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

[nickjr]

Sorry been busy OTL Man I wish I still had school bc now I have to deal w parents lolol

Put the sorbet data into a chi squared goodness of fit (single variable) calculator. Your "actual" is your data and your "expected" is 25% of the total number of sorbets we have.

The chi value that is spur out doesn't mean much (basically only for the curious). You want to look at the p-value. It's a % in decimal form. If it's low (for example, below 0.05) then we have a problem because p-value IIRC is the chance of getting a data sample that is at least as extreme as our "actual" assuming 1) random sampling and 2) that our "expected" values are correct. If the p-value is relatively high then we're good

[Pelicandreams]

Sorry been busy OTL Man I wish I still had school bc now I have to deal w parents lolol

Put the sorbet data into a chi squared goodness of fit (single variable) calculator. Your "actual" is your data and your "expected" is 25% of the total number of sorbets we have.

The chi value that is spur out doesn't mean much (basically only for the curious). You want to look at the p-value. It's a % in decimal form. If it's low (for example, below 0.05) then we have a problem because p-value IIRC is the chance of getting a data sample that is at least as extreme as our "actual" assuming 1) random sampling and 2) that our "expected" values are correct. If the p-value is relatively high then we're good

I just wanted to reach out to you and let you know I can help with statistical tests and the sort. I can also help with looking up more pet ranges if you believe we need more data.

The Chi square is 1.44, the DF is 3
This makes our P-value roughly 0.75. This essentially means that we can reject the null hypothesis and conclude that the sorbets are distributed evenly. So, that means that the sorbets are released evenly (even if we currently have some more green and yellow).

It has been a few years since I took statistics in college, but luckily these equations were drilled into my brain! I'd love to help more if you need it. Just let me know!

[depressedanonymous,]

^im not even out of high school, so I have no idea what any of that means. But, if I do get permission to help count toxics, I'll try to research it!

[Solloby]

I used Excel's CHITEST function to set up dynamic Chi-Square tests for our data.

This is what we have so far:
Chi-Tests
Sorbets 0.097668866 0.05 Equal Rarity
Toxics 0.001806855 0.05 Different Rarity
Galaxies 0.199737241 0.05 Equal Rarity
Balloons 0.159036636 0.05 Equal Rarity
Tribals 0.002305113 0.05 Different Rarity
Aug Rats 1.57054E-45 0.05 Different Rarity

Sorbets, Toxics, Tribals & Aug Rats are expected.
Galaxies & Balloons are unexpected. Even though CEGs slightly outnumber BEGs, the difference isn't enough yet.
Nonballoons have caught up to the Balloons. It's possible they are the same rarity but we may need more data to find out.

It's up to you guys if we keep counting or release the data now and move onto July or Sept 08. Personally I would like to keep counting but we do have to stop at some point. There are a few pets I would not know how to classify their rarities as they are not statistically significantly different from the rarer or less rare outcomes they sit between.

[depressedanonymous,]

While doing my set, I found that there were a lot more nonballoons than Balloon pups^

[Fizzlebee]

We could finish that list of IDs on page 17 and see where we stand after. It's a solid number of IDs, so any underlying ratios should hopefully become a bit clearer. We also have the 08 rerelease data later in the project that we can compare with. If they agree, we're good. If they don't, we can always count more. ^^

I can update Crestalia's list if we decide to continue counting. I think we're somewhere around halfway done with it.

[Pelicandreams]

I'm up for counting till ID 30,000. I'm not sure if ratios will change, but it's always worth a try! ( I believe there's only 9 data sets till 30,000?)

On the flip side, the number of "other pets" has increased significantly with the introduction of the butterfly wolves. So you're not getting at much usable data in each set counted.

[kee;]

I'm so sorry it took me so long to get this set finished. I've had a lot going on this week and haven't had the chance to sit down and work on it until now.
https://1drv.ms/x/s!AomMmkWw1P0odexWPvrCmFRuzsQ

-kee

[Solloby]

Thanks Kee, that data helped the sorbets a little c:
Sorbets
Pink Sorbet 117 22%
Blue Sorbet 119 23%
Yellow Sorbet 148 28%
Green Sorbet 144 27%
Total 528 100%

There might still be IDs in the 21-22k range that haven't been done but I'm not entirely sure? I sort of lost track of which one's we've done ^^; I have been running duplicate checks in the master spreadsheet though and so far we haven't had any which is great.

[Fizzlebee]

After doing some post searching, it looks like none of the IDs in 21-22K have been done. 21001-21250 was given out a week ago, but the volunteer said their computer was being fixed and we haven't heard back. We can go ahead and take care of it for them, and they can jump back in whenever their computer situation is sorted out. c:

21001-21250 (Fizzlebee)
21251-21500 (Pelicandreams)
21501-21750
21751-22000 (NightFlame22)

I'll go ahead and do 21001-21250. ^^

Claim away!

Edit: Finished counting 21001-21250