|
RobynJoanne -> RE: =AQ= Stat and Training Overhaul (3/24/2022 13:38:35)
|
Note: There's a TL:DR for those too busy to read the whole thing.
To preface this inevitably long post, I'd like to first address one point that's pervaded this thread since its conception: the exact nature of Cha's overperformance.
Cray provided the basics of what's happening. quote:
Guests provide an output that is well above what they pay for. They should be paying 48% of a spell per turn, but cost 17.5% of one instead. I'd like to give a quick primer so that we may all be on the same page. Spells and skills are worth 200% Melee. A mage's turn is worth 75% Melee, so a mage must pay 125% Melee to bring their damage up to that 200% Melee level. This is the very principle that governs how the MP bar works. We have just enough MP to cast 4 spells, which is 1 per 5 turns under the 20-turn model. That means that mages are assumed to be attacking for the other 4 turns. 4*75%+1*200%=500%. Thus, mages have an early burst followed by a dropoff later in battle to on average equal the output of warriors. Now, guests provide 60% Melee with full stat investment. Since a spell costs 125% Melee, 48% of that 125% is 60%. This is what Cray means when he says they should be paying 48%. In AQ, to gain power, you must pay an equivalent amount. Guests do not do that. Guests currently cost an upkeep of 21.875% Melee, which is 17.5% of 125%. Cha is indeed overperforming, but more specifically, guests are the problem. Legendary Ash explained the reason for this overperformance, but I'll provide another explanation. The source of the difference between a guest's output and its upkeep is Cha weapons. Cha weapons are worth 75% Melee. 25% Melee is sacrificed to be given to the guest. Because one needs a guest to deal the full 100% Melee damage, guests get a further decompression bonus. This decompression bonus is currently 13.125% even though decompression is usually only worth 5%. 21.875%+25%+13.125%=60%. This is how a guest works.
So, there are a couple of problems here.
First, the decompression bonus is wrong. If this could be fixed (as Cray has stated, this can't be fixed), then Cha would be balanced if not for the existence of pets. Cha also provides benefits for pets. This is the reason why Legendary Ash earlier brought up the idea that pets could scale with mainstat. This way, beastmasters would not be any different from investing in Cha and the mainstats. Unfortunately, this solution comes with its own problems, one of which is related to the overall problems with the assumptions required for guests to exist. The other problem is that this damages beastmaster as an identity far too much. Pets are for beastmasters. That is how things are and how they should be, or the name would be quite the misnomer.
The primary problem is that all of this hinges on the idea that Cha is a mainstat, so there are a ton of Cha weapons for beastmasters to use. As I'm sure everyone who's played the game for an extended time will know, Cha weapons are rare beyond belief. Cha weapons justify the output of guests, but assuming the use of Cha weapons is inherently absurd when we do not even have enough of these to cover all 8 standard elements.
There are no easy solutions to the guest problem since as Cray alluded to, guest upkeep cannot be universally changed with a simple fix. It must be manually changed per guest. One solution is to make all weapons one uses while a guest is out pay 25% Melee damage. This would change the assumption of Cha weapons to just be something that is always true, fixing the absurdity of that assumption. We would need to fix the decompression issue as well, which could involve either paying a further 8.125% Melee, reducing guest damage by 8.125%/60%=13.54%, or reducing SP regen by 8.125% Melee. Another would be to just reduce guest damage output accordingly and make guests deal 25% Melee more damage while one's wielding a Cha weapon. Neither solution is ideal, and I leave the devs to their discretion for choosing one of these or creating a solution of their own.
Now, with all that out of the way, I can address some points.
Cha weapons are considered "Magic" because both are worth 75% Melee, so it's convenient for players to see immediately that Cha weapons are worth 75% Melee if classified as a Magic weapon. It'd be quite annoying for a Cha weapon to deal Melee damage but still be worth only 75% Melee. However, this still means that Cha weapons and Magic weapons are not linked inherently. Thus, SapphireCatalyst's suggestion that Cha weapons use Int for BtH is based upon a misconception.
There's also the idea that Cha should be reclassified as a Mainstat. That would have to get into what exactly defines a Mainstat, but for me, it is the ability to use that stat regularly to attack normally. Cha weapons would make Cha close to that definition, the closest of the three secondary stats, but frankly, as the assumption Cha weapons has gotten us into the mess, I'd like to avoid this.
Now, I'll address the other elephant in the room: the nature of Dex as it stands now based on the plans the devs have provided so far. Dex is too strong, which I think is apparent to everyone. It provides almost the same amount of damage as Str but also has MRM, more Initiative, and Status Potency. It needs to be brought in line with the other Mainstats.
First, MRM. The current assumptions have monster hit rate be 85% and the complement of that is 15% for our dodge rate. All items that proc off blocking an attack and getting hit are based off this. I'll use the dodgelasher's favorite tool Titan's Fall as an example. Titan's Fall has the following effect:
quote:
You take +(15/1.4)% = ~+10.7143% more damage but deal (15/0.15/2)% = 50% melee for each blocked enemy attack as "Counter Damage".
The shield makes you take 15% Melee more damage to get a 15% Melee effect. That effect is modified by a /0.15 for the 15% dodge rate and a /2 for the assumed number of hits a monster does per turn. Slime Barrier, which has the opposite kind of effect, has the following effect:
quote:
MC effect is that at start of your turn, you get hit with a fake attack similar to the Titan's Fall shield. This attack is worth (5*0.85/0.85/2)% Melee per hit that you took in the prior turn, and is Heal element.
This effect is 5% Melee modified by *0.85 for being an auto-hit healing effect, /0.85 for requiring you to be hit, and a /2 for the assumed number of hits a monster does per turn.
As we can see, both items work off the assumption that monster hit rate is 85% and dodge rate is 15%. However, this assumption only works if one has investment in Dex and Luk, as those two stats are required to reach the 15% dodge rate. Without Dex, one's dodge rate plummets to below 0%. This isn't a problem right now since all builds are assumed to run Dex due to its nature as one of the primary sources of BtH, but this won't be the case when the stat revamp comes. Unless we change a whole bunch of items to reflect the necessary change in assumptions, I think it's better if Dex just stops playing a role in dodge (Luk can still be assumed due to its nature as a supplementary stat for all builds).
To solve this, SapphireCatalyst brought up the idea of splitting MRM based on Mainstat, but as KorribanGaming pointed out, this solution would effectively neuter dodge playstyles since they cannot choose what type of damage a monster does. I propose a solution with two different implementations that could fix this. Since all builds are assumed to run a Mainstat, we can instead just make all Mainstats provide MRM. That is, instead of Dex/10, we could just make it Mainstat/10. The two different implementations are that this takes Max(Str, Dex, Int) or that it takes Min(Str + Dex + Int, 250). The former is more similar to Dex/10, except it just lets any of the Mainstats take the role. The benefit of the former is that players can overcap. The latter instead caps MRM to what one would get with 250/10 but allows for the mixing-and-matching of the three mainstats to support builds that do not run full investment in any single Mainstat. Both solutions allow for the current assumptions to exist to let dodge be a viable strategy for all builds that run Mainstat.
Second, Initiative. This is fine even if stronger than Str's damage boost. It lets Dex keep its identity of being quick. This, as Zork Knight noted, is worth ~10% Melee right now while Int and Str get ~5% Melee since +105 Initiative is valued at 5% Melee.
Third, Status Potency. As DoTs and status effects in general are unaffected by lean, this benefits FD Rangers. This also plays into Neko's strengths, and Neko is generally regarded as one of the best armors for FD Rangers. Unfortunately, I don't think this benefits FO Rangers as much for the very same reason of lean. It does let FO Rangers have some inherent extra versatility as a unique identity, but there are other solutions to that problem. This also pushes Dex higher than Str since +5 Status Potency is 5% Melee when Dex already has 5% Melee more Initiative. Personally, I like the idea of giving End Status Potency and along with its other buffs recently allows the possibility of reducing the crazy amount of HP End gives currently.
I think these rein in the power of Dex, but they do not really fix its identity issue. For that, I agree with Dardiel in letting Dex have a different damage graph. I prefer the upward-sloping line since that lets Dex contrast Int while Str is in the middle and because I just dislike inconsistency in damage. I'm also hesitant about the excess accuracy idea since with the stat revamp, Dex is no longer inherently more accurate than the other Mainstats. That lies solely in the purview of Ranged weapons, and while Ranged weapons have traditionally been more accurate than their Melee counterparts, the devs have strayed from this recently. I'd rather Dex's damage graph be inherent to the stat rather than be dependent on equipment.
As a final note, I'd like to correct the misconception that Int should also get a buff. Currently, Str and Int are balanced to deal the same amount of damage as I had explained near the beginning of this post. Str is getting a slight buff because Int is inherently more versatile, as mages can use their MP bar that costs 25% Melee weapon damage for anything they choose while warriors are locked to dealing damage. This slight damage buff makes the two stats inherently unbalanced to give Str a leg up for this sacrifice of versatility. Ideally, the two stats would be balanced, but warriors just do not have nearly enough options for this to be reasonable currently. To give Int a buff would be to return to status quo when the entire goal is to fix the problems with the status quo.
This is a massive post, so this is a TL:DR
1) Cha is overperforming because guests do too much for too little. This is due to assumptions of Cha weapons.
2) Cha weapons are valued like Magic weapons, as both pay 25% Melee damage for something, but the two are not the same thing.
3) Dex is overperforming under the current stat revamp plans due to its MRM, Initiative, and Status Potency together.
4) MRM could instead be something that all Mainstats share. This could be done by taking Max (Str, Dex, Int)/10 or Max(Str + Dex + Int, 25).
5) End is similarly overperforming, so we could shift the Status Potency effect to End and reduce the HP End gives.
6) Dex's identity needs to be addressed. Dardiel has good ideas to let it have its own unique damage graph whether that be inconsistent damage or a steady increase over time as the opposite of Int's burst followed by reduced output for the rest of battle.
7) Int should not be buffed. Str is buffed because it's inherently less versatile despite the two being mathematically equivalent. Buffing Int returns the two to mathematical equivalence when that is the opposite of the intent.
|
|
|
|
