James Newport

Director @ Swift Centre. MSc Decision Science. Former HMT policy advisor. My life makes little sense.

[James Newport @NwprtNarrative]

Nothing highlights the UK’s demographic problem better than every car park having 4-5x more disabled bays than child and parent bays…

[James Newport @NwprtNarrative]

Another one bites the dust.

At the @swift_centre, we are always looking for base rates and reference classes to inform our scenario analysis, foresight, and forecasting work.

As we look to welcome our 5th PM in 4 years, I thought I should check out the base rates of PM terms. Below are three infographics which illustrate the risks of the highest office in the land, based on PM stats from 1945 until today:

🟰 If you are party leader and your party wins a General Election, you have a 52% chance of seeing the electorate again, and a 48% chance the men and women in grey suits get rid of you.

✅ If you inherit the job of PM following someone else's demise, you have a 91% chance of making it to the next General Election.

‼️ But here's the kicker: based on the last 10 years in isolation, you have a 83% chance that your party chucks you to the curb before the democratic process decides your fate.

[James Newport @NwprtNarrative]

Article: https://t.co/xGEr2Fdgdn

[James Newport @NwprtNarrative]

Lots of talk about the under 16 social media ban the last week.

My problem is this is another example of treating the symptom and not the cause of a wider problem.

In the below article, @GeorgeHarrison makes a better argument about the need for accountability than I’ve seen many politicians make. Details are important but the core philosophy of this push seems positive. Keen to see where it goes.

[James Newport @NwprtNarrative]

I’m keen for views on the recalling of Fable.

I’ve been out today with my boys so haven’t looked into it much. What’s people’s read?

It naturally highlights the obvious issue the UK/non-US countries have when it comes to dependency on a non-sovereign AI.

But more fundamentally, is it a political stunt (and if so why?), an actual real vulnerability which warrants such a serious intervention, or something else?

[James Newport @NwprtNarrative]

Most funds don’t beat the market

The @swift_centre on the other hand has destroyed #polymarket yet again

This isn’t surprising - we use scientifically based methods that consistently outcompete others

We all try to predict the future. The Swift Centre helps you do it better.

[Michael Story ⚓ @MWStory]

Here are @swift_centre's forecasts against Polymarket- we smashed it so far!

[James Newport @NwprtNarrative]

A great example of an issue the PM shouldn’t be spending his time thinking about

This is obviously an issue each local authority should decide what they want to do

If Starmer gets frustrated at the slowness of Whitehall - it’s because of distractions like this

[Politics UK @PolitlcsUK]

🚨 NEW: Keir Starmer says the Government will allow pubs to serve outdoor pints and screen matches in their beer gardens for the World Cup

[James Newport @NwprtNarrative]

Another classic case of institutional failure, this time by @HMLandRegistry.

We’re trying to sell our flat. The title needs a simple amendment. All paperwork and new lease drawings are done. Because our entire chain is at risk of collapsing, our request to expedite was accepted.

The case was referred to their in-house legal team, where despite the chain being at risk, they’ve said the queue for the legal team can take an indeterminate amount of time - from "a few days" to "months”.

But because an original caseworker looked at it and passed it along, @HMLandRegistry considers the expedited request "dealt with" within their 10-day window.

@HMLandRegistry: The reason to expedite an application is because a decision is needed urgently, not to see how fast the paper can be pushed elsewhere.

[James Newport @NwprtNarrative]

The idea a model trained on everything that exists can’t discover something new is demonstrably wrong.

Just look at AlphaGo that started using playing new moves that had never been done.

Just ask an LLM to produce a poem about a pirate made of cheese in the style of Jane Austin.

The sooner we recognise that AI’s ability to be creative is not fundamentally different to ours, the sooner we can actually think sensibly about the risks!

[James Newport @NwprtNarrative]

It’s also showing another problem in politics - that politicians think this media is relevant at all.

The vast majority don’t get their news from newspapers!

[Dr Lawrence Newport @lawrencenewport]

There is a huge problem in politics - one this government seems particularly prone to - in thinking that you're doing a great job but all that's required is "better comms".

Early preview of Starmer's speech takes this to it's final point. A PM on the fences, announcing no radical policy action... just a speech about how the gov needs to speak about "hope".

It's exemplified best in that awful Steve Reed video where he complains about the amount of "bad news".

Just because you think you're doing "sensible things" doesn't mean it's actually the right stuff, nor that it's changing people's lives for the better! Comms isn't the problem, doing (the right) stuff is!

[Cyril @cyrilnotes]

“We are in danger of becoming a party of the well-off, not working people.”

Of course, Labour could consider being the party of “making working people well-off”.

Now that would be quite the mission. Growth, anyone?

[Swift Centre @swift_centre]

Our team showcases their quality once again!

Shame this was hypothetical money… but the point was showing the theory: professional forecasting is still the pinnacle of prediction.

Our round 2 forecasts vs Polymarket are in.

This is not financial advice…

[James Newport @NwprtNarrative]

Full article and our next set of forecasts against Polymarket: https://t.co/NdUNwNttfC

[James Newport @NwprtNarrative]

There is a lot of hype around prediction markets given the rise of Polymarket.

These platforms are an effective way of benefiting from "wisdom of the crowds", a phenomenon where you can increase the accuracy of a prediction through aggregation of many people's views. Fundamentally, this wisdom comes from the participants errors averaging each other out.

Here's the challenge - we know from scientific research that wisdom of the crowds loses against the techniques employed by professional expert forecasting organisations, such as the Swift Centre.

Talk is cheap though. So at the Swift Centre we went out to prove it.

In January, our team identified several of the Polymarket questions with the highest volume. We asked our forecasters to produce their own probability estimates on the events, to see how much money we'd win or lose if we were to bet against the market.

The results were definitive: a hypothetical $400,000 portfolio across four questions would have yielded a 32% return (+$127,148) in just over three months.

From not over-updating on the Trump-Greenland hype to anticipating the February US strikes on Iran, our team provided a clear signal through the noise, proving that specialised expertise remains the ultimate hedge against market volatility.

Read the full article below 👇

[James Newport @NwprtNarrative]

Article: https://t.co/iNwIlXZGQ4

[James Newport @NwprtNarrative]

It seems I like to be behind the trends.

I started a podcast in 2025. This year I've started a substack to record my musings. Maybe next year I'll use this thing called ChatGPT everyone keeps raving about...

Do check out my latest piece on how we're failing the next generation (nice post-bank holiday reading).

[EA Forum Posts @EAForumPosts]

"Why Forecasting Fails Decision Makers" by JamesN (@NwprtNarrative)

[James Newport @NwprtNarrative]

Good job. Someone good should do it.

Don’t like the “national tour of duty” narrative as a replacement for a good salary. Just compounds the problem.

Also ironically with AI and the future of work you may not be able to make money after this job…

[Looking for Growth @lfg_uk]

"We are guilty in the UK of trying to prevent anything bad from happening."

— Stephen Fitzpatrick on Britain's risk-averse regulatory culture

Watch the full podcast 👇

[Looking for Growth @lfg_uk]

"Under 10 employees [there's] one set of rules, then you get to 10 employees and you've got a completely different set of rules."

— Stephen Fitzpatrick on the marginal cost of hiring at certain "cliff edges".

Watch now! 👇