100 Overland Chest results

krachall

Many years ago, I noticed that I seemed to get fewer stam (medium) set pieces from overland chests than mag (light) or tank (heavy). So I started keeping track, 100 chests at a time. I did this mainly in base game zones and the numbers validated my observations. In fact, with sample sizes that well exceed the requirement to determine randomness, Light and Heavy sets drop at about double the rate of medium sets.

Today, I found myself in need of a specific piece so I set off to loot some chests. I headed to Deshaan and looted about 100 chests. With the new loot system in place, I paid attention to my collection status. Fortunately, all were really close, with Night Mother's having 16 pieces, Mother's Sorrow having 17 pieces, and Plague Doctor having 19 pieces. If anything, you'd expect the new loot system to skew this in favor of the medium set as I had collected the fewest pieces. Here are the results:

Mother's Sorrow (Light): 33 (36%)
Night Mother's (Medium): 19 (21%)
Plague Doctor (Heavy): 40 (43%)

Looks like not much has changed with Light and Heavy dropping about 2x that of medium. And the most common single item? Mother's Sorrow staff. I had more of those than all other weapons combined. It really Elder Staves Online! Least common item? Night Mother's weapons in general. I had exactly one total, a sword.

redspecter23

The curated set system doesn't affect chests so the set pieces you received should be a random assortment from the available loot table, which is what it appears to be, within margin of error for your sample size.

CrashTest

The new loot system doesn't apply to overland chests. It's only bosses from trials and dungeons, and final arena bosses.

CrashTest

Curated Item Set Drops

Bosses and reward chests from arenas and incursions will now preferentially drop set items that have not yet been unlocked in your Item Set Collections!

Most bosses in the game, along with reward chests from arenas and incursions, will now preferentially drop set items that have not yet been unlocked in your Item Set Collections. These sources will only drop the set items that they can normally drop; the ways to acquire a particular item are unchanged, but it should be much easier to find items that you have not yet unlocked.

Set items looted from treasure chests, containers, or from non-boss monsters are generally not curated in this manner.

Ippokrates

The new system doesn't apply to chests HOWEVER during few dungs and trials I have noticed weird tendency to repeating the same stuff in drop from chests and kills.

Kaelthorn_Nightbloom

You might want to loot more chests before drawing any conclusions. At 100 sample size you have a 99% confidence that your data will have a 12.88% margin of error. That's huge.

To get below a 5% margin of error with 95% confidence, you need to loot at least 384 chests. (667 chests w/ 99% confidence)

krachall

I can't recall the data but the sample size to just show randomness was far smaller than what I had. I'm not trying to predict an outcome, just prove or disprove randomness.

redspecter23

krachall wrote: »

redspecter23 wrote: »

The curated set system doesn't affect chests so the set pieces you received should be a random assortment from the available loot table, which is what it appears to be, within margin of error for your sample size.

Mathematically incorrect. Based on the sample size, the distribution is not random at all. In fact, when I ran the numbers the last time, a sample size much lower was all that was needed. If I recall...and it's been many years, it was something like 48. I had 92 samples in this instance and have had over 200 previously.

100 chests is not a significant number to determine any weighted rng based on your numbers. Sorry.

krachall

And, lastly, even if we were trying to predict outcomes instead of just prove randomness, I actually have 400 chests in total that all give the same results. Heavy and Light occur 2x that of medium.