Classes

The classes sure needs to be rebalanced:

• Within each class, the skills need to be balanced. The OP skills should be nerfed, while the underused skills should be given more punch.
• Each class should not become weaker or stronger - but the skills should be averaged, to encourage trying different skills (instead of relying on the same skills, resulting in an "optimal way to build")

That's the spirit, and I think everyone agrees with that. However, the first real question is :


Question 1 : How to accurately measure the effectiveness of a skill ?

I see several possible solutions :

Solution 1 : Personal experience.

That's what each and every modder has been doing so far. The major problem of that approach is bias. Maybe the modder will be disregarding 2 or 3 skills because he does not like them, without even knowing it. Or maybe the modder did not realize that some people had found a way to use the skill effectively.

The biggest example of it is how Salan modified the Embermage lightning tree. He modified what were for him the least used skills, in particular Shocking Orb - instead of a slow moving orb, he modified it to be a mini turret attached to your character, firing at enemies coming close to you. The result was that the skill was then even less effective than it had ever been, with only limited use. A friend of mine liked the vanilla skill a lot, using a Chaotic build with it.

There is also the problem of overpowering a class. It's so easy to overpower something that is already good enough, and do not need overpowering.

This is why I don't think one should try to balance a class by himself. This is just impossible - you would need to try every damn possible builds before you're able to say "I know what is objectively wrong with this class".

Solution 2 : Gathering the thoughts of everyoneor

This looks like the logical follow-up of the problem raised in solution 1. However, you need a very big and mature community for that. If the community is not big enough, you end up with only a limited vision of the problem. If the community is not mature enough, you have to filter through biased writings, which can consume a lot of time for limited conclusion.

One way to remedy that would be to have a poll, with questions like "What are your favorite skills for the <class> ?", and "What are your worst skills for the <class> ?". If we gather enough inputs on that (which might be worth to try on the steam forum for example ?), then we would have a list of various strong and weak lists of skills - the skills identified as very strong would have to be nerfed, while the skills that are underused would have to be pumped up.

Of course it's not perfect - informations like synergies between the skills cannot be emphasized with such a poll 

But i think that someone experienced would then be able to identify the general directions that balancing should be taking.

Question 2 : How to modify a skill ?

This should also be taken into account - indeed, you could turn a skill into a complete BS one, either OP or completely underpowered.. Or worse, you could end up with the skill producing bugs in the game, perhaps even crashes.

I think the right way is to do simple tweaks. There is no point in completely overhauling a skill - way too much work (just look at how much time it took you guys to port the TL1 skills into TL2 - and you were many people working on it). However, maybe slight damage/knockback increase might just do the thing. Actually, the question to ask is : "why is this skill underused ?". Maybe it does not deal enough damage. Maybe the Tier effects are not strong enough. Maybe the cooldown is too high.

When one is not sure, it would be useful to ask the question directly on a forum, and ask for people's opinions - in this case, the bias would be quite low, since we would speaking of only one skill, and why this or that people don't use it. It would be about gathering actual user experience, not direct balancing.

Question 3 : How to maintain it ?

That's the biggest problem. It's easy to ask the vanilla guys what they think of this or that skill.

But once you have modified them, you need some actual feedback, because obviously the first iteration will never be great. And if only a few people play your mod, you won't have enough information to know if your balance was effective, or if it's actually worst.

I have no answer to that question unfortunately. I guess the only true way is by becoming popular.. But you only become popular if you have a good balancing.. Oh wait, didn't I just bite my queue ?

Gear

Right now, I think the gear by itself is fairly balanced. Maybe the legendaries should be given more importance.. But there is nothing utterly wrong with that.

However, the three biggest problems in my opinion is twofold:

• You pretty much need to socket all your gear with %DR skull, and then with health/missile reflection/dmg skull
• You need to enchant all your gear with a particular character (I think Boris is the optimal enchanter in 99% of the use cases..)
• You can equip lvl 100 gear on a lvl 1 character by abusing the stat condition

Problem 1 : Farming skulls

You have to farm a lot before you're able to get enough skulls. Which can take a significant long and boring time.

I will describe here all the solutions that I have thought in order for it to become less painful

Solution 1 - Change the skulls to have very unique effects

Instead of a plain 5% DR, you would get for example "5% chance to summon <random> on hit" (<random> being
rolled on the item generation)
Instead of damage reflection, you would get maybe a slight chance to increase damage reflecting enemies (again, which one would be selected at random when you roll the gem)

Maybe there's some other mechanisms that would be cool to look at.
The problem though with extra summons, is that you have to design a new spell for each summon, and also
you can have balance/cluttering issues..

It's kind of the same argument that I was stating in the above post. It would be nice to get new complicated
effects on the gems, but it would take a lot of time.

Solution 2 - Make all the skulls available to a merchant in Mapworks lvl 100

If the gems are all available in an end-game merchant, then there would be no more tedious farming involved :-)
You would just have to get enough gold to buy them. Of course, only the end-game skulls would be on that merchant, and this merchant would only appear on Mapworks lvl 100+ (so maybe NG+ or NG++ ?) so that the
player has no access to it while leveling.

Problem 2 : Enchanting

I think the same solution could be applied - make all the enchanters available in Mapwork lvl 100+.
This way, you don't have to look after them in each and every dungeon.

The ideal solution would be to transform the enchanters into quest guys - you encounter them in Mapwork only
if you have met them in a lvl 100+ dungeon. But I have no idea how to do that, for two reasons:

• How to make so that a guy is both a quest guy and an enchanter ? (maybe by having two copies of the same guy, with only one being visible at a time ?)
  
• How to trigger something based on the lvl of the dungeon ?

Also, I don't really know how the enchanter thing has been designed to be honest. It only appears in some dungeons, so do they have a list somewhere of dungeons where it spawns, or did they just **** it, meaning I would have to modify each and every dungeon to insert my special conditions ?

Problem 3 : Overleveled gear
Just today I met a guy in multiplayer who was breezing Frosted Hills with a lvl 8 character. No wonder why, he had a full lvl 100 gear by abusing the stat mechanism..

The simple adjustment that would disallow this, would be to change :
Strength 80 or lvl 100to
(Strength 80 and lvl 95) or lvl 100
This way, the stat cap still has a sense (since you would be able to equip weapons/gear of 5 more lvl than you), but you would not be able to abuse it.

The question I have though, is that I don't know if this is part of the engine or the item. Is it possible to change the boolean formula within GUTS ?

Problem 4 : On the topic of Boris..
I can't remember how game breaking Boris is. It sure allows your gear to be much much more powerful.

I would not want to nerf him, because maybe you actually need him to go to the very high NGs.
But maybe it should be restricted to only lvl 100 characters. I don't know.

I remember having times where a lvl 50 guys with full Boris was literally breezing on a lvl 55 area, but this was so long ago..

Pets
Right now I think the pets are useless. Unless you socket tons of %pet_dmg, your pet is just an aggro that can carry some items for you to town. It does not play any central role.

Fishing

In Torchlight 1 I have fond memories of my pet becoming super-powerful when he was an elemental one. I never had these kind of feelings in Torchlight 2, and I think the fish should provide stronger bonuses to your pet.

I think there should be different kind of fishes, each one transforming your pet with different abilities

• Common fishes. It would transform your pet into one of the base monsters of the game. Which monster can drop should depend on which level you are. At lvl < 20, you are likely to be in Act 1, and only Act 1 monsters should drop.
• Rare fishes. These would transform your pet into one of the champions of the game.
• Unique fishes. Your pet is transformed into one of the minor bosses of the act you're in. (Minor boss = boss of a side quest)
• Legendary fishes. Your pet is transformed into one of the major boss (bosses that you have to defeat in the main story).

Legendary and unique fishes should be very rare of course.

For each category, there should be a roll to determine the duration. You roll 1d20, and :

• If the result is 20, the duration is permanent
• If the result is 19-18, the duration is one hour
• If the result is 17-14, the duration is 15 minutes
• Otherwise, the duration is 5 minutes

It would be possible to get a permanent boss pet. But the odds of getting it would be very very low - if you get one during your playtime, you can consider yourself lucky.

This would need some testing to ensure that the odds are indeed very low..

Also, I can't remember what's the mechanism - if you get a permanent transformation, but feed another fish to your pet. Does he lose his permanent transformation, or does he go back to it after a while ?

Spells

Right the only usefulness of a pet is to carry the droppable spells. I kind of like this aspect of the game, even though some of the spells are really useless.. I will be coming back to it later when it comes to the vanilla spells.

I just wanted to mention that I had played a mod one day which was making every class-spell droppable. In particular, you could feed them to your pet, and he would then be able to cast stuff like Tremor or Forcefield or Blazing Pillars.. It was very fun to see.

I don't think this should be implemented - the pet was way too OP, and I believe some cooldowns were messed up, resulting in funny spams. Besides, if pets can be transformed into bosses, you should not need this.

Gear

I think the pet gears are quite balanced right now. The percentage to do something on hit will make more sense when your pet is transformed into a monster that can use abilities to proc it.

http://torchmodders.com/forums/mtaur's-mod-pack/

The main things that I did were to give a lot of skills substantial cooldowns, as well as massively nerf potions while buffing their durations.  This differentiates burst from DPS skills, and makes potions more about sustain than a really weird OP sustain mechanic that messes up the whole game.

Unfortunately, the mod needs more work.  Enemy damage seems to outscale my nerfed sustain numbers by a lot as you go forward, and I haven't nerfed the enemy damage graph enough to compensate.  It would take a lot more work to zero in on the appropriate numbers, and I might not ever get around to doing that.  The balance is just kind of messed up in vanilla across the board.

Yes I saw your work - but actually I've got a slight different view on the potion mechanism.

At least in Early Game, and playing Elite, I didn't find that the potions were that overpowered when you're playing for example a melee engineer. It's more of a "I have to drink loads of potions otherwise i die" - but if you weren't drinking these potions, you would be saving money. So if you make a more sustainable character.. Then you get more profit in terms of money.

Changing the potions to have a long duration would mean that you lose this aspect of the game, and grant everyone a significant boost in income during Early Game. It also slows the flow of the game, since you have to wait until your health/mana is replenished, and makes you more endangered against the monsters as the damage grows. So you have to rebalance the damage graphs.. And this is not trivial to do at all - it would require either a full-time job or spending all your free time on it to re-balance properly. Just to change the potion mechanism.

But I'm interested to see how your mod changes the gameplay in practice - I will try it soon and post in your thread :-)

The potions are fast enough that you can generally start walking to the next place soon, but still a lot slower than vanilla.  I forget the numbers exactly.  You could achieve similar just by level-locking the vanilla potions, but it would still be a gold sink, and once every 8 seconds got a bit tiring for me.  They also tend to outscale their costs.

The 15-minute potions are only about half as strong as the same level 30-second potion, but are more gold efficient overall (they also tend to drop more often than once every 15 minutes).  They also fizzle when you die, so when you're having a hard time, you have double incentives to use the 30-second ones anyway.

Actually I this "gold sink", because it rewards you whenever you play without drinking potions. It's a way to passively reward you for being a good player/having a good build - because you spend less gold in potions.

In Elite, after some lvl, every monster is two shotting you most of the time. If your regeneration is low and consistent, you cannot survive without drawing away from the fight at every hit. The vanilla potions are not particularly overpowered - they are just necessary to survive considering how hard the monsters hit.

What matters here is the amount of health/sec regenerated rather than the duration. If your potion regenerates less health/sec, your characters will have a harder time. When you press '1' to regenerate health, you really need it to be quick, or you just die. And sometimes it's not even sufficient with the vanilla potion.

I'm talking of Elite here - I did not precise it, but if I do a mod, it will be focused at being balanced for Elite. Of course, for having played Normal and Veteran, potions are overpowered there. But in Elite it's barely enough, and it's actually very damn hard to have enough money to sustain your potions if you don't play well enough.

I never got very far in Elite.  Just couldn't get the DPS out enough to get far even in Act 1.  It may have been a pacing and farming issue.  I don't remember whether I was playing modded or not, but it sounds like unmodded would be more manageable (assuming you can farm enough gold)

Actually unmodded is a lot harder. Every single mod out there (even Synergies !) is actually easier than vanilla. It's always "more challenge but more OPness" - except that the OPness overcomes the challenge overhead, so you get an easier playthrough.

I'm not a big fan of one-shotting or two-shotting.  "When everything is OP, nothing will be" is kind of the feeling I have about that sort of arms race.  Damage amounts tend to be 0 or 100 in the scheme they have going.

Yes, these are the main issues I had with this game. It requires you to focus your gear on :

• Get enough block/dodge/missile reflection to give you more breathing
• Get 75% DR
• Get the +2k health gems
• Have either a super-shield (Forcefield) or a way to regenerate health very quickly (Berserker) or strong crowd control (some engi skills, knockback..) or just kill trash mobs before they hit you (glass cannon build)

In particular, the mechanism I hate the most is the crits. HCE (without intense farming) is just impossible because of that - you have to consider that any damage done to you can get a random 1.5x bonus.. And this is just madness.

There's also some deadly skills that have been poorly designed - you don't see it coming, and it one shoots you. For example, the Regent kind-of-emberquake literally one shoots you if you're facetanking him, and the only clue that he's about to use it is a slight glow around the ground.. If there are already many particles, chances are that you won't even see it. And yet it one shoots you.

  But it might be very hard to get equivalent balance just with graphs and potions if equipment has its own ideas about everything.  Especially with so much flat damage reduction in the game, which tends to be extremely delicate.  It tends to be OP or useless...

Yes.. I'm aware of the problem, but I think it's just impossible to solve it with mods. First, you would need to get rid of the crit system (that I think is hard coded within the engine). But then, you also have to scale down the damage graph. But then the health potions become OP, because you have non-lethal damage everytime - so you have to nerf them as well. In the end you end up with a very hard balancing task.

You could balance based on the assumption that a glass cannon (someone building only on damage, no survivability at all) should be two-shotted. So you would do a playthrough with a glass cannon character (so that you also have an estimation of his gear), and then at each zone you would change the damage graph so that your character is two-shotted by most of the enemies. Then, once you have that, you would have to nerf the health potion so that it's still a challenge to beat the game.

This would work for most of the classes, but then the Berserker would be even more OP since he doesn't really need any potion to get his health back to full in seconds. There's also all the Life Steal items that would not be balanced anymore.. Ever heard of the Embermage hammer build ? With a particular set (emberweave maybe ? can't remember, it was long ago) I would regenerate my entire health pool with a single swing of my Magma Hammer. This means the only way to kill me is with two or one shot abilities.

So yeah, working to eliminate the one-shot mechanism is an interesting idea, but the amount of rebalancing it needs is quite huge I think. If you just nerf the potions, the game becomes impossible - but if you also nerf the damage graph, all the life recovering abilities become OP as hell.

It's kind of odd how the resistance potions grant a flat amount, and there is only one version.  They're really good in the first two acts of Normal when you run into that sort of enemy, and then they fall off...

Yes, definitely odd. I never really understood what was the design decision behind.

If you look at the Elemental Protection (or something similar) spell it's the same issue - it gives you like +10 to all elemental armor. And it's not even a passive, so you have to activate it everytime.. Anyway, I will cover that when I write the paragraph about spells

  At least there are only five kinds.  It drives me nuts when modern RPGs make you maximize fourteen different varieties of damage resistance just to be viable...

Titan Quest (and by extension, Grim Dawn) has a lot of damage types.. For me it's a lot of complexity for nothing.

Heyya @potterman28wxcv Potty! Really sorry i havent engaged you yet on your planning stage, RL is killing me atm

But i wanted to tell u, if u want ur own subforum for your mod just let me know