Mabinogi World Wiki is brought to you by Coty C., 808idiotz, our other patrons, and contributors like you!!
Want to make the wiki better? Contribute towards getting larger projects done on our Patreon!


From Mabinogi World Wiki
Jump to: navigation, search


This page has some tests and info about speed boosts in Mabinogi.

If you're concerned with real-time DPS testing, see User:Rydian/DPS instead for info and comparisons.

On this page, "melee" is anything you physically whack the enemy with up close, including using wands/cylinders as beat sticks.

Attack Delay Reduction

  • Certain things like set effects, certain wings, buff potions, and Vivace can reduce your attack delay by a certain percentage.
    • This reduces the down time after a hit before you can attack again, meaning more rapid attacks.
  • Equips and buffs all stack up to the cap (400ms normally, 300ms under Relentless Assault).
    • For example a Very Slow weapon under Final Hit with Grandmaster Wings would have a delay of (1000 * 0.8 * 0.7) = 560ms.
    • Delay is integer math, any remaining portions under 1ms are truncated/dropped.


Set Effect Example Items Reduction Does Not Work On
Attack Speed Increase Saint Guardian Set
Chinese Dragon Set
Languhiris Chaser Set
Various Enchant Points
20% Fast / Very Fast Melee1
Chains (Near)2
Increases Attack Delay3 Abyss Dragon Horns (M)
Abyss Dragon Horns (F)
10% Fixed-Delay Skills (Final Strike, etc.)
Bows / Crossbows
Alchemy / Magic
Possessed Eiren Wings 10%
Champion Wings 10%
Master Wings 20%
Grandmaster Wings 30%

1 Melee weapons are generally swords, axes, or blunts.

2 Chain blades have different attack speeds depending on how close you are. The further/slower speed is buffed, the nearer/faster speed is not.

3 Nexon derp, actually reduces delay.

4 Aim speed for normal ranged attack & Support short are increased


Buff Reduction Does Not Work On Notes
Attack Delay Reduction Potion (10 min) 15% (Untested) -
Fallen Fairy Summon 30% Bows/Crossbows -
Final Hit 20% Anything non-melee -
(Mini/Fine/Finest) Lavender Scented Candle 1% / 3% / 5% Bows/Crossbows -
Magic Speed Potion 10% Anything non-magic -
Might of Ladeca 1 Weapon Speed Tier
i.e. Very Slow -> Slow
Anything that's not a Close Combat Weapon
Very Fast weapons (as there's no faster tier)
Relentless Assault 5% to 20% Bows/Crossbows Reduces the speed cap to 300ms
for all buffs while active.
Vivace x% Alchemy/Magic (Speeds up charge time instead)
Bows/Crossbows (Used to increase aim speed, currently bugged)
Way of the Gun 50% Anything non-gun -
Wind Guard 1 Weapon Speed Tier
i.e. Very Slow -> Slow
Anything that's not a Close Combat Weapon
Very Fast weapons (as there's no faster tier)

Tests - Single Buff

  • Attack speed appears to cap at 400ms at the fastest unless Relentless Assault is active (300ms then).
  • Dual-wield (swords) or dual-attack (guns) are tested as an attack cycle, not individual hits.
    • Dual-gun attacks are considered one attack but hit twice on purpose (1=2).
    • The game considers a set of dual-wield hits as one "attack cycle" (1+1=2) because dual-wielding was a hack.

Intended Values

  • Found via packet-logging (shhh don't tell).
    • Represent what the server decides directly.

Tested Values

  • Tested by recording the game at a high framerate and measuring the time an attack cycle took, but it's a rough estimation.
    • Testing was mainly done to confirm speed caps and show that the number given by the game is not what you should expect as a player unless you have like 0 ping.
  • Testing annoying/hard and results in laggier attacks.
    • Inconsistent latency requires repeated testing and averaging and/or dropping outliers.
      • Mabi demands all data to be processed in the order it was sent/received and takes multiple steps to do combat tasks.
        • The effects of latency are multiplied because tasks are broken up into multiple pieces, each of which has separate latency.
      • FPS issues can skew test results too, especially where combat is concerned.
        • Mabi only reads input on rendered frames, so lower FPS = longer delays before you can attack again.

Single Melee

  • Scythes, Lances, knuckles, and anything you physically hit with has a delay that depends on the weapon's speed.
    • The extra hit from Lance mastery affects knockdown gauge, not attack speed, so it doesn't matter for this.
Item Intended Delay (ms) -30% Intended Delay (ms) Tested Delay (ms) -30% Tested Delay (ms)
Very Slow 2H (Normals) 1000 700 - -
Very Slow 2H (Final Hit) 800 560 918 603
Slow 1H (Normals) 800 560 - -
Slow 1H (Final Hit) 640 448 757 543
Normal 2H (Normals) 600 420 - -
Normal 2H (Final Hit) 480 400 544 470
Fast 1H (Normals) 520 400 - -
Fast 1H (Final Hit) 416 400 - -
Very Fast 1H (Normals) 450 400 - -
Very Fast 1H (Final Hit) 400 400 - -

Dual Melee

  • Each dual-wield hit is the same delay as a normal hit (calculated normally), then multiplied by two.
    • Single hits still won't go below 400ms normally, which is why the lower bound for dual-wielding is 800ms in these tests.
Item Intended Delay (ms) -30% Intended Delay (ms) Tested Delay (ms) -30% Tested Delay (ms)
Slow 2H (Dual Wield Normals) 1600 1120 - -
Slow 2H (Dual Wield Final Hit) 1280 896 - -
Normal 2H (Dual Wield Normals) 1200 840 - -
Normal 2H (Dual Wield Final Hit) 960 800 - -
Fast 2H (Dual Wield Normals) 1040 800 - -
Fast 2H (Dual Wield Final Hit) 832 800 - -
Very Fast 2H (Dual Wield Normals) 900 800 - -
Very Fast 2H (Dual Wield Final Hit) 800 800 - -


  • Bow speed affects recovery/reload time, but attack delay reduction bonuses don't seem to affect bow recovery time.
Item Intended Delay (ms) -30% Intended Delay (ms) Tested Delay (ms) -30% Tested Delay (ms)
Slow Bow (Normals) 800 800 - -
Slow Bow (Magnum Shot) 800 800 - -
Slow Bow (Arrow Revolver) 800 800 - -
Normal Crossbow (Normals) 600 600 - -
Fast Bow (Normals) 520 520 - -


  • Each instance here is for the pair of shots.
Item Intended Delay (ms) -30% Intended Delay (ms) Tested Delay (ms) -30% Tested Delay (ms)
Slow Guns (Normals) 1000 700 - -
Slow Guns (Way of the Gun) 500 400 - -
Very Fast Guns (Normals) 530 400 - -
Very Fast Guns (Way of the Gun) 400 400 702 700


Item Intended Delay (ms) -30% Intended Delay (ms) Tested Delay (ms) -30% Tested Delay (ms)
Fast Shuriken (Normals) 700 490 - -
Very Fast Shuriken (Normals) 530 400 833 687

Chain Blades

  • Chain Blade attack speed depends on how close you are to the target.
    • The faster speed is if you're within less than half the max range.
    • The slower speed is used if you're over half the max range to the target.
Item Intended Delay (ms) -30% Intended Delay (ms) Tested Delay (ms) -30% Tested Delay (ms)
Fast Chain Blade (closer) 470 400 - -
Fast Chain Blade (further) 720 504 - -

Magic / Alchemy

  • Attack delay reduction buffs have no effect on charging speed or delay when firing.
    • Recovery after firing skills seems locked at 500ms.

Tests - Stacking Buffs

  • So now that we've established what buffs there are and how they work, let's see what happens when we stack them.
Buff Theoretical Delay (ms) Actual Delay (ms)
None 600? 600
Abyss Horns (-10%) 540? 540
Speed Set (-20%) 432? 432
Vivace (-16.5%) 360? 400
Final Hit (-20%) 288? 400
Finest Lavender Scented Candle (-5%) 273? 400
Relentless Assault (-7%) 253? 300

Movement Speed


Based on comparison with various skill radii, each race has the following in-game base movement speed.

(Note that these measurements are rough and could use re-testing with more granular measurement later)

Race Units/Sec Speed %
Human ~380 100%
Elf ~500 131.58%
Giant ~450 118.42%

Equipment / Commerce

I measured the walk cycle of the elephant commerce mount at a high FPS (~120).

  • Nothing: (8.275-9.000) 0.725 seconds.
  • Fleet Feet: (10.383-11.050) 0.667 seconds.
  • Alpaca Robe: (8.617-9.217) 0.600 seconds.
  • Commerce Level 17: (8.008-8.483) 0.475 seconds.

Combining them...

  • 17+Fleet: (5.758-6.192) 0.434 seconds.
  • 17+Robe: (4.400-4.875) 0.475 seconds.
  • 17+Fleet+Robe: (15.125-15.558) 0.433 seconds.

So Fleet Feet are about 10% here, the robe is about 19-20%, commerce level 17 is about 52-53%, and the alpaca robe does not actually stack with the reforge.

I didn't bother with March Song or Shylock or other buff combinations because they don't work on the commerce mount anyways.