originally posted by hardicon The real question I always wanted to ask is why you guys let Jack Emmert push Enhancement Diversification through right after a huge nerf to all powers. That nearly killed the game and in my personal opinion was the killing blow that caused the games demise although it did take a long time for the game to bleed out.
Wow. The game's been dead for almost a whole year and people are STILL mad about Enhancement Diversification, which happened in like 2004. I lived through that and I have to say, at the time as after, I didn't think it was a big deal. It was pretty obvious at the time that people were making a mockery of the enhancement system, a la "I know, I'll put ALL damage SOs in my attacks, I'm a GENIUS." and for there to be any reason why a person would put anything else in an attack power, they had to do something. It had gotten ridiculous. For what it's worth, I thought what they did was fair. Not even "tough, but fair", just fair. Whatever they do to tweak a game in any way, it will usually cause people to have to rebuild toons that were optimized for the last wave so as to be re-optimized for the new wave. So what, deal with it. It was a minor inconvenience for the sake of better overall game play. It was necessary. What I find comical is the implication that once one has optimized a toon, they then feel they have the God-given, sacrosanct, etched-in-stone, legally-binding, non-negotiable RIGHT to that build being "teh uber" for the rest of their life, and their children's, and their children's children's. I would have expected those same people to be just as mad when they rolled out Invention Origin Enhancements. THAT'S causing one to have to scrap all your builds and start over, but nobody seemed to complain about that. Why? Because this time they weren't so much "taking all our toys away", they were making newer, shinier toys that made us want to abandon the old ones. I guess I understand a certain amount of "how did they not see this coming" but really, companies don't have infinite time to tweak the game before they release it. I think CoH actually had MORE play-testing than a lot of games today get before release, and if you recall, the Task Forces needed some TLC in the first month or two. Back in 2004 people might have had higher standards, but nowadays, I don't think anyone expects a game not to have some balance issues in the first year or so. The fact that there was as much content as they had actually impressed me. I disagree that ED killed CoH. Those of us who stayed generally stopped complaining about it in like 2005, and those who still complained about it were complaining about it in CoH chat channels in between "LF TF" requests and costume contests. Inventions, which were awesome, made it largely moot anyway (and by largely moot I mean that after IOs, nobody was using SOs anymore anyway, so the original complaint about ED was no longer anything anyone cared about anymore). One thing I found almost funny was that they managed to redo the HamiO's like RIGHT before the shutdown was announced. It was as if the game HAD to live long enough to see that loophole closed, THEN it could die.
I like more intuitive AI. It's one thing I really liked about Guild Wars - the mobs pretty much played like a simple player rather than a predictable script.
I know the pro-trinity guys hate the shot taken at the end of the article, but it's true. Trinity AI is predictable. It has to be, or the trinity breaks.
And just to point out, you can have clearly defined roles (healer, damage sponge, control, support, etc) and still not have the trinity or have mobs with simple trinity AI. Again, Guild Wars, a non-trinity game, has all those roles and a superior AI for its generation. I'm really hoping EQ:N will have interesting mob behavior.
@Matt Miller(even though it's unlikely to be read by you):
Why did you capitulate to the players demands for the more simplistic AI? Why didn't you stick to your original instinct and have the mobs move out of the fire? It makes sense and even other trinity games I've played (LotRO is the example coming to mind) have made changes so mobs don't stand in pools of roiling death.
You don't have to answer directly, but I hope you will address this in a future column. It's not meant as a pot shot either as I never played CoH. I'm just really curious what leads to that conclusion over the one you wanted? I think your original idea was great.
edit: I did see your post above (if you're MMODesigner). So to clarify, why not make the entire game AI more intelligent rather than go with the simplistic trinity behavior? Is there really that much of a risk involved with improving it? Is there a real fear that the backlash could be huge (NGE/CU style)? Are MMO players that inflexible? That would seem odd to me in a genre where evolution is foundational.
Originally posted by Torvaldr
@Matt Miller(even though it's unlikely to be read by you):Why did you capitulate to the players demands for the more simplistic AI? Why didn't you stick to your original instinct and have the mobs move out of the fire? It makes sense and even other trinity games I've played (LotRO is the example coming to mind) have made changes so mobs don't stand in pools of roiling death. You don't have to answer directly, but I hope you will address this in a future column. It's not meant as a pot shot either as I never played CoH. I'm just really curious what leads to that conclusion over the one you wanted? I think your original idea was great.
I'm no Positron, but I'll give that question a try anyway. First, the badguys DID try to get out of AoEs (eventually). Some/most AoE damage powers got added "and it slows the affected badguys in the area" added to theor effects to compensate though (at least Rain of Fire, my fire blaster's one AoE got that). Second, if you completely ignore the response of the player base enough times or for too long, or as a general policy, you won't have one for very long, and that means going out of business. The fine people who made CoH as GREAT as it truly was wanted to keep doing that for a living, and I don't blame them. Third, simply making the thugs not stand there and burn to death like morons wasn't really enough. If your Tanker could hold and entire map's aggro indefinitely, it was still unrealistic that all the badguys would be falling all over each other trying to follow him around. The aggro cap or whatever it was made things more realistic, I thought. The only real crime in the Taunt + Burn = I win nerf was that Burn, the small, immobile DOT power, was made almost completely useless, at least for Tankers. A Blaster or Defender could have used it to keep melee mobs at arms length when shooting at them, so as a defensive stand-off kind of deal it would have worked. But Tankers didn't need or want that, so that Burn power became a thing nobody ever took anymore. I don't remember if they replaced it altogether later on or not (I never made a Fire Tanker).
City of Heroes was just a bad game, sorry. The mobs just stood there, the combat was clunky. You could play it for an hour and tell it wasn't going to do well without changes. WoW had nothing at all to do with it.
To say trinity = simple AI and nontrinity = advanced AI is where You loose me. That is not a question of trinity at all. Even the simplified trinity of today is much more advanced than what You describe, and old "trinity" models in for example EQ were much more complex, although still mostly based on aggro numbers, was also calculated on distance, faction, burst threat (afaik) and other factors. What I am saying is trinity is just roles to play, and is not clashing with the making of intelligent AI, in fact several games based on trinity has intelligent AI such as an enemy switching to ranged to attack a caster. Also the attempts at non-trinity so far, has not more intelligent AI than Your trinity combat, not any I have seen anyways
But I do get the point, and an excellent example with the fire-patch kiters. I wonder if it would be possible to combine the need for expected behavior from some or the need for interesting behavior from others, or it is a matter of serving the right AI for the right customers. Personally I would like to see more games that focuses more on one type of players, instead of trying to please everyone. Niche is what we need, and I think it is possible, because when You go niche You also simplify all the game concepts and hence development cost (totally a theory that I cant back up in any way haha).
"I am my connectome" https://m.youtube.com/watch?v=HA7GwKXfJB0
I can understand where the firetankers are coming from. It is like giving them a bunch of powers then making the enemy immune to them and doing so after they have invested time into that characters. : For those read that literally and are about to complain about it, pull your head out of your butt before replying.
IMO, it is bad dev work. Not making needed changes for the good of the game but not creating the proper thing during the design phase and also not catching it during testing. Smoke was broken and should never have been in beta in the state it entered release in. It is often too simple for devs to adopt the we can fix it later rather than doing it right mindset. Poor planning leads to poor results.
You have to remember, it is easier to give than take away.