Yes, I think this ought to be possible. Have no experience with scraping myself, willing to try it out though given that is a usefull skill.
Iāve found this company that says they offer a opensea scrape bot for a price:
https://web.axiom.ai/recipe/scrape-opensea-data
Thatās awesome!
And looking at the pricing, if we go that route, Iād be okay paying for that.
It would be interesting to see if gem.xyz rankings are close to what we get. Although I donāt understand why these websitesā rankings are messed up, it should be straightforward.
On another point, opensea doesnāt have all 10,000 NFTs lifted, and neither does gem.xyz which is weird. So that may mess up our attempts a bit.
During my examples I did above, if a trait said 511 NFTs have that trait (random example), then I said it was a 0.0511% to have that trait. While the percent listed is slightly higher because it is out of something like 9,900 instead of 10,000.
Also, more NFTs might have that trait once the remaining NFTs are listed on the sites.
I think we could still make a pretty accurate rarity list though with the information we have
Yeah I was wondering about that too, why doesnāt OS have all 10k Cypherpunks?
Yup, if we get our hands on a spreadsheet containing these details its a āwalk in the parkā from there on. Yeah they donāt charge that much and we get 2 hours of run time for free, perhaps these two hours is enough. Will you try this bot out or would you prefer not to?
Wouldnāt it be nice if it were possible to pay a couple of ZEDs and receive spreadsheets of interesting data 
Would be so cool if Zcash and Filecoin did a collab 
CPZNFT_RAR_MATRIX_v6.pdf (4.1 MB)
I extracted the rarity matrix in Excel and did the calculations according to the example of particlmike33 (but Iām not sure myself if it is calculated correctly).
I have attached the result as a PDF. Personally, I donāt agree with this ranking, I think there are much nicer CPZs than the ātop ranksā in my calculation, even if they are not that rare in terms of rarity/probability. At the end of the day with NFTs, it all comes down to aesthetics and āless is sometimes moreā.
Scratch that entire original comment
The presented table completely coincides with my calculations:
The raw data input gave me 10000 entries (checked by excel count formula), #9999 is definitely on the list at position 1533 (by total rarity calculation).
I will upload the excel file later.
EDIT Excel file uploaded on external website (XLS upload not supported on this forum) - link for download: Easyupload.io - Upload files for free and transfer big files easily.
Thanks for doing this in pdf and excel!!
I wonder why these ratings are so different than any other rarity ranking online.
And prices seem to have been influenced by the gem.xyz rankings online which isnāt great since they arenāt correct!
I think your ratings might be wrong, but thank you so much for working on it!
Maybe we can correct the math somehow and it will auto rank the rest, so most of the work is hopefully done already, thanks for doing so!
So gem.xyz ranks #16 as rarity #1. Yours ranks #2483 as rarity #1.
For #16:
0.0202 * 0.1449 * 0.0576 * 0.0028 * 0.4977 * 0.8277 * 0.0623 * 0.0294 * 0.1600 * 0.3686 * 0.4065 * 0.9731 * 0.9600 * 0.5501 * 0.4684
=2.055 * 10^-12
I think it should be this for #2483:
0.1460 * 0.1157 * 0.0211 * 0.0864 * 0.4977 * 0.1391 * 0.4687 * 0.0452 * 0.0316 * 0.4065 * 0.8277 * 0.9731 * 0.9600 * 0.9700 * 0.3149
=1.370 * 10^-10
So NFT #16 is definitely rarer than NFT #2483
Edit: missed 2 traits in doing my hand calculations for NFT# 2483, so Iāve fixed it. It didnāt change the results though, NFT# 16 is still rarer
You are right, there is a bug in the calculations.
Since I used this link (Cypherpunk Zero NFT Rarity Matrix) for the raw data, I did not consider the traits Sidekick, Bonus and Spirit.
Not sure how to get the distribution of those traits 
Where did you get these coefficients? In my table, all coefficients are calculated using formulas from the original list. There is no manual counting anywhere. In my life, I have considered much more complex models than this one.
The methodology is not given anywhere. My table is made based on the suggestion that each characteristic has an equal specific gravity. Pure statistical non-weighted methodology. However, if we propose, for example, some other methodology in which, for example, the specific weight of the ācharacterā characteristic will be two or three times more valuable than all other characteristics, then the results will be completely different. In general, it all depends on the methodology. At the same time, two different people got the same numbers. I didnāt think to post it at all until I saw that my calculation coincided with @spectre. How all these sites calculate I donāt understand at all.
Yes, indeed, I did the same thing.
We have the exact same results because the exact same traits are missing, since our raw data is not complete! It is shown in the pie charts, but not in the matrix itself.
We need the distribution of the last 3 traits and then we can make a correct comparison to the gem.xyz rankings.
Yes, now I see whatās the matter.
The only option to catch the missing characteristics is data parsing from OpenSea. However, this is outside my area of expertise.
So Gem gives Cypherpunk Zero #940 rarity rank 485 while the current community ranking gives it rarity rank 4521 
Friendly reminder that personal attacks is against the forum CoC.
Letās go to other projects to interact with giant whales to promote our CZNFT and how to improve the value of CZNFT. This is a topic worth discussing.
Giant whales are already here and new whales will come because Cypherpunk Zero NFTs are dope AF.
