Sign in

Tristan Snowsill

@tmsnowsill.bsky.social
504 followers 790 following 192 posts

Health economist at the University of Exeter. Also: stats, maths, politics, Star Wars, programming, NFL New York Jets, Formula 1, pixel art...

PostsRepliesMedia
Tristan Snowsill @tmsnowsill.bsky.social · 09/09/2026
I am probably going to melt my computer. It's been running at >90% CPU core usage without a break for 4 hours, with another 4+ hours still to go. @mc-stan.org running with 10/12 threads...
030
Tristan Snowsill @tmsnowsill.bsky.social · 29/08/2026
"Yes" "No" "That would be an ecumenical matter"
010
Tristan Snowsill @tmsnowsill.bsky.social · 27/08/2026
Great thread about the University of Exeter redundancies #DefendExeterUni
011
Reposted by Tristan Snowsill
Tim Bale @timbale.bsky.social · 24/08/2026
"Government has got to stop wasting money on extraordinarily high-margin generic intermediaries to buy in teaching it could commission directly from the starving universities that already provide it."
ft.com
Letter: Don’t pay consultants if universities can do the job
From Amanda Goodall, London EC1, UK and Andrew Oswald, Coventry, Warwickshire, UK
05112
Tristan Snowsill @tmsnowsill.bsky.social · 24/08/2026
Cool article, have now added Monsters (2010) to my watchlist!
011
Tristan Snowsill @tmsnowsill.bsky.social · 11/08/2026
The EV charging flap (door) on our VW was very temperamental - couldn't close it when it was open, couldn't open it when it was closed. Luckily it was suitable for DIY repair, so after buying the spare part from a local dealer/service shop for under £40 and following a YouTube video, it's fixed!
010
Reposted by Tristan Snowsill
Rachel Cunliffe @rmcunliffe.bsky.social · 10/08/2026
Fun (not fun) fact: when the coalition government was changing the student loan system, Vince Cable said the aim of the Plan 2 changes would result in the state funding 40 per cent of a degree, with 60 per cent coming from the individual, which he thought was a fair balance...
23310
Reposted by Tristan Snowsill
Peter Geoghegan @petergeoghegan.bsky.social · 03/08/2026
Behind the bewildering transformation of British universities, are management consultants being paid millions Really proud to have this vital piece on our unis are being gutted from the inside by @jheathershaw.bsky.social on Democracy for Sale. Do read democracyforsale.substack.com/p/inside-bri...
democracyforsale.substack.com
Inside Britain's university crisis
As teaching staff face job loss, their bosses have netted pay rises and bought advice from a “vampiric” management consultancy, writes Prof. John Heathershaw of the University of Exeter
24555319
Tristan Snowsill @tmsnowsill.bsky.social · 30/07/2026
"Risk differences are so great. It's just such a shame we can NEVER get them from our logistic regressions..."
010
Reposted by Tristan Snowsill
Simon Evans @drsimevans.carbonbrief.org · 30/07/2026
🧵UK media has been duped about net-zero (again) The playbook is identical: 1️⃣ Find a big scary number 2️⃣ Speed-dial non-specialist hacks 3️⃣ Do say “serious economic damage” 4️⃣ Don’t say “we pretended gas is free” 5️⃣ That's it! Many papers still fell for it 🤷
14401142
Reposted by Tristan Snowsill
Gavin Buckingham @drgbuckingham.bsky.social · 21/07/2026
The Cognition Group in the Psychology dept at Exeter are at risk of redundancy because "basic cognitive psychology research does not align with the current priorities of funding bodies" If you think this is overtly wrong and misguided, please sign this open letter docs.google.com/forms/d/e/1F...
docs.google.com
Open Letter in Support of the Cognition Group, University of Exeter
To Executive members of the University of Exeter, including Vice Chancellor Prof. Lisa Roberts, Senior Vice-President and Provost Prof. Dan Charman, Pro Vice Chancellor and Executive Dean Prof. Sallie...
128591
Tristan Snowsill @tmsnowsill.bsky.social · 06/07/2026
This surely demands a joint response from @ucu.org.uk @universities-uk.bsky.social @ucea.bsky.social and NUS Asking students to pay more to prop up universities is one thing. Asking them to pay more to prop up the Department for Education and Office for Students is something else entirely.
010
Tristan Snowsill @tmsnowsill.bsky.social · 06/07/2026
Ok this is just ridiculous. How can you see a sector apparently on the brink of collapse and think the right thing to do is undermine them further? Sue Gray had universities collapsing on her shit list in 2024. The government seems to think that was a goal rather than something to avoid.
010
Tristan Snowsill @tmsnowsill.bsky.social · 03/07/2026
I think an underpriced scenario is that Trump fights for and wins the GOP nomination to go for an unconditional third term, he is then hammered in the election, and spends his final days escaping accountability by claiming his advanced dementia makes him unfit to stand trial or give evidence.
000
Tristan Snowsill @tmsnowsill.bsky.social · 01/07/2026
I just love the idea that confounders are floating around there, readily measured, and it's really just a birthday paradox problem or something.
090
Reposted by Tristan Snowsill
Exeter UCU @exeterucu.bsky.social · 01/07/2026
We just had a meeting with over 700 (!!) of our members and OVERWHELMINGLY voted in favour of a motion to calling for a strike ballot! 🩷 We're doing this together. We are the University! #DefendExeterUni.
EXETER UCU VOTES IN FAVOUR OF MOTION CALLING FOR STRIKE BALLOT
04629
Tristan Snowsill @tmsnowsill.bsky.social · 30/06/2026
Wish I had more time to engage with your excellent Bluesky posts? Then get the University of Exeter to withdraw its proposals to cut 200 staff by signing this petition! #DefendExeterUni c.org/ySvZyTfKcf
c.org
Sign the Petition
Stop Redundancies at the University of Exeter
000
Reposted by Tristan Snowsill
Francesca Stavrakopoulou @proffrancesca.bsky.social · 24/06/2026
I’d love it if you’d sign this. It might not prevent me and my colleagues losing our jobs, but it will draw attention to the ideological vandalism being perpetrated by the wealthy execs running our university. Stop Redundancies at the University of Exeter - sign the petition below! c.org/NpvY4sR9gH
c.org
Sign the Petition
Stop Redundancies at the University of Exeter
12154129
Tristan Snowsill @tmsnowsill.bsky.social · 24/06/2026
So my employer, the Uni of Exeter, is proposing to cut up to 200 jobs (153 full time equivalent). The axe is mostly falling on our highly esteemed Humanities, Arts and Social Sciences... and we can't even get a clear answer why? Please sign this petition to get them to rethink! c.org/pfjK7HLjkH
c.org
Help Anonymous Academic grow this petition! 🚀
Stop Redundancies at the University of Exeter
020
Reposted by Tristan Snowsill
Dr Francis Young @drfrancisyoung.bsky.social · 24/06/2026
If you asked me, off the top of my head, to name a non-Oxbridge university in England synonymous with excellence in the Humanities - I would probably think of Exeter first. This feels like a very deliberate, targeted and ideological Bonfire of the Humanities
11367131
Tristan Snowsill @tmsnowsill.bsky.social · 24/06/2026
Some serious birther shit here from Tesco. Trump would be pleased.
Two oven pizza boxes from Tesco, one labelled "The Hawaiian", the other "The American"
000
Tristan Snowsill @tmsnowsill.bsky.social · 22/06/2026
I'm guessing "I'm a white man"?
010
Reposted by Tristan Snowsill
Phillips OBrien @phillipspobrien.bsky.social · 17/06/2026
Hi All, we now have an official text of the USA-Iran MOU from the US point of view. It is extraordinary. The US is in a far weaker position than before it started this war fewer than four months ago. Here is a free, blow-by-blow description why. open.substack.com/pub/phillips...
open.substack.com
Midweek Update #15: The MOU Is Out
Hi all,
39365171
Tristan Snowsill @tmsnowsill.bsky.social · 12/06/2026
Perfection. No notes.
010
Tristan Snowsill @tmsnowsill.bsky.social · 12/06/2026
"I wonder if a big part of this is class" said every Marxist when confronting any phenomenon ever
010
Tristan Snowsill @tmsnowsill.bsky.social · 12/06/2026
I wonder if a big part of this is class (rather than exactly income) which enjoys a strong association with educational attainment and "mating" between different classes has long been and continues to be limited.
110
Tristan Snowsill @tmsnowsill.bsky.social · 12/06/2026
*Child whining* Why can't I have 0.6??? *Parent, gruffly* We've got 0.6000000000001 at home!
110
Tristan Snowsill @tmsnowsill.bsky.social · 11/06/2026
Given one of the most common complaints from former and current government ministers is that when you pull on the levers of power, nothing happens, I think we should rapidly discount anyone whose professional experience involves orders being followed without question. It's no preparation at all.
010
Tristan Snowsill @tmsnowsill.bsky.social · 11/06/2026
Yes it should be axed - here is what I would replace it with to ensure it is sustainable and doesn't lead to pensioners once again slipping into poverty as they did before the triple lock. bsky.app/profile/tmsn...
000
Tristan Snowsill @tmsnowsill.bsky.social · 10/06/2026
It is absolutely frightening. Lots of money and social media clout is behind a far right vision of this country that would see it ethically cleansed and its politicians and civil servants imprisoned for having carried out lawful duties. They are also dragging the mainstream in their direction.
000
Tristan Snowsill @tmsnowsill.bsky.social · 08/06/2026
Ah yet another HTA model I'm involved in which would be so much quicker to build in R but, for reasons I cannot divulge, must be made in Excel. In fairness, when I have been able to do the models in R, it usually results in more complexity to fill the time rather than increasing leisure...
000
Tristan Snowsill @tmsnowsill.bsky.social · 04/06/2026
Decorate in haste and Repaint at your leisure
000
Tristan Snowsill @tmsnowsill.bsky.social · 04/06/2026
Then the characteristic function for k degrees of freedom follows by simply raising the 2df CF to the power k/2
000
Tristan Snowsill @tmsnowsill.bsky.social · 04/06/2026
I've had a little bit of fun recently with the chi-squared distribution. It is fairly easy to derive the density function for 1df, and then you can use convolution to derive the PDF for 2df (and perhaps establish an induction rule?). From the PDF with 2df, the characteristic function is unlocked!
100
Reposted by Tristan Snowsill
The Empty City @davidallengreen.bsky.social · 03/06/2026
This, on the death of Henry Nowak, is an outstanding piece of journalism by @stephenkb.bsky.social www.ft.com/content/aa94...
ft.com
What Henry Nowak’s murder does — and does not — tell us about policing
Kemi Badenoch says his killing needs to be seen as a ‘Stephen Lawrence moment’ but the failures in both cases differ
12482220
Reposted by Tristan Snowsill
Jess Morley @jessrmorley.bsky.social · 03/06/2026
On Monday the HSJ named me as the "arguably the single most influential voice on ethics and AI policy" on their list of top 50 influential people for NHS tech in 20216 www.hsj.co.uk/news/the-50-... is (a) very kind; (b) most likely untrue. BUT FWIW here are my current top 10 AI messages 🧵:
hsj.co.uk
1137
Tristan Snowsill @tmsnowsill.bsky.social · 02/06/2026
#stats
"Let's see who you really are!" Scooby Doo meme
The chi-squared distribution with two degrees of freedom is revealed to be the exponential distribution with rate 0.5
120
Tristan Snowsill @tmsnowsill.bsky.social · 01/06/2026
Oh wow that looks a great alternative to Aseprite!
110
Tristan Snowsill @tmsnowsill.bsky.social · 31/05/2026
I think for me it was when I learnt LaTeX that I first developed my unhealthy obsession with selecting the correct dash for each situation...
020
Tristan Snowsill @tmsnowsill.bsky.social · 30/05/2026
Thanks for sharing all this esoteric wisdom!
010
Tristan Snowsill @tmsnowsill.bsky.social · 30/05/2026
110
Tristan Snowsill @tmsnowsill.bsky.social · 30/05/2026
As someone who spent a lot of time in other languages before coming to R, this is borderline triggering behaviour for me! How does it do so? Do environments have different hashmaps for closures versus non-closures? When does it make these decisions? NSE is so unpredictable.
210
Tristan Snowsill @tmsnowsill.bsky.social · 30/05/2026
Yeah I prefer .data or data, but so long as noone will think you mean degrees of freedom, df is fine, especially if you won't accept a matrix or list but only a data.frame
010
Tristan Snowsill @tmsnowsill.bsky.social · 29/05/2026
As surprised/impressed as I am that works, I don't imagine anyone would defend it as clear to a reader of the code. You can even do > dt <- seq(0, 10, by=0.1) > sapply(dt, dt, df = 10) I truly expected R to fall over rather than deliver, but it surprised me.
120
Tristan Snowsill @tmsnowsill.bsky.social · 29/05/2026
Dear #Rstats users, please do not name your main data table/frame "dt" or "df". One day, you will thank me when you need to use the probability density function for the Student's t or F-distribution! ("d" or "dat" are safe 😊)
4190
Tristan Snowsill @tmsnowsill.bsky.social · 28/05/2026
Mind you, start putting these rules in and suddenly tactical voting is resurrected...
010
Tristan Snowsill @tmsnowsill.bsky.social · 28/05/2026
I probably wouldn't do exactly this. I think if you win a clear majority (>50%) in a system without tactical voting, you should definitely win. And if you have a very low share of the vote (maybe <5%) you should be excluded.
100
Tristan Snowsill @tmsnowsill.bsky.social · 28/05/2026
Interestingly, this ONE WEIRD TRICK can achieve proportional representation with single member constituencies... Everyone casts a single vote for the person they would like to win, and a vote is drawn at random from all the votes cast. Probability of winning is exactly proportional to your vote.
100
Tristan Snowsill @tmsnowsill.bsky.social · 28/05/2026
As a rule, if an electoral system involves electing a single person per geography, it will not be proportional. It can still be an improvement on first past the post (plurality voting), but for proportional representation you need to have multi member constituencies or "top up" seats.
000
Tristan Snowsill @tmsnowsill.bsky.social · 28/05/2026
Which is why many policymakers stop thinking as soon as they find out that (under assumptions they never bother to verify) the market maximises total welfare (but not in a necessarily good way) (and let's just ignore that a monopolist with perfect price discrimination also maximises welfare)
120