[RELEASE] Golden Sun Reloaded

[Caledor]

With the forum back up, i'm leaving a note that TBS has been updated around a month ago to tone down a bit Hydros Statue and Manticore up a bit.

Moreover the detailed changelog, the one that was found in the former download section of the forum, is now shipped with the patches

[Dark Mits]

Hey Caledor! I haven't posted in a long time. It's good to see that the project is still going on!
I am replaying (3rd or 4th time) through Reloaded. Some notes I have while running a different set up:

- Some Psynergy that can be used in the overworld is not available as "base". This necessitates switching djinn around out of combat often to solve the puzzles. For example in TBS, Whirlwind is only available if Ivan has Venus djinn (but not triclass), or Isaac has Jupiter (without triclass) or Garet has Jupiter (again without triclass). I remember that you explained that this was intended and that you did not want to have them as base on the characters. My suggestion would be, if possible, to add all Psynergies used for overworld puzzles as equippable items (except Move and Retreat of course)

- Killer Ape in Mogall Forest can kill a character from full hp if it uses Ransack twice on a row on them. If intended, you can disregard this :p  After all we need a reason to use Guard and Protect

- Manticore in Lamakan Desert is the first boss that makes extensive use of 5-target damage abilities. This is the first fight (after the Hydros Statue boss change) that requires Wish almost every round. The available djinn up to this point "unlock" Wish with 4 Mercury djinn on Mia, Isaac or Ivan, but not on Garet. No other class can be unlocked up to this point with Wish. At the same time, Resist, which would massively help here to reduce the need of Wish, is only available for Ivan monoelement or Garet with 4 Jupiter djinn but high level, or Ivan with Mercury (which already unlocks Wish anyway). This is where I'm stuck now, I'm trying to find a way to defeat it without having to switch djinn around :p
My suggestion would be to make Resist available at 18+ levels to open up that strategy for this boss.

Looking forward to further updates!

PS: I STILL haven't defeated Dullahan on Hard Mode with Reloaded!

[Caledor]

Hi Dark Mits, sorry for the late reply

- Some Psynergy that can be used in the overworld is not available as "base". This necessitates switching djinn around out of combat often to solve the puzzles.

This has always been the case, even in the base game. It's not a bad idea to add items for every psynergy but at this point it's very low priority for me, not gonna lie.

- Killer Ape in Mogall Forest can kill a character from full hp if it uses Ransack twice on a row on them.

It literally can't do anything else, it's second best move is... Douse. A Dull cripples it. I can't take aways a boss' only strength.

- Manticore in Lamakan Desert is the first boss that makes extensive use of 5-target damage abilities.

Manticore always felt like a joke to me, especially after Hydro Statue. Did i perhaps overtune it while trying to balance them? I'll have to check.

Thanks!

[Dark Mits]

No need to apologize; if anything I should be the one apologizing since I act as a spoiled child ^^ . Continuting with my notes on my latest playthrough. For a change I have switched to tri-element classes once the 7th djinn of each element became available:

- Regarding Manticore, it's not overtuned as a boss. It's mostly that any party setup that doesn't have 4 Djinn set on Mia or Ivan for Wish requires some other way to handle its party damage. The 2nd best solution is to use Breeze and Granite, which would allow for Spritz to be used as party healing in place of Wish. A class with the Resist psynergy would allow for mitigating magic damage without needing to "downclass" by unleashing Breeze, but Resist is learned at a higher level than what the party is expected to be when fighting it.

- Enemies between Tolbi and Venus Lighthouse seem to be somewhat low on Agility. Up to and including the Kraken fight, each turn was usually Ivan and Isaac attacking first, then a fast enemy, then Mia and Garet, and then the slower enemies. Some fights had all 4 party members go before any enemy without any equipment that affected Agility. Granted, both Garet and Mia had their class be one of the faster available for them (Garet was Mars/Mercury, Mia was Mercury/Venus). Enemies after Suhalla Gate "restored" the order.

- Storm Lizard in Suhalla Desert may also be facing the "I ran out of PP" issue that other bosses experienced in long fights. Around 10-12 rounds or so it stopped using any Psynergy.

- Ply Well says that it restores 200hp on the tooltip even though it's 220hp for Reloaded. I forgot if you have mentioned if changing the tooltip causes issues.

- On the subject of healing, TBS has: Cure 70hp, Ply 100hp, Cure Well 150hp, Ply Well 200/220hp, Potent Cure 500hp, Pure Ply 700hp. This makes for a very nice exponential curve but with a gap: Ply ~= 150% Cure, Cure Well ~= 150% Ply, Ply Well ~=130-140% Cure Well, [Vial ~= 150% Ply Well], Potent Cure ~= 150% Vial, Pure Ply ~= 140% Potent Cure. But this assumes Vials will be used for the "gap" between Ply Well and Potent Cure. Do you think that Potent Cure would be more "appropriate" at around Vial's strength of 350-400hp (with its PP cost adjusted accordingly) and Pure Ply reduced to about 600hp? Or if it is necessary for the 3rd stages of Cure and Ply to be at high values to be worth in TLA, would it be interesting to make the percentage differences between the different healing psynergies around 65% like this? Cure 70hp, Ply 110hp, Cure Well 175hp, Ply Well 275hp, Potent Cure 450hp, Pure Ply 700hp. This could allow the Boon series to slot in between them like that: Boon 90hp, Nature Boon 225hp, Vital Boon 575hp

- Deadbeard proved to me that Break is the 2nd worst enemy ability after Djinn Storm  Casting Break followed by Spark Plasma/Freeze Prism, and on the following round starting with Spark Plasma/Freeze Prism always killed 1 player, sometimes 2. (note to self: maybe I should have used Defend when Break was followed by a tier-3 psynergy?)

- Psy Crystals seem to be overpowered, pretty much acting as a Full-PP restore for characters below level 30 (except maybe Ivan with a high-PP class), while Mountain Water loses any value it can have mid-battle when characters learn Psynergy that costs more than what it restores or even when equipped with items that passively restore per turn almost as much as it does. What if Mountain Water restored 30-40pp, and Psy Crystal 100PP?

- On the subject of PP restore options and how powerful they are in end-game, what if items with PP restore only appeared on mid-game equipment and maybe shirts/rings/boots, and every end-game item only had HP restore and maxHP/maxPP in its place? That would increase the value of Psy Crystals, Mountain Water but also of Ember, Ether and Aroma.

Hopefully before next weekend I will have completed TBS and have started TLA!

[Caledor]

Regarding Manticore, it's not overtuned as a boss. It's mostly that any party setup that doesn't have 4 Djinn set on Mia or Ivan for Wish requires some other way to handle its party damage. The 2nd best solution is to use Breeze and Granite, which would allow for Spritz to be used as party healing in place of Wish. A class with the Resist psynergy would allow for mitigating magic damage without needing to "downclass" by unleashing Breeze, but Resist is learned at a higher level than what the party is expected to be when fighting it.

There has to be a power spike in the Hydro Statue - Manticore area, and that's exactly cause you get Wish. It makes an enormous difference. I can only try to distribute this power increase between the two bosses compared to the ones before which is what i tried to do in the latest update. So as long as Manticore doesn't feel overpowering like Hydro Statue felt before, it's fine for me.

BTW, hydro statue was like that cause i always reflexively used the double brute setup which is the best by far, and i tuned it down cause bosses shouldn't force player into a specific class setup.

- Enemies between Tolbi and Venus Lighthouse seem to be somewhat low on Agility.

I don't think i changed agility of random enemies that much, especially in TBS. Are you sure you didn't get agility boosting boots in tolbi and/or swapped classes around?

- Storm Lizard in Suhalla Desert may also be facing the "I ran out of PP" issue that other bosses experienced in long fights. Around 10-12 rounds or so it stopped using any Psynergy.

That I need to check. I don't like bosses running out of PP

- Ply Well says that it restores 200hp on the tooltip even though it's 220hp for Reloaded. I forgot if you have mentioned if changing the tooltip causes issues.

IIRC the text compression prevents me from writing 220 in TBS

- On the subject of healing [...]

I wouldn't set values for that reason. Those mostly depend on what the element can do on other fields. For example, Water has the best healing and the worst offense. Jupiter has the best offense but lacks team healing so gets second best single target offense. Venus has second worst offense but has good team healing so gets worst st healing.
The power spike in the third tier is to make it endgame viable, where team healing is much more important.

What if Mountain Water restored 30-40pp, and Psy Crystal 100PP?

Mountain Water isn't really supposed to be used mid battle after a certain point in the game but retains its value out of battle so i don't see an issue if it remains low. On the other hand Psy Crystal must be balanced with other PP restoring options due to being an endgame item. This to say that while MW won't probably see an increase in power, Psy Crystal might potentially see a decrease if i wanted to make other PP restoring options more relevant.

- On the subject of PP restore options and how powerful they are in end-game, what if items with PP restore only appeared on mid-game equipment and maybe shirts/rings/boots, and every end-game item only had HP restore and maxHP/maxPP in its place?

the duality of PP restoring and PP increasing equipment is integral to the balance of the game after the changes to base psynergy damage from the PP multiplier, which is the cornerstone of the mod. Coupled with high costing mass healing, it makes dedicated healers prefer PP regeneration (and thus losing in offense) and offensive casters like Fateweaver, Incantatrix, Inquisitor etc prefer max PP increases, while classes that can do both, like Miko and Arcanist, have to choose.

Hopefully before next weekend I will have completed TBS and have started TLA!

Hope you'll enjoy your run, keep me updated! ;)

[Dark Mits]

Yep, thank you for the explanations! Completed TBS and transfered data to TLA. Very few notes since last time:

I don't think i changed agility of random enemies that much, especially in TBS. Are you sure you didn't get agility boosting boots in tolbi and/or swapped classes around?

I had switched classes around for Kraken so that Garet went with full Mercury djinn and Mia full Venus, which make both of them faster than their mono-element classes. I definitely didn't use any Agility-increasing items on either of them.

- Tri-element classes seemed to make the run slightly easier compared to duo-element, characters were more durable. Or I remember the duo-element run wrong. Or I became better player since then (doubtful)

- Saturos also ran out of PP. Around 15 rounds in and right after defeating Menardi, he only used Attack and Heat Flash, no Psynergy at all until the end. Possibly Menardi would also face the same issue, but I didn't test it. Fusion Dragon didn't seem to have that problem (does it use Psynergy?)

- For some weird reason, Garet fully replenished his PP when transitioning from Saturos+Menardi to Fusion Dragon.

- Earth Shield, which can be used to restore 150hp, is not usable in combat. I don't remember if you had explained if they were one of those items that wouldn't break when used in-combat and you kept them only off-battle.

- Is Kikuikimonji's drop chance boosted? I got 2 of them without using djinn. Stroke of immense luck?

And yep, I definitely enjoy these runs! I wouldn't be completing them otherwise! 

[Dark Mits]

I found some weird bug with item classes, unless it's intended behaviour. When at 2nd tier class (Jester, Trainer, Necrolyte) then some specific element level combinations show that the character will undergo a class change even though the name remains the same and there are no multiplier changes. I found at least 3 different examples; for one of those I attach screenshots of Felix starting at 0 Venus, 1 Mars, 1 Jupiter, 2 Mercury and given a 3rd Mercury djinni. Same was found with 3rd Jupiter djinni but NOT 3rd Mars djinni. Similar behaviour with Sheba. I didn't test the other characters and item/djinn combinations. I can upload my save if you want me to test it, I'm at Taopo Swamp right after defeating Flower.

Is it related to the change for the item classes where you need 3 element levels of the first 2 elements, then 3 of the 3rd, then 3 of the 4th?


The rest of my feedback is once again not of particular value, since it's on Hard Mode which the mod is not balanced around. But I'll go with it anyway, maybe something might also affect Normal Mode too :p :

- Blitz in Yampi Desert takes 2 actions per turn. Sour a bit further down the line (and the party being 1 level higher) takes 1 action however. Intended because Sour's Psynergy is stronger?

- Sleep and Stun have the same effect, right? There isn't any difference between them except that Sleep can break on damage which makes it "less bad" than Stun, correct? Would it make it interesting (if it is even possible) to make Stun act as the opposite of Seal, where the character can use Psynergy and Defend but not Attack and Item?

- Item classes seem to be amazing at PP efficiency.

- Kindle in Gondowan Cliffs restores 11PP per turn. Possibly too much? It's good for enemies to eventually run out of PP when they can spam 2 MT Psynergies per turn :p

- The bandits outside Madra after meeting Karst are easier relative to regular enemies in Kibombo Statue

- Aqua Hydra seems slow as a boss, even slower than the enemies on the ship. This makes it act always last, which "breaks" the strategy of having 2 MT-healers with one acting between Aqua Hydra's actions.

- Fear Puppet from Necrolyte class may be too powerful with its Stun effect having amazing chance to apply.

---
Some other notes:

- In this run I have to once again mention that losing overworld Psynergy breaks the "fun" a bit. Right now at Taopo Swamp for example, I had to both Remove items and put djinn on Standby to access Growth and Whirlwind, then set them back again in case combat initiated.

- Regarding PP recovery on enemies: What if the amount of PP recovered per round by enemies (bosses and djinn) is such that it allows spamming its strong Psynergy at the start of the battle, but after a while it becomes soft-exhausted and has to rely on other attacks? For example right now I fought Flower in Taopo Swamp which uses Mother Gaia (19PP cost) and Punji Trap (17PP cost). You had mentioned that enemies have their initial PP pool at 5x their level (unless it's inherenty higher) so Flower has definitely over 100PP when the fight starts and can last at least 4 rounds before exhausting it with 13PP/round that it currently recovers. On Normal mode that's the estimated battle time against djinn, so this works great. But "true" bosses have longer fights of over 10 rounds, possibly over 20. So I thought if it would make for "interesting" boss fights to have them resort more on 0PP or entirely physical attacks when they exhaust their PP pool, which means that the player as well needs to also switch Resist for Protect, and a small PP/round recovery rate would allow them to use Psynergy once every 2-3 rounds.

- Are the empty chests at the beginning of Treasure Isle hard-coded to be empty? Can we put even just 1 herb in them? It breaks my heart opening them and getting nothing

- Would it be possible to add heal-over-time Psynergy, like the secondary effect of Coatlicue but of course much weaker?

- In this run I am not picking up any Summon tablets before defeating Valukar in Yampi Cave, to prevent him from buffing himself or wrecking my party with status effects or (the absolute worst) getting to summon Coatlicue. From my 2 previous runs I remember that the difference in fight difficulty is immense between having and not having those extra summons available. Would it be worth looking at it to avoid this "exploit" by either allowing every summon (except Coatlicue!) regardless if they have been picked up by the player or simply limiting him to the basic 16 ones at all times?

- Djinn Stun is also a nice move by enemies that can force a strategy change mid-battle. Do you think it should be given to more enemies and/or made even worse by stunning 2 djinn per character? Or would it be possible to set djinn immediately on recovery instead of standby?

[Caledor]

I need to check many of these, especially balance questions, but for know i can answer:

Are the empty chests at the beginning of Treasure Isle hard-coded to be empty?

Dunno but the point of those is that it has just been raided by another party (Champa pirates IIRC) so they're staying like that to respect the lore of the game.

Valukar

Never considered that he's limited by the specific summons the player has obtained. I don't think i'd change it but if i were to enable everything i'd definitively include Coatli as well, no reason to handpick some. Also it's very easy to play around Valukar's skill by summoning before he does.

Djinn Stun

I think it should be extremely limited cause it's actually frustrating to play against. Enemies messing with your gear shouldn't be the norm.

Would it be possible to add heal-over-time Psynergy

It's something i considered many times but always decided against. It's either irrelevant or favors turtle strategies too much

Stun & Sleep

Yeah they differ mainly by application and recovery chance, otherwise they're very similar. Your idea of Stun sounds like a powered up version of Delude though, so if i change it it's gonna be different from that.

Tri-element classes seemed to make the run slightly easier compared to duo-element, characters were more durable

Tri element classes have the highest stat totals + the gap is often due to defensive stats so, perfectly normal.

I found some weird bug with item classes [...] Is it related to the change for the item classes where you need 3 element levels of the first 2 elements, then 3 of the 3rd, then 3 of the 4th?

yes. I implemented it by creating new identical class lines with different requirements. I didn't consider it'd be possible to swap between different version of the same class step and now i'm wondering if it can be avoided somehow.

[Dark Mits]

Valukar

Never considered that he's limited by the specific summons the player has obtained. I don't think i'd change it but if i were to enable everything i'd definitively include Coatli as well, no reason to handpick some. Also it's very easy to play around Valukar's skill by summoning before he does.

Unfortunately it isn't, due to Valukar getting 3 actions per round on Reloaded mod. I remember on the Normal run I had done (ie. the intended way to play TLA with Reloaded) that Valukar could use Djinn Stun, then Djinn Stun and, if those 2 Djinn Stuns managed to put 3 Mercury and 3 Jupiter djinn on standby, he'd cast Coatlicue. This would become even easier if I had some Mercury or Jupiter djinn already on standby from previous actions.

The second issue is that Coatlicue's hp regen effect does not have a cap. After unleashing Coatlicue, Valukar will regenerate more than 8000 hp per round for 3 rounds. It's effectively a full-heal. If Coatlicue's hp regen effect had a cap at 999 hp per round, it would be manageable.

Regarding Coatlicue:
If I remember correctly, Coatlicue restores a total of 200% hp on its targets, split as: 20% at cast time, 60% at end of first round, 60% at end of second round, 60% at end of third round. Would it make sense to nerf it to a total of 100% hp restored, or to stagger its healing over more rounds? It would still be catastrophic when summoned by Valukar however.

Stun & Sleep

Yeah they differ mainly by application and recovery chance, otherwise they're very similar. Your idea of Stun sounds like a powered up version of Delude though, so if i change it it's gonna be different from that.

I didn't consider Delude, true.

Another idea I have is to change Sleep to always wake on damage taken but increase damage taken by the hit that will wake you (sort of coup de grâce or 100% chance to be critically hit) and also make it last very long if not hit (like 7 rounds), and Stun to have a static duration of 2 or 3 (or 4 or 5...) rounds. This would make Sleep more "dangerous" than Stun, but on average the player would be expected to have control of the character earlier than if Stunned and without having to waste a turn using an Elixir or Restore.

When used by the party on enemies, it would make Stun more preferable of course, except if attempting to execute an enemy with a Djinn of its weak element.