Lychee Dragon Cookie Toppings Build — ATK Becomes Cooldown

Lychee Dragon Cookie's Toppings have a job the other Cookies' don't: whatever ATK they give is also paid out as Cooldown. The skill's Violet Scales line reads Cooldown -0.8% per 1.0 ATK gained from Toppings, capped at -60.0%. The cap is the same 60% that the whole game puts on Cooldown reduction, so this one passive can fill the entire allowance on its own.

Our v7.10 update guide listed that line and stopped, saying a real build had to wait. This guide does the arithmetic it left open: how many seconds each Topping setup takes off a 16-second skill, how likely a six-piece set is to reach the cap, and which bonus effect to keep. Everything about the skill comes from cookierun.wiki, whose Lychee Dragon skill box still has only its Base column filled (the Max column reads "?" in the revision of November 1). The Topping tables are cookierun.wiki's and Devsisters' own support pages.

Who Lychee Dragon Cookie is

The skill's own numbers are in the v7.10 update guide, so only the parts that touch gear are repeated here. Their blink, scythe and Violet Dragon hits each combine a percentage of ATK with damage proportional to Max HP against enemy Cookies, and an ATK-only version against everything else. Violet Scales also hold the survivability side: Max HP +30.0%, DMG Dampening 37.5%, Debuff Resist +30.0%, and a once-per-battle Invulnerable window of 5.0 seconds that cuts the remaining Cooldown by 30.0%.

The passive: Topping ATK in, Cooldown out

The line is the last one in the Violet Scales block of the skill box: Cooldown -0.8% per 1.0 ATK gained from Toppings (capped at -60.0%). Two things need pinning down before the math.

The unit. The box says "1.0 ATK", with no percent sign. A flat-ATK reading looks unlikely, since the neighbouring Dragon boxes on cookierun.wiki use percentages. Its Pitaya Dragon box phrases its sibling passive as "+3.0% per 1.0% of increased ATK", and its Ananas Dragon box says "-3.0% per 6.0% of ATK gained from Toppings". Read the same way, Lychee Dragon converts 0.8 points of Cooldown for every 1 point of ATK from Toppings, so the cap is reached at 75% ATK from Toppings. That is an inference from how the wiki words its neighbours. The game text does not print the percent sign.

The cap. cookierun.wiki's Cooldown page gives the maximum Cooldown reduction a Cookie can have as 60%, and its Cooldown/Data page says the same limit applies to total Cooldown. So the passive can fill the whole allowance by itself. Anything else that shortens Cooldown, such as a Beascuit Cooldown line or a Treasure like a Jelly Watch, shares that 60% and adds nothing once the passive has used it up.

Cooldown becomes seconds through the wiki's formula: total Cooldown is Base x (1 - reduction), rounded up to the nearest frame (one sixteenth of a second). For a 16-second skill, every point of reduction is worth 0.16 seconds, and a full 60% leaves 6.4375 seconds after rounding. The same data page lists Lychee Dragon among the Cookies that have an extra multiplier on their initial Cooldown, but the multiplier cell is still blank, so this guide gives no first-cast time.

What each build reaches

The ATK sources are all documented. A Searing Raspberry M gives 3% ATK plus 0.5% per upgrade, so 9% at +12; five give 45%. The Equip 5 set bonus is +10% ATK and Equip 6 is +12%. An Epic Searing Raspberry Tart rolls 7.2% to 12.0% ATK, and our Topping Tarts guide covers the odds. Each +12 Topping carries three bonus effects, all different stats, so it can roll ATK only once. On a regular Epic that roll is 1.0% to 3.0%, and ATK is one of ten effects, so a given Topping carries it 3 times in 10 if every effect is equally likely. On a Resonant Topping the roll is 2.0% to 3.0%, and ATK is one of six effects, so it shows up 3 times in 6 on the same assumption. Every value inside a range is equally likely (Devsisters' support page). The equal-odds-per-effect assumption is mine: the Devsisters chance tables are images I could not read, so the probabilities below inherit it.

The table converts ATK to Cooldown at 0.8 per point, then to seconds with the wiki's formula. The "average" rows use the expected number of ATK lines (1.5 across five regular Toppings at 2.0% each, 2.5 across five Resonant ones at 2.5% each) and an average Tart of 9.6%.

Reaching the cap takes a six-piece set and good rolls. I ran the odds from Devsisters' tables with a uniform 7.2% to 12.0% Tart: with regular Epic Toppings a six-piece Raspberry set reaches 75% about 2.5% of the time, and with Resonant Toppings about 26%. The middle 80% of outcomes runs from about 66% to 73% ATK on regular Epics and from about 69% to 77% on Resonant. Five pieces alone cannot reach the cap: 45% + 10% + 15% is 70% even with every roll perfect.

If the passive turns out to count only main stats and bonus effects, and not set bonuses or the Tart, the picture is lower: five +12 Raspberries plus average regular-Epic rolls give about 48% ATK, or -38.4% Cooldown, 9.875 seconds. Neither reading is confirmed. The page says "gained from Toppings" and does not say more, which is the same gap our Ananas Dragon build works around.

The Lychee Dragon Cookie build

Why a Swift Chocolate hybrid loses. A maxed Swift Chocolate M gives 3.0% Cooldown (1.8% plus 0.1% per upgrade). The Raspberry it replaces gives 9% ATK, which this passive turns into 7.2% Cooldown and also adds to every damage line. Dropping to four Raspberries also costs a set step: from the 10% Equip 5 bonus to the 8% one before a Tart, or from 12% to 10% with one. The same arithmetic that favours five Raspberries on Ananas Dragon applies with a steeper exchange rate.

The Tart is the biggest single step. A Tart adds its own roll (9.6% on average) and moves the set from 10% to 12%, so it is worth about 11.6 points of ATK, which at 0.8 is more than 9 points of Cooldown. That beats any Resonant upgrade on the other four slots, and a Tart is what takes the set past 55%, since five pieces top out at 55% before bonus effects. If the Tart does not count toward "ATK gained from Toppings", you still keep its set step for damage.

Which bonus effect to keep

On a regular Epic, an ATK line averages 2.0% ATK, which the passive converts to 1.6 points of Cooldown while also adding damage. A Cooldown line averages 1.5%. On a Resonant Topping, ATK averages 2.5%, worth 2.0 points of Cooldown, against a Cooldown line of 1.5% to 2.0%. So on the wiki's reading an ATK line beats a Cooldown line on both counts, and it keeps working past the point where a Cooldown line would be wasted against the 60% cap.

DMG Resist comes second. Lychee Dragon is prioritised to the Middle and the skill blinks them onto the enemy, so defence has some value, and Epics roll DMG Resist at 1.0-6.0% (Resonant 4.5-6.0%). Keep a Cooldown line only when you are certain to fall short of the cap and have no ATK roll to take its place.

Dragon's Sweet Prank Resonant Toppings

The official notice lists five Cookies who can equip Dragon's Sweet Prank Toppings: Lychee Dragon, Moonlight, Rambutan, Jax and Pomni. The notice says it comes in 10 kinds of Toppings, so a Dragon's Sweet Prank Searing Raspberry exists. Resonant Toppings draw their bonus effects from six stats (ATK, ATK SPD, CRIT%, Cooldown, DMG Resist, CRIT Resist), so on the equal-odds assumption ATK is a 1-in-6 draw per line and a Topping's three lines carry it half the time. Our Resonant Toppings guide covers where they come from, and the Rambutan build shows the opposite use of the same pool: she wants the Cooldown line this build can do without.

The upgrade is real but modest. Five Resonant Toppings average about 3.25 points more ATK than five regular Epics (6.25% against 3.0%), worth 2.6 points of Cooldown, and they lift the chance of a six-piece set reaching the cap from about 2.5% to about 26%.

The Crispy Beascuit slot

Beascuits are locked to Cookie type, so an Ambush Cookie takes Crispy, which our Beascuit tier list puts in the S tier. A Legendary Beascuit has four bonus-effect slots, each with a 1-in-11 chance of any stat (excluding Element DMG), duplicates allowed, and a Legendary Cooldown line rolls 2.0% to 6.0%. The chance that a Legendary Crispy carries at least one Cooldown line is about 32%, before Attunement.

Treat the Cooldown line as filler. Beascuit ATK does not feed the passive, since the box says "from Toppings". Once the passive sits near -56% (70% ATK from Toppings), one average line of 4% reaches the 60% cap, and everything past it is wasted.

What the sources do not settle

• The unit on the passive. "1.0 ATK" has no percent sign. The Pitaya Dragon and Ananas Dragon boxes suggest percent points, but that is an inference.
• What counts as "gained from Toppings". Set bonuses and a Tart are the gap between a build that reaches 55% and one that reaches 45%. The wiki does not say.
• Whole steps or pro rata. If only whole points count, a build loses at most 0.8 points of Cooldown.
• The Max column. Skill Powder values for most of the kit are still "?" on cookierun.wiki. The ATK-heavy build is chosen on the passive alone.
• Max HP against ATK. The kit's damage has a Max HP component against Cookies, and no source gives Lychee Dragon's stats, so no post can weigh ATK Toppings against HP Toppings for damage.
• First-cast time. The Cooldown/Data page flags a multiplier for Lychee Dragon's initial Cooldown but has not filled it.

Quick Action Checklist

1. Build five Searing Raspberry M Toppings and take them to +12: 45% ATK, worth 36 points of Cooldown on the wiki's reading, 10.25 seconds.
2. Add the Equip 5 bonus, then a Searing Raspberry Tart M for the six-piece bonus: a full set sits near 64-69% ATK before bonus effects.
3. Keep ATK bonus effects first and DMG Resist second. Cooldown lines are filler, because the passive already fills the 60% allowance.
4. Use Dragon's Sweet Prank Raspberries if you have them: ATK rolls are 2-3%, and a six-piece set reaches the cap about 26% of the time against about 2.5% on regular Epics.
5. Put a Legendary Crispy Beascuit on them and keep a Cooldown line only if you are still short of the cap.
6. Expect numbers to move. The wiki's Max column is empty and the unit and set-bonus questions are open.