Sign in

Dr Will Ball

@wpball.com
3.3K followers 1.2K following 2.6K posts

Interested in working out how to reduce unfair and avoidable differences in health. Mostly youth mental health. Chancellor's Fellow at Robert Gordon University | Social Epidemiologist | Pro Cycling | Rstats | Causal Inference 🇮🇪🏴󠁧󠁢󠁳󠁣󠁴󠁿🍉

PostsRepliesMedia
Dr Will Ball @wpball.com · 27/09/2026
Best result of this guy's career btw
000
Dr Will Ball @wpball.com · 27/09/2026
I love bike racing. So good!
001
Dr Will Ball @wpball.com · 23/09/2026
'No longer' There was no golden age where paper count was a useful metric.
010
Dr Will Ball @wpball.com · 22/09/2026
That makes sense. We were just wondering about parents of Juniors trying to book last minute leave to travel and watch!
110
Dr Will Ball @wpball.com · 22/09/2026
Listening in from Scotland 🏴󠁧󠁢󠁳󠁣󠁴󠁿 How long in advance do these riders know they've been selected?
110
Dr Will Ball @wpball.com · 21/09/2026
A quick prototype.... or just explaining what the hell this stupid regex string is supposed to mean!
000
Dr Will Ball @wpball.com · 21/09/2026
I am a 'question first' evangelical. Very happy to acknowledge that my chosen set of methods/data may not be optimal though. It's not really feasible to only do analysis once you have the perfect data. Triangulation for the win.
000
Dr Will Ball @wpball.com · 21/09/2026
This is possibly the difference between a RQ and the objectives of a study, where it makes more sense to clearly state the methodological limitations you're working with
110
Dr Will Ball @wpball.com · 21/09/2026
I'm with you on this. Also why I think every causal analysis should really have 2 DAGs. 1) Conceptual DAG outlining the theoretical causal structure you're interested in 2) Analytical DAG showing your operationalisation/measurement of those concepts
120
Dr Will Ball @wpball.com · 21/09/2026
How many hours in total do we think it takes for 21.5k applications to be prepared? Can we think of better ways for scientists to spend their time?
050
Dr Will Ball @wpball.com · 21/09/2026
I am reminded of that bloke on Question Time strenuously arguing that his £80k salary 'wasn't even in the top 50%' when it was comfortably within the top 5% of household incomes at the time
010
Dr Will Ball @wpball.com · 16/09/2026
Does this mean that an apple crumble, or 'top crust only' pie is actually upside down toast?
000
Dr Will Ball @wpball.com · 16/09/2026
Also calzone pizza is a sandwich
110
Dr Will Ball @wpball.com · 16/09/2026
But a taco is a sandwich.
110
Dr Will Ball @wpball.com · 11/09/2026
Why does the UNO-X rider have red numbers on? Messaging from Scotland 🏴󠁧󠁢󠁳󠁣󠁴󠁿
001
Dr Will Ball @wpball.com · 10/09/2026
The only way to win the bike race is to go faster than everyone else. Solid strategy from Wout here.
000
Dr Will Ball @wpball.com · 10/09/2026
The certainty is that things are financially bad, especially for university employees.
030
Dr Will Ball @wpball.com · 04/09/2026
So chonk! Much cute!
120
Dr Will Ball @wpball.com · 31/08/2026
He gets paid to write this stuff huh?
030
Dr Will Ball @wpball.com · 28/08/2026
Also, wasn't there lots of polling under Starmer that showed Reform voters still hated Labour?
110
Dr Will Ball @wpball.com · 28/08/2026
If you only view politics through the lens of strategic game playing, it sounds like a logical assessment. In reality, how can another party implementing the policies you want be a nightmare?!
120
Dr Will Ball @wpball.com · 25/08/2026
9 to 5 is a working class anthem. Well done Dolly
020
Dr Will Ball @wpball.com · 13/08/2026
Six of the ten hottest Tour de France stages on record happened this year. The average stage temperature was 29.4°C, nearly 3.5°C above the previous hottest edition in recent seasons. Link: willpball.substack.com/p/was-2026-t...
Title: 'Was 2026 the hottest Tour de France ever?"
Subtitle: Checking claims against a century of temperature records
By Will BallA bar chart showing that the 2026 Tour de France was the hottest of recent memory.
110
Dr Will Ball @wpball.com · 29/07/2026
🏴󠁧󠁢󠁳󠁣󠁴󠁿
000
Dr Will Ball @wpball.com · 27/07/2026
I imagine top speeds where the danger is highest, e.g. flying down the side of a mountain or in a sprint finish, haven't actually changed all that much in recent years. The speed/safety conversation is misguided if you only look at average speeds for stages or the whole race.
100
Dr Will Ball @wpball.com · 27/07/2026
Yes, average speeds have gone up but that's not purely down to physiology. If it was possible to look at the distribution of speeds during the stages that would likely tell the true story - more time spent at higher speeds because the racing starts earlier & there are fewer soft breakaways.
100
Dr Will Ball @wpball.com · 23/07/2026
It would have to be ridden very passively. I suppose Alpecin will have a dual-threat plan with MvdP trying to get ahead.
020
Dr Will Ball @wpball.com · 23/07/2026
Remember what happened last time? www.independent.co.uk/news/uk/home...
independent.co.uk
Tunnock's teacake dancer could lose job after sick claim
Amy McIntosh's bosses said they were "surprised" when they saw her
020
Dr Will Ball @wpball.com · 23/07/2026
He must do now. Philipsen cannot win the Montmartre Paris stage surely?
130
Dr Will Ball @wpball.com · 23/07/2026
Behind the scenes with the Commonwealth Games opening ceremony planning team. "Fuck it, let's just have a massive Tunnocks tea cake with sparklers"
100
Dr Will Ball @wpball.com · 22/07/2026
Send the reply!
120
Dr Will Ball @wpball.com · 15/07/2026
As much as I enjoy England and their coach/tactics being the reason for their loss, it seems a bit 'Brits at it again' behaviour to ignore that Argentina may have influenced the result somewhat too.
130
Dr Will Ball @wpball.com · 15/07/2026
Surprisingly not on penalties though
000
Dr Will Ball @wpball.com · 15/07/2026
Ha! They'll be insufferable regardless.
020
Dr Will Ball @wpball.com · 15/07/2026
Irish and Scots fans can legitimately support Argentina because of Alexis Mac Allister right?
000
Dr Will Ball @wpball.com · 27/06/2026
Also, perverse that private hotels are also likely to make a profit (from public money) from the situation.
120
Dr Will Ball @wpball.com · 27/06/2026
We collectively normalise a lot of awful stuff don't we? IMHO (equitable) population health should be the guiding measure for a successful society.
120
Dr Will Ball @wpball.com · 25/06/2026
As this is unconsented, identifiable patient data being used for non-clinical purposes, surely they should be receiving a massive fine for this?
010
Dr Will Ball @wpball.com · 25/06/2026
Fitba aye?
000
Dr Will Ball @wpball.com · 24/06/2026
Sorry to hear that Richard. I presume there's going to a response from the UCU branch?
100
Dr Will Ball @wpball.com · 23/06/2026
Meet my new friend.
A hospital glove and apron dispenser that looks like a 3 eyed monster sticking it's tongue out.
020
Dr Will Ball @wpball.com · 21/06/2026
Up the Da!
000
Dr Will Ball @wpball.com · 19/06/2026
Maths in heresy.
static.klipy.com
Monty Python: Burn The Witch Scene
Alt: Monty Python: Burn The Witch Scene
020
Dr Will Ball @wpball.com · 19/06/2026
Perfect weather for midges
000
Dr Will Ball @wpball.com · 19/06/2026
Scottish Summer: wearing shorts in the rain.
100
Dr Will Ball @wpball.com · 17/06/2026
Not All Flags.
010
Dr Will Ball @wpball.com · 16/06/2026
Reducing poverty: one weird trick - capitalists hate it!
0110
Dr Will Ball @wpball.com · 14/06/2026
Does that mean eternal torment is coming up for the USians?
000
Reposted by Dr Will Ball
Lorenzo Fabbri @epilorenzo.bsky.social · 08/06/2026
I've been building the etverse: R packages for causal inference in epi. The idea is triangulation: come at the same question from angles that fail for unrelated reasons (different methods, designs, assumptions) and trust it more when they agree. etverse.github.io #episky #causalsky #RStats
etverse.github.io
etverse — triangulating causal evidence in epidemiology – etverse
173
Dr Will Ball @wpball.com · 09/06/2026
Instead of funnelling yet more cash to tech companies who over promise and under-deliver, can we perhaps trying to reducing social and economic inequalities first?
"Evidence shows that one to one tutoring can accelerate a pupil’s learning by around five months, but access to tutoring is deeply unequal, with children from wealthier families far more likely to benefit. The government is running a tender for industry to co-create AI tutoring tools with teachers, with the goal of bringing these tools to a similar level of quality, so that we can offer, at scale, the kind of personalised one to one support often only available to a privileged few.

By adapting to individual pupils’ needs, the tools could provide extra help when they get stuck and identify where they need more practice to master their lessons, with the potential to help them catch up with their peers.

From Summer term this year, there will be teacher-led co-creation of AI tutoring tools with industry."
040