#Randomisation
LightNews LightNews
leicesterctu.bsky.social
Weekly CTU Trial Update! 📢

Huge milestones this week:
@easyas-trial.bsky.social welcomed its 1,000th UK participant and completed an SIV with Erasmus MC, Netherlands!

Momentum continues for @coloprevent.bsky.social with a new Resveratrol trial randomisation. Onwards! 📈
November 10, 2025 at 8:40 AM
stevensenior.bsky.social
I understand Mendelian randomisation as a form of instrumental variables.

And like all instrumental variables, it's black magic and not to be trusted.
November 6, 2025 at 7:51 PM
elife.bsky.social
New Mendelian randomisation study finds no strong evidence that higher genetically predicted adiposity increases head & neck cancer risk. This suggests obesity itself may not be a major causal driver, after accounting for smoking.
buff.ly/G0cu6i0
November 1, 2025 at 2:59 PM
dasneilpferd.bsky.social
#BabyPferd couldn't decide which game to play, so I modified a simple randomisation code to decide for us. It chose "Tricky Wave", which made her realise she'd like to play Snap. Which basically made me feeling exactly how a field partner deciding they won't randomise my study, after all, does.
A spreadsheet photoshopped form Stata, which applies a random value to a load of rows which denote, either, Tricky Wave or Snap (two kids games we have). The next row assigns a random value to each observation of tricky wave. The next is a count by order of those values. The next is the mean of the order of Tricky Wave and the last, of Snap. The value for Tricky Wave is greater than Snap, meaning we should have played Tricky Wave. But we didn't.
November 2, 2025 at 9:08 PM
qwertdwlrma.bsky.social
Path to Nowhere CN Look-Ahead (October 2025)

01. Optimising In-Battle UI/VFX Obstruction
02. Record Player Optimisation
03. New Sinner Tag Filter
04. Carousel Theme/Sinner/Scene Randomisation Support
October 29, 2025 at 4:16 PM
paulbaldowski.bsky.social
I love @kholbatar.bsky.social art, so wanted to continue the flow of reposting, even though what I create is... I'll say "unskilled" in comparison 😂

I have my moments.

Reskeet with yours.
A green sticky note with a cartoon panel drawing. The upper panel shows a fly in flight and the whole planet Earth with the fly saying, "I can fly! I could go literally anywhere." The second panel has the fly buzzing in a circle around a light fitting. Simple drawing style. A book with a Hieronymous Bosch painting as a cover, titled Ship of Fools, a Sanction Genre Setup. A folded origami bishops hat in red and gold sits on the cover in the middle of the image. On the right, a bookmark featuring an image of a 14th century pope wearing the same hat. An image of an old style milk bottle with a red cap sitting on a stone front door step. Shades of white and grey throughout, with the red top a highlight. A hole with a spray of cracks on the side of the bottle, milk in a puddle on the step. Unrealistic, in the style of old school espionage covers. Written in the middle: "Cold Shoulder. The Last Broadcast." with a reviewer quote "deliciously pathetic and rife with roleplaying possiblities". Bottom right corner: " A Mission Collection for The Agency." A comic collage style image of an enormous foot stepping down in the midst of a simplistic sand coloured wasteland with mountains in the distance. Above, a 19th century style hot air balloon flies at an angle with a rope beneath trailing an unrealistically large grey house cat. The image is printed on a neoprene mat, pinched at the corners to make a tray. Two many-sided dice - one red and one white - sit in the tray. Around the tray edge are the words "Sanctioned Polyhedral Interactive Randomisation Kettle" and the logo for All Rolled Up. The background is a desk topper with a starry image in dark blue streaked with pale blue lightning.
October 27, 2025 at 4:02 PM
eicathomefinn.bsky.social
This is one reason why a number of funders (including but not only the Swiss version of UKRI) are increasingly using partial randomisation: applications that meet the threshold requirement go into a pool and are randomly accepted. Quicker, fairer and upfront about the fact that it's a lottery.
October 24, 2025 at 7:35 AM
mecfsskeptic.bsky.social
3) The authors used Mendelian randomisation but several people pointed out problems with their methodology on PubPeer. They did not correct for multiple comparisons, for example, an approach increases the risk of false positives.
pubpeer.com/publications...
PubPeer - Causal Relationship Between Diet, Lipids, Immune Cells, and...
There are comments on PubPeer for publication: Causal Relationship Between Diet, Lipids, Immune Cells, and Chronic Fatigue Syndrome: A Two‐Mediation Mendelian Randomization Study (2025)
pubpeer.com
October 22, 2025 at 1:35 PM
situ-oxford.bsky.social
Fantastic day at the Oxford Science + Ideas Festival! Teaching randomisation using a bead-moving trial using chopsticks or a spoon. Thanks to for the brilliant finger model for great discussion on surgery vs non-surgery in nailbed repair
October 20, 2025 at 3:43 PM
diabetologiajnl.bsky.social
New guidelines for Mendelian randomisation studies submitted to Diabetologia: guidance for authors and reviewers #MendelianRandomisation #DiabetesResearch link.springer.com/article/10.1...
Raising the bar for publication of Mendelian randomisation studies in Diabetologia - Diabetologia
Diabetologia -
link.springer.com
October 10, 2025 at 7:29 AM
michelleklupton.bsky.social
Our newest preprint is now available, using Mendelian randomisation to investigate the causal relationship between Frailty and Alzheimer's disease. www.medrxiv.org/content/10.1...
Awesome work from our visiting PhD student Jonny Flint #NGDgroup
Domains of Frailty as Early Risk Factors for Alzheimer Dementia - Genetic and Causal Evidence
Background Alzheimer’s dementia (AD) is a leading cause of disability and dependency in older adults. Frailty is a phenotypic risk factor for AD, but its causal role remains unclear - partly due to re...
www.medrxiv.org
October 9, 2025 at 2:41 AM
boardgamefeast.bsky.social
Witchpop from @havlark.bsky.social is a match 3 with floating and flicking physics but what really popped for me was the 2D animations full of character. Furthermore, randomisation is just as much about customizing difficulty modifiers as power ups.
Tutorial screen showing a 2d cartoon witch explaining the mechanics, 5 vegetables with animated faces are overlaid on a soup pot. Screenshot of a large pot with vegetables dropping in the top with a large number of burnt chunks on the bottom. A cartoon with and bunny are on either side
October 1, 2025 at 1:06 AM
dov.bsky.social
Nothing was created, nothing was produced, someone made money from the fact that someone else essentially hit a randomisation button on how much money their family has to spend on food. That cannot be right. Or morally tolerable.
September 26, 2025 at 4:29 PM
jamescummings.bsky.social
Wait.... what?! With them now doing partial randomisation granting trials doesn't that have... ya know, problems?!
eve.gd Martin Paul Eve @eve.gd · Sep 18
Wow, someone at the AHRC has a brass neck for treating their own grantees as "research quality".

Each HEI’s formula score is determined by two variables:

research staff (within arts and humanities), this is a proxy for research capacity

AHRC grant income, *this is a proxy for research quality*
Training and support within the Future Doctoral Provision Programme
AHRC has commissioned an engagement exercise to develop thinking around the future of doctoral provision in the arts and humanities.
www.ukri.org
September 18, 2025 at 5:01 PM
tomstafford.mastodon.online.ap.brid.gy
Reliability, bias and randomisation in peer review: a simulation

https://osf.io/preprints/socarxiv/4gqce_v1

new preprint from UKRI #metascience Unit
For a variety of reasons, including a need to save time and a desire to reduce biases in outcomes, some funders of research have started to use partial randomisation in their funding decision processes. The effect that randomisation interventions have on the reliability of those processes should, it is argued, be a consideration in their use, but this key aspect of their implementation remains under-appreciated. Using a simple specification of a research proposal peer review process, simulations are carried out to explore the ways in which decision reliability, bias, extent of decision randomisation and other factors interact. As might be expected, based on both logic and existing knowledge, randomisation has the potential to reduce bias, but it may also reduce decision reliability as inferred from the F1 score and accuracy of a simulated binary (successful, rejected) decision outcome classification process. Bias is also found, in one sense and qualitatively, to be rather insensitive to partial randomisation as it is typically applied in real- world situations. The simple yet apparently effective specification of the simulation of reviewer scores implemented here may also provide insights into the distribution of merit across research funding proposals, and of assessment of them.
September 11, 2025 at 8:49 AM
jesterofdiamonds.bsky.social
Tetris purist here
Are you doing the randomisation right?

Tetris doesn't give you pieces randomly. It puts the seven pieces in a random order and gives that set. You won't get a second one of a piece unless you have gotten every piece once. Only get a third once you have had everything twice etc.
January 1, 2025 at 2:28 AM
yamiku.bsky.social
'it's only fair that everyone votes for someone. I name you '

jap, totally fair 🤣
randomisation would have been way more fair

#themole
July 13, 2024 at 8:55 AM
hetanshah.bsky.social
We are trialling partial randomisation as a way of giving our smallest grants and finding so far this is giving us a more regionally & ethnically diverse pool of applicants (which then translates into who is awarded the grant)...
November 22, 2024 at 10:14 AM
dankayur.bsky.social
I would argue Snowboard Kids has far better balance than Mario Kart in general and a less infuriating item set. While DKR's lack of item randomisation puts it in a subtley different, less party oriented category.
December 3, 2024 at 6:36 PM
marnamai.bsky.social
I have experimented making the entire building out of a single complex tileset, but I had to drop this due to persistent crashing. The basic mesh is a simple cube, each face is filled in by a tile, that can have conditions and randomisation.
October 1, 2023 at 9:32 PM
freshfleisch.bsky.social
my most mundane controversial opinion is that hiring should include more randomisation. too many resumes? random sample. too may good candidates for interview time? random sample. too many good options after interview? random sample.
September 17, 2024 at 8:52 AM
load-dependent.bsky.social
I think the apparent benefits of moderate drinking has largely been debunked as an artefact of confounding by studies leveraging better causal inference methods like Mendelian randomisation already.
December 21, 2024 at 10:31 PM
hetanshah.bsky.social
Bit of coverage in @timeshighered.bsky.social about the @britishacademy.bsky.social trial of partial randomisation for our small grants scheme
www.timeshighereducation.com/news/grants-...
One of the most important things which I had not anticipated was that it is not just the randomisation changing who is given the grant, but the signalling effect of doing something radical like this has changed the applicant pool,” said Shah. “In other words, people who historically may not have applied to the academy – and thus ruled themselves out – have now started applying.” Anecdotal evidence also suggested early career academics preferred the lottery system to the old system, in which proposals were ranked on excellence, because those who met the quality threshold – about 50 per cent did so – but missed out on a grant felt more comfortable reporting back to their departments, he added. “They could go back to their universities and say, ‘I passed the quality test but missed out in the lottery’,” explained Shah on how the lottery system helped to reframe what might otherwise be seen as a failure.
January 23, 2025 at 12:27 PM
eterm.bsky.social
Here's a fiddle to demonstrate: dotnetfiddle.net/zJMQ3D

Both "runs" have consistent randomisation because the same seed(s) are used.
C# Online Compiler | .NET Fiddle
Test your C# code online with .NET Fiddle code editor.
dotnetfiddle.net
March 29, 2025 at 9:17 AM
eicathomefinn.bsky.social
Quite. Several national funding bodies/research charities implementing/piloting partial randomisation, including in Austria, Denmark, NZ & Switzerland. We used it at the RHS during Covid pandemic for the Hardship Fund.
January 23, 2025 at 11:56 AM