Beginner Guide

This guide sticks to what we can source. Where something is unconfirmed, we say so instead of guessing — that is the whole point of this site, and it matters most when you are new and cannot yet tell good advice from confident noise.

The shape of the game

Cookie Run: Crumble is an idle deck-builder. You assemble a squad, it fights largely on its own, and your job is deciding who is in it and what they upgrade into. You can field up to twelve Cookies at once, and three pets. Both of those are slot caps that unlock gradually as you clear Main Stages rather than being available from the start.

Because combat is automatic, roster composition is where nearly all of your agency lives. That is why we put the most work into the Team Builder.

Stats are not how you pick Cookies

This is the single most useful thing we can tell a new player, and it is not obvious.

Base stats in this game are class templates, not individual values. Every SSR Ranged Cookie has exactly the same block:

| Rarity + class | ATK | HP | DEF | |---|---|---|---| | SSR Ranged | 155 | 2,170 | 60 | | SSR Charge | 146 | 2,635 | 85 | | SSR Support | 133 | 2,790 | 60 | | SSR Tank | 124 | 3,255 | 120 |

Milky Way, Espresso, Scorpion, Melon Soda, Twizzly Gummy and Cream Puff are all 155 / 2,170 / 60. Identical.

The consequence: any ranking sorted by stats or by Combat Power is sorting noise. Two Cookies of the same rarity and class are separated entirely by their skill — what it does, how often it fires, and which synergy it scales with. Judge on that.

(Base stats for rarities below SSR are not published anywhere we could find, which is why most Cookie pages on this site show stats as unverified.)

Synergies are the real constraint

Every Cookie's skill scales with a synergy — Range, Duration, Volley, Chain, Pierce, Rapid Fire or Projectile Speed. But only 18 of the 71 Cookies we track supply one.

That asymmetry is the thing to understand early. Eighteen Cookies of the roster provide the buffs that the other fifty-three are built to consume. A skill that scales with Volley is doing a fraction of its job on a team with no Volley provider.

Supply is uneven, too:

| Synergy | Providers | Cookies wanting it | |---|---|---| | Range | 4 | 18 | | Duration | 3 | 14 | | Rapid Fire | 2 | 12 | | Volley | 3 | 9 | | Projectile Speed | 2 | 8 | | Pierce | 2 | 6 | | Chain | 2 | 5 |

Four providers covering eighteen consumers is tight. Two providers covering twelve is tighter. Before you invest heavily in a Cookie, check whether you can actually feed it — our synergy pages list every provider, and the Team Builder flags unsupplied boosts on a team automatically.

Cooldowns vary enormously

Skill cooldowns across the roster run from 0.8s (Red Pepper) to 10s (Pomegranate, Rockstar, Tiger Lily), against a roster average of about 5.2s. That is more than a tenfold spread, and it interacts directly with the buffs above: a long-cooldown skill gets far more out of Duration effects, while a fast one benefits from being able to fire repeatedly inside a buff window.

Cooldown is listed on every Cookie page, along with how it ranks against the roster and against its own class.

Cover your classes

Cookies are Charge, Ranged, Tank or Support. A team missing an entire class has an obvious hole, and the Team Builder will tell you which one. This is the least subtle failure mode and the easiest to fix.

Redeem the codes first

Codes are not redeemed in-game. You enter them on the official coupon site at coupon.devplay.com/coupon/cc/en, together with your Member ID — found by tapping your profile avatar in the top-left corner of the game. Rewards then arrive in your in-game inbox, and they expire if unclaimed.

Current codes are on our codes page, flagged automatically by expiry date.

What we are not going to tell you

You will find beginner guides confidently ranking the best starter Cookies as though the rankings were measured. They are not — nobody has published in-game performance data for this game.

Our tier list does carry grades, drawn from community consensus and cross-checked against creator lists and community threads. But every grade says exactly that on its face, because "the community broadly agrees" and "we tested this" are different claims. Treat it as a starting point, not a measurement. See the methodology.