The Sad, sad, state of MMORPG in this brave new era...

prescient

Silver Knight of the Realm
97
5
I haven't done "end-game" anything in a really long time, but isn't the answer to people posting strats online and ruining the fun for everyone to simply imbue the end game mobs with AI and/or randomness? If you don't know what you are getting or you don't know what the mob will do you can't plan for it. Sure it'll go horribly wrong sometimes (too easy/too hard), but maybe that's the fun in having to stay on your toes rather than every fight being a well choreographed dance.
 
  • 1Like
Reactions: 1 user

a c i d.f l y

ಠ_ಠ
<Silver Donator>
20,060
99,460
I haven't done "end-game" anything in a really long time, but isn't the answer to people posting strats online and ruining the fun for everyone to simply imbue the end game mobs with AI and/or randomness? If you don't know what you are getting or you don't know what the mob will do you can't plan for it. Sure it'll go horribly wrong sometimes (too easy/too hard), but maybe that's the fun in having to stay on your toes rather than every fight being a well choreographed dance.
Jesus Christ, no. There's enough random shit that happens already that causes things to go horribly wrong too often (especially in Mythic raiding). Orchestrated fights are enough challenge to add any more randomness. AI you get a fight like the pvp-esque encounter in ToC from Wrath which still allowed for a controlled manner of taking it down, or the shit we have on Islands.
 
  • 1Like
Reactions: 1 user

Dandai

<WoW Guild Officer>
<Gold Donor>
5,904
4,481
I haven't done "end-game" anything in a really long time, but isn't the answer to people posting strats online and ruining the fun for everyone to simply imbue the end game mobs with AI and/or randomness? If you don't know what you are getting or you don't know what the mob will do you can't plan for it. Sure it'll go horribly wrong sometimes (too easy/too hard), but maybe that's the fun in having to stay on your toes rather than every fight being a well choreographed dance.
That’s an answer, but it’s definitely not a good one. The reason some games are ball-bustingly difficult but still fun and commercially popular (Dark Souls, Cuphead, etc) are because they’re consistent and by extension “fair” and not random.

Adding randomness and/or unpredictability would certainly achieve what you’re stating, but it would be at the expense of the fun of the game.
 

DickTrickle

Definitely NOT Furor Planedefiler
12,863
14,744
A good mechanic would be enemies that learned from what you specifically did to them and adapted and then made you adapt (though they'd only change if you could actually kill them). Figuring out the details of that and the timing of the adaptations while still making it fun would be incredibly difficult, but something like that would allow for changes in combat tactics without seeming fundamentally unfair. Something like the Mordor games have that at the most simple of level but MMOs are a long way off from having anything like that.
 

Ukerric

Bearded Ape
<Silver Donator>
7,886
9,471
A good mechanic would be enemies that learned from what you specifically did to them and adapted and then made you adapt (though they'd only change if you could actually kill them).
You already do that. You kill the boss in normal. Then you go and kill it on Heroic. Then you go and kill it on Mythic. Most of the time, there's a new mechanic that gets added to deal with.
 

fred sanford

<Gold Donor>
1,559
4,382
To keep it more simple you could have the usual normal and heroic mode with different abilities/strats entirely but then mythic would be a combination of the two with randomness. Phase 1 of a boss would, upon instance creation, choose random abilities from the normal or heroic phase 1. The same would go for phase 2, etc. I'd rather have that than having to grind for better gear to make a character simply stronger to take on the mythic boss. Just a thought, I haven't played in years :)
 

prescient

Silver Knight of the Realm
97
5
A good mechanic would be enemies that learned from what you specifically did to them and adapted and then made you adapt (though they'd only change if you could actually kill them). Figuring out the details of that and the timing of the adaptations while still making it fun would be incredibly difficult, but something like that would allow for changes in combat tactics without seeming fundamentally unfair. Something like the Mordor games have that at the most simple of level but MMOs are a long way off from having anything like that.

With the amount of player interactions that these mobs have I suspect you could create some really cool AIs with reinforcement learning. Give the mobs some sort of cost function and see what happens. You could have all sorts of cost functions: staying alive, killing healers, protecting treasure, guarding rooms, corpse camping you etc. It would be fun to see what sort of emergent tactics a mob might come up with given enough interactions with players.

People complain that they don't like having encounters spoiled but don't want randomness. You can't have your cake and eat it to.
 

Kyougou

<Gold Donor>
1,520
4,939
Then the strategy would be to teach it to be really dumb then kill it.
Anyways, you already have that experience available in most MMOs, just go PvP.
 
  • 1Like
Reactions: 1 user

prescient

Silver Knight of the Realm
97
5
Then the strategy would be to teach it to be really dumb then kill it.
Anyways, you already have that experience available in most MMOs, just go PvP.

So that is interesting. If the cost function is based on protecting treasure for example (no idea how to design that function), and you want the treasure how would you teach it to be dumb?

Also I'm already there on the PvP. Destiny is the only game that's really hooked me in recent years.
 

Kyougou

<Gold Donor>
1,520
4,939
If the cost function is getting to the treasure, how would you evaluate how close a team got to opening the treasure? Dupe the boss into thinking some stupid strategy is great by purposely getting further away from opening the treasure, then just walk in and grab it.
 

prescient

Silver Knight of the Realm
97
5
If the cost function is getting to the treasure, how would you evaluate how close a team got to opening the treasure? Dupe the boss into thinking some stupid strategy is great by purposely getting further away from opening the treasure, then just walk in and grab it.

The moment you grab the treasure there is going to be a big negative hit to the cost function. I don’t think that would work like you are thinking about it. Additionally, you can design around this sort of stuff. mob ai’s are laughable now. Better AIs would make for more fun and unpredictable games is all I’m really getting at.
 

DickTrickle

Definitely NOT Furor Planedefiler
12,863
14,744
You already do that. You kill the boss in normal. Then you go and kill it on Heroic. Then you go and kill it on Mythic. Most of the time, there's a new mechanic that gets added to deal with.

That's not adaption. The idea with be adaptation is that you would have to figure it out as would slowly change over time. You wouldn't necessarily be able to use the same tactics for eternity, but the changes wouldn't be random either. Theings you mention are set in stone.

Again, seems eminently doable in future single player games, but not sure it would be feasible in a non instanced MMO.
 

DickTrickle

Definitely NOT Furor Planedefiler
12,863
14,744
Then the strategy would be to teach it to be really dumb then kill it.
Anyways, you already have that experience available in most MMOs, just go PvP.

No, it wouldn't change until you killed it. So you couldn't train it in the way you're talking about.

If you wanted to metagame, you'd have to kill it in the most inefficient and sloppy way possible at first.
 

Punko

Macho Ma'am
<Gold Donor>
7,912
12,564
I don't know why MMO's don't just hire asians to control certain mobs and provide AI.

Its far cheaper then designing a good AI short term and longer term things can be switched around so the content doesn't get stale.

I am positive it is better to have an MMO with 10-20 new and relevant encounters, then one like EQ.
 
  • 1Like
Reactions: 1 user

pharmakos

soʞɐɯɹɐɥd
<Bronze Donator>
16,306
-2,239
That's not adaption. The idea with be adaptation is that you would have to figure it out as would slowly change over time. You wouldn't necessarily be able to use the same tactics for eternity, but the changes wouldn't be random either. Theings you mention are set in stone.

Again, seems eminently doable in future single player games, but not sure it would be feasible in a non instanced MMO.

would probably just end up being several strats that you have to rotate between... use strat A then next time you need strat B, then you need strat C, then maybe strat D and back to strat A.

it would be difficult to code enough variables in the logic to require more than a handful of strats.
 

Big Phoenix

Pronouns: zie/zhem/zer
<Gold Donor>
44,511
92,908
Also I'm pretty sure Valve's problem is the opposite of milking it for profit, it's that they have a bunch of wealthy game nerds who can work on their pet project with no pressure to actually achieve anything, pretty much the George RR Martin problem.
Pet projects? They have made anything original since Portal 2 in 2011.
 

prescient

Silver Knight of the Realm
97
5
would probably just end up being several strats that you have to rotate between... use strat A then next time you need strat B, then you need strat C, then maybe strat D and back to strat A.

it would be difficult to code enough variables in the logic to require more than a handful of strats.

The whole point of these kind of AIs is that you don't code the logic. The AIs learn from playing the game (something along the lines of Alpha Go). I don't know how tractable this is given the current state of reinforcement learning, but a learning mob should be the ultimate goal.
 

Chanur

Shit Posting Professional
<Gold Donor>
26,482
38,336
Until it becomes self aware and builds a robot version of itself to destroy humanity to prevent it from being raided thus ensuring it's continued success.
 
  • 3Like
  • 2Worf
Reactions: 4 users

Ukerric

Bearded Ape
<Silver Donator>
7,886
9,471
but a learning mob should be the ultimate goal.
Nope. An Interesting Mob is the ultimate goal. As some dev said once, making a boss that slaughters your raid is trivial (kill healers first, then debuff, then dps, THEN the tank). Making it so that the fight is challenging/interesting is much, much harder.

And learning mobs would almost certainly learn how to slaughter raids, not how to entertain them. Which is the real goal.