Helping the mage class continued

[ValdainTheKing]

http://www.soldak.com/forums/showthr...ow+your&page=2

Nobear-Middle page. Single class mage. Uses a variety of spells from all 3 spell schools of wizard.

[Crisses]

Ok, so one example so far of a pure mage who made it to end-game.

Skills

Fire Maelstrom lvl 1
Frost Nova lvl 1
Permafrost lvl 1
Shatter lvl 28
Cold Mastery lvl 30
Evasion lvl 2
Teleport lvl 6

Still basically a 1-tree character -- but was able to pour enough points into Teleport to make it a defensive possibility. So everyone's saying there's one spell on the 3 trees that's very effective.

I tried several games in a row (and abandoned them all) with my Conjurer character and I'm really getting tired of dying. I went from Level 36 Very Fast to trying a Level 32 (same as character) Low Stress Normal Pace -- even summoned a champion chaos lord (and raised only champion critters) and it's painful trying to get through the dungeon. The XP debt started rivaling the US National Debt. (support spells are Tornado and Blight to knock the hitpoints out of the masses to help my pets, and good DPS on my own weapon). I'm calling it a night. If I'm working that hard to just get to level 2 of a dungeon, something's up.

[Bluddy]

I posted this in another thread but it really belongs here, as warlocks and necromancers are also mage types.

The reason raise dead is so powerful is that you raise a monster of the exact level of one you kill, as opposed to all the other create pet spells, where you create monsters whose level corresponds to a fraction of your level + a bonus based on however many spell levels you have. I'd say this is a pretty serious imbalance. A necromancer needs only to invest once in raise dead. He'll then have a monster that keeps up with the dungeon level. All other pet spells require constant investment or as they go up in level, their monsters will be less and less effective. Thus, the other spells fall into the category of 'spells you need to put all your points into if you want to keep up'.

Let's take a peek at the fury spell. To keep a fury at the level of your character, you would have to invest in the fury spell every 4 levels. You get around a total of 1124 skill points in the game, of which about 674 need to be used to take the fury all the way to level 30. Half of your skill point would have to be used just on the fury. If we account for the fact that you get 1/4 of a fury every skill level, you would need to upgrade the fury skill every 5 levels to keep up. This would cost you about 1/3 of your skill points.

The necromancer's raise dead spell is more along the lines of what the stronger classes have. Other classes generally don't need to keep investing in specific skills unless they want to, to boost their damage output.

I would change it so that, rather than getting monster levels per spell level, you get a percentage of the player's level count. Thus, monster would start out at 50% of a player's level and would gradually get more and more percentage points per level. If you keep investing, they'd surpass the player's level. Also, to balance out the necromancer, raised monsters should get a % level drop when you raise them. Keep investing in the skill and that level drop would disappear.

[Bluddy]

I proceeded to do the same analysis for shadow bolt. Assuming that I would feel a need to upgrade my bolt once monsters require more than 3 hits to die, I need to upgrade it about once every 4 levels. This is actually a conservative estimate, because it doesn't account for monsters' elemental resistance (which is around 10%, or 5% in my formula) or for the monsters' damage curve which accelerates. 3 hits to kill a monster may not be fast enough for a mage when the monsters can kill you with 1 hit. In any case, once every 4 levels means that 40% of the skill points in the game (my calculation in the post above was a little off) must go towards shadow bolt alone -- otherwise it'll be useless.

I don't know if more skill points could even fix this. I think what's needed is a non-linear curve for all offensive skills that don't base themselves on your current DPS. Instead of gaining a set amount every level, shadow bolt should gain more and more with every level. Of course the amounts have to be balanced, but I think this is the only way the mage classes will become balanced. Also, what has to be accounted for is area effects. Area effects are non-linear by their nature (you hit multiple enemies for the same amount) so they need to accelerate much more slowly.

I'll start working on what I think are proper, accelerating values for the skills.

[Crisses]

Originally Posted by Bluddy

I proceeded to do the same analysis for shadow bolt.

Yes see my thread on "scales per level"--this in what I'm talking about.

I looked at my Hunter/Trickster character last night and on those trees there are few skills that don't scale to level.

Immolation Trap she has to lvl 6. It doesn't scale-- but it still is useful to explode things, and open doors. Monsters just laugh at it. She could just have it at lvl 1 to do what it does for her except the HPs of barrels and doors go up too. Freezing trap has the same problem without the benefit of catching things on fire...

ALL the spells on the Trickster tree seem to scale. Either they scale through DPS escalations as one acquires better weapons, or they don't really seem to need to scale, like slight of hand which just makes monsters MORE angry with one another when you raise it.

Hunter traps need to scale to level, a damage &/or duration increase. However, Lure for example already does (though I think it needs to act like an immobile pet so you can walk through it and it's immune to your own spells--but don't fix one without fixing the other!).

I don't have trees in front of me, but a fire bolt purchased at lvl 100 still does a measly 15ish damage. That shouldn't even get the monsters attention at that point.

I've been thinking about it and this is by far the worst problem. Its not how many skill points...this is why some skills just aren't worth keeping in the endgame, and why spell casters are so narrowly specialized that 1-2 trees seem wasted. You actually can't throw enough skill points at this problem to fix it, since prices escalate so quickly. There should be the ability to spread your points out a little, and more than 1 or 2 spells that are effective at endgame per tree.

Perhaps for most skills SOMETHING should scale to level. For many the change could be gradual and subtle unless you pour a few levels into it, and there are exceptions like Teleport which are just fine as they are, but say Hardiness duration increased, the healing spells heal a little more depending both on the spell level and the level of the caster, cooldown decreases or duration increases on some spells by a percentage depending on the level of the caster, etc.

Most importantly, if the spell affects something that scales to the level of the dungeon, then it should scale to the level of the character. Monster hit points scale, so anything that does direct damage needs to scale. And that goes for every tree, because there are Rogue & warrior skills that also suffer from this problem. Assassins poison for example.

[Bluddy]

Originally Posted by Crisses

Yes see my thread on "scales per level"--this in what I'm talking about.

I looked at my Hunter/Trickster character last night and on those trees there are few skills that don't scale to level.

Immolation Trap she has to lvl 6. It doesn't scale-- but it still is useful to explode things, and open doors. Monsters just laugh at it. She could just have it at lvl 1 to do what it does for her except the HPs of barrels and doors go up too. Freezing trap has the same problem without the benefit of catching things on fire...

ALL the spells on the Trickster tree seem to scale. Either they scale through DPS escalations as one acquires better weapons, or they don't really seem to need to scale, like slight of hand which just makes monsters MORE angry with one another when you raise it.

Hunter traps need to scale to level, a damage &/or duration increase. However, Lure for example already does (though I think it needs to act like an immobile pet so you can walk through it and it's immune to your own spells--but don't fix one without fixing the other!).

I don't have trees in front of me, but a fire bolt purchased at lvl 100 still does a measly 15ish damage. That shouldn't even get the monsters attention at that point.

I've been thinking about it and this is by far the worst problem. Its not how many skill points...this is why some skills just aren't worth keeping in the endgame, and why spell casters are so narrowly specialized that 1-2 trees seem wasted. You actually can't throw enough skill points at this problem to fix it, since prices escalate so quickly. There should be the ability to spread your points out a little, and more than 1 or 2 spells that are effective at endgame per tree.

Perhaps for most skills SOMETHING should scale to level. For many the change could be gradual and subtle unless you pour a few levels into it, and there are exceptions like Teleport which are just fine as they are, but say Hardiness duration increased, the healing spells heal a little more depending both on the spell level and the level of the caster, cooldown decreases or duration increases on some spells by a percentage depending on the level of the caster, etc.

Most importantly, if the spell affects something that scales to the level of the dungeon, then it should scale to the level of the character. Monster hit points scale, so anything that does direct damage needs to scale. And that goes for every tree, because there are Rogue & warrior skills that also suffer from this problem. Assassins poison for example.

Right. I think I remember reading that thread, but until now I hadn't realized how serious the problem is. You can pretty much say that all mage spells are broken as they are right now.

There are 2 ways I can think of to fix this: either do as you say, and link the spell damage to something in the character, like intelligence. But that would be a complete reworking of the system.

The other option is just to make the spells go up non-linearly ie quadratically: every spell level, you get more bang for your buck than you did last time you upgraded. The down side to upgrading would be quadratic mana costs. Essentially what happens with skills is that while the monster HP constantly increases in a straight line, spells are like stairs: you get jumps in power every time you upgrade. Those jumps can't be linear, because you're not constantly upgrading the same skills. They have to be big jumps relative to monster HP increase.

The reason I hadn't noticed this problem in my own character (necromancer) is that raise dead does scale, as I explained above.

I'm trying to work on the proper fixes to make this option work. The nice thing about a quadratic function is that it should be possible to make only small modifications per spell to fix the situation.

[Crisses]

Originally Posted by Bluddy

I'm trying to work on the proper fixes to make this option work. The nice thing about a quadratic function is that it should be possible to make only small modifications per spell to fix the situation.

(I have the skill trees in front of me now)

I don't think that 100% of the change should be dependent on the number of skill points you spend. It should be a balance of both. 1 point in Firey Blast does 5-15. I think if I were level 90 and decided "Hey, I want this spell" it should NOT start at 5-15. Perhaps even just 50-150 (though potentially 500-1500 at that level! It would take several of them to kill a monster at your own level -- I'm not sure what the immolation effect is so it should be buffered with that in mind). It's a difference that matters because a level 90 warrior can take Power Strike at level 1 and it does a HECK of a lot more than that (+20% physical damage).

Call it the "No more mosquito bites" campaign for fair skill effectiveness.

Some durations could increase by as little as 1% per level -- by level 100 they're just twice the duration. Something that lasted 30 seconds now lasts 60. It's a nod to the fact that you've become a world-class Warrior/Rogue/Wizard etc. Your capacity to control magic has simply increased, period -- aside from the actual spells you've practiced and acquired.

I'm trying to picture what the current system would look like in real life. A level 100 fire mage with a level 20 fireball and a level 1 firey blast: Cast one spell and everything's a little hurt. Cast the other and you can barely light a cigar. Hardly warrants the name "Firey Blast" right? If I were a monster, I'd be thinking "Did someone say lunch?" Definitely not the awe-inspiring all-mighty-powerful-wizard.

An Assassin with Viper Venom 1 does 2.0 poison dps (effectively 16 damage over 8 seconds). I think a level 100 Assassin with Viper Venom 1 should perhaps do 100.0 poison dps (and consider a duration shift per level too -- but this would be 800 damage total -- it's a dent but not deadly at level 100, and takes its time to do it). It's not that it shouldn't scale significantly if it becomes VV2 -- I think that adding in new skills even late in the game should have SOME level of effectiveness against your level of creature.

And if a level 100 ANYCLASS goes slumming with level 20 characters, they're going to wipe the place quickly, even with poorly chosen skills or as a spellcaster "jack of all trades". And I think they should! Right now a warrior has that privilege. I'd like to enjoy playing any class -- and I think it would make the game far more re-playable. Otherwise just grab a warrior, finish the game -- maybe play a rogue a couple times for variety, then retire the game. Those of us hashing this out in the forums are the ones who want more replayability from it, I think.

[Crisses]

I'm taking a quick glance at Din's Additionals -- just to see an example of skill modifications.

Yeah, what can be modified is hard-coded It's not formulas in the game that we can simply modify on our own. Not that I can tell -- Shadow would have to confirm.

Skellys have one of the stats that's a "Per User Level" stat:

MonsterLevelPerUserLevel 0.5

I'm trying to think whether anything other than things that are summoned have a "scale per user level" effect....so I can look at that skill in the files to see what is done about it -- but I can't think of any.

[Crisses]

Skill-adjustment wishlist:

I would want the ability to modify the following on a PerUserLevel basis:

Duration
ReuseTime
ProjMinDamage
ProjMaxDamage
Power (?)
TotalTime
ProjRadiusMinDamage
ProjRadiusMaxDamage
HateChange
MinDamage
MaxDamage

Can we add floating-point "PerUserLevel" modifiers to each of these formulas on the back-end so that we can play with skill/spell balance?

Such that a FieryBlast on a Level 10 character and

MinDamagePerUserLevel 0.5
MaxDamagePerUserLevel 2

Would do 10-35 damage.

And on a Level 100 character would do 55-215 damage.

It's still linear, but it helps. I think it will help a lot.

Is this something that can be added to the formulas on the back-end so that we can adjust skills and report back on them?

[Bluddy]

I also initially thought that it was necessary to relate spells to user level -- but I don't really think so. Why?

Suppose you're a level 100 wizard and you invest in 1 level of lightning. If it were related to your level, you'd get a tiny burst of static electricity to do tricks with your opponents' hairs. This is the equivalent of one level of lightning. A level of a more powerful spell would be a little better. Because the spells are cheap, and you're high level so you get many spell points at this point, you can use your many spell points to quickly get the lightning spell up to a level where you can smell burning scree. To get it to the point where you can fry chaos lords might take more points than you have, but scree meat you can do.

This is how the system is supposed to work. Unfortunately, because of the linearity of the spell strengths/level, you can't really do it. But if the jumps per skill level got bigger and bigger, you'd need a relatively small amount of points to get the skill up to speed.