CS Rarity List [WIP] - status update on page 75

[wonderlandfox]

If there is any help needed, I'm available and open to help.
I'm good at organizing things if that helps in anyway ♡

[Solloby]

^ We'd probably be better asking a math/stats site if we wanted outside help. We can certainly do that if we don't get this worked out over the holidays.

I've reuploaded the Master Data spreadsheet, which now has all July/Aug/Sept pets in one sheet, and some rerelease data for URs in another. Unfortunately Stash zips up spreadsheets. If the contents of the zip are not a single excel file, don't unzip it - just open the .zip directly into Excel, it should open. From there you can save it as a .xls file: https://sta.sh/01kqpb1xuqll

Since rereleases are different to original release percentages, and much smaller amounts compared to the original releases, my current thinking is we ignore the rereleases for now as well as month boundaries and try to do a full comparison of 2008 pets. This will allow us to compare pets within litters, across litters and across months. I will try to run some stats based on this if everyone else agrees this approach is best?

[Erythro]

^ We'd probably be better asking a math/stats site if we wanted outside help. We can certainly do that if we don't get this worked out over the holidays.

I've reuploaded the Master Data spreadsheet, which now has all July/Aug/Sept pets in one sheet, and some rerelease data for URs in another. Unfortunately Stash zips up spreadsheets. If the contents of the zip are not a single excel file, don't unzip it - just open the .zip directly into Excel, it should open. From there you can save it as a .xls file: https://sta.sh/01kqpb1xuqll

Since rereleases are different to original release percentages, and much smaller amounts compared to the original releases, my current thinking is we ignore the rereleases for now as well as month boundaries and try to do a full comparison of 2008 pets. This will allow us to compare pets within litters, across litters and across months. I will try to run some stats based on this if everyone else agrees this approach is best?

I am having issues with the excel document again. I would recommend looking into another data sharing website as I can't even open the zip into excel.

Are we ignoring all rereleased pets or just unreleased pets? Because total rereleases completely effects the rarity and shouldn't be ignored, and I am curious why you think they should.

[Solloby]

Never used this site before but it will only host the file for 30 days: https://ufile.io/rrm8d
Since it's just the raw data I doubt we'll need it longer than that, I am hopeful we can get some analysis done in that timeframe.

Ignoring all rereleases aside from URs, though I'm not sure how to treat URs since they are always rarer their first year due to having been released less.

Based on the 2011 etc data, rereleased pets are only released in small double digits (around 10-30) for many list pets. It doesn't appear that they follow the original release stats, and can change year to year. We only have data up to 2011 but if we had more I am still not sure it would give us confidence if we tried to include the data. Overall it just seems like an awful lot of work to include a very small amount of data.

Edit: Sorry I think I was looking at the wrong data, was thinking of some URs/Zebras. Numbers are around 70 or so, but seem fairly even across litters. Unsure how they can be incorporated into the original release data though. For example the 2008 rerelease doubled the amount of Dark Heart Rats in circulation. Around the same amount of Sunjewels were released as Dark Heart rats, but more Sunjewels were originally adopted.

If we want to incorporate rereleases, we may need to count more of them and add those numbers to the original 2008 adopt numbers. Thoughts?

[chunkypeanutbutter]

I could offer some help with this ^^ mark for later <3

[nickjr]

Do we want to run analyses on just the original adoptions and then run the same analyses on the original adoptions + all/most rereleases?

With Law's script I think it's possible to get a census of all of the pets we're looking at in a reasonable ish timeframe? I think Law just took samples of later rereleases because we thought we'd be able to get started on analyses back then and I think we were also only wondering what the ratios were like per rerelease

Thinking out loud: Could we estimate a full count based on the rerelease ratios and the number of pets adopted Dec 18? (+ dec 24 2011)

Side thought: I wonder how the pets on inactive accounts would affect the results. I think we'll just have to assume that there'll be no difference in our analyses. Probably will be fine?

edit: @Solloby First sheet of the excel file doesn't have any data; just has pet IDs, names, and litter names

[Solloby]

Looking at the data I think we may need to include rereleases after all. Without the rereleases the top rarity pet at double the rarity of Sunjewel is the Dark Heart Rat, which as we know dropped back to VR while Sunjewel remains OMGSR. So the rereleases did that. My guess is in the original release, more people adopted Sun pups than Heart rats due to species demand, sun pups having more outcomes and sun pups having a very rare outcome.

So far we have rerelease data for 08 - 11. We need 12 - 17. The master sheet doesn't contain the rerelease data though, as I wasn't sure what to do with it.

Nothing we can do to identify inactive accounts so best to just ignore that factor.

Does the excel sheet have filters on? Try clearing the filters if so. If not I will try uploading another version, and maybe remove the filters in case they are causing issues with different versions. I am using Excel 2007 on PC, I imagine others are using newer versions.

[Law]

About the rerelease for 2010 and 2011, note that the numbers are taken out of a sample of around 20k pets, and a lot of those pets are monthly adopts. :o

[Solloby]

Ok the pets that the rerelease affected are Dark Heart Rat vs Sunjewel, so here are the numbers for those pets.

Original numbers:
Dark Heart Rat - 64
Sunjewel - 110

Rereleases:
08:
Dark Heart Rat - 74
Sunjewel - 76

09:
Dark Heart Rat - 33
Sunjewel - 32

10 sample:
Dark Heart Rat - 32
Sunjewel - 15

11 sample:
Dark Heart Rat - 32
Sunjewel - 11

Total at 2011 end:
Dark Heart Rat - 235 (+ uncounted rereleased)
Sunjewel - 244 (+ uncounted rereleased)

So the uncounted rereleases would increase the amount of Dark Heart Rats above the amount of Sunjewels. It seems that 2010 was the first year the rarities changed. So it sounds like we are going to need the full data sets from 2010 - 2017, to get the most accurate picture. Without the rereleases, Dark Heart Rat goes to the top of the Rarity List (without considering Zebras & URs, the former of which would likely sit above). With the rereleases, it sits below Sunjewel which is should because it's only VR for a lot of the time, not OMGSR all the time like Sunjewel.

Which is a pain because I was hoping we could get away with ignoring the rereleases, but they've changed the rarity of some pets considerably :\

[nickjr]

Does the excel sheet have filters on? Try clearing the filters if so. If not I will try uploading another version, and maybe remove the filters in case they are causing issues with different versions. I am using Excel 2007 on PC, I imagine others are using newer versions.

I don't see any filters or hidden rows/columns on the sheet D:

@Law can we get a census of the 2010 - 2017 rereleases, maybe?

edit: Has anyone heard from Lya recently?