Metin2 Forum

Metin2 drop rate files: how PCT numbers work, the level penalty, and the mistakes that break a drop economy

Private server admins edit the same five files over and over, and the most common disaster is a drop table where nothing ever drops or everything drops. Both come from the same misunderstanding.

The five files that control loot

• mob_drop_item.txt, per-mob drops on kill. This is the file people edit most often.
• common_drop_item.txt, shared drops for all mobs within a level range.
• etc_drop_item.txt, drops linked through the drop_item field in mob_proto.
• special_item_group.txt, the contents of chests and boxes.
• drop_item_group.txt, drop groups for stones and specific entities.

The PCT numbers are not percentages

A group block in mob_drop_item.txt looks like this, in the layout used by the community tooling:
Each item line carries five numbers, and the last two are the probability range. The engine picks a random number between pct_min and pct_max and divides it by 1,000,000. That is the base chance.

Worked through, pct_min 10 and pct_max 30 average to 20, which is 0.002 percent, or roughly one drop in 50,000 kills. A value of 100 does not mean 100 percent. It means 0.01 percent. To target a 1 percent drop rate you need a value around 10,000, so pct_min 8000 and pct_max 12000.

The level-difference penalty

The base chance is not what a player experiences, because the game applies a multiplier based on the gap between the player's level and the mob's level:

• Player level at or below mob level plus 5: full drop rate, no penalty.
• Player 5 to 15 levels above the mob: linear reduction from 100 percent down to about 1 percent.
• Player more than 15 levels above the mob: drop rate approaches zero.
• Player below the mob's level: a small penalty also applies.

That is why a level 90 player farming a level 70 mob gets almost nothing, even when the PCT values are set correctly for a level-appropriate character. The penalty is steep enough that it reads as a bug, and the usual fix is to test with a character at the mob's own level rather than to lower the numbers.

Other multipliers stack on top

• Boss mob, rank BOSS or KING: times 1.5
• Premium account: times 1.5
• Double item event: times 2.0
• Thief gloves: times 1.1

These are multiplicative. A premium player wearing thief gloves during a double item event against a boss gets base times 1.5 times 1.5 times 2.0 times 1.1, which is base times 4.95. Nearly five times the intended rate, which is the other half of the "everything drops" problem.

Mistakes worth avoiding

• Setting pct to 100 expecting 100 percent. It is 0.01 percent.
• Testing drops with a character above the mob's level and then assuming the table is wrong.
• Forgetting that each item in a group rolls independently on every kill, so one item's rate does not consume another's.
• Reading kill_drop as an item rate. It scales yang drops, not item drop rates.
• Using the same PCT values for bosses and normal mobs, since bosses already carry the times 1.5 bonus.
• Editing item_proto or mob_proto without a versioned backup. Duplicate VNUMs cause silent overrides or a server crash.
• Expecting the client to change after a server-side edit. The client binaries have to be regenerated, and the binary item_proto.bin and mob_proto.bin are encrypted with TEA and compressed with LZO, so they are never edited by hand.

One caution on the source

The figures and the file descriptions above come from a third-party admin tooling site rather than from an official Gameforge document. The formula it describes is internally consistent and matches the format of the group blocks, but treat it as community documentation and check a live drop table before changing a production server.

Sources
Metin2 Drop Rate Explained: how PCT values work in mob_drop_item.txt
Metin2 private server admin checklist: the configuration files and the proto formats