The reason I use "they think" up there is they have given me so fucking many different IQ tests in my life, and I score in about a 40 point range. That "2 stdev above yours" discards the 6 highest outliers. I *could* claim 3.5 stdev, but I don't. Because IQ is bullshit, especially at the margins.
January 12, 2026 at 11:20 AM
The reason I use "they think" up there is they have given me so fucking many different IQ tests in my life, and I score in about a 40 point range. That "2 stdev above yours" discards the 6 highest outliers. I *could* claim 3.5 stdev, but I don't. Because IQ is bullshit, especially at the margins.
I really don't need a binary win/lose call. Give me an estimated vote margin and stdev
August 12, 2026 at 5:28 AM
I really don't need a binary win/lose call. Give me an estimated vote margin and stdev
We're looking to control the mean opinion in online content. We have absolutely no way of controlling the stdev. You just have to accept that up front
September 8, 2026 at 1:02 AM
We're looking to control the mean opinion in online content. We have absolutely no way of controlling the stdev. You just have to accept that up front
#splcensus #compsplatoon #splatoon
Day 3!
Today on the age distribution of the Splatoon community
90% are 27 or younger
About half are 19 or younger
with minors accounting for 31% of those surveyed
Median: 20
Mean: 20.78
StDev: 5.75
again, again, graphic by the incredible @zapfisk.splstrong.com
Day 3!
Today on the age distribution of the Splatoon community
90% are 27 or younger
About half are 19 or younger
with minors accounting for 31% of those surveyed
Median: 20
Mean: 20.78
StDev: 5.75
again, again, graphic by the incredible @zapfisk.splstrong.com
September 20, 2025 at 7:11 PM
#splcensus #compsplatoon #splatoon
Day 3!
Today on the age distribution of the Splatoon community
90% are 27 or younger
About half are 19 or younger
with minors accounting for 31% of those surveyed
Median: 20
Mean: 20.78
StDev: 5.75
again, again, graphic by the incredible @zapfisk.splstrong.com
Day 3!
Today on the age distribution of the Splatoon community
90% are 27 or younger
About half are 19 or younger
with minors accounting for 31% of those surveyed
Median: 20
Mean: 20.78
StDev: 5.75
again, again, graphic by the incredible @zapfisk.splstrong.com
High five, my fellow "I only care about knowing that I measure over 5 standard deviations on "IQ" tests because it makes people who measure 3 stdev very mad when you explain why IQ is fucking bullshit" friend!
January 12, 2026 at 12:33 PM
High five, my fellow "I only care about knowing that I measure over 5 standard deviations on "IQ" tests because it makes people who measure 3 stdev very mad when you explain why IQ is fucking bullshit" friend!
yeah. if you build a bunch of relatively thin majority districts that are safe against like, 2 stdev events, you get your clock cleaned when a 3stdev happens
September 16, 2026 at 10:50 PM
yeah. if you build a bunch of relatively thin majority districts that are safe against like, 2 stdev events, you get your clock cleaned when a 3stdev happens
My understanding is that the +2 stdev gamblers are a really big part of the profit margin for operators
September 8, 2025 at 11:24 PM
My understanding is that the +2 stdev gamblers are a really big part of the profit margin for operators
--at my elementary school, there were two of us who hit the 5 stdev threshold of *whatever* that test measures, and he and I had absolutely NOTHING in common and barely liked each other but clung to each other anyway, because we were the only people who could understand each other 99% of the time.
December 10, 2025 at 5:43 AM
--at my elementary school, there were two of us who hit the 5 stdev threshold of *whatever* that test measures, and he and I had absolutely NOTHING in common and barely liked each other but clung to each other anyway, because we were the only people who could understand each other 99% of the time.
The entire crypto industry could go away and it would amount to a 1 stdev monthly move in the S&P500 in terms of total value lost. Said different, it doesn't matter from a macro perspective.
bobeunlimited.substack.com/p/crypto-cra...
bobeunlimited.substack.com/p/crypto-cra...
February 7, 2026 at 6:13 PM
The entire crypto industry could go away and it would amount to a 1 stdev monthly move in the S&P500 in terms of total value lost. Said different, it doesn't matter from a macro perspective.
bobeunlimited.substack.com/p/crypto-cra...
bobeunlimited.substack.com/p/crypto-cra...
We also assume, because the correlation never fails, that anyone who fetishizes their frankly mid "not even a full 3 stdev" score is contaminated with Hitler particles, because the only grown-ass adults who throw around their IQ the minute they aren't met with fawning adoration --
January 12, 2026 at 11:25 AM
We also assume, because the correlation never fails, that anyone who fetishizes their frankly mid "not even a full 3 stdev" score is contaminated with Hitler particles, because the only grown-ass adults who throw around their IQ the minute they aren't met with fawning adoration --
A new map I added [left] compares the ensemble 500-hPa height standard deviation to ERA5 climatological 500-hPa height standard deviation for each location and time of year.
Where these values are >=1 is where ensemble stdev is greater than climatological stdev.
Where these values are >=1 is where ensemble stdev is greater than climatological stdev.
January 26, 2025 at 3:37 PM
A new map I added [left] compares the ensemble 500-hPa height standard deviation to ERA5 climatological 500-hPa height standard deviation for each location and time of year.
Where these values are >=1 is where ensemble stdev is greater than climatological stdev.
Where these values are >=1 is where ensemble stdev is greater than climatological stdev.
I do not give a literal fuck what my "IQ" is (or any of the sub-scores they've taken to reporting instead) except to throw it in the faces of people who think their just-barely-3 stdev score is special! Well, that and because I will die mad they never identified the ADHD
January 12, 2026 at 12:30 PM
I do not give a literal fuck what my "IQ" is (or any of the sub-scores they've taken to reporting instead) except to throw it in the faces of people who think their just-barely-3 stdev score is special! Well, that and because I will die mad they never identified the ADHD
Day 25: The #MathArtMarch prompt is "chaos". Add a bit of chaos into a Sierpinski triangle. Rather than creating the triangular holes via the midpoints of the sides of the triangles, chose points normally distributed about the midpoint with stdev=5% length of the line. Inspired by Matt Enlow.
March 25, 2025 at 12:25 PM
Day 25: The #MathArtMarch prompt is "chaos". Add a bit of chaos into a Sierpinski triangle. Rather than creating the triangular holes via the midpoints of the sides of the triangles, chose points normally distributed about the midpoint with stdev=5% length of the line. Inspired by Matt Enlow.
wonder how many stdev more photos this pigeon has had taken of it than the average pigeon
April 26, 2025 at 1:33 PM
wonder how many stdev more photos this pigeon has had taken of it than the average pigeon
Apple’s having a 2 stdev day on the price increases
June 25, 2026 at 1:51 PM
Apple’s having a 2 stdev day on the price increases
When you go in you tell them the mean and stdev amount of alcohol you want per drink and every drink you get is determined by a normal distribution that takes the mean and stdev you requested and multiplies them by two thirds
January 30, 2026 at 12:04 AM
When you go in you tell them the mean and stdev amount of alcohol you want per drink and every drink you get is determined by a normal distribution that takes the mean and stdev you requested and multiplies them by two thirds
‘OpenAI projected revenue of $145 billion in 2029. The data reveal that no public company has grown this fast for five years in the last three-quarters of a century. The forecast implies a roughly 9.5 stdev outcome which is extraordinarily unlikely.’ www.morganstanley.com/im/en-us/ind...
February 25, 2026 at 5:18 PM
‘OpenAI projected revenue of $145 billion in 2029. The data reveal that no public company has grown this fast for five years in the last three-quarters of a century. The forecast implies a roughly 9.5 stdev outcome which is extraordinarily unlikely.’ www.morganstanley.com/im/en-us/ind...
My @iclr-conf.bsky.social batch of 13 papers has the following stats:
Min avg rating: 2.5
Max avg rating: 5.5
Avg avg rating: 3.87
Stdev avg rating: 0.96
4 = marginally below the acc thresh
6 = marginally above the acc thresh
From my experience the ratings will rise during the rebuttal.
Min avg rating: 2.5
Max avg rating: 5.5
Avg avg rating: 3.87
Stdev avg rating: 0.96
4 = marginally below the acc thresh
6 = marginally above the acc thresh
From my experience the ratings will rise during the rebuttal.
November 12, 2025 at 7:17 AM
My @iclr-conf.bsky.social batch of 13 papers has the following stats:
Min avg rating: 2.5
Max avg rating: 5.5
Avg avg rating: 3.87
Stdev avg rating: 0.96
4 = marginally below the acc thresh
6 = marginally above the acc thresh
From my experience the ratings will rise during the rebuttal.
Min avg rating: 2.5
Max avg rating: 5.5
Avg avg rating: 3.87
Stdev avg rating: 0.96
4 = marginally below the acc thresh
6 = marginally above the acc thresh
From my experience the ratings will rise during the rebuttal.
🦠🧠🟦 Severe psychomotor slowing in #LongCovid:
*Cognitive slowing was evident even on a 30-sec task measuring simple reaction time.
*Patients with post-COVID-19 conditions responded to stimuli ~3 stdev slower than healthy control (!!!)
(n=257, huge effect size replicated in a 2nd cohort!)
*Cognitive slowing was evident even on a 30-sec task measuring simple reaction time.
*Patients with post-COVID-19 conditions responded to stimuli ~3 stdev slower than healthy control (!!!)
(n=257, huge effect size replicated in a 2nd cohort!)
"These results robustly demonstrate pronounced cognitive slowing in people with PCC [#LongCovid] which distinguishes them from age-matched healthy individuals who previously had symptomatic COVID-19 but did not manifest Long Covid." medrxiv.org/content/10.1...
December 6, 2023 at 3:10 PM
🦠🧠🟦 Severe psychomotor slowing in #LongCovid:
*Cognitive slowing was evident even on a 30-sec task measuring simple reaction time.
*Patients with post-COVID-19 conditions responded to stimuli ~3 stdev slower than healthy control (!!!)
(n=257, huge effect size replicated in a 2nd cohort!)
*Cognitive slowing was evident even on a 30-sec task measuring simple reaction time.
*Patients with post-COVID-19 conditions responded to stimuli ~3 stdev slower than healthy control (!!!)
(n=257, huge effect size replicated in a 2nd cohort!)
Sobre o uso de linkedlists vs arraylists...
LinkedList lutou com 10k bravamente
Dominou com 1M (2x)
Mas perdeu de LAVADA com 5M
LinkedList lutou com 10k bravamente
Dominou com 1M (2x)
Mas perdeu de LAVADA com 5M
![Result "jeffque.bench.StreamList.oldForLinkedList_List": 10k
26806.911 ±(99.9%) 1058.370 ops/s [Average]
(min, avg, max) = (21374.039, 26806.911, 29549.493), stdev = 2137.959
CI (99.9%): [25748.542, 27865.281] (assumes normal distribution)
Result "jeffque.bench.StreamList.oldForArrayList_var": 10k
25912.637 ±(99.9%) 760.091 ops/s [Average]
(min, avg, max) = (20165.753, 25912.637, 28871.771), stdev = 1535.422
CI (99.9%): [25152.545, 26672.728] (assumes normal distribution)
Result "jeffque.bench.StreamList.oldForArrayList_List": 10k
26951.109 ±(99.9%) 778.651 ops/s [Average]
(min, avg, max) = (19320.944, 26951.109, 29862.408), stdev = 1572.915
CI (99.9%): [26172.457, 27729.760] (assumes normal distribution)](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:22ytg5gvocx2igjmauxxmt3j/bafkreic2tmyzbuetku3iwwzl4q5eoglaxxfztutfxy5gzxth5qt5tg5qae)
![Result "jeffque.bench.StreamList.oldForLinkedList_List": 1M
105.989 ±(99.9%) 9.123 ops/s [Average]
(min, avg, max) = (52.615, 105.989, 125.810), stdev = 18.429
CI (99.9%): [96.866, 115.112] (assumes normal distribution)
Result "jeffque.bench.StreamList.oldForArrayList_List": 1M
62.481 ±(99.9%) 2.246 ops/s [Average]
(min, avg, max) = (39.800, 62.481, 66.480), stdev = 4.536
CI (99.9%): [60.236, 64.727] (assumes normal distribution)
Result "jeffque.bench.StreamList.oldForArrayList_var": 1M
61.374 ±(99.9%) 2.047 ops/s [Average]
(min, avg, max) = (39.230, 61.374, 66.836), stdev = 4.136
CI (99.9%): [59.327, 63.422] (assumes normal distribution)](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:22ytg5gvocx2igjmauxxmt3j/bafkreigkpxrc2umuqg76inz7eletkd66jdlqom36trymleveuyuugxg46e)
![Result "jeffque.bench.StreamList.oldForLinkedList_List": 5M
2.741 ±(99.9%) 0.243 ops/s [Average]
(min, avg, max) = (1.925, 2.741, 4.213), stdev = 0.491
CI (99.9%): [2.498, 2.984] (assumes normal distribution)
Result "jeffque.bench.StreamList.oldForArrayList_List": 5M
10.767 ±(99.9%) 0.353 ops/s [Average]
(min, avg, max) = (8.220, 10.767, 11.613), stdev = 0.712
CI (99.9%): [10.414, 11.119] (assumes normal distribution)
Result "jeffque.bench.StreamList.oldForArrayList_var": 5M
10.187 ±(99.9%) 0.596 ops/s [Average]
(min, avg, max) = (6.093, 10.187, 11.590), stdev = 1.203
CI (99.9%): [9.591, 10.782] (assumes normal distribution)](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:22ytg5gvocx2igjmauxxmt3j/bafkreiahmqwlyygfigusidnqrtbiluzk3kkcwq4t3hw72muoeu4oc3rneq)
August 21, 2025 at 7:31 PM
Sobre o uso de linkedlists vs arraylists...
LinkedList lutou com 10k bravamente
Dominou com 1M (2x)
Mas perdeu de LAVADA com 5M
LinkedList lutou com 10k bravamente
Dominou com 1M (2x)
Mas perdeu de LAVADA com 5M
‘FMS investors are net 26% underweight energy. Current reading is 1.8 stdev below its long-term average.’ www.dailychartbook.com/p/703 Via @dailychartbook.bsky.social
June 18, 2025 at 2:12 PM
‘FMS investors are net 26% underweight energy. Current reading is 1.8 stdev below its long-term average.’ www.dailychartbook.com/p/703 Via @dailychartbook.bsky.social
hahahahhahahahahah
First one is the original app in Elixir
second one is the same in Rust
yeah no
First one is the original app in Elixir
second one is the same in Rust
yeah no
August 4, 2025 at 9:43 AM
hahahahhahahahahah
First one is the original app in Elixir
second one is the same in Rust
yeah no
First one is the original app in Elixir
second one is the same in Rust
yeah no
it's interesting that the NLRB's election process seems be continuing unfucked.
Distribution of days from filing to first election
For RC cases filed since January 1, 2010. Negative durations or durations greater than 365 are truncated. | year | median | mean | stdev |
|-|-|-|-|
|overall | | | |
| 2010 | | | |
| 2019 | | | |
|...
observablehq.com
July 7, 2025 at 7:23 PM
it's interesting that the NLRB's election process seems be continuing unfucked.
Taking your poster stats and building rpg character around it
Engagement stats (incoming likes rp comment) the overall level
Average Posts per day STR
Stdev posts per day DEX
Portion of days posting CON
Portion of posts with images INT
Portion of posts with quotes WIS
Followers/follows RIZ
Engagement stats (incoming likes rp comment) the overall level
Average Posts per day STR
Stdev posts per day DEX
Portion of days posting CON
Portion of posts with images INT
Portion of posts with quotes WIS
Followers/follows RIZ
February 3, 2024 at 10:09 PM
Taking your poster stats and building rpg character around it
Engagement stats (incoming likes rp comment) the overall level
Average Posts per day STR
Stdev posts per day DEX
Portion of days posting CON
Portion of posts with images INT
Portion of posts with quotes WIS
Followers/follows RIZ
Engagement stats (incoming likes rp comment) the overall level
Average Posts per day STR
Stdev posts per day DEX
Portion of days posting CON
Portion of posts with images INT
Portion of posts with quotes WIS
Followers/follows RIZ
Which is why those of us who score over that threshold and understand the limitations of assessments tend to just go with "number of standard deviations". There is no way to accurately distinguish between 190 and 195, but you can say "six stdev" and be in the right-ish neighborhood.
January 12, 2026 at 1:17 PM
Which is why those of us who score over that threshold and understand the limitations of assessments tend to just go with "number of standard deviations". There is no way to accurately distinguish between 190 and 195, but you can say "six stdev" and be in the right-ish neighborhood.