Dublin International Film Festival, CC BY 3.0, via Wikimedia Commons
A bit about me…
Hi… i’m Phil.. I live in Northamptonshire, married with two kids, I work in the construction industry, or at least adjacent. I’m a long time lurker (early days of Guido Fawkes here!), never read the comments here but always read the articles. Sorry it took so long to write an article, I finally decided to do it now my kids sleep at night affording me the time to sleep and finally operate as normal, or at least as I was before kids, in some way this and hopefully the following updates can help feed the machine.
I may be a long time lurker but I am an even longer time geek, nay-naysayer, boyscout, prepper and doom mongerer who , unfortunately keeps being right, mainly about the gloomy stuff. I decided I wanted to be more certain on the doom, so I put this together (with the help of AI to be sure, but may as well as it shaved off hours and hours of research.
I grew up on Asimov’s Foundation Trilogy. For those who haven’t read it, Hari Seldon is fictional. He was a character Isaac Asimov invented, a psychohistorian. His idea was that you could use math’s to predict the future of civilizations, not people, but whole systems. Not what you will have for breakfast, but whether people will have food next year. He used big populations and behavior’s, statistical laws of large numbers, to model the fall and rise of empires. If you haven’t read the book(s) they are fantastic, do it.
It was science fiction. Or it was for me, until 2022.
That’s when I was involved with business risk modeling and went down the wormhole of how long and when it started only to realize the government does it too on a much bigger scale (they just don’t show the population the workings, or in most cases even the question or the answer). Governments use it to drive civil servant policy and direction and planning, in much the same way a business does to forecast, plan and drive policy to make the business better.
Now there’s a big gap between Psychohistory predicting the rise and fall of empires and governmental risk, but the crossover is there, everything boils down to a number in the end. In one night of sleep addled fog i wondered, could you predict the future of the UK in the same way?
I made a basic 10 variable (that has now expanded to 100-variables) resilience model for the UK, now scored as of 6 October 2026. It is not a prediction that the lights go out tomorrow. It is not prepper fantasy. I am not a serious prepper, I have a small 3 bed house in but i have a family to look after and support so knowing the future would come in handy in making sure i look after them all. Knowing in advance of everyone else means I can prep before it’s too late, stock up on food, diesel, petrol, water, whatever it may be.
The model is a series of calculations of where we are as a country on the curve from stability to disorder.
Governments have run these models since the Cold War whole offices of hundreds of people were involved. In the UK it lives in the Cabinet Office as the National Risk Register, in MOD as Defense Intelligence, in the Bank of England as its Financial Stability Model, in NESO as the winter outlook. They never publish the full math. They publish a 40-page glossy summary with 4 colours: green, amber, red, black. Voters get edited colours. Ministers get the 200 variables, breakdowns, scenarios, from both public and confidential information sources. Think information such as GDP, to RPI to family debt to NHS waiting times, etc, all this information is categorized analyzed and graded.
For my model I reversed this process. Instead of starting with classified briefs, I started with public data and built proxies for the classified bits. 60 variables you can check on from places such as the ONS, BoE, BRC, UK Finance. 40 variables you can infer from Commons committees, NAO reports, Royal United Services Institute, and think tanks that get briefed privately then published later. The data may be from a year ago, however year after year builds up a picture you can then predict with near certainty what may be written next year.
My model uses a total of 100 variable, 100 scores from various aspects of the UK. Each variable is scored 0-10. 0 is 1950s UK stability. 10 is total state failure, think fall of Rome.
How the government uses it and why…
The government version is called Dynamic Risk Modelling. It exists for one reason: to know when 5 small 6/10 problems combine into one 9/10 crisis.
Example: On their own, UK gas storage at 2 days is an 8/10 risk for stability (low risk would be 1 month or more like in the 50’s) , the number of UK diesel refineries is down to 3 (from 7 in 2010) this is 8/10 risk, 30-year gilt at 6.07% is 8.5/10 risk for UK finaces, NHS beds at 94% for the time of year is 7/10 risk. Separately, civil servants can manage each risk within their own departments, i.e. import diesel from USA, reduce elective surgeries raise interest rates, put off big spending announcements.
Together however for instance this coming January, they multiply exponentially the risk. Gas for 2 days means no backup if power margin of 4.2GW fails and the interconnectors or gas pipes from Norway are interrupted. Diesel storage of 40 days means supermarkets Just in Time 1 day fresh transport systems have no lorries, if Dover blocks after EU EES checks further risk factors. Gilts at 6.07% means councils can’t borrow at PWLB gilt+1% = 7.5% to fix it, so government spending is blown out the water on debt. NHS 96-97% bed forecast means no ambulances to cover emergencies and social unrest and angst grows, imagine if a Section 34 disorder was ordered in 20 towns, to control diesel queues, NHS queues, uncollected bins due to bankrupt councils as central government has no money.
That multiplication is what the model calculates. That’s why some variables have weighting of up to x3 compared to others. A good positive would be if Real wages grow by +3.5% this year, this gets x3 weighting because if real wages stay positive, it fixes 6 variables at once: council tax arrears £6bn, water arrears £4bn, rental arrears, unsecured debt (worst since 1998), supermarket volume 2%, even trust in government19%. One stabilizer holds the whole floor together.
Without weighting, the UK looks like its on the way to anarchy as the model spits out 6.25/10, this is equivalent to the miner strikes and blackouts of the 70’s. With weighting however the risk drops to 5.2/10. The gap is hope and perception if those things can be quantified, UK society is on the whole still fairly law abiding, they pay their taxes, they work, they spend, they buy, they behave. This model in another country could look very different.
Asimov’s fictional idea, applied to real Britain is however my own doom mongerer prediction. I can ask AI to look at the variables and see what has changed, i can ask the risk for the next week, or the next month or 6. So far the model is proving fairly accurate, over the last 2 years that is. Its currently still not predicting the end of the UK in the next 6 weeks.
Now for the end of the world prophecy… there are now 2 variables which could tip everything to the 70’s outlook or even worse and they are coming up on us fast. One is the Autumn statement, the other is the cost of diesel. If the Autumn statement causes gilts to hit over 6.1% especially 30 year, then combined with inflationary aspects from diesel being over 205p in most places these 2 variables will cause that overall risk value to hit the 6.5/10 mark. This is when society as a whole in the UK becomes so fragile a single event can cause the house of cards to collapse.
Asimov invented Seldon in the Foundation Trilogy as a fictional psychohistorian. He was not real, but his method was based on real statistical mechanics. I borrowed/adjusted/modified/invented one for myself, populations are predictable if they are large enough (think sheeple) and if they don’t know they’re being modelled they cannot act in a way as to noticeably change the prediction.
We are 70 million (officially… i would dispute this but hey that’s for another article). Individual choices cancel out, but flows don’t. There is a mortgage cliff with 700k households moving from 2.17% mortgage deals to 5.03% in the next year, this is a flow of £1.7bn per year out of consumption into banks. That’s +£2400 to £3800 per family. UK Finance says that moves default rate from 3.5% to 5.5% nationally, 15% in subprime. That’s not a prediction that’s math’s about 700,000 households.
Youth unemployment is another example, NEET (not in education employment or training) 1.2m 16-24 year olds, which is 14.5% youth unemployment, added to school persistent absence of 22% (10% for most of the 00’s). Those 1.2m don’t enter the tax base. That’s £11bn headroom of no taxes or economic activity gone at exactly the moment the UK debt is £3.03tn, refinancing £3tn at 6% costs £18bn extra. The government can’t fix it because 30yr gilt at 6.07% blocks and cheaper fiscal options.
I want people to be cautious, not scared, which is why i write this article. Caution is to keep 14 days of non-perishable food, because supermarket JIT is only 3 days ambient then done. Keep a small amount of cash because small business use cash, keep your car above half because diesel refinery capacity is down to 3 after Grangemouth closed, keep spare batteries for torches and radio and few 5litre mineral water bottles.
The model itself – what the numbers mean…
0-2/10: 1950s stability.
3-4/10: Managed decline 2010-2019.
5-6/10: Brittle – society has lost trust, (we are here as they say)
6-7: Localized Severe Disorder (Currently the disorder in France but not contained to just students)
7-8: Society Breakdown. (Zimbabwe with hyper inflation)
8-10: Collapse/Total Anarchy/War
My model gives me a value of 5.2/10. Like any forecast this can be expanded as follows:
12-18% chance severe disorder in the next 6 months,
3-6% localized state failure in next 6 months…with a chance of rain
I built this because i think maths can predict the future (to a degree), but the value of it is only realized if the sum is shared. I hope you all enjoy this and I will be writing an update to this after big events if SB is willing. Im curious to see how well this works and happy to share this with you all!
© Phillip of meekle 2026