Spiral Knights - UV probability analysis

UV probability analysis

8 replies [Last post]

Sun, 05/01/2011 - 16:47

Ouros

Legacy Username

UVs are all the rage, and for good reason! Just how much money and time are they worth? The answer depends on their rarity.

The question we aim to solve in this thread is this: Just how rare are the good UVs?

It's data mining time! Anytime you craft items, please list here what you crafted and in what quantities and what UVs resulted. Please edit your posts rather than making new ones in order to save space.

As for me:

8 brandishes
1 low slime
1 low beast
1 low construct
1 low fiend

5 skelly shields
1 low normal
1 medium shadow

Sun, 05/01/2011 - 18:24

ElementLord

Legacy Username

UVs I've gained.

My great defender has Low Elemental defence, and the flourish i recently crafted has a low damage bonus to undead.

Both of them i got on first try, but i don't farm for UVs.

Sun, 05/01/2011 - 17:30

Ouros

Legacy Username

The goal here is to note how

The goal here is to note how many flourishes it took before crafting that one, so that over time we can discover what % of crafted weapons feature low/medium/high/vh variants.

Sun, 05/01/2011 - 17:36

Icee

Already tested, 10%

http://forums.spiralknights.com/en/node/1315

Over hundreds of tries, the UV rate was determined to be about 10%. I suggest you either add your analysis to that thread or at least copy the data from there to here, so all of the numbers are collected in one place and no data is missed.

Sun, 05/01/2011 - 17:38

Lymojo

Legacy Username

People who happened to get a

People who happened to get a UV on their one personal copy of an item should NOT post here. If you do, you're introducing huge sampling bias (think of all the people who did not get a UV and are thus not posting here), and will throw off the stats.

The only people who should post here are people who are specifically farming for UVs and have lots of sample points.

Sun, 05/01/2011 - 17:49

Pupu

Legacy Username

Yeah

To actually get a decent number we need someone who crafted at least 1000 or 10000 of the same item, everything else will give bad numbers.

Sun, 05/01/2011 - 18:23

Icee

All data is useful, but to be accurate you must report all

We don't need one person who crafted 1000 items themselves, but we do need people to report as many points as they know, including the one-off attempts that failed. We'll have a bias that indicates a higher than actual UV rate if people only report when they get UVs, but knowing that someone made a single cutter without a UV is still useful.

Here's the data of mine that I know:

Spiral Demo Helms: 0/8
Fused Demo Helms: 0/1
Heavy Demo Helm: 1/1 (sleep resist medium)

I haven't really farmed anything else, but here's other items I have made that I am sure of:
Crystal bomb: 0/2
Cobalt helm: 1/2 (low fire resist)
Solid Cobalt helm: 0/1
Cobalt armor: 0/1
Blast bomb: 1/1 (med dmg to slime)
Jelly shield: 0/3
Brute jelly shield: 0/2

My totals, so far: 3 UVs / 22 alchemies or about

Sun, 05/01/2011 - 19:28

Garbeld

I've crafted 12 Static Edges

I've crafted
12 Static Edges
13 Haze Bombs
2 Firotech Alchemers
1 Spur
Arc Razor
Cutter
Cobalt Helm
Vitasuit Plus
Defender
Swiftstrike Buckler
with 0 UVs total.

I've crafted 1 Firotech Alchemer with Dmg. Vs. Gremlins: Low

... am I just unlucky?

Sun, 05/01/2011 - 20:21

Pauling

Thanks for linking to the

Thanks for linking to the wiki thread, Icee. That is indeed the best place to post data, as people over there have reported ~500 trials so far. Data is also collected for variant size and type.

In the future, we should also try to collect data on how much damage a given weapon does across tiers, to get a sense of how much "bonus" damage is provided by variants vs special damage types. Again, the wiki forum is the best place to do that. (For one thing, posts don't get buried within hours)

Search

UV probability analysis