Ancient Mage
Goal: Damage per second
One Above All ranks first at 457.78B DPS and 13.33 attacks per minute, range 44.6.
Berserker is second here, +84.2% DPS against the bare panel versus +122.2%.
Anime RNG Defense accessories are the one slot with no published numbers behind it, so this page ranks the ways of running it against something the game does show: what your unit costs to place. The trait side sits underneath, applied to that unit's own stats.
By Jim LiuLast updated 2026-07-11
Provenance: what is known about Anime RNG Defense accessories here is a single line of gameplay commentary, and the unit stats and trait effects come from panels read off gameplay footage. All of it is the list published on this site, reconstructed from public community sources. It covers 3 units out of the 125 the game's Collection counter reports, and 5 trait effects out of a panel that scrolls past both ends of what was captured. It may be incomplete or wrong.
Read as written, the whole of it: Two kinds: cosmetic-only accessories, and stat accessories that grant Luck/Money boosts. Upgraded with duplicate accessories + materials farmed from the matching world map.
That sentence is the entire record, and it is held at medium confidence: it comes from a creator describing the system during a playthrough, not from the game's own UI like the trait and stat figures further down. Two things in it are stated plainly, that there are cosmetic and stat kinds, and that upgrading runs through duplicates plus materials from the matching world map. Everything else a reader usually wants is absent: what a Luck or Money boost is worth, whether boosts stack, which accessories exist, and how many materials an upgrade takes.
So the comparator cannot rank accessories against each other and does not pretend to: it never takes an accessory as input. What it ranks is three ways of running the slot, ordered by one number the game does display, the placement cost on the unit's own panel, against a cutoff set on this page. See the items page for the six currencies and boxes that are documented, and the world pages for what is recorded about each map.
Both rankings below are computed, not looked up. The accessory slot is ordered by this unit's placement cost against the cheapest placement cost on file. The trait side applies each modifier to the unit's own damage, cooldown and range, scores the results against each other for the goal you pick, and states how far each row sits behind the best of the five on damage per second, on attack rate and on range.
Scored on damage divided by cooldown and nothing else (100% damage per second). That is not the same as ordering the traits by their damage modifier, because each one also moves the cooldown.
Ancient Mage with no trait: 1.03T damage, 5s cooldown, 37.2 range, 6.17K to place. Read off the Collection tooltip card. For the trait list on its own, what each one does and how a Trait Roll works, the traits page and reroll calculator is this site's page for that; this section only shows what those effects do to the numbers on the unit you picked.
Top score for damage per second, at +122.2% damage per second against the same unit with no trait. Best damage per second of the 5. Gives up 5.6% of the best attack rate and 7.7% of the best range.
Ranked 2 of 5 for this goal, at +84.2% damage per second against the same unit with no trait. Gives up 17.1% of the best damage per second and 10.5% of the best attack rate and 19.2% of the best range.
Ranked 3 of 5 for this goal, at +26.3% damage per second against the same unit with no trait. Gives up 43.2% of the best damage per second and 10.5% of the best attack rate and 19.2% of the best range.
This trait also carries +100% Dot Damage. Nothing on file says how damage over time is applied in this game, so it is excluded from the score rather than guessed at.
Ranked 4 of 5 for this goal, at +23.5% damage per second against the same unit with no trait. Best attack rate of the 5. Gives up 44.4% of the best damage per second and 19.2% of the best range.
Ranked 5 of 5 for this goal, at +10.5% damage per second against the same unit with no trait. Best range of the 5. Gives up 50.3% of the best damage per second and 10.5% of the best attack rate.
Nothing on file says whether you can hold the Luck and Money boosts at the same time, so this is a suggested order to chase them, not a claim that you have to pick one. It is ordered by placement cost because that is the only number in play that the game actually displays.
1Stat accessory, Money boost (suggested first)
Ancient Mage costs 6.17K to place, 4.5x the cheapest placement cost on any verified stat panel. That is the expensive end of the placement costs read off this game, which is what puts this row first. The order is this page's suggestion, not the game's: the Money boost is suggested first when that ratio reaches 2x. No source gives the size of either boost, what either one acts on, or whether you can hold both at once, so this is not a claim that you have to choose.
2Stat accessory, Luck boost (suggested second)
The source names Luck as one of the two boosts the stat kind grants and gives no figure for it. It sits second here purely because the unit you picked is at the expensive end to place. The order is this page's suggestion, not the game's: the Money boost is suggested first when that ratio reaches 2x. No source gives the size of either boost, what either one acts on, or whether you can hold both at once, so this is not a claim that you have to choose.
3Cosmetic accessory
The source describes the cosmetic kind as cosmetic-only, so it is ranked last and nothing else is claimed about it.
Next step: a trait is bought with a Trait Roll, and no odds for landing a specific trait appear anywhere in the data on file, so before you spend anything put your own estimate of those odds through the trait reroll odds calculator, then check what the currencies and items on file actually pay out.
Each card below is the comparator itself, run with the inputs named on the card. Change the picker above and you will reproduce these exactly.
Goal: Damage per second
One Above All ranks first at 457.78B DPS and 13.33 attacks per minute, range 44.6.
Berserker is second here, +84.2% DPS against the bare panel versus +122.2%.
Goal: Attack rate
Godspeed ranks first at 46.81B DPS and 18.58 attacks per minute, range 43.1.
One Above All is second here, +122.2% DPS against the bare panel versus +23.5%.
Goal: Range, with damage as the tiebreak
One Above All ranks first at 5.82B DPS and 12.12 attacks per minute, range 32.9.
Sniper is second here, +10.5% DPS against the bare panel versus +122.2%.
Cursing a unit is the one action mentioned on this page that can move a unit backwards. Nothing in the source data says how cursing and traits interact, so the two are kept separate here. The patch notes on file describe it as one random buff and one random nerf, and they stop there: no list of which stats are in the pool, no sizes, no odds. So the comparator does not price the curse and does not model it. All it does is show where a trait concentrates its stat movement, so you can see how much of a loadout rides on one stat before you decide whether to curse that unit.
The trait ranked first for your unit and goal, One Above All on Ancient Mage, puts 77% of its stat movement into damage. That describes the trait, not the curse: the game does not publish which stats a curse can touch or by how much, so nothing here says a nerf is more likely to land there or what it would cost.
This is where a trait's value sits, not a probability. Read the cursed finger page for where the fingers come from, and the King of Curses tracker for the quest that pays them out.
The trait effects are read out of the effect lines stored in this site's data file rather than retyped, so a change to the source changes the ranking. For a selected unit, each trait produces a modified damage figure, a modified cooldown and a modified range; damage divided by cooldown gives damage per second and 60 divided by cooldown gives attacks per minute. Those three numbers are min-max normalised across the 5 candidates for that unit and combined with the weights printed under the goal picker. The gaps each row reports are its shortfall against the best of the five on each of the three measures, not only the measure that decided the goal, so a row can top the ranking and still show what it costs elsewhere. Unit stat panels are checked against the evidence text in the source file at load time, and a unit whose figures no longer appear there is dropped from the picker rather than shown. The Anime RNG Defense accessories ranking runs separately and on one input only: this unit's placement cost divided by the cheapest placement cost on file, with a 2x cutoff set on this page rather than by the game, and no boost magnitude anywhere in it because none has ever been published. Compiled by Jim Liu; the standard this site holds itself to is set out on the about page.
What is known about Anime RNG Defense accessories is one line of creator commentary, held at medium confidence, and its video is the first link below. The trait effects and unit stat panels were read off the game's own UI. All of it is stored in this site's data file and the page renders only what that file contains.