Woynich
@woynich.bsky.social
He/Him, born 1996, East Coast US, Digital Art (Toon), emphasis on Furry/Macro/Muscle. 🔞
www.furaffinity.net/user/woynich
www.furaffinity.net/user/woynich
I'd say to go with kink just to be safe. The rule of thumb as its been explained to me is only post in sfw chats stuff that would be fine to show up in the background if you opened the chat up on your work laptop before a presentation by accident.
November 24, 2025 at 11:26 PM
I'd say to go with kink just to be safe. The rule of thumb as its been explained to me is only post in sfw chats stuff that would be fine to show up in the background if you opened the chat up on your work laptop before a presentation by accident.
I could be off by as much as 3 or 4 orders of magnitude, which to be clear: is still very much a small enough p-value to reject the null hypothesis and conjecture that the representation of rolls on the show do not match a fair die— but there are many explanations why that may be, as you said.
November 17, 2025 at 3:51 AM
I could be off by as much as 3 or 4 orders of magnitude, which to be clear: is still very much a small enough p-value to reject the null hypothesis and conjecture that the representation of rolls on the show do not match a fair die— but there are many explanations why that may be, as you said.
I was working on exactly that this morning but messed up and used the wrong formula, got the p-value for rolling exactly the number of 20s they got as 7x10^-24 but I need to calculate the standard dev and z scores for rolling 1255/19139 nat 20s or higher.
November 17, 2025 at 3:48 AM
I was working on exactly that this morning but messed up and used the wrong formula, got the p-value for rolling exactly the number of 20s they got as 7x10^-24 but I need to calculate the standard dev and z scores for rolling 1255/19139 nat 20s or higher.
I get that to a certain extent, it’s folks being awkward around how the fetish is focused on bigness so that’s where you go, that’s how you escalate things. I’ve made my own commentary in the past about trying to ask clients how specifically we want to represent a character getting bigger.
November 16, 2025 at 9:20 PM
I get that to a certain extent, it’s folks being awkward around how the fetish is focused on bigness so that’s where you go, that’s how you escalate things. I’ve made my own commentary in the past about trying to ask clients how specifically we want to represent a character getting bigger.
I don't think I'm gonna be able to finish this problem before I have to get ready for work
November 16, 2025 at 6:57 PM
I don't think I'm gonna be able to finish this problem before I have to get ready for work
trying to fix my math now, you've got me reminding myself how to consult a z-table.
November 16, 2025 at 6:53 PM
trying to fix my math now, you've got me reminding myself how to consult a z-table.
I did the math wrong! This is the formula for getting exactly this number of 20s and not that number of 20s or higher! Ottersparx pointed this out to me: bsky.app/profile/otte...
This is the probability of rolling *exactly that many 20s, when I'm sure what you're interested in is the probability of rolling *at-least that many 20s.
Because the probability of rolling exactly *any number of 20s is going to be low in a sample size that large.
Because the probability of rolling exactly *any number of 20s is going to be low in a sample size that large.
November 16, 2025 at 6:47 PM
I did the math wrong! This is the formula for getting exactly this number of 20s and not that number of 20s or higher! Ottersparx pointed this out to me: bsky.app/profile/otte...
even if I had done the math right right, and had a convincing p-value: that does not seriously prove that anyone cheated, there are other explanations.
November 16, 2025 at 6:44 PM
even if I had done the math right right, and had a convincing p-value: that does not seriously prove that anyone cheated, there are other explanations.
there could be plenty of 2s that were part of multiple roll checks where only the sum was announced, or were part of high/low checks where they just had to say it was lower and moved on, or just other incidental nonrecorded rolls which the stats lists a lot of those.
November 16, 2025 at 6:42 PM
there could be plenty of 2s that were part of multiple roll checks where only the sum was announced, or were part of high/low checks where they just had to say it was lower and moved on, or just other incidental nonrecorded rolls which the stats lists a lot of those.
I still think its a low chance and there's some bias going on, but clearly I did bad math and should leave this up to someone more facile with statistical analysis. Also need to keep in mind, its a show, we do have the built in bias of how these are the exciting rolls more likely to be announced.
November 16, 2025 at 6:41 PM
I still think its a low chance and there's some bias going on, but clearly I did bad math and should leave this up to someone more facile with statistical analysis. Also need to keep in mind, its a show, we do have the built in bias of how these are the exciting rolls more likely to be announced.
Another person did better math than me and still got a low probability, but its also been pointed out that D&D has the advantage/disadvantage rules which filter out to higher/lower numbers in each case, and a lot of tables use "taking 10" on checks to treat a unimportant check as 10 without a roll
November 16, 2025 at 6:39 PM
Another person did better math than me and still got a low probability, but its also been pointed out that D&D has the advantage/disadvantage rules which filter out to higher/lower numbers in each case, and a lot of tables use "taking 10" on checks to treat a unimportant check as 10 without a roll
Oof. Yeah, that's a regrettable mistake.
November 16, 2025 at 6:38 PM
Oof. Yeah, that's a regrettable mistake.
There's probably a similar bias you'd get across every tabletop podcast or show just by the nature of the medium where some rolls aren't recorded vs. a real game where all the rolls are experienced, but I suppose we don't really have a way to factor that out.
November 16, 2025 at 2:13 PM
There's probably a similar bias you'd get across every tabletop podcast or show just by the nature of the medium where some rolls aren't recorded vs. a real game where all the rolls are experienced, but I suppose we don't really have a way to factor that out.
That is something to consider, yes. It is a show and the dataset we're working from here is definitely biased to favor what die results are likely to be recorded. I still think that factors into the feeling Rick Griffin expressed that die rolls on the show end up feeling less fair.
November 16, 2025 at 2:10 PM
That is something to consider, yes. It is a show and the dataset we're working from here is definitely biased to favor what die results are likely to be recorded. I still think that factors into the feeling Rick Griffin expressed that die rolls on the show end up feeling less fair.
Angelhood did some math in the replies here that might be a bit tough to navigate on bluesky so here's a link to the end of the thread, but yeah: chicanery is afoot. bsky.app/profile/ange...
Beyond a shadow of a doubt. But it becomes clearer that what they're likely doing is skewing some of their low rolls to nat1 for a more cinematic effect, and their high rolls to nat20 for the same. I'm positive the peak in 10s is because of Reliable Talent on a rogue or a similar class feature.
November 16, 2025 at 1:30 PM
Angelhood did some math in the replies here that might be a bit tough to navigate on bluesky so here's a link to the end of the thread, but yeah: chicanery is afoot. bsky.app/profile/ange...
and ~908 nat2s. Even just eyeballing it, I'm confident the p-values are close to 0. We have, for 1s: 902(expected) and 935(result). for 2s: 908(e) and 648(r). for 20s: 1011(e) and 1255(r) without doing the calculations, I'm guessing the p-value would be again around the order of 10^-20.
November 16, 2025 at 1:09 PM
and ~908 nat2s. Even just eyeballing it, I'm confident the p-values are close to 0. We have, for 1s: 902(expected) and 935(result). for 2s: 908(e) and 648(r). for 20s: 1011(e) and 1255(r) without doing the calculations, I'm guessing the p-value would be again around the order of 10^-20.
I thought they replaced Taking 10 with Passive Perception in 5e. Either way though, this is also filtered for rolls specifically, so a d20 was rolled and produced a 10.
November 16, 2025 at 12:13 PM
I thought they replaced Taking 10 with Passive Perception in 5e. Either way though, this is also filtered for rolls specifically, so a d20 was rolled and produced a 10.
Is there some narrative reason it would be good for the podcast to roll a 10 on a perception check? Like because it's right in the middle, it creates tension for the audience because it could go either way: Did the player not see the thing or is there really nothing there?
November 16, 2025 at 11:54 AM
Is there some narrative reason it would be good for the podcast to roll a 10 on a perception check? Like because it's right in the middle, it creates tension for the audience because it could go either way: Did the player not see the thing or is there really nothing there?
You can also filter by the type of roll it is, which reveals some things: like there's this massive outlier for the number of times players roll 10 on perception, which definitely contributes to that peak.
November 16, 2025 at 11:45 AM
You can also filter by the type of roll it is, which reveals some things: like there's this massive outlier for the number of times players roll 10 on perception, which definitely contributes to that peak.
I thought I linked this earlier: lookerstudio.google.com/u/0/reportin...
This page has all the statistics I'm using, you can filter to only have the player's d20 rolls on the dashboard, but I'm sure if you imported the data into excel you could get a count for if the null rolls account for theweird
This page has all the statistics I'm using, you can filter to only have the player's d20 rolls on the dashboard, but I'm sure if you imported the data into excel you could get a count for if the null rolls account for theweird
Main Campaign All Rolls
Looker Studio turns your data into informative dashboards and reports that are easy to read, easy to share, and fully customizable.
lookerstudio.google.com
November 16, 2025 at 11:38 AM
I thought I linked this earlier: lookerstudio.google.com/u/0/reportin...
This page has all the statistics I'm using, you can filter to only have the player's d20 rolls on the dashboard, but I'm sure if you imported the data into excel you could get a count for if the null rolls account for theweird
This page has all the statistics I'm using, you can filter to only have the player's d20 rolls on the dashboard, but I'm sure if you imported the data into excel you could get a count for if the null rolls account for theweird