iiinova — where your digital self keeps evolving

Zeb @zebm.bsky.social · 07/10/2026

27 stochastic games playable with cards, dice and boards. Some of my own invention. stochanswers.com/education/

stochanswers.com

Education

Educational examples.

020

Zeb @zebm.bsky.social · 04/10/2026

And then Captain FitzRoy left to set up the Met Office.

010

Zeb @zebm.bsky.social · 04/10/2026

If he ever said it's growing exponentially then he failed both basic Maths and basic literature review. Would that make him more qualified for Congress?

000

Zeb @zebm.bsky.social · 04/10/2026

Neither do the modelling community. e.g. The staple of Biology textbooks that is the Lotka-Volterra model of predator-prey uses the Malthus model of exponential population growth which was debunked by Darwin in his seminal work. It also seems that every epidemic modeller has their own model...

020

Zeb @zebm.bsky.social · 04/10/2026

Ah, I recognise the big ball of mud architecture pattern in code I've had AI generate for myself.

010

Zeb @zebm.bsky.social · 03/10/2026

#EpiSky #EduSky #MathSky #StatsSky In my rush to derive the standard model using basic Maths I initially made the mistake of not subtracting one from the denominator and thus fixing an inconsistency with the definition of the basic reproduction number, zenodo.org/records/2308...

010

Zeb @zebm.bsky.social · 29/09/2026

#MathSky #StatsSky #EduSky The Galton–Watson process as a game stochanswers.com/education/ex...

stochanswers.com

Extinction

The card and die game

000

Zeb @zebm.bsky.social · 18/09/2026

I'd be interested to see what an AI code review makes of their software.

010

Zeb @zebm.bsky.social · 18/09/2026

My view is that ABMs should be able to reproduce classical ODE models where that model has been validated against real world data. Thus the ABM is validated to an extent. That is the case here.

000

Zeb @zebm.bsky.social · 18/09/2026

The Standard SIR model is said to hold for a well-mixed population. In the base ABM model each agent can tag each other with equal probability, so a complete graph. How far you can go about removing links and still get the "same" average behaviour I never looked into.

100

Zeb @zebm.bsky.social · 15/09/2026

#EpiSky #MathSky #StatsSky #EduSky A trivial ABM and basic probability theory used to derive the standard SIR model of epidemiology. Now to find a journal that doesn't mind debunking. github.com/stochanswers...

github.com

papers/minimal.pdf at main · stochanswers/papers

Contribute to stochanswers/papers development by creating an account on GitHub.

000

Zeb @zebm.bsky.social · 02/07/2026

#MathSky #EduSky #EpiSky An online card game / stochastic simulation of an epidemic stochanswers.com/education/ep...

Zeb @zebm.bsky.social · 30/06/2026

Just finished some more JavaScript. The same algorithm but as a very boring (educational) card game. stochanswers.com/education/ep...

stochanswers.com

Epidemic

The card game

010

Zeb @zebm.bsky.social · 28/06/2026

#MathSky #EpiSky #EduSky I've just implemented a JavaScript version of my epidemic simulation game (which averages to the standard SIR model) so that people don't need to run around. stochanswers.com/education/ta...

stochanswers.com

Epidemiological tag

Maths in the playground

010

Zeb @zebm.bsky.social · 28/06/2026

Yes you are correct. To get exponential growth you have to actively ignore everyone but the never infected and hunt them down without mercy. How a virus would know this is never explained. The game, at scale, when averaged matches the standard SIR model which has logistic growth.

200

Zeb @zebm.bsky.social · 26/06/2026

I've just added a JavaScript game to this page on how epidemics grow logistically resulting in 'herd immunity'. Much less exhausting than running around. stochanswers.com/education/ta...

stochanswers.com

Epidemiological tag

Maths in the playground

110

Zeb @zebm.bsky.social · 23/06/2026

Typo. Should be number of Y type items times the number of not Y type items

100

Zeb @zebm.bsky.social · 23/06/2026

The logistic function is y = (1 +e^(-x))^(-1). It arises from the rate equation dy/dt = y(1-y) i.e. the rate of change is proportional to number of Y type items numbers the number of not Y type items. Which comes from 'selection without replacement'. e.g. An infected picks a susceptible to infect.

100

Zeb @zebm.bsky.social · 23/06/2026

#MathSky #EduSky A JavaScript implementation of the Avalanche board game by Per Bak. stochanswers.com/education/av...

stochanswers.com

Avalanche

The board game

000

Zeb @zebm.bsky.social · 16/06/2026

It's Maths not biology. What happens and when to cause a finite variable to leave a curve to infinity? Applying Occam's razor - the simplest solution is that it was never growing exponentially in the first place and yet all these impostors get published when combinatorics gives logistic growth.

100

Zeb @zebm.bsky.social · 11/06/2026

It could just be marketing. Years ago Jaguar developed a great 2 valves per cylinder engine but the market had moved to 4 valves per cylinder so they axed the engine in favour of an inferior performing one.

000

Zeb @zebm.bsky.social · 11/06/2026

It does seem to be rather a scattergun approach. I would support an end to welfarism such as string theory and the retirement of spaghetti Fortran code such as the climate model NASA GISS ModelE. There are lots of cuts that could be easily justified.

010

Zeb @zebm.bsky.social · 11/06/2026

I suggest that you read some epidemic modelling papers. Quite a lot that get published after peer review claim that epidemics grow exponentially as though the authors know better than Farr (1840), Darwin (1859) and Kermack-McKendrick (1927).

100

Zeb @zebm.bsky.social · 11/06/2026

So a forecasting error? You don't give the number of standard deviations that allow people to make that judgement call. There are lies, damned lies and statistics.

010

Zeb @zebm.bsky.social · 11/06/2026

I noticed lots of basic Maths and literature review errors during Covid. i.e. All the papers that had passed peer review whilst claiming that epidemics grow exponentially. This was implicitly prebunked by Farr (1840) and explicitly by Darwin (1859) when he used Reductio ad Absurdum on Malthus (1798)

000

Zeb @zebm.bsky.social · 03/06/2026

#MathSky #EduSky With the school holidays approaching it is important for children's development that they experience sheer and utter boredom. This game of traffic congestion may help. stochanswers.com/education/tr...

stochanswers.com

Traffic congestion

The bored game

010

Zeb @zebm.bsky.social · 15/05/2026

Alternatively you could just transfer unused votes to a neighbouring constituency. github.com/stochanswers...

github.com

010

Zeb @zebm.bsky.social · 15/05/2026

I had a look at your website and found basic Maths errors so I've emailed you.

000

Zeb @zebm.bsky.social · 10/05/2026

#EpiSky 1798 - Malthus - populations grow exponentially 1840 - Farr - data shows non-exponential for epidemics 1859 - Darwin - disproves Malthus using basic Maths 1927 - Kermack-McKendrick - logistic growth 2020 - Why the failure in basic education of public health officials, etc?

000

Zeb @zebm.bsky.social · 31/03/2026

Modelling on two edges of the Lumley triangle is much easier than general anisotropy. The traditional Reynolds stress models fail a constraint equation there. #MathSky #CFD stochanswers.com/studies/axis...

stochanswers.com

Modelling of axisymmetric turbulence

Where RSMs fail

010

Zeb @zebm.bsky.social · 07/12/2025

Some of us aren't so sad as to allow alerts from social media to distract us from life.

000

Zeb @zebm.bsky.social · 07/12/2025

You must be anti-Newton then because Einstein was less wrong. Flat earth can be an appropriate level of science though, obviously, not as accurate as QED which itself is wrong.

100

Zeb @zebm.bsky.social · 05/12/2025

Nice to read about an actual epicentre for a change. I got bored hearing about that secret lab under the market in Wuhan.

000

Zeb @zebm.bsky.social · 05/12/2025

I always ask LLMs how do epidemics grow exponentially and they always respond with a dumb answer just like most people.

000

Zeb @zebm.bsky.social · 05/12/2025

Why relativity? And why turbulence?

010

Zeb @zebm.bsky.social · 05/12/2025

The ballistic equation. Rayleigh-Benard convection. Stress analysis of giant structures sagging under their own weight. All analysed using a uniform vertical gravity field implying that the Earth is flat.

200

Zeb @zebm.bsky.social · 05/12/2025

So terse a reply that it is meaningless

100

Zeb @zebm.bsky.social · 05/12/2025

Pseudo science. You mean like epidemics grow exponentially?

100

Zeb @zebm.bsky.social · 14/11/2025

Only the second company to land an orbital booster. No national space agency has achieved that feat.

010

Zeb @zebm.bsky.social · 14/11/2025

I only knew the term from old radios

010

Zeb @zebm.bsky.social · 14/11/2025

I've seen research that has failed critically because they assumed that the random variable was normally distributed. Perhaps in tribute to a famous Belgian that was removed from the Index of history? (General knowledge crossword clue)

000

Zeb @zebm.bsky.social · 07/11/2025

Three instances of the same Mathematical model. Tag, cards, NetLogo #EduSky #MathSky #EpiSky stochanswers.com/company/abst...

stochanswers.com

Mathematics is merely abstraction

Three models that are the same

010

Zeb @zebm.bsky.social · 06/11/2025

Oops, should have said exponential growth. Obviously exponential decay is natural.

000

Zeb @zebm.bsky.social · 06/11/2025

I hope it includes the word "exponential" and says that it has nothing to do with epidemiology whilst referencing Farr (1840), Kermack-McKendrick(1927), etc. etc.

100

Zeb @zebm.bsky.social · 06/11/2025

I've seen several stochastic epidemic models that break the laws of Maths. The bridge between the two subjects is probability theory. You can use it to calculate the odds that an epidemic will grow exponentially. stochanswers.com/education/ep...

stochanswers.com

Epidemic

The card game

000

Zeb @zebm.bsky.social · 06/11/2025

This software may help, stochanswers.com/downloads/ep...

stochanswers.com

Minimal SIR model

in NetLogo

000