1. Hello!

    First of all, welcome to MapleLegends! You are currently viewing the forums as a guest, so you can only view the first post of every topic. We highly recommend registering so you can be part of our community.

    By registering to our forums you can introduce yourself and make your first friends, talk in the shoutbox, contribute, and much more!

    This process only takes a few minutes and you can always decide to lurk even after!

    - MapleLegends Administration-
  2. Experiencing disconnecting after inserting your login info? Make sure you are on the latest MapleLegends version. The current latest version is found by clicking here.
    Dismiss Notice

New Mage damage formula

Discussion in 'Suggestions' started by brunandes, Dec 1, 2023.

  1. brunandes
    Offline

    brunandes Windraider

    466
    38
    296
    Aug 29, 2021
    7:35 AM
    The other thread has became a mess and as promised, I will post a separate suggestion with my thoughts on the new formula.

    Firstly, let's start with why a change was needed.

    Old damage formula for mages:
    MAX = ((Magic²/1000 + Magic)/30 + INT/200) * Spell Attack
    Breaking down into its simplest components, it is basically some version of:
    TMA^2+TMA+INT, with some other constants thrown in.

    For comparison, let's look at the NL damage formula since it is the simples:
    MAX = (LUK * 5.0) * Weapon Attack / 100
    Breaking down into its simplest components, it is basically:
    LUK * Weapon Attack

    Hence the main difference is just whether it is an addition or a multiplication.
    Older players are already aware of the issue that is caused by the old mage formula, which is the poor scaling resulting from the additive formula at higher levels. Intuitively, you can already tell that the more base LUK a NL has, the more each weapon attack adds to overall damage.

    To address it, the new formula developed is:
    NEW MAX = {(Z²/100000 + 23 * Z/2000 - 8) + [(Magic - Z)²/18750 + (Magic - Z)/5] } * Spell Attack
    Similarly breaking it down into the simplest components, the new formula is:
    Player Level ^2 + Player Level + TMA^2 + TMA
    Clearly, it is still a additive, so intuitively you can tell it will be difficult to achieve the same effect as the NL formula.

    Thankfully, we can test it by doing some simple simulations:

    To start off, the NL formula is extremely simple. The chart below shows the marginal damage from 1 additional weapon attack at each level, assuming the NL only has base LUK and nothing else. The increasing trend shows the scaling in effect.
    upload_2023-11-30_18-27-51.png

    It is slightly more tricky to calculate the marginal damage from 1 tma increase due to the square present in the formula, but it can be done calculating the difference in damage from 2 separate simulations of 1 tma difference (can assume the tma comes from INT, not much difference).

    For the illustration, we shall assume that the mage 200 MA from other sources, is pure INT and adds 5 INT every level up. You can play around with other values of additional MA it doesn't change the trend.
    upload_2023-11-30_18-36-58.png
    Interestingly, the old formula actually leads to some form of scaling by levels due to the squaring (tma)^2.

    However, the new formula offers completely zero scaling by base AP and levels. Instead, the scaling is with respect to the current excess tma you have i.e. the better your current gear, the more dmg the next marginal tma adds.
    Note: this is also the converse with how weapon attack works. Assume same LUK, the NL gains slightly less proportionate damage with each additional WA relative to the last total WA because each WA adds the exact same amount of damage.

    This means that the new formula DID NOT solve the underlying problem of poor scaling at high levels. Instead, it changed the way scaling works to purely reward gear and opened a whole new can of worms. Hence, to the endgame mage mains celebrating your increase in damage and defending the new formula on the other thread: Congrats on your power spike, well deserved, BUT YOU ARE DEFENDING THE WRONG PROCESS TO GET THERE.

    Using the example from above, assuming a lvl 130 Mage has 300 excess TMA and a lvl 150 Mage has 200 excess TMA. This brings their TMA to be the same at 967.
    In this situation, the lvl 130 mage's damage will be a whopping 41% higher than the lvl 150 mage (if we want to assume that both sources of INT includes MW30, i.e. the lvl 150 mage will have additional 13 INT without adjusting gear, the lvl 130 mage is still 33% stronger).
    In fact, the lvl 130 mage is 23% stronger than a lvl 200 mage with the same excess 200 tma.

    To illustrate, let's use the same baseline lvl 130 mage with excess 200 tma and see what happens when he adds incremental INT (up to the max of 350 i.e. lvl 200):
    upload_2023-12-2_8-56-20.png
    In essence, leveling up to gain INT is almost irrelevant in terms of gaining damage now. What you really need is to obtain obscene gear, which gives you the same powerful scaling regardless of level.

    You must be thinking, isn't this pretty bonkers? Can you imagine an attacker popping an onyx apple and getting in effect more than 100 wa scaling, depending on how high their current wa is? Well, the staff's method of 'balancing' end game mages so that the damage don't get out of hand is compensate by shifting the intercept down and cutting into the base damage of the other end of the gearing spectrum.

    This brings me to 2 main issues brought about with the new formula.


    1.Base AP is the basis of damage in old school Maplestory.

    The new formula takes away the agency of players in terms of AP allocation, which is a key feature of OS maple. With all things being constant, 2 players with the same INT and MA should have the same damage. Same goes for STR for warriors, DEX for archer etc etc.

    The new formula completely goes against this principle by penalizing damage upon level up. Let's do some stress testing. The chart below shows what happens when a naked mage failed to add INT and levels up (assuming he has been pure INT up to this point):
    upload_2023-11-30_21-5-20.png
    Congratulations, your mage starts doing negative damage from lvl 125.
    For anyone curious on how it works in reality, I actually did some testing on my lvl 126 bishop.
    upload_2023-11-30_21-14-10.png upload_2023-11-30_21-14-32.png
    Sadly, we don't get to see any negative damage since there appears to be some baseline formula built in to prevent it (and can sometimes do above 1 dmg interestingly)
    Shining Ray damage for comparison:
    upload_2023-11-30_21-15-44.png

    For those people whose response is "just add INT", that is not the point. The point is that we need to respect the fundamental game design. If we go with this, might as well add an "auto allocate AP" button for all mages since the manual selection is obsolete. Guess what, that is modern Maplestory, where AP is prescribed and there is no longer secondary AP.

    Fortunately, this issue can be easily addressed. Just switch Z in the formula (which ignores any input from the player) with actual Base INT. This way, there will not be a deduction in damage just from leveling up. Of course, it does not solve the scaling issue mentioned earlier, but we shall go by the premise that the structure of the new formula is going to be kept.

    2. The new formula places too much weightage on the 'damage from gear' portion.
    Refer to the chart of the baseline lvl 130 mage I have shown above. The reason for this is that the first part of the new formula (Z²/100000 + 23 * Z/2000 - 8) actually results in an extremely small number compared to the 2nd part.

    Because the new scaling depends on current tma from gear, it can lead to some crazy results if the damage of current under-geared mages are maintained. Based on the numbers presented, I presume staff just looked at the damage at TMA cap and adjusted the curve downwards from there. As a result, ungeared mages does no damage, and semi-geared mages get a serious nerf.

    In the other thread, a staff member cherry picked an example of a lvl 130 mage to illustrate that even 'ungeared mages got a small buff'. This was pretty smart way to mislead readers because lvl 130 is actually one of the most buffed levels post change as it is the last lvl most mages get a huge tma bump from changing to a new wand.

    You know what, I can cherry pick too!
    upload_2023-12-1_8-43-43.png
    This is assuming average gear realistic for a new player (all scrolled with 60%, excluded shield because too expensive). We already know a naked mage will be doing negative damage from the first issue i mentioned.

    A lvl 129 mage gets a 32% nerf on their CL, paralyze. Is it not significant? And don't bring in the 'low level mage should only be using ult' argument. We should not be prescribing how a new player should play. More playing diversity is positive for the gaming experience and helps in retention.

    Again, going by the premise that staff has a reason to follow, we can also tweak the weightage to ensure that a realistic scenario as above do not get nerfed as much. It is just a matter to tweaking the constants to ensure that the TMA cap damage still follows what staff wants.

    Based on a quick testing on my own, I came to the following:
    NEW MAX = {(BASE INT²/35000 + 23 * BASE INT/1000 ) + [(Magic - BASE INT)²/76000 + (Magic - Z)/5] } * Spell Attack
    Of course, this still does not solve the scaling issue raised above and INT from gear always beats base INT.

    So where does that lead us?
    My suggestion to staff is either
    revert the changes and go back to the drawing board for a better scaling solution (I hope)
    OR
    review the current formula to at least remove some of the counterintuitive features.

    Feel free to take the suggestion with what you will, I have decided that I am no longer going to continue playing much. The changes that have happened over the last 2 years has resulted in a server that is far removed from the OS Maplestory experience I was looking for back in 2021 and the constant deterioration of new player experience is not something that bodes well for retention and long term future. So, thanks for helping to break my addiction, I guess.
     

    Attached Files:

    Last edited: Dec 2, 2023
    • Great Work x 20
    • Informative x 2
    • Like x 1
    • Disagree x 1
    • Creative x 1
  2. fartsy
    Online

    fartsy Skelosaurus

    1,272
    683
    442
    Jun 29, 2017
    Male
    5:35 PM
    Fartsy
    F/P Wizard
    Pasta
    that looks like more damage than my night lord at 130
     
    • Informative Informative x 2
    • Agree Agree x 1
    • Disagree Disagree x 1
  3. DrJason
    Offline

    DrJason Windraider

    429
    188
    301
    May 30, 2018
    Male
    Canada
    6:35 PM
    DrJason, Zrar, WhyDaggers
    Islander, F/P Arch Mage
    Oddjobs
    why not just add a single line of logic:

    damage=maximum(oldDamage,newDamage);

    so we keep the buff while the off-meta builds don't have to complain.

    Good for you for recognizing you need to step away from the game though (hard to phrase this without sounding sarcastic, which I am not), this 20 year-old poorly-designed nostalgia-fueled game is best enjoyed when played as casually as possible imo
     
    • Agree Agree x 5
    • Informative Informative x 1
    • Creative Creative x 1
  4. Subterlabor
    Offline

    Subterlabor Wolfspider

    552
    95
    315
    Mar 13, 2023
    4:35 PM
    SweatLord
    Bowmaster
    177
    Toilet
    your NL almost has more tma than my bishop though
     
    • Agree Agree x 2
    • Great Work Great Work x 2
    • Like Like x 1
    • Creative Creative x 1
  5. Subterlabor
    Offline

    Subterlabor Wolfspider

    552
    95
    315
    Mar 13, 2023
    4:35 PM
    SweatLord
    Bowmaster
    177
    Toilet
    I just don't get why when people have been complaining about early game experience we once again punish them and buff deep end game mages with the highest ROI being 180+

    Why are we punishing mages that are MP washing, underfunded or making mistakes by leveling wrong stats? Wands are already ridiculously overpriced yet they are even more valuable now since tma scales better than raw int.
     
    • Agree Agree x 3
  6. OP
    OP
    brunandes
    Offline

    brunandes Windraider

    466
    38
    296
    Aug 29, 2021
    7:35 AM
    This is a band-aid solution instead, I hope staff can do better. Doing a cap cup just makes damage plateau for a number of levels. Yes it is better than negative, but not that much better.
     
    • Agree Agree x 1
    • Disagree Disagree x 1
  7. OP
    OP
    brunandes
    Offline

    brunandes Windraider

    466
    38
    296
    Aug 29, 2021
    7:35 AM
    To be accurate, the biggest buff is actually not to high levels, its is to best gear. Levels matter very little under the new formula now. If you can find away to get insane gear as a lvl 130, you will doing almost the same damage 70 level ups later.
     
    • Agree Agree x 1
    • Great Work Great Work x 1
    • Informative Informative x 1
  8. ScarletMoon
    Offline

    ScarletMoon Orange Mushroom

    37
    12
    48
    Oct 5, 2020
    Male
    7:35 AM
    There seems to be 3 problems with the magic damage function, with one problem being introduced in the most recent function given.

    1st problem is level as a parameter of damage function. All other classes damage is a function of primary stats, secondary stats and weapon attack. While mage damage is a function of primary stats, magic attack and level which is why it does not look intuitive.

    2nd problem is int to magic 1:1 conversion. Staff mention that the reason magic damage formula needs to be changed is because mages scales too well with level. This is to be expected because int >= magic attack. If there is a stat for other class that not only increases primary stat but weapon attack as well, you will see the exact same issue. Especially since mage scrolls are not that different from other classes, where weapon 30% scroll increases 3 primary stat and 5 magic attack which is about equivalent to 1.6 level. Whereas for other classes the same weapon 30% scroll with similar stats is about equivalent to 5.6 levels (assuming 1 level = 5 primary stat).

    3rd problem is magic damage function is a quadratic function. Even though other classes damage function multiplies (primary stat + secondary stat) and weapon attack, their damage scales more linearly as compared to mage's magic ^ 2. This also makes balancing difficult because quadratic function scales very well, it has to be divided by a large constant to keep higher level mage in check which causes lower level mage with lesser stats to suffer.

    To resolve the magic formula problem, maybe these 3 methods can be adopted:
    1. Untie Int to magic 1:1 conversion and make them separate like every other class or
    2. Rework mage scrolls so that stat gain from the scrolls are higher and more in line with other classes or
    3. Make magic damage function nlogn (it doesn't have to be log base of 10, lone or any others works as well) so that it scales slower and difference between higher level mage and lower level mage are not so apparent.
     
    • Great Work Great Work x 4
    • Agree Agree x 2
    • Like Like x 1
  9. beegoratto
    Online

    beegoratto Skelosaurus

    1,204
    338
    426
    Sep 22, 2021
    Male
    3:35 PM
    leetoratto
    Bowmaster
    1
    Nimbus
    I don't have any personal input on this debate but I just wanted to call out this post really quick as legitimately great work. You've identified problems and offered easy to understand analysis and viable solutions without the use of subjective and impossible to quantify or overly emotional arguments.
     
    • Agree Agree x 5
    • Friendly Friendly x 1
  10. ScarletMoon
    Offline

    ScarletMoon Orange Mushroom

    37
    12
    48
    Oct 5, 2020
    Male
    7:35 AM
    Thanks~

    I thought a bit more and the 2nd problem identified is off. I think the problem lies in magic = int + magic attack. But the rest of the point still stands so I think I will leave it.

    That being said, maybe the first method initially suggested can be broken into 2 different method.
    1. Untie int from magic ie magic = magic attack and use a formula similar to other classes.
    2. Make magic = int + magic attack * 5 so that the difference between level (ie stats gained) and magic attack from equip can be treated separately
     
    • Like Like x 1
  11. OP
    OP
    brunandes
    Offline

    brunandes Windraider

    466
    38
    296
    Aug 29, 2021
    7:35 AM
    Honestly I don't see mage's original formula as too much of an issue. It was just part of neckson's design to make magic unique from physical attack. What should have been done to make mages better bossers should probably have a rework to have bigger impact from wdef/mdef and more features that might require the use of ults and/or freeze/poisoning.

    If the aim is really to address the high level scaling issue, the most obvious way would be to copy the normal weapon attack damage formula. But I do acknowlwdge staff's point that this is very difficult and can have wide economic repercussions e.g. earring/shield scrolls that add ma suddenly becomes even more valuable.
     
  12. DrJason
    Offline

    DrJason Windraider

    429
    188
    301
    May 30, 2018
    Male
    Canada
    6:35 PM
    DrJason, Zrar, WhyDaggers
    Islander, F/P Arch Mage
    Oddjobs
    The solution you are describing is essentially the post-Big Bang magic formula. Where instead of a quadratic factor and the 1 INT = 1 magic conversion, post-BB uses a familiar multiplier*INT+LUK*mastery formula. Obviously Neckson realized that the 2002 formula wasn't working and changed it to line up with the other classes in 2010. This update also of course changed all magic skills to have % based damage (i.e. instead of 570 power, Meteor would be like 2000% power) and mages had visible damage ranges like physical attackers. This would require a *HUGE* re-work of the database, all skills, all mage gear, etc. imo not in the scope of an 8 year old nostalgia server (but it is undoubtedly a better way to handle mage balancing).
     
    • Agree Agree x 2
  13. Edann
    Offline

    Edann Slimy Retired Staff

    240
    122
    235
    Oct 24, 2020
    1:35 AM
    Edann
    I/L Arch Mage
    200
    Pasta
    Any reason you started a new thread other than having to deal with the criticism your opinions have gotten in the previous one? A thread isn't a mess only because you're being disagreed with you know. I'm asking since you seem to ignore those entirely both back there and in this new post.
     
    Last edited: Dec 1, 2023
    • Agree Agree x 8
  14. OP
    OP
    brunandes
    Offline

    brunandes Windraider

    466
    38
    296
    Aug 29, 2021
    7:35 AM
    This is a suggestion thread.
     
    • Great Work Great Work x 2
    • Creative Creative x 1
  15. UnknownCode
    Offline

    UnknownCode Nightshadow

    677
    229
    350
    Jun 6, 2021
    California
    3:35 PM
    Islander, Bishop, Bandit
    0
    Speenies
    We're having the league of legends "mid lane meta" chat in maplelegends now.

    aka, buff mages
     
    • Disagree Disagree x 1
    • Great Work Great Work x 1
  16. Tobi3
    Offline

    Tobi3 Slime

    19
    4
    15
    Nov 22, 2023
    Male
    3:35 PM
    He wasn't understood. In similar to beegoratto's post, they're trying to be productive. I think these discussions are good, and it helps the community come together on ideas.
     
    Last edited: Dec 1, 2023
    • Like Like x 3
  17. UnknownCode
    Offline

    UnknownCode Nightshadow

    677
    229
    350
    Jun 6, 2021
    California
    3:35 PM
    Islander, Bishop, Bandit
    0
    Speenies
    The thread was already created with a similar title, it feels more like he's avoiding the criticism that he received within the previous thread so he created a new one so that it looked more cleaner, it also spams the forums which I also used to do and got in trouble for doing so, so not sure why they're doing the same thing.
     
    • Agree Agree x 1
    • Disagree Disagree x 1
  18. OP
    OP
    brunandes
    Offline

    brunandes Windraider

    466
    38
    296
    Aug 29, 2021
    7:35 AM
    Which criticism did I try to avoid other than from misinformed people who didn't look at the numbers?
    Numbers that are now shown in this thread.
     
  19. fartsy
    Online

    fartsy Skelosaurus

    1,272
    683
    442
    Jun 29, 2017
    Male
    5:35 PM
    Fartsy
    F/P Wizard
    Pasta
    show with mw20/30, wizard pots and echo
     
    • Agree Agree x 1
  20. OP
    OP
    brunandes
    Offline

    brunandes Windraider

    466
    38
    296
    Aug 29, 2021
    7:35 AM
    To embarrass the new formula further? Wow you are harsher than me. Will do, wait for an update.
     
    • Great Work Great Work x 1

Share This Page