EQ Never
- Thread starter Convo
- Start date
tad10
Elisha Dushku
- 5,518
- 583
tad10
Elisha Dushku
- 5,518
- 583
They said that, but they also said raid sizes would be "smaller" only way (edit: I can think of) for the second to happen is through instancing.
Mr Creed
Too old for this shit
- 2,383
- 276
Unless you catch it by surprise with nets and shit, any dragon with an ounce of self-preservation will take off and flee when it sees an overwhelming zerg approaching. Dynamic raid caps thanks to AI, griefing opportunities included.
tad10
Elisha Dushku
- 5,518
- 583
Did you mean gald you're not the only one checking out the gonewild subreddit or something else? I assumed every healthy straight male checks that.
/derail I don't know how anyone makes money with porn these days with shit like that.
They have already stated that at the end of an open style dungeon there will be a raid boss which will be in an instance. IE. You get to the end of a diablo3 level and zone into the boss fight. I like the idea that everyone can fight the boss, but that means he's not as epic like bat phone 4am week day epic where ppl will actually wake up and log in. If you make stuff that special people will do it and it makes it that much sweeter when you have to compete with other guilds for the goods. They just need to make them hard as fuck so people cant just kill him with 10 ppl you need 100 people working together to bring some shit down, that is epic as fuck. Screw the 10 man fights that shit is wack and played out. Bring back the carnage where a tail swipe takes out 30 people if you got hit by it and theres no fucking red circles telling you its coming. I have always been for in combat ressing with penalties. Maybe like 1-2 resses ur out and in that time healers can't be healing.. but wait theres no healers wtf am I talking about.
I think one of the original reasons for using threat based mechanics was that they are incredibly cheap in terms processing power required on the server (this was a huge benefit for early MMOs). And after they made that choice they designed the rest of the games around it. So if you just take a regular MMO and remove the ability to tank mobs it will definitely suck. However, if you add stuff like attacks of opportunity, collisions, more effective CC and less effective in combat healing it might work. You probably want to reduce mob hp as well. So mobs would be unwilling to break away from the melee and go for the casters since that would open them up for very damaging attacks from behind. On the other hand, casters/stealthed mobs could go for casters without getting hit in the back... This would be a lot closer to how it usually works in a pen & paper rpg.
Malakriss
Golden Baronet of the Realm
- 12,402
- 11,810
PVP might be more balanced if they actually overhauled defensive capabilities in a non-aggro based environment. The issue with that is the amount of reflexive/reactionary abilities or situational skills would be far more than 8 on a hotbar. Would not scale that well with intelligent AI either. But I would still applaud any game that tries to add mitigation, avoidance, shielding, etc so cloth casters can actually cast something useful besides fireball x1000.
tad10
Elisha Dushku
- 5,518
- 583
http://aigamedev.com/insider/coverag...ides-takeaway/
Dave Mark slide on the right side (have to sign up to see - not worth it IMHO). Heavy, heavy into weighted randomness. Basically, the trick of Mark's AI in these slides is that it will not always do the same thing. It will often do the same thing (ranged attack, melee attack, retreat, whatever) but will occasionally do something totally different.
I don't know how innovative this is - my knowledge of AI programming approaches zero. But would be amusing the one time the Giant Iron Golem actually runs from the lone ranger as he approaches.
Dave Mark slide on the right side (have to sign up to see - not worth it IMHO). Heavy, heavy into weighted randomness. Basically, the trick of Mark's AI in these slides is that it will not always do the same thing. It will often do the same thing (ranged attack, melee attack, retreat, whatever) but will occasionally do something totally different.
I don't know how innovative this is - my knowledge of AI programming approaches zero. But would be amusing the one time the Giant Iron Golem actually runs from the lone ranger as he approaches.
omg. weighted randomness... i'm having flash backs of making my miragent legs /puke 300000 bloodstains later.
I have not read the actual article, but that sounds very much like the AI I wrote for a hobby project. Basically, evaluate a bunch of possible actions. Randomly choose which action that is performed with a heavy bias towards the ones with the best scores. It's also easy to add modifiers to make mobs prefer to not switch target, assist each other, favor targeting low hp players etc. In terms of performance this is significantly more expensive than a threat based AI, but should still very much be viable with current hardware.
Zehnpai
Molten Core Raider
- 399
- 1,245
I, for one, say fuck it. If people want to zerg shit, let them zerg it. Never understood why "Throw a ton of fucking people at it" isn't considered a valid strategy anymore. I've done the 12 man elite hit squad and the 200 person zerg swarm and I can honestly say both felt equally as rewarding. Don't even have to scale most encounters honestly. If EQ was any indication about 40% of a zerg raid force will just die on pull anyways because it's hard enough getting 10 people to pay attention let alone 200.
"But...but...the integrity of the fight and world firsts and and..." blah blah blah. It's 2013, who the fuck cares.
tad10
Elisha Dushku
- 5,518
- 583
That's how I read it - an early point in the powerpont was that AI can be too good (for example go straight for the healer) so they wanted the AI to occasionally make mistakes, or get stubborn (keep targeting the guy in front). Just like a human might.
Goalish AI is pretty easy. My last ludum dare game I did a basic surround and melee enemy, and a nuke / heals / buffs shaman type. It would stay near a melee (which had the effect of running away if it was the last one alive), heal friendlies, and throw fireballs.
The companion AI would buff if it saw more than 3 enemies at once, AE nuke if stuff was clumped in the distance, heal injuries, always stay in sight of the player.
Fancy systems use decision trees or somesuch, but since I had 48 hours I just had a big bunch of ifs with a bit of randomness.
The companion AI would buff if it saw more than 3 enemies at once, AE nuke if stuff was clumped in the distance, heal injuries, always stay in sight of the player.
Fancy systems use decision trees or somesuch, but since I had 48 hours I just had a big bunch of ifs with a bit of randomness.
Yeah, screw the integrity of the fight. If the creatures can blow up the world to murder you, you should be able to do whatever it takes to win, within the constrains of the world/skills. The penalty for throwing 200 bodies at something 40 people can do is obvious - less chance of an individual getting compensated for their efforts. If people want to play that way, who cares.
I'm trying to look at this new-AI thing from a different perspective: if we think what came out of the AI/skills (not very smart/not very many) in Everquest was fun/entertaining, and people wanting trains, kiting, feign splitting, etc., imagine how many new and fun ways to do those things will come from this system + destructible/modifiable terrain. If the AI/game itself is fun and doesn't suck, there's going to be exponentially more ways to kill shit in EQN than there was in EQ.
I'm just glad that we will be able to ditch our starting class when we find another...I haven't had the best luck in the last few games i've played picking a class that wasn't god-awful. And I guess part of the appeal of the class system they're going for, is a newly-found class could actually make some class skills that you thought sucked from a former base class, really freaking good.