-
Link: http://www.runicgamesfansite.com/vbdownloads.php?do=download&downloadid=649
The download link includes a lot of my thinking, but I wasn't sure it directly addressed what @Phanjam had to ask about them. So below is more or less what I was thinking. For now there is only one combined mod, but in theory they could be separated along certain lines, so those have subforums of their own. I probably won't separate them anytime soon, if at all. If I do separate them, it would be after significant balance feedback has been taken into account. For example, my extra classes use skills from the base classes, and I don't like making every change twice in the mean time.
Cooldowns - I don't like everything to be spamable. When everything is spammable, you can end up with swiss army knife skills like Shield Bash where you can right click endlessly, performing multiple functions with no restrictions (assuming you have mana and mana potions). Cooldowns can allow other power levers to be moved to compensate. So instead of spamming a 40% stun chance, maybe you can get a 100% chance stun every 12 seconds instead, and much better damage. Spamming just one skill shouldn't be the answer to everything. Some of this is inspired by League of Legends to some extent, but PvE is slightly different, so some of the extremely long cooldowns aren't quite necessary here. 60 seconds is about the longest that seemed reasonable.
Potions - My experience in vanilla Veteran was extremely binary. Either you get one-shotted, or you heal to 100% in just a few seconds with a health potion. Gameplay shouldn't hinge on whether they can burst through the sustain barrier, with damage completely failing to stick or accumulate otherwise. Mostly, I nerfed potions. They're diluted over a longer duration (30 seconds, but not strictly an 8:30 ratio necessarily). There are also Philosopher potions, which last 15 minutes, but are about half as strong - these are farming-oriented, and can be used whenever you don't need that bit of extra healing rate. All potions are now level-restricted.
Health, damage, and XP graphs - I thought that Normal monsters were too trivial on Veteran and below, so they got a 60% health buff. Since I nerfed potions, I nerfed all enemy damage by 12.5% (I think that's the number I used). Also to compensate for the extra 60% health, Normal monsters grant 10% more XP. The 12.5% figure feels a little low as you get into the final acts of Veteran NG0, but maybe the game is just supposed to get harder regardless.
The classes - Pirate and Ascetic are remixes of existing classes with some custom passives cooked up on the side. They fit in with the above approaches to the game rework without changing the main idea too much. Pirate is a mid-range melee+ranged fighter who passively heals if they've killed something recently, 1-2% max health per second for 1-5 seconds depending on passive level. Ascetic uses either bow or polearm. The Ascetic's health rises or falls passively toward 50-60% or so, and has a 70% penalty to potions. They start with a 20% AS penalty and a 20% damage bonus, losing up to 15 of the AS penalty as they level up the passive. They are slow and deliberate, and benefit a lot from pacts, which help them avoid early deaths when hovering at lower health amounts. Sacrifice lets them double down on their passive, largely for the purpose of lowering their downtime, but in a very double-edged way: They lose 50% current health over 3 seconds, but restore 75% maximum health over 15 seconds. (These numbers could be lowered a bit, but beta testing was way too slow-paced before adding this skill in)
Both classes have some hyperscaling - HP and VIT scale off each other by 10-50%, and DEX and FOC scale off each other by 10-30%. Additionally for Pirate, ranged and melee damage scale off each other by about 20-30%. This last bonus is a clear analog to the usual dual-wielding passives, but designed for Pirate. -
-
Hi MTaur -
Played a bit with the mod, probably will have more thoughts, but here are my initial impressions:
Overall your changes give quite a different feel/pace to the game. With your mod, gameplay requires attentiveness and thought from the player because most abilities have a decently long CD. It won't be to everyone's tastes, but it's definitely something different.
However, as you noted before, it probably won't scale well beyond level 40-50 as the monster damage curve grows more quickly than player Max HP, even with the reductions you've put in place.
Some half-baked ideas to address scaling with respect to one-shots at higher levels:
-- A cheap but easy way would be to either make Max HP also scale quadratically like monster damage, or make monster damage scale linear-ish like Max HP.
-- Have some skills reduce monster crit chance and/or crit damage bonus. Toning down spike damage would go a long way in reducing one-shots.
Also, Heal Bot's numbers seem unchanged but not sure if that's an oversight?
-
-
Hi MTaur -
Played a bit with the mod, probably will have more thoughts, but here are my initial impressions:
Overall your changes give quite a different feel/pace to the game. With your mod, gameplay requires attentiveness and thought from the player because most abilities have a decently long CD. It won't be to everyone's tastes, but it's definitely something different.
However, as you noted before, it probably won't scale well beyond level 40-50 as the monster damage curve grows more quickly than player Max HP, even with the reductions you've put in place.
Some half-baked ideas to address scaling with respect to one-shots at higher levels:
-- A cheap but easy way would be to either make Max HP also scale quadratically like monster damage, or make monster damage scale linear-ish like Max HP.
-- Have some skills reduce monster crit chance and/or crit damage bonus. Toning down spike damage would go a long way in reducing one-shots.
Also, Heal Bot's numbers seem unchanged but not sure if that's an oversight?
Those are some nice thoughts. I probably left Heal Bot alone because it felt weak in vanilla, making it a more noticeable support/tank skill generally. But I might give it a look again and nerf it now. I also don't remember what exactly I did with Forcefield other than that it scales with Charge better now.
The damage reduction for enemy damage could be made to scale. Instead of 87.5% of the base graph, the graph (divided by the original) could be made to trend downward, but I chose the simple linear graph as a first guess and because it's easy to implement, revert, and communicate. Are there certain percentages you have in mind for different level ranges? It could be piecewise linear, or something smoother could be cooked up that hits the same points. -
-
Seeing that I only recently nerfed Stone Pact, there's a chance that Heal Bot is also OP in this mod setting and it simply escaped my attention.
One thing I noticed about my mod is that bosses are more kiteable with the 12.5% damage reduction, but the "trapped in a room" arena challenges can be brutal with +60% HP to normal monsters. The lack of skill spamming combined with enemies that refuse to evaporate can make it a real challenge to juggle your mobility skills with your damage skills to stay alive.
I also find the game getting harder in the forest and in the mines, markedly so. -
-
Are there certain percentages you have in mind for different level ranges? It could be piecewise linear, or something smoother could be cooked up that hits the same points.
I imagine this would require some trial-and-error to figure out what kind of scaling works, especially for a fundamental balance change like this.
The monster damage graph would definitely have to be steeper than the Max HP graph, but it can't really be quadratic like it is now, otherwise it would be fine (maybe even OP) at lower levels but being unplayable at endgame.
Piecewise linear may be more optimal than just simple linear, since simple linear you run into a reverse problem - enemies being proportionally too powerful at lower levels. But piecewise takes more work to get right, IMO.
But I think the most optimal curve would probably be a hyperbola, since it has all the optimal properties - it is concave upwards like a quadratic, but unlike the unlimited growth rate of the quadratic, the hyperbola has an asymptotic growth rate that never exceeds linear. Specifically something like this:
y = (A * (sqrt(1 + x^2) - 1)) + B
You can tweak the coefficient on the sqrt to get something to suit your needs. Some guidelines:
A is the hyperbola's asymptotic slope, and should be greater than the growth rate of Max HP. Something like 60-100 seems reasonable to start with, but should vary by difficulty.
B is the baseline damage of units at any level. This should roughly equal the baseline damage currently (10-26 depending on difficulty).
I think if you manage to get an optimal A and B, this can provide relatively smooth consistent difficulty increases as levels increase, since you'll have control over how powerful monsters proportionally are to players at any given level. -
-
y = (A * (sqrt(1 + x^2) - 1)) + B
This seems like it doesn't take long to straighten out. I think
y = (A * (sqrt(1 + cx^2) - 1)) + B
should allow finer control over the rate of convergence of the derivative.
(The original graph hits 97% slope by level 4 or so, if my quick sketches are correct.)
I don't necessarily want to do something this complicated with the numbers. It seems like more of a NG+ problem regardless.
EDIT: y = ((A/c) * (sqrt(1 + cx^2) - 1)) + B
(It's the same family of functions, but the asymptotic slope is A in this case. I'm also not sure if B was supposed to be Bx + D or something else with a slope at "level 0". Anyway, overkill maybe.) -
-
This seems like it doesn't take long to straighten out. I think
y = (A * (sqrt(1 + cx^2) - 1)) + B
should allow finer control over the rate of convergence of the derivative.
(The original graph hits 97% slope by level 4 or so, if my quick sketches are correct.)
You're of course right about the derivative convergence - those were "off the top of my head" thoughts in the general direction of a hyperbola, but didn't really consider specifics (having not actually implemented it myself
).QuoteI don't necessarily want to do something this complicated with the numbers. It seems like more of a NG+ problem regardless.
At its root, it's not actually a NG+ issue - it's level scaling related. If you stay in mapworks in NG0 past level 50, you will encounter the same problems. Playing in normal/casual may take longer to manifest scaling issues, but they will surface eventually.
But I don't know if you are leaning towards treating this as a by-design thing (like Synergies does).
As for complexity of implementation:
If you are redoing the graphs beyond simple scaling, you would probably anyway have to use/write some generator to create new graph files. -
-
Another idea, probably simpler, is to take the derivative of the existing curves at 40-50 (or wherever), and hold it constant for the rest of the way to 100. Basically piecewise quadratic + linear.
A quick napkin calculation for veteran shows that if you hold past 50, at 100, monster damage would be roughly the same as it currently is for normal. Doing this for elite makes monsters at 100 about 10-15% weaker than their current level 100 veteran counterparts. -
-
Another idea, probably simpler, is to take the derivative of the existing curves at 40-50 (or wherever), and hold it constant for the rest of the way to 100. Basically piecewise quadratic + linear.
A quick napkin calculation for veteran shows that if you hold past 50, at 100, monster damage would be roughly the same as it currently is for normal. Doing this for elite makes monsters at 100 about 10-15% weaker than their current level 100 veteran counterparts.
What is it about vanilla that makes the quadratic scaling not a problem? I hesitate to do something so sweeping. -
-
What is it about vanilla that makes the quadratic scaling not a problem? I hesitate to do something so sweeping.
Basically, damage mitigation stats are kind of broken in TL2.
Players often get around it by farming and stacking DR/MaxHP and block/dodge up to max to avoid being one-shotted. Because if you can survive hits, then recovering in vanilla is easy (as you noted with potions and other defense/healing skills).
Without such farming, vanilla elite would be near unplayable at 100. Even on normal, it's a rather noticeable difference between having decent DR and not bothering with it.
Of course the same is true with your mod too, but it's much harder to deal with since potions and primary defense skills were nerfed so significantly.
Edit: I suspect TL2's design intention was that at high level incoming damage would be overwhelming, but you also had equally powerful abilities to overcome it (shadow burst, forcefield, ice shield, etc.)
So nerfing those abilities may require some counter-balancing on the monster side (IMO of course) to keep it playable beyond a certain level. But if 50+ game balance isn't a huge concern for your design/vision, then apologies for derailing the discussion
-
-
What is it about vanilla that makes the quadratic scaling not a problem? I hesitate to do something so sweeping.
Basically, damage mitigation stats are kind of broken in TL2.
Players often get around it by farming and stacking DR/MaxHP and block/dodge up to max to avoid being one-shotted. Because if you can survive hits, then recovering in vanilla is easy (as you noted with potions and other defense/healing skills).
Without such farming, vanilla elite would be near unplayable at 100. Even on normal, it's a rather noticeable difference between having decent DR and not bothering with it.
Of course the same is true with your mod too, but it's much harder to deal with since potions and primary defense skills were nerfed so significantly.
Edit: I suspect TL2's design intention was that at high level incoming damage would be overwhelming, but you also had equally powerful abilities to overcome it (shadow burst, forcefield, ice shield, etc.)
So nerfing those abilities may require some counter-balancing on the monster side (IMO of course) to keep it playable beyond a certain level. But if 50+ game balance isn't a huge concern for your design/vision, then apologies for derailing the discussion
I think maybe I'll sit on it for now. If we ever get a clearer idea on how to fix high levels, we can do it.
I've never tried an entire NG+ playthrough, and I usually assume that it's like D2 and you just need awesome gear after NG0 and that's the real reason I'm dying. So I usually hesitate to make changes. Is there a way to link damage reduction to NG+ levels, though? I don't remember, and I haven't tried before. (Titan Quest did this kind of thing with pet health once, and it was really jumpy and bad, so maybe that's a horrible idea. But I don't like non-linear modifications, because then everything can go bananas.) -
-
I've never tried an entire NG+ playthrough, and I usually assume that it's like D2 and you just need awesome gear after NG0 and that's the real reason I'm dying. So I usually hesitate to make changes. Is there a way to link damage reduction to NG+ levels, though?
TL2's NG system isn't like D2. NG+ is just a level bump for the areas in the game, so a level 100 NG+ enemy is basically the same as a level 100 NG0 one. Loot and map layouts can be tweaked for NG-specific logic, though.
Linking DR to NG+ may be possible since there's a predefined stat for it ("PLAYER NEW GAME PLUS"). I don't know if this is an integer or a string since I never used it. But if it's an integer (0 = NG0, etc.), it would be easy to make a DR passive based around that.
-
-
I took a look at healbot. The time between pulses shortens too much per level for what I'm going for. At player level 50 and skill level 10, it's 900 every 7.5 seconds, out of 2000 or so base health. Maybe 3000 or 4000 with stats and gear?
It seemed ok at skill level 1, but maybe even then one-pointing it is already pretty strong. Maybe 22-15 seconds would be more in line, or 17-10 seconds. I could have it heal the same amount over 10 seconds instead of five, or something like that.
I looked at the health potions at the same level, and right now, it's a bit stronger than that. Cutting total healing per second down to 1/2 or 1/3 of where it is now would leave it as a skill worth spending points in.
(Apparently I nerfed Forcefield, but it's still very strong. Something like it needs 5 charges just to get its old base value, and which still brings it up to 2000 at rank 10, level 50. And if the wiki is correct, it never used to have a cooldown timer at all... Anyway, losing all of your charge for a 40% max health shield is probably fair.) -
-
I'm remembering one reason why I didn't lengthen the time between pulses in the past... it's tricky to find. I never used to mess with .LAYOUT files. But finally, it's under SKILLS\MONSTERS\SUPPORTBOT\SUPPORTAURA
That was a pain to dig for. A lot of times related files are called by a NAME tag, and the name is apparently registered and fetched automatically without the need to give the actual location in the files. -
-
I almost want to make the damage graph 15*level, which briefly cuts above Elite at low levels and gets overtaken by Easy around level 80 or so. This isn't that great, because Veteran is mostly balanced, but gets harder around 30-50 in the current state. The described change would make the tundra stage a lot harder, and possibly nerf enemies a bit too much going into the last two acts.
Balance is kind of a pain. It doesn't sound like I'd enjoy NG+ very much based on the mechanics that are supposed to keep the player alive. If I mess with the graphs, I'll probably make it a separate mod with higher loading priority that I can disable at will. The amount of testing needed would be large. -