Tree of Savior Forum

Musings on iToS Balance - Part I of V - A hollow game of numbers

General Discussion
#41

Not only you, I feel the same too, I mean, I’m aware that the game still under development, in my mind I take that IMC still putting up the pieces together, then when everything is mounted, they will start to polish it the way it should be.

But for now we can only wait and see if that will really happen, the thing is, it is tiring all this wait, and even more when we don’t get any concrete response from the developer itself.

It’s like you buy a new apartment with all the stuff in, but you don’t have gas, water and electricity to use the stuff you have, and when you go on the Real Estate you find it with the doors closed and the telephone unplugged from the wall.

5 Likes
#42

Besides boss fights, wow is also killing a group of things instead of 1 target most of the time.

#43

@Nekorin Yes, the “cheesemelts” probably made earth tower seem like nothing. It also made every other content seem like nothing as well. It’s just that most other content don’t even require any strategy.

But even without ausrine + meltis, a lot of earth tower content are dps checks that are based on gears and not strategy. Rest were like kiting and using “barrier” nonstop. I don’t see much strategy involved at all. Most people treat it as a grind I think.

I think there is no way to change all that now since it’s the base design of this game (gathering lots of mobs and aoe them down). What they can do right now is to make classes more balanced towards this design and also tweak the OP combos that make everything trivial. I hope they will abandon adding more floors to earth tower starting with R9 and start making new and creative dungeons for the new end game content.

#44

I (compared to other classes) greatly enjoyed leveling up my Templar C2 (solo) because the DPS is so low that the gameplay focus turns to all the other tools provided by each attack. If I want to fight monsters 1v1, it is slow and monotonous. But if I want to fight a pile of 20 mobs or so, then I’m forced to juggle and parse enemies with knockbacks and stuns. The leveling process is greatly sped up and it becomes a lot more fun.

All of my other characters have been boring to play. Cryomancer is slow and boring no matter the number of enemies you face (chronomancer didn’t help it either); it’s even more boring in a party. Quarrel shooter (although hilarious) is boring to play because you just kite everything or slap a pavise on the floor and don’t kite. Pyromancer, just plant your trash and dance around it. Any cleric tree, don’t even need to stress dps race.

I’ve played a ton of classes and combinations, but the best one has been Sword1>Pelt2>Rodel3>Templar2 explicitly because the nature of the monsters AI lends itself to the gameplay of my dumb templar build. There needs to be more for other classes. Monsters shouldn’t be smushing themselves against Icewall, or tripping over Caltrops, or getting pissed off at a pavise, or slowly roasting themselves in flameground. The healer monster archetype should not be plopping heals down for flying monsters (why), nor should it be eating all of a melee’s attacks. etc

A monster might stun you, or freeze you, but it’s only coming from one specific monster type and at the point when you get hit with that stun or freeze, you may be chain-locked until death. I don’t know how you guys feel about this sort of AI, but my computer refuses to render monster’s Ice Pike, so I can’t ever react and I hate it.

#45

The game is too antagonizing to melee(i know i sound like a zealot saying this everywhere),that’s why you feel more compelling to play your Templar, the vast majority of basic mobs are designed around the low tier Swordsman classes. I always try to use Swordie’s knockdown and pushbacks to contribute to the party(interrupt enemies that go for squishy party memeber/drop enemies inside fireballs,etc) but the meta simply doesn’t allow for it to be appreciated,things that can be pushed back die in less than one second

->About AI

I doubt “better AI” alone can solve the elephant in the room that is Melstis+Ausrine. If you give human AI to the bosses, with their current skillsets, you can be sure they would create forum accounts and petition for a nerf/change to it.

1 Like
#46

I don’t know about antagonizing, but It definitely feels like it’s balanced for what I was playing.

So what if the AI just avoided players with invincibility? If they’re melee, just stay out of reach; if they’re ranged, sprint off the screen? Mostly I just think Ausrine (I’m not familiar with Melstis) shouldn’t be providing 18 seconds of damage immunity. Maybe AOE immunity, maybe poison immunity, maybe knockback immunity, or a mix of those. etc … a rebalance.

Also I feel I need to stress this, I don’t mean to give bosses or mobs human AI. I mean the AI should have more options than aim > attack/use skill; There’s a lot of simple behaviors that can be programmed in that make a dynamic AI.

#47

I agree that mobs AI should be revamped, especially bacause ALL of them are identical ones apparently. However, I still feel the need of monsters ramming Ice Walls or roasting themselves on Flame Ground simply because PvM on the fields are supposed to be relaxing.

My personal experience is that when me and my buds want to “play” then we’d pop in a dungeon and play in a coordinated party. When we’re out on the fields, we usually roam around the map while chatting and killing stuff. I wouldn’t want extra tension or difficulty on those maps. In fact, the difficulty tier of any MMO at all should be:

  1. Field (super easy, very grindy or questy, won’t die unless you do something stupid, required for story progression)
  2. Daily dungeons (mild difficulty, rewarding payout, need a degree of focus and the common ability to not die, required for character progression)
  3. Challenge dungeon (hard difficulty, end-game tier loots, need a coordinated party play and an understanding of the game’s mechanics, required for optimizing only)
  4. PvP environment (balanced, luxury item rewards like costume or headgears, player skill based competition and NOT blind DPS race, required for bragging rights)

Also, I don’t think Ausrine should be nerfed. If anything, maybe we should break that combo with Melstis. Ausrine was supposed to be a stationary invincibility skill with considerable down time, and it works okay. It’s Melstis that broke it.

1 Like
#48

You did it, you found how to rebalance both skills. They’re just like Safety Wall now.

Anyways, the way I feel about Field content is it should be just as hard as a Challenge dungeon for solo play. Calm and relaxing comes in on the pace and ability of those participating even within these dungeons. If you want to fight one monster at a time, fight one monster at a time.

Maybe I just don’t like losing my time to “calm and relaxing”.

#49

About the arrows… Calculate projectiles trajetory is very cost. I remember ppl talking about it when Elder Scrolls Online was in betas. The server will need to computate every arrow trajetory for every player in the screen.

#50

Man, the way you put things I bet you’d enjoy playing a hunter falconer. Managing your pet, making sure the enemies come at the right angle to fall for your pet, pushing them to your pet, recalling your pet, calling the hawk, deciding the best time for circling etc etc.

But yeah, as stated previously, the enemies of this game are too bland and stupid, they just will mindlessly follow and attack.

1 Like
#51

And Sterea Trofh too! My point is Clerics making absolute safe zones from time to time is okay. But some skill granting infinite invincibility is not okay.

Maybe ToS should just make unlimited solo dungeons for hardliners like you. A win-win solution? Maybe make some maps hard as to cause people to enter with caution. But the “story-related highways” between towns shouldn’t have head-bashing difficulty mobs.

#52

I was there for those discussions and I disagree with most of what was said. The Battlefield servers can calculate 10’s of thousands of trajectories within a number of different instances. The trajectory information is shared to clients within the instance and the information sent is looking at the damage fall-off, the bullet drop, the speed. If 32v32 clients can do this with this much data (probably more data), then ToS can do it with 5v5 in a PvP setting.

(based on what I know from ESO from launch to 18 months later) ESO lags incredibly bad in PvP already and that’s because their netcode is horse s***. AOE’s on the ground draw thousands of calculations because the server decides that every client needs to be updated on every other client’s vicinity to any one AOE (we figured this out by tracking packets when playing in Cyrodiil). Stealth and siege weapons was also a huge issue that lagged other clients.

This is another thing where implementation is key. Besides, ToS would only need to look at the angle on one plane, and use hit detection instead of a trajectory (if it’s arrows and bullets, who cares; the screen is too small).

#53

Ehm, EA, Dice with its pedegree of freaking creating game engines? A lot of cash? 60 dollar game?

Sadly, We’d need someone of that caliber to bring hope for us.

#54

Zenimax Online Studios had a large budget and they couldn’t pull it off, so I’m sure that money is a none issue when it comes to figuring out how to code something.

I mean, in the simplest way possible, you would just unhitch the lock on and make the right analog (or mouse) draw a line in a direction, if a target lands in that line, they are considered your target. There, I did it. Didn’t need to do anything except change the behavior of the targeting system.

#55

If you’ve ever played anything based on freaking bethesda’s work you couldn’t possibly compare them to dice in quality. Freaking ladders been missing since for ever.

#56

Zenimax Online Studios (the creators of Elderscrolls Online) is not Bethesda

#57

I highly doubt they didn’t ask for a hand, the taint is real.
Regardless of that. You’d need real good servers and real good quality coding.

The algorithm has to be efficient.

#58

I Just made it efficient.

#59

That’s like a requirement, vague, with an objective. not the algorithm. But regardless, a requirement is the start line.

#60

But you already know what happens after a target is established.