# Guild Wars 2 And Magic Find ~ Tue, 03 Feb 2015 03:24:41 +0000

**Luck**^{1}

Account-wide magic find is increased by consuming Essence of Luck. Essence can be acquired by salvaging fine or masterwork items. Magic Find is also increased through achievement rewards.

**Magic Find**^{2}

Magic Find is both an account bonus and a character attribute that increases the chance to receive higher-quality loot from slain foes.

The above descriptions are for game mechanics in Guild Wars 2. Basically, as Luck increases so so does Magic Find (MF). But MF is confusingly reported in-game as a percentage, e.g. 22% or 130% (all the way up to 300%).

If you're like most people, you'd assume that at 100% MF every piece of loot you find would have magic properties. That notion is as incorrect as the representation. A character's MF percentage is a weight, similar to a weighted average, that affects which loot table will be used to determine the loot "found" by that character.

So, to better understand this weird "100% Magic Find doesn't mean every found piece of loot is magical" situation, let's go through a plausible scenario for how it works.

Assume you have a GW2 character with 100% MF and you've just induced a loot decision that involves the following tables:

- Junk
- Common
- Rare
- Exotic

Tables *1* & *2* contain zero items with magic properties, table *3* contains only items with "average" magic properties, and table *4* contains items with very desirable magic properties (i.e. you **really** want to draw from table *4*). Each table contains ten total items. Once a table is chosen, a random item will be picked from that table as the loot item. We'll now theorize a possible mechanism by which MF could work in GW2 under this scenario.

We want to draw from a single loot table so we need to eliminate three of our choices. A simple way to do that is to merely use a PRNG to pick a number between 1 and 4. But let's try to be a little more "fair." We have 4 choose 2 possible combinations of tables, for 6 combinations in total:

A. 1 & 2 B. 1 & 3 C. 1 & 4 D. 2 & 3 E. 2 & 4 F. 3 & 4

Now let's assign a weight to each of these groups based on our assumed 100% MF:

A. 10% B. 15% C. 20% D. 20% E. 20% F. 15%

Now we can use some PRNG to pick one of our groups, but also factor in the given chance that group will be chosen (e.g. 10% for group *A*). Whatever scheme we devise, it is clear that there is still a significant chance the group we choose will include a loot table with zero magical items in it. Regardless, once we have picked a group we will then eliminate one of the two tables and choose from the remaining table.

Let's assume that our algorithm landed on group *B*. Factoring in our MF again, we could assign a weight of 30% to table *1* and 70% to table *3*. Now we would use our PRNG again to pick the table to use, again factoring the weights (.3 and .7). And then, finally, we would use a simple, non-weighted, PRNG to pick an item from the chosen table.

It should be clear that there is still a very significant chance that we will pick from a table that has zero magical items. Thus our assumed 100% MF **does not** guarantee that the loot item will be magical.

Personally, I think "Magic Find" should be labeled "Luck" in the game. It would make a little more sense than 100% MF resulting in non-magical loot drops.