frequency values

[digdug]

Hi George,
just a quick question, are the frequency values still as in this table ?
* Common 20%
* Uncommon 10%
* Rare 5%
* VeryRare 3%
* Notrandom

I'm asking because I don't know how to test it in a simple way.

[schilcote]

frquency values?

[george moromisato]

Hi George,
just a quick question, are the frequency values still as in this table ?
* Common 20%
* Uncommon 10%
* Rare 5%
* VeryRare 3%
* Notrandom

I'm asking because I don't know how to test it in a simple way.

They are (and, I believe, have been for a while):

Common: 20
Uncommon: 10
Rare: 4
VeryRare: 1

[Technically, they are not % values--they simply relative values: "common" is twice a frequent as "uncommon".]

[Arisaya]

Thanks george!
So it's like:
(this-rarity/sum(all-rarities))=chance <- like that?

[Prophet]

Code: Select all

Common: 20 
Uncommon: 10 
Rare: 4 
VeryRare: 1

Common: 20/35=	57%
Uncommon: 10/35=29%
Rare: 4/35=		11%
VeryRare: 1/35=	3%

One further question: what would be the effect of creating 1000 new items, all with veryRare frequency? would that skew the curve or would those 1000 items all be competing for that 3% chance of some screen time?

Just out of curiosity, can transdata output a summary of the number of items in each frequency?
eg:
Common: 200 items
Uncommon: 100 items
Rare: 40 items
VeryRare: 10 items

[george moromisato]

Thanks george!
So it's like:
(this-rarity/sum(all-rarities))=chance <- like that?

Yes, that's right.

[george moromisato]

Code: Select all

Common: 20 
Uncommon: 10 
Rare: 4 
VeryRare: 1

Common: 20/35=	57%
Uncommon: 10/35=29%
Rare: 4/35=		11%
VeryRare: 1/35=	3%

One further question: what would be the effect of creating 1000 new items, all with veryRare frequency? would that skew the curve or would those 1000 items all be competing for that 3% chance of some screen time?

Just out of curiosity, can transdata output a summary of the number of items in each frequency?
eg:
Common: 200 items
Uncommon: 100 items
Rare: 40 items
VeryRare: 10 items

Yeah, that looks right.

Creating 1000 very rare item types would skew the curve.

The frequency is used most commonly in <RandomItem> (which generates random items for stations). The algorithm iterates over selected item types and adds up their total frequency value. The percent chance of any one item type is its frequency value divided by the total frequency value.

So, imagine that the criteria selected 1 common item and 80 very rare items. The total frequency value is (1 * 20) + (80 * 1) = 100. The chance of getting the common item is (20/100) = 20%. The chance of getting a specific very rare is (1/100) = 1%. The chance of getting any very rare is 80%.

The best way to get stats on anything is to have TransData output a full table and use Excel to slice the data.