Soldak Home   Drox Operative   Din's Curse   Depths of Peril   Zombasite  

Go Back   Soldak Entertainment Forums > Din's Curse > Demon War
Register FAQ Members List Calendar Search Today's Posts Mark Forums Read

Reply
 
Thread Tools Display Modes
  #31  
Old 02-28-2011, 04:02 AM
Bluddy Bluddy is offline
Legend
 
Join Date: Jan 2011
Posts: 1,802
Default

Made some corrections to the skill point count -- for some reason both you and I got it wrong initially. Also added a sheet to compare the offensive spells.

The nice thing about a linear spell system is that it's easier to do the math. 20 levels mean you should need to upgrade about every 5 levels, 15 levels mean every 6.66 levels. The chart looks at monster HP every 5 levels ie. it assumes 20 levels is the max you should need to get your spell ready for the end game, and compares it to the spell power at that level (more or less). For more expensive spells, 16-17 would probably be a more reasonable max. Notice that the table is split into max and min values (which are at the bottom). Some spells have extra effects which are harder to take account of but need to be considered when balancing.

What you can see in the table is that if we consider 20 to be more or less the maximum spell level, with anything beyond that being real dedication to a spell ('overload' level), the max values are pretty decent compared to a generic monster's HP. The problem is the low min values, which drag the average values down (should probably make another table for those).

The shatter and bone shatter spells are insanely powerful as you've noticed. They should probably be toned down a little, or nobody would want to use the other spells even if they're boosted.

BTW I've come to your side of seeing things on this spell topic. Essentially, with the current system, if I'm a mage at level 50 and I want to put just 8 points into something, I have no choices whatsoever. If I'm level 50 and I put 1 point in holy bolt, I should have about 10-15% of monster HP in damage, which I could slowly improve with skill level if I so desire. Instead I get about 5%, which is completely ineffective. Of course it gets worse as I go up in level. To try any spell when my character is at a high level, I need to invest a serious sum of points into it -- otherwise I shouldn't bother. A quadratic function may make things somewhat better, but using intelligence/spirit as a base level for the spell would probably be best.

Because of this weakness of the current mage spell system, maybe we should compensate by making the optimal maximum spell level (ie one that can kill monsters of level 100 with 1 or a few hits) to be even lower than 20. 20 levels for a skill that starts at 1 is around 18% of total points. For a skill that start at 8 it's 30%. A skill that requires only 18 levels to be optimal and starts at 1 takes 15%, and if it starts at a cost of 8 it's 25%. A 16 level skill takes 12% of total points, or 22% if it starts at 8.
Reply With Quote
  #32  
Old 02-28-2011, 07:42 AM
Crisses's Avatar
Crisses Crisses is offline
Elite
 
Join Date: Jan 2011
Location: Middletown, NY
Posts: 636
Default

Too early in the morning for me to wrap my brain around the numbers, but if we're thinking of basing the power of the spells on INT or SPR, perhaps we should also figure out how stat points would build.

There's 2 possibilities to consider: pour all your points into the stat vs. pour part of them into the stat. heh i.e. putting all 5 points into INT until it's completely pumped for those whack folks who want their mage maxed out vs. putting 1/5 to 4/5 into it (average 2.5?) -- for hybrids, or characters who want to actually get dressed in the morning and carry a big stick, for example. I personally think the game is painful without dropping at least a few points into STR/DEX/VIT, but +stat items can help a lot.
Reply With Quote
  #33  
Old 02-28-2011, 02:01 PM
Bak's Avatar
Bak Bak is offline
Elite
 
Join Date: Dec 2008
Posts: 259
Default

Going to toss in my two cents.

- If mages need help (not necessarily agreeing with the "if"), that's probably due to Depths of Peril, with the many threads about how spellcasters had a huge advantage over the fighters. And the recent change of immobilizing spellcasters when they use spells was an attempt to balance the advantage they had on fighters.

- INT increases the chance of criticals, so any damage comparison using the STR bonuses for fighters also has to also account for spell criticals.

- Spell attacks should *never* have their damage match the fighters (IMO). Reasons? Spell casters have many AOE. Pets/summons allow more attacks and draw away attackers. Most spells are auto-hit (so weapon attacks should be multiplied by 70-90% to account for misses). And a spell caster will get several attacks before most enemies get a single swing (except for those %#$%& dark elf priests ).
Reply With Quote
  #34  
Old 02-28-2011, 02:39 PM
Crisses's Avatar
Crisses Crisses is offline
Elite
 
Join Date: Jan 2011
Location: Middletown, NY
Posts: 636
Default

We're really playing a numbers game here.

SOME of the spells are area of effect, or have splash effects. Not all.

Shatter is a great spell to use as an example: That's a mage to be feared. Ok, so not ALL mage spells have to be quite that good, but they shouldn't be so bad that every mage that succeeds has to use Shatter.

For one, when a warrior picks up a new attack skill at level 80, it actually still DOES something. If it doesn't then it should ALSO be tweaked.

If a mage picks up a new attack spell at level 80, it actually does nothing. A level 80 mage needs skills at about 16 to keep up with the monsters -- an investment of 136-280 skill points. How many warrior attack skills need to be at level 16 to keep up with the monsters? Oh, 16 is definitely nice, but once you hit 200-300% crit chances it's not really NECESSARY to keep raising it, right? So a warrior can throw their skill points around more, actually diversify their portfolio. A spell-caster has to very narrowly specialize. Has to. To remain a viable character at those levels, the spellcaster has no choice.

Role play is about choices. About trying on the shoes of some fantasy-land character and taking them for a spin. I want to know why my Wizard can't specialize in fire attacks and ice defense & have teleport even at 1 point and survive. Why bother having 3 trees if I can't take 3 skills on the Fire tree, 2 on the Ice tree and 1 level-1-skill on the Magician tree? And that's only 6 skills! It's not asking too much. But I can only get away with 3 skills that keep up with the monsters, because by level 100 you can only have 3 skills at level 20+, and that's what you need to have for them to even feel it.
Reply With Quote
  #35  
Old 02-28-2011, 03:12 PM
Bluddy Bluddy is offline
Legend
 
Join Date: Jan 2011
Posts: 1,802
Default

Quote:
Originally Posted by Bak View Post
Going to toss in my two cents.

- If mages need help (not necessarily agreeing with the "if"), that's probably due to Depths of Peril, with the many threads about how spellcasters had a huge advantage over the fighters. And the recent change of immobilizing spellcasters when they use spells was an attempt to balance the advantage they had on fighters.
Depths of Peril didn't need any proper character balancing, because you always had a buddy with you to help you out. Also, the huge variety of spells per class made it even harder to balance.

Immobilizing spellcasters was a fine idea in my opinion. Also, the problems with spellcasters don't reveal themselves until later levels. The fact is, very few mages have made it to the end game, and almost all chose shatter as their main spell. Why? Because it's the most powerful spell (one of the few truly quadratic damage spells), and the only one not affected by the bad slope that affects mage spells with level. Even those mages that use shatter, can barely afford any other spell.

Quote:
- INT increases the chance of criticals, so any damage comparison using the STR bonuses for fighters also has to also account for spell criticals.
It's true. However, notice the difference: as STR goes up, your damage keeps climbing and climbing, and every hit is affected. To simplify just a tad, you can get to 250 * 0.6% = 150% EXTRA damage (ie. 250% damage) for the mage classes, if you packed all your attributes into strength. If you have a thief or hunter and use only strength, you can get to 250 * 0.8% = 200% EXTRA damage (300% damage total). If you have a warrior class, you can get to 250 * 1.2% = 300% damage (400% damage total)! This doesn't account for strength raising items or warrior skills that increase damage. Remember that out of the classes listed above, warriors will have the most strength to exploit this bonus. They'll also have the weapons with the highest DPS -- and the top weapons get VERY high DPS. Basically, STR is a super stat, that creates a near-exponential damage output. Let's not forget that strength ALSO gives crushing blows, so that has to be added in as well.

Compare that to INT, which at most, can increase the likelihood of doubling the output of weapons. That's it. That's the most damage it can do. It can increase the chance of doubling of DPS. Much weaker than STR.

Quote:
- Spell attacks should *never* have their damage match the fighters (IMO). Reasons? Spell casters have many AOE. Pets/summons allow more attacks and draw away attackers. Most spells are auto-hit (so weapon attacks should be multiplied by 70-90% to account for misses). And a spell caster will get several attacks before most enemies get a single swing (except for those %#$%& dark elf priests ).
The point about AOE is correct, and AOE spells need to be improved with great care.

Pets -- not all mage classes have them. The ones that have them, have a formula that is very punishing, except for the necromancer who has raise dead and can therefore pick his exact creature level. As I mentioned elsewhere, the summoning formula needs to be improved. To keep summoned pets at my level, I have to keep investing in the skill every 3-4 levels. The fact that the pets multiply doesn't help much because they keep getting weaker and weaker relative to the player and the dungeon. A level 37 pet is completely useless in a level 60 dungeon, but unless you invest 20 levels in the skill, that's what you'll get. The warlock has an even worse deal, as his demons can't even co-exist with each other. I have ideas how to improve this -- creatures should be at most a couple of levels below the player's, and slowly make their way up.

I agree with the spells auto-hitting part. Still, even if you say spells should hit for 70-90% of fighter DPS at high levels, spell damage should be much higher than it is. DPS at high levels can easily reach 2000. Almost all spells can't reach 700 even at spell level 40, which would essentially means putting every skill point you have into 1 skill!

I'm not saying the spell trees are completely broken, like I did before. Having played with it a little, I just think that if the adjustments aren't made to the spells, mages will have an impossible time reaching the end game without using the shatter spell, which is the one really powerful spell they have. Fighters, in comparison, have many ways of reaching the end game and when they get to it, are overly powerful (that's a different balancing story).
Reply With Quote
  #36  
Old 02-28-2011, 03:46 PM
Bak's Avatar
Bak Bak is offline
Elite
 
Join Date: Dec 2008
Posts: 259
Default

Part of Shatter's advantage is that it is instant damage. If the usual "cast time" (like other spells) were applied, something like half a second between when the spell is started to when it does damage, then shatter will not be so overpowered (compared to other spells).

Edit: Shatter is only quadratic if each "grenade" can 1-hit its neighbors, otherwise it is linear (with normal AoE). If that is the case, then the character would have enough skill points to make almost any single spell "1-hit".

Bluddy, you still are not accounting for the advantage of getting in several shots (from a distance) before the creature gets its first swing at the wizard. That's most of what I meant by the wizard damage output should not be equal to a fighter, who is more likely to get stunned or critical-ed while in close quarters.

Quote:
Originally Posted by Crisses View Post
For one, when a warrior picks up a new attack skill at level 80, it actually still DOES something. If it doesn't then it should ALSO be tweaked.
Rupture Armor, Enrage, Gut, Parry, Block... Level 1 skill for level 80 fighter does nothing.

I think you need to refine your argument. It is not "fighter versus spellcaster", at least not directly. It is "fixed amount per level" skills compared to "% increase per level". Wizards happen to have a lot of the former, fighters get more of the latter. Similar to the argument being made about pet summons being underpowered compared to raise dead. It's also true that at low levels, spell casters have a huge advantage over fighters, when their "fixed amount" is better than a % increase for low level weapons.

Maybe this will work: all skills that have a fixed increase per level (like fire or frost bolt) should have that amount reduced, but then add back an amount that scales with skill level *and* character level. This would not be limited to the wizard class, but any skill that currently scales linearly with skill level.

As an example, frost bolt does (1-3)+(4-12)S, where S is the skill level. So a level 5 frost bolt does 21-63, and a level 10 frost bolt does 41-123. Now replace that with (1-3)+(3-8)S+(0.1-0.2)SL, where L is the character level. A 10th level mage with 5th level frost bolt does 21-53 (less damage), but a 40 level mage with 5th level frost bolt does 36-83, a respectable power bump. And a 40 mage with 10th level frost bolt would do 71-143. (Pet summons would need some work, the increase in number of pets, with the increase in pet level per character level.)

Obviously there would need to be balancing, as I pulled these numbers out of a hat, but it would be a way to scale damage as levels go up.

Last edited by Bak : 02-28-2011 at 04:03 PM.
Reply With Quote
  #37  
Old 02-28-2011, 04:18 PM
Crisses's Avatar
Crisses Crisses is offline
Elite
 
Join Date: Jan 2011
Location: Middletown, NY
Posts: 636
Default

Quote:
Originally Posted by Bak View Post
I think you need to refine your argument. It is not "fighter versus spellcaster", at least not directly. It is "fixed amount per level" skills compared to "% increase per level".
Hi, yeah, I'm the person that started the thread here:
http://www.soldak.com/forums/showthread.php?t=3500

about ALL skills needing to "Scale to level". Across the board. Things like Teleport work as expected no matter what level you pick them up at, and thus don't need changes. However there's a lot of spells AND skills that don't scale to level. Already said elsewhere. I tried to drag conversation into a less spellcaster-oriented thread. Because ALL the trees have this issue on at least a few of their items.

Hunter's Lure scales but has the same lagging behind the character that other summoned creatures do. Hunter's Immolation Trap, Freezing Trap -- they're near-useless unless pumped. Etc. I know there's problems on the warrior trees. But there's still more stuff not broken than broken. On the mage tree, there's more broken than not, so that's where we first saw the issue. The issues started in DoP -- but got so much more complicated in DC it was hard to tell what the issue was. People chose hybrids with trees that had skills that work, and other trees fall by the wayside.



Quote:
It's also true that at low levels, spell casters have a huge advantage over fighters, when their "fixed amount" is better than a % increase for low level weapons.
That advantage quells in the 20s and by the 30s it's really easy to tell there's a problem.

Quote:
Maybe this will work: all skills that have a fixed increase per level (like fire or frost bolt) should have that amount reduced, but then add back an amount that scales with skill level *and* character level. This would not be limited to the wizard class, but any skill that currently scales linearly with skill level.

As an example, frost bolt does (1-3)+(4-12)S, where S is the skill level. So a level 5 frost bolt does 21-63, and a level 10 frost bolt does 41-123. Now replace that with (1-3)+(3-8)S+(0.1-0.2)SL, where L is the character level. A 10th level mage with 5th level frost bolt does 21-53 (less damage), but a 40 level mage with 5th level frost bolt does 36-83, a respectable power bump. And a 40 mage with 10th level frost bolt would do 71-143. (Pet summons would need some work, the increase in number of pets, with the increase in pet level per character level.)

Obviously there would need to be balancing, as I pulled these numbers out of a hat, but it would be a way to scale damage as levels go up.
Ok, so now 2 people are with me on the stuff I was saying in the other threads on the subject.... heh.

Before you start citing numbers, take a look at the spreadsheet with real numbers on it. For example, how many times should you have to hit something with a Level 5 frost bolt before it dies? The chart helps us out because at Dungeon level 40 the monster has about 200-250 HPs if I read the chart that Bluddy made correctly. So a 10th lvl mage with a 5th lvl frost bolt would have to hit about 10 times to kill a critter. That could be acceptable. We figure that -- at optimal -- you want to improve a spell every 5 levels, so a "good" level for the frost bolt would be 8 at level 40 mage.

Here's the spreadsheet


Here's the thread where I pointed out this Scaling to Level issue:
http://www.soldak.com/forums/showthread.php?t=3500

Last edited by Crisses : 02-28-2011 at 04:18 PM. Reason: add link
Reply With Quote
  #38  
Old 02-28-2011, 04:34 PM
Bluddy Bluddy is offline
Legend
 
Join Date: Jan 2011
Posts: 1,802
Default

EDIT: while I was writing this, Crisses submitted a post with very similar points. My post may help people visualize the problem better so hopefully I'm not just repeating what she said. Plus I think we can make non-scaling skills work if we adjust them properly, though the full solution is probably to make them scale.

I'd like to try and make by point by images. All these graphs are in my dropbox excel file, which can be found here.

The first graph asks, is it possible for a mage's skills to keep up with monster HP until the end of the game, if we upgrade the skill to level 16, or once every 6 levels or so. The graph shows the average damage of a couple of spells. To get the character level, multiply the spell level by 6. The purple line at the top is monster HP (it was red before being converted to JPEG). The other lines are various spells. Notice how, if we try to keep up our spell power by upgrading only to level 16, we just won't keep up. Initially it might seem ok, but as we keep going, our spells will get relatively weaker and weaker. Upgrading a regular spell to level 16 costs 12% of our total points. Even spells that cost 8 skill points initially won't scale properly, even though they'll cost 22% of our skill points! Interestingly, volcano, which is the strongest spell on paper (shatter is really the strongest), manages to almost keep up with monster HP.

The 2nd graph asks, what if we try to upgrade our skills to 20th level by the end of the game? Will they be able to keep up (ie. not get weaker and weaker) if we upgrade this skill every 5 player levels? The answer can be seen in the graph. Aside from volcano, the slope of each spell is such that with every level they get weaker and weaker relative to the monsters. Not that it's cheap to upgrade every 5 levels: upgrading a skill that starts at a cost of 1 will cost us 19%, while a skill that starts at a cost of 8 will take 30% of our skill points, and still won't keep us in the fight up to the end game.

The 3rd graph tries to answer the question: what if I upgrade my skills to level 30? Will I then be able to fight in the end game? The answer is in the graph. Volcano keeps up again, but other spells still get weaker and weaker relative to the monsters. The cost of our spells, however, is now very significant: upgrading to level 30 costs 40% of my skill points for skills that cost 1 to begin with, and 59% of my skill points for skills that cost 8 points. And I STILL can't really keep up with the monsters! (Except for volcano).

This is why mages can't keep up. The slope of their lines is such that they keep getting weaker relative to the monsters. And that's just the regular monsters -- elite/champion monsters are much harder, not to mention bosses. So mages do fine around level 10 (as you can see in the graphs -- just multiply the spell level numbers by the 6, 5, or 3 for 16, 20, or 30 respectively). It gets really challenging around level 25 as the monsters start opening a gap, and then the monsters just wallop wizards for the rest of the game. Plus, just to try keep up, mages have to invest 1/3 of their skill points or more per skill! No wonder everybody runs to shatter, which is the strongest spell. Shatter's curve is more similar to fighers' curves, which, just for comparison, don't look like straight lines, but like a parabola going up. They easily exceed and trounce the monsters.

What should be done? One problem is that mages can't put only a few skill points into a skill for it to be effective (as other classes can). Mages have to fully invest in any skill. They therefore need to be given more skills to play with: skills should reach end game level by the 20th skill level. Even this costs 20% of skill points, which means by the end game they'll have only 5 skills that can keep up.

Secondly, the skills need to follow a slope that more closely matches that of the monsters' HP. It's fine if the skills diverge from the monsters' HP line, so long as they don't diverge too far away from it. Proper lines for the skills would actually be a lot easier to build if they weren't linear but slightly curved upwards (ie. they multiplied by a certain amount each skill level, say 1.1 or something.)

Finally, as a bonus, check out the last graph. If you thought you were just imagining that fire got deadlier and deadlier with level, you were wrong. Look at how the fire's damage slope (which is what matters) completely dominates the monsters' HP slope with increasing dungeon levels. If we don't want monsters (and pets) to die instantaneously when touching fire, we need the slopes to match almost perfectly.
Attached Images
File Type: jpg spells 16.JPG (40.0 KB, 5 views)
File Type: jpg spells 20.JPG (40.0 KB, 4 views)
File Type: jpg spells 30.JPG (41.6 KB, 5 views)
File Type: jpg fire.JPG (35.1 KB, 6 views)

Last edited by Bluddy : 02-28-2011 at 04:39 PM.
Reply With Quote
  #39  
Old 03-01-2011, 05:48 AM
Bluddy Bluddy is offline
Legend
 
Join Date: Jan 2011
Posts: 1,802
Default

After finding some inconsistent values, I realized I neglected to account for the effect of increasing difficulty levels (normal, champion, elite, legendary), all of which have multipliers that increase monster HP and resistance among other things.

The picture that comes out of this update is very grim. The new graphs are attached. This time, the real monster levels are along the bottom. As you can see, even if we upgrade our spells to level 30, they become completely worthless as we advance along the levels to new difficulties. Being a fire mage gets challenging at level 25 or so. At elite level it's impossible. I also accounted for increasing monster resistance with difficulty.

If my data is right, this situation is unfixable given the current tools. The only way to 'fix' this right now would be to adjust the spell lines so they reach somewhere below the end point of monster effective HP. But this would make mages invincible for the first 3/4 of the game, since spells would be higher than monster HP for that region. The only solutions are either to make spells relative to caster level, or to make the spell graphs quadratic by increasing them proportionally at each level.

BTW if you look at fire, now that I adjusted for difficulty, you can see why it's got the graph that it has. Monsters actually get stronger than fire at legendary level.
Attached Images
File Type: jpg spell 16.JPG (41.9 KB, 4 views)
File Type: jpg spell 20.JPG (43.2 KB, 2 views)
File Type: jpg spell 30.JPG (45.9 KB, 4 views)
File Type: jpg fire.JPG (39.9 KB, 4 views)
Reply With Quote
  #40  
Old 03-01-2011, 06:01 AM
Maledictus's Avatar
Maledictus Maledictus is offline
Champion
 
Join Date: Jun 2008
Location: Haarlem, NH, The Netherlands
Posts: 163
Default

Does the explanation/discussion and the graphic representation take into account the boost (secondary skills that boost/support primary attack) skills?
Same for the charlevel+x factor when creating a new town. I.e. what is the context for the measurements.

Last edited by Maledictus : 03-01-2011 at 06:19 AM.
Reply With Quote
Reply


Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -4. The time now is 10:48 AM.


Powered by vBulletin® Version 3.6.7
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.
Copyright 2007 - 2016 Soldak Entertainment, Inc.