Reading an item card before you take it

The small marks in the corner of an item’s icon are its rarity. That is the question almost every new player asks first, and it stays worth asking, because plenty of people who have played for hours never find out. But rarity is the least useful line on the card. The lines that decide whether you should take the thing are further down.

Here is the whole card, in the order that actually helps.

Rarity: five steps, and not a power ranking

Rarity runs Common, Advanced, Rare, Legendary, Eternal. It moves the price a merchant asks and it tells you roughly how uncommon the card is. In 1.0.23 you will only ever see the Eternal mark on a tablet — no artifact reaches it.

What rarity does not tell you is whether the card is good for your run. A Common artifact that belongs to the category you are stacking is contributing to a threshold; a Legendary from a category you hold one of is contributing nothing but a filled slot. Rarity is a tiebreaker between two cards that are otherwise equally relevant, and nothing more than that.

Item type: Artifact or Tablet

These are two different machines and they are the first thing to separate.

An artifact produces the effect. It works simply by being in your inventory, and its strength is set by its level. A tablet produces nothing on its own — it changes the level of artifacts in the direction it points. Tablets have no level of their own, so a Legendary tablet is not a stronger version of a Common one, it is a different shape with a different reach.

Anything else you pick up — potions, food, throwables, quest pieces — is neither, and the arithmetic for those is simply one slot each, with a stack counting as one slot.

Categories: the line that decides the run

The Categories line lists the combos the card belongs to: Sturdy, Precision, Glacier, Academy, Companion, Shadow and the rest of the twenty. This is the single most important line on an artifact card, because combos pay at counts, not at rarity.

Two Precision artifacts already give Critical Chance +4%; six give +8%; ten swap the whole thing for Critical Damage +30%. Sturdy climbs through flat Physical Damage and pays out at ten with Physical Damage Amplification +15%. Wind Song hands out Attack Speed on the way up and finishes at ten with Weapon Damage +15 and an extra dash. Lake works in threes instead of twos. Every category has its own ladder, and the ten-step of a category you have committed to is usually the largest single upgrade available to a run.

So the comparison to make at a reward screen is not “which of these is rarer” but “which of these is a step on a ladder I am already climbing”. Artifact combos covers the spread-or-stack decision properly.

Maximum level and Unique effect

Maximum level is the artifact’s ceiling. It matters for two reasons: an artifact already at its maximum cannot be enchanted, and an artifact whose maximum level is 0 cannot be enchanted at all — so at an altar the ceiling tells you in advance which of your cards can even accept the offer. Tablets can push an artifact past its ceiling, and when that happens the card starts showing you how far over it is. Tablets can also push in the other direction: an artifact dragged below level 1 stops working, which is the usual explanation for an effect that has silently gone quiet.

Unique effect is a yes/no line. A Unique effect is the card’s own named effect rather than a stat you can add up, and the clearest thing the game says about it comes from a talent perk that forces every artifact to carry one: while that perk is active, picking the same artifact up again merges the two and raises the artifact’s level by one instead of handing you a second copy. Without it, a duplicate is simply a second card asking for a second slot. Level is the thing that compounds without costing you space, which is why tablets and enchanting matter more than duplicates do.

The condition lines

This is where cards stop being comparable and start being placement puzzles.

Some artifacts only trigger from certain positions, and they say so in plain words: triggers when placed in the top row, when in the lowest slot, when in the inner slots, when in an outermost slot, when placed in either of the end columns, when both adjacent slots are empty, when all surrounding slots are filled, when there is an adjacent grimoire, when at max HP. Read that line before you take the card, not after you have already filled the row it wants.

Tablets carry two kinds of line of their own. A placement condition — only active in the marked slots — means the tablet does nothing at all unless it sits where the pattern says; a failed condition applies none of the tablet’s effects, not a reduced version of them. And a rotation line: most tablets can be rotated, and the ones that cannot say disable rotate, so their pattern is the only pattern you will get. Tablet placement takes the condition-versus-effect distinction apart in detail.

The effect a tablet applies is written in the same plain language: increase artifact level, reduce artifact level, disable artifacts, and the one that causes the most confusion in 1.0, ignore artifact constraints. That last one waives placement rules, not an artifact’s own gameplay condition — ignored constraints draws the line exactly. A worked example of a plain, strong tablet card is Conviction: Legendary, rotatable, no condition, and it raises one neighbouring artifact by five levels — you rotate it to choose which.

The ten-second read

At a reward screen, in this order:

1. Type. Artifact or tablet — am I buying an effect or buying levels?
2. Categories. Is this a step on a ladder I am already on?
3. Condition line. Does it want a position I can actually give it?
4. Maximum level. Is there room to grow it, or is it finished on arrival?
5. Rarity. Only if two cards are still tied.

Everything on the card is also on its page in the item database, and if you only half-remember a name from a run, the search will find it from a fragment. When you want to test whether a card fits before you commit a slot to it, lay the grid out in the bag planner.