Suggestion: Overall Skill Rebalance

• Rydian wrote on 2016-08-05 22:21
  Overall Skill Rebalance

  
  
  Mabinogi is an old MMO that's been under development (more or less) for many years. With the changing landscape of games in general and the multiple changes in developers/directors, a lot of Mabinogi's skills aren't well-balanced (even with multiple targeted attempts).
  
  Balanced skillsets should provide choice. Multiple skillsets should all be viable decisions with different upsides, downsides, and capabilities. Sometimes existing skillsets or meta will need to be changed to make changes in other skills viable or attractive alternatives once they're functioning properly.
  
  In this thread I'll go over every skill available and discuss their downsides (or broken aspects) as well as suggesting changes/fixes to many of the game's skill-centered problems. I will attempt to be as objective as I can, but feedback and suggestions are welcomed. I plan to post this for more feedback on the official forums after polishing it up too.
  
  A few things to note...
  

  1. I'm covering skills available to humans because it's where the majority of my experience is, feel free to chime in on the elf/giant skills I didn't touch.
     
     
     
  2. Many skills need rebalanced training. A major part of Mabinogi gameplay is skill training because many skills require you to put yourself in specific scenarios or accomplish specific objectives in order to train them. Unfortunately a number of skillsets were released with horrible training. When Grapple Shot at rank A is slower than Windmill at rank 5, there's a major design problem.
     
     
     
  3. The skill order is from my skill lists and focuses on use so excuse the ordering.

- Life
The art of being a village NPC in an MMO. Life skills are the most varied within a skill tab, and in general either need minor changes, or additions to crafting in general to be more desired skills. Wearing the master title for a production skill should also increase the minimum roll for goods. All production skills should have an entrust option (like with Enchant), given the rarity and value of many materials nowadays.

[spoiler=""] Tailoring
Design: Give players the ability to create unique and high-quality clothing.
Reality: Fails on both counts. Most unique clothing comes from other sources, and the quality bonuses on many objects mean little to nothing.
Changes: All the unavailable tailoring manuals need to be given sources in-game (such as fishing). In addition, tailoring manuals for clothing that's no longer available need to be added as well. Defense bonuses should be increased and non-body items need better stat bonuses.

Rest
Design: Let players sit down and increase their passive recovery speeds.
Reality: Increases some speeds, but is capped so cannot increase others.
Changes: Uncap the skill and even out the recovery increases (even increase at all ranks).

Campfire
Design: Make a fire for additional bonuses with other skills.
Reality: Rest bonuses mean little to nothing and cooking is better done at cooking NPCs.
Changes: Resting near a campfire should greatly increase wound recovery. Campfires should also either have specific meals that can only be cooked on them, or should give general bonuses to cooked food depending on the wood quality.

First Aid
Design: Let players heal wounds using bandages.
Reality: Works, but is capped (so players can't get the full potential) and is awkward to use on the fly.
Changes: Uncap the skill and remove the requirement to be sitting for full effect.

Weaving
Design: Allow players to make intermediary materials for other skills.
Reality: Works, more or less, but the RNG with quality is aggravating for training and producing goods.
Changes: Divide fabric/silk qualities into their own recipes.

Handicraft
Design: Let players create various tools and random objects.
Reality: Most items in the list don't serve a purpose.
Changes: Tools made should have high durability bonuses, and more tools should be craftable.

Refining
Design: The skill players use to turn ore into bars.
Reality: It's a skill that turns ores into bars.
Changes: -

Blacksmithing
Design: Allow players to create high-quality and special weapons.
Reality: Many weapons cannot be crafted and attainable stat bonuses are small.
Changes: Stat bonus ranges need to be increased (like up to 10 per stat), and like with Tailoring we need access to the many manuals we do not have access to.

Cooking
Design: Combine foods for hunger restoration and stat buffs.
Reality: Hunger restoration is not needed in general. Stat buffs can take a lot of effort and might hit the stat cap for endgame players. Eating dishes normally causes unwanted weight changes. Many types of cooking take too long to be worth it.
Changes: Make flour and other high-effort ingredients buyable from normal NPCs for gold. Make weight values a part of character looks/creation. Make temporary stat increases from cooked food ignore the stat cap.

Herbalism
Design: Pick herbs for use in other skills.
Reality: You pick herbs for use in other skills. Many herbs are available easily, others not so. The requirement to pick a white often gets players stuck for a long time. Oranges to be picked also really slow down progression without the help of others.
Changes: Make more herbs (white/orange) spawn in more places weaker players can reach. Herbs other than base herbs should randomly spawn in "wild" locations in the overworld, maybe even use up some of that empty Iria space...

Potion Making
Design: Make healing and recovery potions for use.
Reality: Craftable potions have not scaled to modern stat values. Making anything above 100 is a labor-intensive process using unstackable materials.
Changes: Bump all crafted potions up a tier. 50 as the base, 300 craftable without clovers, etc.

Fishing
Design: Fish up fish and random trash or goodies.
Reality: Most players have little to no need for fish and there's better ways to get random items. Can be played manually but there's little incentive to do so.
Changes: Manual fishing should have much better chances at rare things. Any other suggestions? Some use for fish that would consume them, like making them into high-stat dishes or something?

Production Mastery
Design: Control skill that grants higher success rate overall.
Reality: May or may not work with newer skills.
Changes: Make it work on newer skills.

Taming Wild animals
Design: Tame animals to have them join your side.
Reality: Tamed entities are still weak and have little to no use to make up for the difficulty taming them. Tame a monster and it ends up fighting other monsters around the same difficulty as it, meaning it doesn't last long. Also the stat boosts you get from a tames animal are too small to matter.
Changes: Tamed entities should get stat boosts (+50% HP/MP/SP and base stats, +25% defensive stats) so they're helpful for the time they exist. They should be strong enough to take down at least a few of their own kind before getting killed. The stat boosts from tamed animals should be multplied based on your skill rank, up to 5x(?) at rank 1.

Metallurgy
Design: Get ore fragments and gems from sifting through sand.
Reality: Works, if RNG lets you rank it. Gives tiny gems.
Changes: The training needs to be way less reliant on luck. Methods need to have higher total possible counts. Gem sizes should be increased.

Carpentry
Design: Create nice firewood and bows.
Reality: Pretty small list of craftable items, really grindy skill. No minigame so results are RNG.
Changes: Could have more items added to the crafting list. More bows to fill out the ranks, and better possible stat bonuses so players want to craft them... and a minigame so player skill has a part.

Wine Making
Design: ... ?
Reality: ... a few stats?
Changes: Should this even impact the game, and if so, how? They added a minigame and there's some tactics to ranking the skill, but it needs more benefit. Perhaps additional grains could make different types of alcohol that act as food that gives status effects instead of stat changes.

Rare Minerology
Design: Dig up minerals for use in other skills.
Reality: Works.
Changes: -

Hillwen Engineering
Design: Craft guns and other weapons/arms from minerals and such.
Reality: Crafting rolls are out of the player's hands, there's no minigame so player skill has no effect. Stats are rolled individually so objects are often crappy in at least one aspect. The training is terrible (less than 0.01% per creation at higher ranks).
Changes: Consolidate weapon/armor quality so you don't end up with a 95% roll on one stat and a 2% roll on another. Add a minigame so player skill actually plays a part. Change a lot of the training to work off of results and not pure grind.

Shyllien Ecology
Design: Collect magic crystals and monster parts for use elsewhere.
Reality: Works.
Changes: -

Magic Craft
Design: Craft wands and other weapons/armor from crystals and monster parts.
Reality: Same issues as Hillwen Engineering.
Changes: Same as Hillwen Engineering.

Tasting
Design: Control skill to boost food effects.
Reality: Works.
Changes: -

Ingredient Hunting
Design: Kill enemies for leather and special food drops.
Reality: Weird, but it works.
Changes: -

Catering
Design: Allow players to merge cooked food for higher buffs to the party.
Reality: Works, but strong buffs can be impeded by the stat cap for high-level players.
Changes: Like mentioned earlier, make food effects ignore the stat cap to keep it relevant.

Sheep Shearing
Design: Allow players to shear sheep for wool.
Reality: Wool is basic. Master the skill and you produce the same items as a newbie.
Changes: Multiple tiers of wool you can shear as you rank up, with higher tiers being better at making higher-quality fabric. Higher quality wool can be a chance from normal sheep, and/or sheared from some monsters.

Mining
Design: Mine ores for use in other skills.
Reality: Mines ores for use in other skills. Also wants you to mine useless alch stones.
Changes: Change the training to not need the alch stone mining.

Egg Gathering
Design: Gather eggs for use in other skills.
Reality: Same as with shearing, eggs are eggs no matter your rank.
Changes: Players with at least a specific level in Egg Gathering will be able to gather from more exotic creatures, like the wild ostriches. These special eggs will have higher-tier uses in cooking and such, use them instead of normal eggs in recipes and get different, higher-stat food items out of it. Maybe even gather eggs from pacified monsters at high ranks.

Mushroom Gathering
Design: Gather mushrooms.
Reality: Gather mushrooms.
Changes: Should really not be a skill, Production Mastery should take care of random gathering like this. Unless you all want hallucinogenic mushrooms or something.

Milking
Design: Milk cows.
Reality: Milk cows.
Changes: Unless we're going to be milking Glas or something, I don't really see a use for this skill either.

Harvesting
Design: Gather wheat and such.
Reality: Gather wheat and such.
Changes: Could use more products to gather at higher ranks for use in higher-ranked cooking. Maybe different types of wheat and such for use in a Wine Making potential alcohol addition.

Hoeing
Design: Gather random products.
Reality: Gather random products that don't have much of a use.
Changes: Undecided. Should this stay and get additional benefits? Should flowers have a better use? Or should it be merged with Harvesting?

Commerce Mastery
Design: Carry goods from one place to another for profits.
Reality: Uninteresting to train because it's just unloading goods.
Changes: Training needs to be spiced up. Things like "deliver a load without running into bandits" or "defeat at least 5 bandits on the way", for example. Training also needs to automatically count multiples of profit values. Right now if you would make 10K profit you need to sell in two batches to get two 5K training counts. It should do the math to know it's two counts by itself.

Transformation Mastery
Design: Collect monster souls to transform into them.
Reality: Works, is fun.
Changes: Should let you use more skills while in the monster form. We can't use basic bolts or archery or anything like that even if the monster normally can.

Exploration Mastery
Design: Find artifacts, sketch things, and do quests.
Reality: Gives additional benefits when doing these things. However you can only get into treasure dungeons if you're the Treasure Hunter talent, which is crap.
Changes: Allow access to treasure dungeons for players that aren't the paid talent. Bigger changes would be needed to exploration in general to make it more interesting.

Artifact Investigation
Design: Analyze and attempt to restore artifacts for gold and experience rewards.
Reality: Huge cooldown means the skill is only useful on very special/rare things you can't risk losing. Also the cooldown makes training take forever.
Changes: Reduce the cooldown by a huge amount and remove the gold rewards.[/spoiler]

- Archery
The art of remotely installing new pointy sticks into somebody. Bows in general are weaker than comparable melee weapons across the board, so some skill changes could be used instead. In general, ranged attacks need to calculate their range from the outside of a hitbox, not the center. With the way it is now, targeting the center can often put you within melee distance on really large enemies.

[spoiler=""] Ranged Attack
Design: Use bows to attack enemies from a distance.
Reality: Works, but archery skills are prone to missing, something other ranged attacks don't do.
Changes: Normal archery attacks should always hit after 75% aim, with further aim increasing damage (up to 25% more)?

Magnum Shot
Design: Ranged version of smash, high single-shot damage multiplier.
Reality: Works, but is too slow to aim at a distance.
Changes: Magnum should always aim at the same median speed (as if the target was halfway) regardless of distance.

Arrow Revolver
Design: Rapid-fire fast attacks to keep an enemy pinned down and/or chip away at it.
Reality: Works, though mainly for pinning now.
Changes: Potentially copy the changes to normal attacks, but that might be too strong in practice.

Support Shot
Design: Attack an enemy with a fast, weak attack that sets it up for the next melee hit.
Reality: Sort of works, is used in specific situations but not much in normal combat. Also a total pain to raise to acceptable levels in the first place.
Changes: Should buff the next attack from all sources, not just melee. Training needs to be greatly increased to encourage people to rank and use the skill, right now it's a horrible grind where you have to use it instead of normal attacks until your arm falls off.

Urgent Shot
Design: When you miss, you can fire an emergency shot to recover.
Reality: Is annoying to use in emergency situations because it's two actions.
Changes: Change it to be one action. Triggering the skill should fire it.

Crash Shot
Design: A ranged AOE, shoot an exploding arrow at a target.
Reality: Hits limited targets and aims too slow, other players will usually kill things before you fire. Takes a long time to train due to grind as well, and the cooldown makes using it repeatedly annoying.
Changes: Needs to aim much faster and hit more or all targets in range. Increase the points available from killing enemies with it so training is faster. Reduce the cooldown (from 7/6-5.5) to an even 6-4.

Bow Mastery
Design: Mastery skill for bows, increasing their effectiveness.
Reality: Works.
Changes: -

Crossbow Mastery
Design: Mastery skill for bows, increasing their effectiveness.
Reality: Works.
Changes: -

Mirage Missile
Design: Infect enemies with damage over time that spreads.
Reality: Works too slowly and doesn't take enough stats into account.
Changes: Should take the player's damage range (including enchants) into account. Should also tick faster, but have a lower total duration.

Final Shot
Design: Temporarily buff ranged aim speed and add teleporation.
Reality: Has a long loading time left over from previous updates, and doesn't last very long. In addition it doesn't load skills smoothly.
Changes: Remove the loading time and increase the duration (from 20-30 seconds) to 20-40 seconds. make ranged skills load and aim smoothly like they do when the skill isn't active.[/spoiler]

- Lance
The art of making people feel like they got hit by a truck. Lances have unique mechanics, but unfortunately the minimum attack range doesn't work well. Mabi has lots of position lag issues, and since these aren't getting fixed (being inherent in how the game operates), the minimum distance to attack with lances should be removed because it does nothing but unfairly hinder players who use lances.

[spoiler=""] Lance Counter
Design: Special counter for lances, deals more damage but cannot be controlled properly.
Reality: Being locked into it, it going into cooldown even in use, and not being able to cancel it makes this skill just... bad.
Changes: Remove some of the restrictions

Lance Charge
Design: Charge at an enemy, hitting others in the way, with more damage for longer charges.
Reality: Works, but you tend to get your position desynced, the target enemy is not properly knocked down, and it won't hit enemies in the finish state on the way, making training horrible.
Changes: Set the player position to the enemy's current (not previous position facing them) at knockdown, knock down the enemy properly, and hit enemies in the finish state on the way there.

Lance Mastery
Design: Control skill for lances, adding damage and a hit to the combo.
Reality: Works.
Changes: -[/spoiler]

- Combat
The art of sticking the pointy thing into the squishy thing that bleeds. Many combat skills affect other talents, so suggestions will vary.

[spoiler=""] Combat Mastery
Design: Control skill for normal melee, giving health and damage.
Reality: Works, though gives double bonuses for dual-wielding.
Changes: Should only grant bonuses once.

Defense
Design: Block the next attack, taking reduced damage.
Reality: Damage works well, taking both defense and protection into account.
Changes: -

Smash
Design: A strong defense-piercing attack against an enemy with some side effects to boot.
Reality: Works, but the effects are unreliable due to the activation chances. Also prevents an enemy from being attacked by other things that should hit it.
Changes: Increase the chances of the special effects, and increase the prot reduction for axes to 10. In exchange, lower the max bleed of swords (from 20%) to 15% per tick and reduce the axe debuff duration (from 22) to 10 seconds. Let other attacks hit the enemy too.

Counterattack
Design: Counter an enemy's next attack, dealing increased damage and pushing it away.
Reality: Effectively a single-hit combo, but locks archery from loading any skills.
Changes: Allow archery skills to load during the recovery. It's tempting to increase the damage multipliers to make up for the fact that it's just one hit, but that'd be dangerous for players to go against.

Windmill
Design: AOE attack that damages and knocks away enemies.
Reality: Works.
Changes: Could use more training points from weaker monsters.

Critical Hit
Design: Control skill for crits to deal increased damage.
Reality: Works, if slow to train.
Changes: Could use slightly faster training.

Charge
Design: Charge at enemies, shunting off ranged attacks on the way.
Reality: Works well, if a little awkward due to the minimum distance, and requiring a shield.
Changes: Reduce the minimum distance for the skill (from 400) to 300. The cooldown already prevents spamming. Allow charging with no shield either at rank 5 or 1.

Assault Slash
Design: Jump at a downed enemy to attack them.
Reality: Works in most cases just fine.
Changes: Allow it to work in more cases, like the "knockdown" from Lance Charge.

Rage Impact
Design: Stun enemies around you and cause them to take increased damage on the next hit.
Reality: The stun time is tiny and the insane stamina use means hitting the next enemy twice instead is a more efficient use of your resources.
Changes: Stun time should be greatly increased (from 0.2-1.6) to 2-6 seconds for the stamina this skill costs.

Bash
Design: Repeated stacking melee damage combo on enemies.
Reality: OP, sorry. It's simply better than most other skills and it bypasses an important part of Mabi's combat (hit/stun/knockback exchanges). In addition many weapons can "double Bash", which is an accepted bug (it is NOT a proper second use of the skill) that unfairly priorities different weapon speeds with no tradeoff for weapon choice. It also uses a lot of durability and has a low attack range preventing players from comboing into it.
Changes: Reduce the max damage multipliers (from 380-460%) to 180-250% at rank 1. Cut the stun time down (from 3.1-4.5) to 2.5 seconds for all ranks. Make the skill cause double (2.5x?) the current knockback meter increase. Fix the bug allowing two activations with one skill use. Reduce the combo time to 75% of the current. Change durability usage to normal. Make the attack range take the held weapon into account for proper combos.

Final Hit
Design: Trump card skill, gain teleportation and greatly increased melee abilities with a large cooldown.
Reality: Works, but the 400% damage at rank one combined with rapid attacks supercedes other skills, making this skill a central focus when it should be more of a specific use skill.
Changes: Lower the damage multiplier (from 400%) to 300% at rank 1.

Sword Mastery
Design: Mastery for swords.
Reality: Works.
Changes: -

Axe Mastery
Design: Mastery for axes.
Reality: Works.
Changes: -

Blunt Mastery
Design: Mastery for blunts.
Reality: Works.
Changes: -

Shield Mastery
Design: Mastery for shields.
Reality: Works.
Changes: -

Light Armor Mastery
Design: Mastery for light armors.
Reality: Works, but the only upside is the lack of Dex penalty.
Changes: Raise the magical stats. Increase the magic defense (from 16) to 30 and the magic protection (from 3) to 15

Heavy Armor Mastery
Design: Mastery for heavy armors.
Reality: Stronger overall than light with the only downside being a Dex cut.
Changes: Should grant less magic defense and protection. 20->10 and 5->3.

Evasion
Design: Dodge ranged attacks and bolt magic.
Reality: Has down time after you dodge, letting the attacker charge up again. Also is capped.
Changes: Uncap it, and reduce/remove the down time after a dodge.

Dual Wield Mastery
Design: Mastery for wielding two weapons.
Reality: The balance increase isn't too useful since combat and sword mastery already grant bonuses. It also makes the dex cut of heavy armors mean nothing to sword-wielders.
Changes: The balance increase should be changed to something else or removed. Any suggestions?

Gold Strike
Design: Grant non-combat players the ability to expend gold for damage.
Reality: Works, for the very few players it applies to.
Changes: -

Crisis Escape
Design: Attempt to drop aggro from enemies by disguising yourself.
Reality: Works, but looks ridiculous.
Changes: Should use different models depending on location. A hasty paper-mache tree for outside, a rock-printed cloak for other areas, etc.[/spoiler]

- Cleric / Healing
The art of using fartsparkles for good deeds. Clerics need some way to produce or sanctify goods for use, the easiest way would be for Clerics to have some way of producing Holy Water, perhaps only while under the talent.

[spoiler=""] Healing
Design: Restores HP.
Reality: Works, but scales slowly.
Changes: Should get increased HP from magic attack, say 1 per 5 instead of 1 per 10.

Party Healing
Design: Heal the whole group at once.
Reality: Works well.
Changes: -

Magic Shields
Design: Put up a shield that protects you and other players from elemental attacks.
Reality: Cumbersome. Prevents a player from doing anything, so rarely used.
Changes: Make the shield mobile once cast (Yui and such already cause this).[/spoiler]

- Magic
The art of powering up like you're in DBZ, then hitting them with magic instead. Charging up and then firing heavily-damaging skills, works well overall but everything could use attention to detail. One major issue is that speed upgrades on wands/staves seem to do little to nothing, compared to speed increases from other sources. This needs to be looked at.

[spoiler=""] Magic Mastery
Design: Overall control skill, increases MP and general magic stats.
Reality: Works.
Changes: -

Meditation
Design: Passively recover MP faster depending on rank.
Reality: Works, but is capped.
Changes: Uncap it.

Enchant
Design: Enchant items for additional stats/effects.
Reality: An unreliable, damaging mess of a skill.
Changes: Took a multi-paragraph post to suggest fixes for this a while ago.

Elemental Masteries
Design: Increase damage, reduce mana usage, generally increase potency.
Reality: They work.
Changes: -

Bolt Mastery
Design: Control skill for bolt spells.
Reality: Works.
Changes: -

Magic Weapon Mastery
Design: Control skill for magic weapons, increases bolt magic and intermediate charging speed.
Reality: Works.
Changes: -

Bolt Spells
Design: Deal damage with differing mechanics.
Reality: Work.
Changes: -

Thunder
Design: Chain attack that stuns and then deals a multitude of damaging bolts.
Reality: Works, but only chains from the initial target so groups are often only partially hit. Also prevents you from swapping gear during the wait time.
Changes: Make additional targets work off of any existing targets, not just the initial one. Allow the player to swap gear after the initial attack. Fireball already had the whole "stop swapping to different gear from applying different stats to the attack" fix.

Lightning Rod
Design: Charge up a big piercing attack.
Reality: Works.
Changes: -

Fireball
Design: Charge up a large damaging AOE fire spell.
Reality: Works, but without a special wand is often too slow to fully charge in battle, being used via Snap Cast instead.
Changes: Make the spell work from three charges and up with appropriately-scaled damage? Any different ideas?

Meteor Strike
Design: Charge up and then drop a meteor, dealing damage in a large AOE and leaving leftover fire.
Reality: Works, but can be annoying to train. Needs to be fired at raid bosses and dragons, and most of these are not available at night outside of rare passes and completable generation quests. This can leave some players unable to effectively train the skill depending on their play schedule.
Changes: Change training points so that using the skill on a raid boss is not required to rank.

Ice Spear
Design: Freeze and then explode enemies for chained damage and crowd control.
Reality: Works well for the most part, but the short range makes it unusable in some situations.
Changes: Increase the range to match Thunder's range.

Hail Storm
Design: Repeatedly charge up and fire large chunks of ice at enemies.
Reality: Has a charge time and locks you in place, doesn't have enough range for many situations. Multi-aggro is hard to handle due to the down time after tossing one attack before the next starts to load.
Changes: Increase the max range (from 1400) to 1800 to cover bigger rooms and such. Halve the down time after tossing one attack before the next charges. Reduce the charging time per layer (from 1-0.7) to 1-0.5.

Fusion Bolt
Design: Fuse bolts for increased damage and combined effects.
Reality: Works well, an underrated skill in many cases. Does not carry over all desired effects though.
Changes: Fusions involving Icebolt should have the slowdown effect.

Mana shield
Design: Use MP in place of HP as a buffer for attacks.
Reality: Works, though due to stat changes you tend to lose more MP than you would HP at all available ranks, despite the skill claiming otherwise. Also capped.
Changes: Uncap the skill. If the R5+ efficiency isn't enough in practice, increase overall efficiency a bit. Make the skill use your Magic Defense and half of your Magic Protection.

Blaze
Design: Charge up magical energy in a physical form and knock over enemies with it for large amounts of damage.
Reality: Charges too slow and gives little to no choice over target once charged, the player is at the mercy of enemy positioning.
Changes: Increase the charge speed (from 4 seconds) to 2 seconds and increase the range of the initial attack. Let players aim at an enemy to use it on if no enemy is targeted, sort of like how Focused Fist works.

Spellwalk
Design: Allow players to charge magic while moving, and to move with intermediate magic loaded.
Reality: Players use it for moving with int magic because mobile charging is too costly. It seems the stamina drain was geared towards advanced players with high stamina values, and is way too costly on newer players to keep up the skill for any usable amount of time.
Changes: Reduce the MP usage when charging magic (from 200%) to... 175%. Reduce the stamina drain (from 3-10) to 1-10 to make it easier on lower-level mages to use it.

Snap Cast
Design: Periodically instantly charge a magic spell.
Reality: Works when available. Trains way too slowly due to the grindy nature of it.
Changes: Make "use the skill within x minutes" available for training for entire ranks.

Inspiration
Design: Pause to recover a percentage of your MP.
Reality: Works quickly, the percentage means it scales well with stronger mages.
Changes: It's a little too convenient, make the skill take (from 2) 4 seconds to restore the MP when activated.[/spoiler]

- Construct Alchemy
The art of making little clay horses and then losing your mom to an unknown illness. Intended for support and to be used alongside life skills, lots of changes could be used here.

[spoiler=""] Fragmentation
Design: Allow players to break down finished goods to get some of the materials back.
Reality: Very little can be done with this skill. The skill works off of a specific list of items and the vast majority of goods that can be created cannot be broken down.
Changes: Change the skill to work on all the items players can produce. Players should be able to break down weapons, armor, clothing, and all sorts of potions to be able to get at least something back. There's lots of waste in fragmenting so it's not like players can re-craft a Bone Breaker 5 times with the same items anyways, but being able to break down stuff at least once will help a lot.

Synthesis
Design: Combine various items to make new ones, either from a recipe list or randomly.
Reality: Can produce a limited amount of usable items without extreme cost. Ranking it can cost tons of gold, being based on RNG for NPC resale values. No recipe list so you have to check outside sources. Making multiple items is annoying because the skill will eat stacks and only produce one.
Changes: More useful recipes need to be added. For example the ability to make the special music scrolls (which have more uses) by synthing paper with magic powder or something. This skill should be the magical Handicraft equal, not something ranked up purely for dyes and Dorren's Request. Got any other suggestions? Also the training needs to stop relying on NPC sale price of produced goods after R6 or so. needs a recipe list added, and needs auto-production. Higher skill ranks should need less gold value to make good random equipment.

Mana Crystallization
Design: Capture magical spells in crystal form for use by alchemists.
Reality: Bad to train, slow to use, draining, and underwhemling when in effect. No niche use.
Changes: Removal. I can think of no redeeming qualities for this skill, and changes to make it useful would be so extensive that it would be turned into a different skill at the end anyways.

Barrier Spikes
Design: Summon defensive spikes to block enemy attacks and movement.
Reality: Spikes scale poorly and don't defend much.
Changes: Spikes need scaling defense/protection in order to survive any time at all in more difficult situations. Spikes should also be wider to block more movement. I want spikes to be usable anywhere, but the potential for trolling by blocking areas in the overworld is too high.

Rain Casting
Design: Create a cloud that buffs life skills and reduces aggro.
Reality: Works for the most part, but at R1 bugs out with sleeping enemies. Also reliant on Water Cannon to train. If you're hit while it's loading it'll go into cooldown despite not activating.
Changes: Remove the lightning bolts from rank 1 and change training to not rely so much on Water Cannon also being ranked, to give non-combat alchemists the ability to rank the skill. Stop it from going into cooldown if it doesn't complete.

Metal Conversion
Design: Change ores and fragments into other ores and fragments.
Reality: Is incredibly slow, simply mining out the desired ores is faster now.
Changes: Allow the skill to work on batches at a time.

Transmutation
Design: Control skill for various kinds of alchemy.
Reality: Works.
Changes: -

Summon Golem
Design: Summon a golem to help in combat.
Reality: Really awkward to use, golem often doesn't listen, prevents you from taking other actions, doesn't scale with your stats, gives no stats.
Changes: Golem should have an auto-pilot mode where it uses normal AI against the enemies... properly. Very aggressively, hopefully. Should use the modern player versions of skills. Should get stat boosts based on 25% of the player stats. Should give some stats for ranking. The ability to buff golems and maybe heal them is needed.[/spoiler]

- Battle Alchemy
The art of using magical science theory to cause proven physical trauma.

[spoiler=""] Alchemy Mastery
Design: Passive control skill for alchemy in general.
Reality: Would work, if it could be ranked up. Some of the training is literally broken.
Changes: Fix the training to not require impossible things, and make both combat and construct alchemy able to rank the skill up individually.

Elemental Alchemy Masteries
Design: Provide passive bonuses to elemental alchemy.
Reality: Works, but provides relatively low bonuses.
Changes: Increase the max bonus (from 10) to 15 at rank 1 for each.

Life Drain
Design: Suck HP from an enemy to use for yourself.
Reality: Works great, though it doesn't scale well to players since it's static amounts of HP.
Changes: Should be percent HP restored, capping at 75% at rank 1.

Water Cannon
Design: Charge up and shoot water at an enemy for high damage.
Reality: Works well.
Changes: -

Wind Blast
Design: Blow an enemy away while dealing damage.
Reality: Works, but does low damage and only knocks away one enemy.
Changes: Make the skill an AOE that can send multiple enemies flying back, for self-defence. The cooldown prevents spamming anyways.

Flame Burst
Design: Shoot a cone of fire, stunning and damaging multiple enemies repeatedly.
Reality: A little slow and underwhelming, even considering all the multipliers.
Changes: Total number of attacks should scale evenly (from 3-5) to 3-7. Base damage increased by 5 each rank. Attack ticks should have less time between them.

Sand Burst
Design: Blind an enemy, preventing it from attacking.
Reality: Low range, single-target, limited CC. Little to no benefit for ranking, surprisingly.
Changes: The attack should be a cone AOE instead of single target, increasing in distance with rank. Blind time should be changed (from 9-11) to 7-11 seconds, scaling evenly. Should grant some stat for ranking, maybe Will? It'd be thematic to use Luck.

Frozen Blast
Design: Freeze enemies in a line, preventing them from doing anything.
Reality: Useful when it works, but it's slow, has a low range, and many important/annoying enemies cannot be frozen (such as golems).
Changes: Make the skill cast faster, make the attack range larger, and make it work on more enemies people regularly encounter.

Shock
Design: Apply the shock status to you or another player, granting them a periodic stun to enemies.
Reality: Hits enemies a little too slowly, assuming you can get it past N. If you're hit while it's loading it'll go into cooldown despite not activating.
Changes: The pages need to stop expiring and pages need higher drop rates and more sources. Shock should tick slightly faster, 1.5 seconds or so. Stop it from going into cooldown when it doesn't actually fire off.

Heat Buster
Design: Once enough pressure has been built up, fire it off in a big attack.
Reality: Damage is a bit too low, splash is a bit too small.
Changes: Increase the splash radius by 25%, maybe a base damage increase too.

Chain Cylinder
Design: Allow periodic multiple charges on offensive alchemy skills.
Reality: Works when you remember to turn it on.
Changes: Should stay toggled after a relog/CC.

Elemental Wave
Design: Gain elemental bonuses depending on what you cast, and buff one next skill.
Reality: The elemental bonuses have minimal impact.
Changes: Instead of container stat increases, the bonuses should be flat damage increases to the elemental skills themselves. So building it up would be helpful, and then the choice of using it all for a buffed next attack would be tactical. "Do I sacrifice my flat damage bonuses temporarily in order to do the AOE freeze right now or not?"

Golden Time
Design: Activates Elemental Wave and then goes into cooldown.
Reality: Works, but has a small benefit to use an entire skill on.
Changes: Should reduce Chain Cylinder's cooldown while active.

Hydra Transmutation
Design: Summon a hydra to slowly poison an area.
Reality: Little to no effect on moving targets due to the small range. A skill from another skillset (Lullaby) should not be a requirement for this alchemy skill to be effective.
Changes: Increase the max range by at least 50% by rank 1.

Guard Cylinder Mastery
Design: Control skill for dual-wielding fancy arm-mounted potato guns.
Reality: Works.
Changes: -[/spoiler]

- Fighter
The art of punching things until they stop moving. While the skillset was designed to be single-target initially (as balance against melee), the offensive options for it are limited and repetitive and some of the nerfing needs to be undone. Counterattack should have normal stamina usage and Windmill should have a normal radius with a fighter setup. Counter Punch should be a normal skill with two more skills added to make a third full chain to help reduce repetition. Activating chain skills (2 and 3) also does not preload them, refusing to use them if you don't already have a target, and normally choosing an enemy uses a normal attack instead... this needs to be fixed.

[spoiler=""] Knuckle Mastery
Design: Control skill for fighter combat.
Reality: Works.
Changes: -

Chain Mastery
Design: Control skill for most fighter skills, allowing and controlling chaining.
Reality: The timer for chains is often too small to account for enemies moving around in combat, and the skill itself trains terribly. It's not uncommon to have all other skills rank 1, but Chain Mastery still rank 5.
Changes: Increase the timer on chains and speed up the training.

Charging Strike
Design: Initial attack, rushes in and hits an enemy.
Reality: Works.
Changes: -

Somersault Kick
Design: Multi-hit attack that can daze enemies.
Reality: Works, though the duration may be too high to totally lock out other skillsets.
Changes: Potentially reduce the duration of the daze effect.

Pummel
Design: Finishing move that pins and repeatedly hits the enemy.
Reality: Does not work on a large number of enemies (even though Drop Kick tends to work on said enemies), and is incredibly time-consuming.
Changes: Change it to a standing rapid beatdown so it fires off quicker and doesn't take so long to complete (happens in one stage instead of three).

Focused Fist
Design: Charge up an attack to hit and stun an enemy while starting a chain.
Reality: Works, but can take too long to fully charge when other players are also attacking enemies.
Changes: Make it charge quicker.

Spinning Uppercut
Design: Strong hit on an enemy that can reduce their defenses.
Reality: Works.
Changes: -

Drop Kick
Design: Finisher attack that kicks an enemy back into others for area damage.
Reality: Works, if weak in AOE damage.
Changes: Could use a splash damage increase (from 20-30%) to 30-50%. Could use a distance increase (from 300-350) to 400-600.

Respite
Design: Restore HP and SP with a small penalty.
Reality: Waste time to restore jack with an annoying penalty.
Changes: The skill should restore a large percentage of SP (akin to Inspiration with MP) but no HP. It should clear all negative status effects as well. The penalty on magic/alchemy should be removed. In exchange, this skill should only be usable within 60 seconds of finishing a chain 3 skill (to prevent alchs from just swapping to knuckles for stamina regen).

Tumble
Design: Backflip out of harm's way.
Reality: Backflip for not much reason since it doesn't actually dodge anything and fighter skills have mobility as it is. Leaves you stuck for a second after execution as well, unable to load skills longer than needed.
Changes: Tumble should temporarily dodge attacks with the cooldown increased (from 10-7) to 30-15 seconds, with the training increased accordingly. Also should unlock the player slightly quicker.[/spoiler]

- Music
The art of using concussive or compressive force to bend the will of others to your means. Generally consisting of buffs that players can swap to an instrument for, established bards with high-rank singing can buff barehanded (though missing instrument upgrade buffs). One major annoyance is, when performing songs for entertainment, you cannot use buff skills for perfect playing while on a sitting instrument. This should be changed to avoid unrequired wear and tear on score scrolls.

[spoiler=""] Musical Knowledge
Design: Passive skill for increased music abilities.
Reality: Works.
Changes: -

Composing
Design: Allows players to write score scrolls to play msuic in-game, potentially with special buffs.
Reality: Lack of control over score rank makes jam creation annoying, encourages players to keep Musical Knowledge low-ranked, whereas games should encourage players to rank their skills instead (and not have a permanent penalty for ranking). Also the stat buffs and restorations from scrolls are really weak.
Changes: Allow the skill to choose which difficulty (up to the MK rank, R1 including master difficulty) score scroll is made. Perhaps buff the magical music stat effects (on scrolls) too.

Playing Instrument
Design: Allow users to play instruments
Reality: Works.
Changes: -

Dischord
Design: A melee attack with splash AOE that causes slowdown and damage over time effects.
Reality: Not used as a normal means of attack often due to many instruments being relatively-weak, but it has unique mechanics and is actually effective when it is used, which is the important part. Training can be annoying due to the reliance on crits. The slowdown lasts less time than intended due to the stun time.
Changes: Lower the requirements to crit to rank, and add in another requirement to hit at least one (or more) other enemies with the splash. Increase the slowdonw time to 7 seconds.

Lullaby
Design: Allow players to put enemies to sleep as a form of CC.
Reality: Fast to fire, characters can use it even under attack, and it can instantly halt and lock down huge groups. This is the best form of CC, which supercedes all other forms in use. That is unfortunately a balance problem and as long as Lullaby is OP like this, other forms of CC will rarely be used even if made viable.
Changes: Enforce a cap on the number of monsters able to be slept by a single player overall (not just per tick), say 2-5(?) enemies at ranks F-1. Reduce the total sleep duration by lowering the talent duration bonus and raising the duration of other skills to compensate. Lullaby should go from 39 seconds base lockdown at R1+master to 20(?) seconds base lockdown at R1+master. Enemies should be granted temporary immunity from Lullaby instead of permanent immunity after being affected once, one minute?

Battlefield Overture
Design: Apply a buff to increase damage for yourself and other players.
Reality: Grant a massive scaling buff to physical attacks. The buff is beyond pretty much any other bonuses available (by far) and uses up no resources, so it's a crazy scaling damage gain and pretty much considered required use in tons of situations... but only works properly for physical damage (outside of strange damage formulas).
Changes: Lower the damage increases (from 20%) to 10% base at rank 1, and add a 10% MA bonus to it.

Enduring Melody
Design: Grant a buff that adds to defensive properties and increases passive regen for SP/MP.
Reality: Does what it says on the tin.
Changes: -

Harvest Song
Design: Grants production-related success and spedd bonuses to the party.
Reality: Useful, though it doesn't work for alchemy.
Changes: Make it work for alchemy too. Also, I know that the skill caps at 5% and that's not affected by effect increases or performances. This is important to keep, because if it was changed people would only accept the best buff and would replay over and over to only get the best before doing whatever other skill... and then replaying over and over when it wears out each time. That's not fun at all in practice, so it's important to keep the production bonus unalterable (even if it were to be changed).

Vivace
Design: Grant a buff that increases attack and charging speed.
Reality: A little underwhelming, when latency is factored into combat the charging speed buff is barely noticeable.
Changes: Increase the speed buffs (from 11%) to 15% base at rank 1, should be just enough when bonuses are factored in to make it work, or would 20% base be OP?

Enthralling Performance
Design: Lull enemies into following you for a period of time.
Reality: Surprisingly, very useful for all sorts of skill training since you can collect weak enemies easily. Unfortunately has a hard cap on the enemy strength which limits effectiveness in combat and makes it much harder to train for stronger players.
Changes: Make the CP cap scale, anything above 4x the player's CP should be immune (instead of a hard cap of 4400 that it is now).

Encore
Design: Resets the cooldown of other skills, then puts itself into cooldown.
Reality: Not very useful. Definitely seems like a concept from other MMOs pulled into Mabi, but given cooldowns and durations, the only skill this is potentially useful for is Fantastic Chorus... and that's just because the circle doesn't follow you like other buffs do.
Changes: Encore should change in fuction. It should only be usable after a few performances or something, and once activated it should buff the next buff skill used (like Elemental wave activation).

Fantastic Chorus
Design: Generate an area that grants regen to players or debuffs to enemies.
Reality: Regen works fine, but the attack slowdown from the red circle isn't even noticed.
Changes: The attack reductions should be changed (from 15%) to 30% at rank 1.

March Song
Design: Apply a buff to movement speed.
Reality: Works.
Changes: -

Song
Design: Allows singing vocals, buffs music in general, and allows singing with anything in your hands.
Reality: Works.
Changes: -[/spoiler]

- Puppetry
The art of using the uncanny valley in an offensive manner. Puppetry is potentially useful and has various crowd/AOE options, but is simply too slow and cumbersome for normal use. The player and puppet should have less restrictions on moving around (higher max distance), the puppet should move faster, and a skill for the player and puppet to swap places should be added. The skills, overall, need to simply be faster and have less down time after attacks and the skill suggestions below will all assume that the skills perform faster. You should be able to pick up items while controlling a puppet if the item is below you. Also the cooldown bug needs to be fixed.

If the skills functioned a bit more differently depending on which puppet was summoned, it'd add more variety to the skillset as well. Imagine if Act 1 hit and stunned the enemy when using the Pierrot, but shoved the enemy backwards instead if used with the Collosus, for example.

[spoiler=""] Control Marionette
Design: Basic control skill for puppetry.
Reality: Works in most cases. The decreases to skill cooldown don't function properly, the skills appear to be cooled down but are not actually functional until the real cooldown has passed.
Changes: Fix the cooldown bug.

Pierrot Marionette
Design: Basic control skill for the little puppet.
Reality: Works.
Changes: -

Collosus Marionette
Design: basic control skill for the big puppet.
Reality: Works.
Changes: -

Act 1: Inciting Incident
Design: Dash into an enemy to intercept/damage/stun them.
Reality: Works, if a bit weak. Has a lot range for being an interception type of skill. Does not target the attacked enemy.
Changes: Should have a higher damage multiplier at higher ranks. Incease the range (from 700-850) to 800-1000. Should target the attacked enemy.

Act 2: Threshold Cutter
Design: Perform a combo on an enemy, damaging it a few times.
Reality: Works.
Changes: -

Act 4: Rising Action
Design: AOE attack that damage and shoves enemies away.
Reality: Works.
Changes: Assuming puppetry movement changes give more distance between the player and puppet, the AOE should be scaled to how far away the two are. Note that the existing cooldown will prevent rapid spam from a player.

Act 6: Crisis
Design: Grab nearby enemies and yank them on top of the puppet.
Reality: Grabs from a tiny range unless the player is heavily invested in the skill. Grants enemies invincibility for multiple seconds even when the skill appears to be finished, leading to lots of wasted attacks from party members.
Changes: The range needs to increase (from 300-600) to 500-800, and the invincibility needs to turn into basic stun that happens when the enemies land.

Act 7: Climactic Crash
Design: Send the pupet flying, damaging enemies in a large range on the way.
Reality: Potentially powerful skill, with the downside of having the puppet disconnected at a distance after using it. Also it won't fire off with certain things in the way that aren't obstacles, like the Cerberus statues. Also if a merionette is targeting a Cerberus (due to being hit by one of the attacks) it will always aim Climactic Crash at the Cerberus... and not actually fire it, because it can't due to the previously-mentioned issue.
Changes: Should the pupept stay connected and in range with the puppet range increases, or is the disconnecting at a range a suitable downside already? Should fire off through Cerberus anyways. Also should ignore pre-set targeting so it doesn't get in the way of manual aiming.

Act 9: Invigorating Encore
Design: Grant the puppet increased usability and/or disconnect it to let it fight alongside you.
Reality: Meh bonuses, awkward, not very effective AI. Skills the puppet uses don't show the cooldown effect for the player.
Changes: Make the puppet always disconnect upon use, and when disconnected it should stay near the player as a first priority and be very aggressive as a second priorty. Makes skills the puppet uses by itself correctly show the cooldowns.

Wire Pull
Design: Yank one enemy towards you.
Reality: Does what it says on the tin, but isn't too effective in combat.
Changes: The skill should allow you to pull one enemy from one position and place it at another position, adding to Puppetry's position control capabilities.

Puppet's Snare
Design: Tie down an enemy and deal damage to it over time.
Reality: Does absolutely abysmal damage and locks the player up to CC a single enemy.
Changes: Should either not lock the player up (and allow up to three targets to be remotely locked at R1), or the player should be locked up but able to stun and hold groups of enemies (think of thunder chaining).[/spoiler]

- Dual Gun
The art of making pew pew noises until stuff gives up and falls over. Has one or two useful skills, with the rest gimped in usability. One thing that always annoyed me, though, is how Dual Guns turn you into a ballerina the moment you pick them up. Even basic skills like Counterattack suddenly become all flippy and flashy. I think that each type of skill should have three animations. A more awkward/simple one at low ranks, a normal one at medium ranks, then the flashy animations at high ranks. Dual Gun Mastery could affect the animation overrides for other skills.

[spoiler=""] Dual Gun Mastery
Design: Control skill for guns.
Reality: Works.
Changes: -

Flash Launcher
Design: Gunner version of Smash, gets up and close to perform a strong attack.
Reality: Works, but the training is bad.
Changes: Speed up the training.

Shooting Rush
Design: AOE attack that dashes across enemies, damaging them.
Reality: Works, but is spammed due to being the most effective attack with a short cooldown.
Changes: Increase the cooldown, sorry. 12-9 seconds (scaled over rank) should work. The training is mostly okay for what the skill is, since you get increased training for multiple enemies and it's intended for use on groups... but with the increased cooldown, training points should increase as well.

Bullet Slide
Design: Back away from an enemy while damaging it.
Reality: Works, but locks you for too long afterwards unless manually canceled. Also the skill is really slow to train.
Changes: Push the player further, faster, and reduce the lock time. Speed up training since it has a long cooldown at low ranks. What else could make this skill more useful while still keeping it as an escape type of skill? Should it drop aggro from that target via the confusion (sudden low slide and distance)? I'd totally use it if so.

Bullet Storm
Design: Fire rapidly at a group of enemies in front of you for damage.
Reality: Uses tons of bullets and leaves you locked down for a few moments, only hitting things really close to you.
Changes: Bullet use needs to be lowered to 2 per target to keep it in line with other skills. The skill should also stun enemies for 2-4 seconds after finishing to give the player time to recover (and to give it a little CC). Also needs better training. The range should be increased (from 300-350) to 400-600, increasing evenly over the ranks.

Grapple Shot
Design: Hook onto an enemy and yank yourself to them to shoot them.
Reality: Bad in pretty much every way. The minimum range for using it is annoying, the maximum range is way too short for being a hooking skill, you can get attacked by your target on the way, and once you reach them you do a single shot that doesn't even stop them from immediately retaliating.
Changes: The minimum range needs to be cut in half and the maximum range needs to be increased from 1000 to 1500. The enemy targeted needs to be stunned when hooked and the attack that hits the enemy at the end needs to stun as well. Also the training is beyond horrible. Should slow down the monster by 10-25% for 4-10 seconds.

Reload
Design: Reload your guns. Automatically executes when out of bullets in some situations.
Reality: Works, but won't get canceled by trying to use another skill, so if you run out of bullets and try to load Counterattack you'll often just stand there like an idiot until the enemy whacks you in the face. Cancels whenever you're hit by anything even if it's weak.
Changes: Should cancel into other skills like Counterattack and Windmill. Training is slow, reloading while empty should count for more and have more total uses. Should be able to finish reloading unless knocked back or down.

Way of the Gun
Design: Buff yourself with infinite bullets, forced crits on normal attacks, and increased attack speed, cools down other gun skills.
Reality: Works, though cooling down the existing gun skills when activated doesn't seem too useful. Would a buff to halve the cooldown of other guns skills while this is in effect be too much? It'd allow temporary spamming of skills again while active.
Changes: [/spoiler]

- Ninja
The art of convincing people you didn't cut them when you really did. Has a lot of unique and control skills that could be pretty useful... if people could train them reasonably. Also can we please get Ninja's KR updates?

[spoiler=""] Shuriken Mastery
Design: Control skill for normal attacks.
Reality: Works.
Changes: -

Shuriken Charge
Design: Charge up and launch a damaging attack at an enemy.
Reality: Slow to charge and the targeting system is unlike other skills, needing a target ahead of time.
Changes: You should be able to click a target after charging has already started in order to attack it. Also training needs to be better.

Shadow Cloak
Design: Drop aggro and hide for a few seconds, launching an attack when you reappear.
Reality: Useful skill, though the damage seems a little low for a special skill and the cooldown is a bit long.
Changes: Increase the damage to 90-150% and reduce the cooldown to 40-30. Also fix the training.

Explosive Kunai
Design: Set explosive traps that damage and slow enemies.
Reality: The slow effect is so short-lived that a lot of people I've mentioned it to didn't even know it existed.
Changes: The slow debuff should last 5-9 seconds depending on rank. Also better training.

Kunai Storm
Design: AOE attack hitting enemies in range four times.
Reality: Works.
Changes: Better training.

Smokescreen
Design: Area debuff lowering enemy defenses.
Reality: Potentially increases your end damage by up to 20. Points. Flat. If you even rank this skill. Also won't fire off if you click an enemy directly with it.
Changes: Should reduce protection in addition to defense. KR's version reduces 3-7 prot and that seems fine. Also training needs to be increased by debuffing multiple enemies at once. Right now a whole group could be debuffed and it only counts as one. Also make it work if you click on an enemy directly.

Shadow Bind
Design: CC skill that temporarily halts enemies around you.
Reality: Works well in parties, but only starts becoming useful for yourself at like rank 5.
Changes: Increase the binding time (from 3-5.5) to 4-8 depending on rank, and increase the cooldown (from 18-10 seconds) to 30-20 seconds so spamming it doesn't keep things still forever.

Sakura Abyss
Design: Charge up a large AOE multi-hit attack with a long cooldown.
Reality: Works, but takes so long to charge that it relies on broken CC from other skillsets to fully execute.
Changes: Greatly increase the charging speed by lowering the time per charge (from 1 second) to 0.3 seconds.[/spoiler]

- Hidden
The art of getting peanut butter in somebody's chocolate.

[spoiler=""] Study: Potion Lore
Design: A passive chance of a buff to potions and reduction in potion poisoning.
Reality: A small chance of a huge buff on potions means it's unreliable. When you drink potions normally you're either drinking just the one you need (in which case if you expect the base a buff is wasted), or spamming a few of them to refill whatever (in which case you don't get even refill per hotkey).
Changes: Instead of a chance for a refill, the skill should grant you a constant smaller buff. Right now, over a lage period of time you'll get 137.5% Instead, buff the skill a bit to a constant 150% effectiveness at rank 1.

Spider Shot
Design: Shoot a web onto enemies to hold them still and have other attacks deal extra critical damage.
Reality: Works, but requires making the traps in order to rank the skill. This was previously determined as something players simply did not want to do, so the traps were put into NPC shops... but you still need to make them in order to rank the skill. Even if you do buy them from NPCs, they're 10K per a stack of 10. And 1K per shot of a skill is kind of a heavy price to pay for normal use. Also it won't put you into combat mode and target what it's used against if you didn't have it targeted ahead of time.
Changes: Decrease the cost of the traps to 100 gold or less each. Make the skill target the target properly.

Doppelganger
Design: Kill enemies for souls, then do a multi-hit attack on one enemy.
Reality: Despite being a physical attack, it deals flat damage and is not based on your character, only the skill rank. This means it scales horribly with the player.
Changes: Make Doppelganger a percent damage skill like everything else. Each hit should do 30% (x5, 150% damage) at rank F, up to 80% (x7, 560%) per hit at rank 1.

Berserk
Design: Become temporarily enraged, gaining a strength bonus and advanced heavy stander while losing your defenses.
Reality: The stat bonus only applies fully to melee (half effectiveness to gunner/ninja/puppetry and none to other skillsets), is subject to the stat cap, and the loss of defensive bonuses means the advanced heavy stander won't keep you standing long.
Changes: Berserk should grant raw physical damage (up to 64 max at rank 1) and your defensive stats should be halved, not reduced to zero. The skill should grant you the inability to die while active, leaving you in whatever deadly state you're at when it concludes.

Chain: Counter Punch
Design: A counterattack fighter skill that starts a chain. Is strongest when timed perfectly.
Reality: Works if you use it quickly, but the damage drags down too quickly to deal with latency. Also training is too reliant on RNG after normal uses have completed. The skill doesn't actually chain if the enemy waits too long to hit you, which is often outside of the control of the player. It also doesn't actually target the enemy that triggers the counter. And in addition to all that crap, if you're hit while it's loading it'll go into cooldown despite not activating.
Changes: The damage should not reduce for the first two seconds, after that it should decay. Increase the total training counts for using the skill at each rank. Make it so the skill always starts a chain even if hit late. The skill needs to actually target the enemy it hits. Prevent it from going into cooldown if it doesn't trigger.

Shockwave
Design: Defensive skill for mages, sends out a pulse of damage to hurt and stop enemies at the cost of a percentage of MP.
Reality: Works for self-defense, does acceptable damage, but costs too much MP. It actually penalizes higher-ranked mages because it eats more MP for them than others. There should not be a downside to higher stats like that.
Changes: The MP cost should be flat amounts like normal spells. 30 MP per use too low?[/spoiler]

- Crusader
The art of using the power of friendship. While mainly intended for raids, the skills could actually be pretty useful elsewhere. It's annoying to respec the points any time you're on a different role, so a "max points" button would help.

[spoiler=""] Shield of Trust
Design: Generate shielding for you and two allies, granting defenses and additional bonuses.
Reality: Works.
Changes: -

Giving Heart
Design: Increases the range at which you can shield allies.
Reality: Works.
Changes: -

Healing Hands
Design: Restores some HP to you and allies you shield.
Reality: Restores HP before any wound restoration, meaning it could heal 0 HP even if shielding restores wound due to Recovering Touch.
Changes: Change Helping Hands to trigger after Recovering Touch.

Magic Dampening
Design: Grants additional magical defenses while shielded.
Reality: Rather weak even at max.
Changes: Double the max bonuses (from 15 Mdef / 15 Mprot) to 30 Mdef / 30 Mprot.

Guarded Footsteps
Design: Grants increased movement speed for you and your allies while shielded.
Reality: Allies have to be nesting in your armpit hair for the training to count. Seriously, within 35cm? East coast position last is at least triple that distance.
Changes: Double the max range at which you can protect allies for training.

Recovering Touch
Design: Restore a percentage of wound damage for you and your allies.
Reality: Works off of total percentage like First Aid, so it works very well... if you have it. Clearing four weekly orders is difficult and likely to take over a month (due to the ring drop rates).
Changes: The unlock should be changed to two weekly orders cleared, if not one.

Celestial Spike
Design: Generate spikes on enemies, slowing them down and dealing damage over time.
Reality: Works, but has flat damage and does not scale at all, despite being a magical attack (using MP and being reduced by Mdef and Mprot).
Changes: Have the skill's damage take your Magic Attack into account.

Weighted Conscience
Design: Makes the slow effect work on targets near the initial one.
Reality: Not too useful of an effect, and has a tiny radius.
Changes: Increase the radius (from 50-330) to 200-500. Perhaps add or change effects, suggestions?

Deepening Trust
Design: Restores some MP to yourself when you spike an enemy.
Reality: Restores a pitiful amount.
Changes: Should restore 10 per level, like Holy Vitality does.

Spreading Faith
Design: Increases the range you can spike from.
Reality: Works, but isn't a large increase.
Changes: Increase the range bonuses (from 10-210) to 110-310.

Wounding Soul
Design: Increases the base damage of Celestial Spike.
Reality: Works.
Changes: -

Sinful Thorns
Design: Causes enemies near the initial spiking target to take damage over time as well.
Reality: The shared damage over time is too tiny to matter.
Changes: Change it (from a flat 1-35) to 10-30% of your spike damage.

Judgment Blade
Design: Summon a huge blade and slam it down for large damage.
Reality: Works.
Changes: -

Stronger Power
Design: Increases the range of your blade.
Reality: Works. Increases the angle/width at levels 6 and 11 which helps on groups.
Changes: -

Binding Needle
Design: Increase the amount of time Girgashiy is bound for.
Reality: Works, but is a little low.
Changes: Increase it (from 0.1-1.5) to 0.1-2.5.

Breaking Lies
Design: Grants additional apostle defense reduction to your blade.
Reality: Not enough of an overall increase when fully leveled.
Changes: Increase it (from 0.2-3) to 1-7. Also make it work up to 3 prot on non-Girg mobs.

Christening Blades
Design: Increases the damage your blade deals on a critical hit.
Reality: Works.
Changes: -

Holy Vitality
Design: Restores some stamina when you blade something.
Reality: Works.
Changes: -[/spoiler]

54,000 characters later: owmyhands.