tomo,
Yes, Briggs puts it succinctly.

Possible FoI questions: How many times has ONS redefined the expected deaths model? When was the most recent previous change?

I didn't do anything more than glance at the mathematical model earlier (and think "ugh"), but looking at that tweet, age group, sex, region sound reasonable enough, but trend surely sounds like a fudge factor — a place to hide an unwanted ramp in the data. Not wild keen on the cross factors either.

And what's the point of the added factors anyway? If the model is all lumps and bumps from overfitting, it'll tend to obscure real variations in death rates. Well, I guess that *is* the point.

The Microsoft thing: the subdomain possibly just means they've moved the mail server inside another firewall while leaving other services unchanged at the old domain. It does have the air of a rushed change about it though.

tomo,
One amusing thing about the walk-off: we hear a female voice saying "Let's go; let's go". She forgot to say "Brandon".

Listening to one of the John Anderson "Conversations", I was intrigued to hear of a recent policy in Hungary where women who have had 4 or more children are exempt from paying income tax for the rest of their lives. That seems an interesting incentive to lift the fertility rate. Reading the EU's pdf on it didn't enlighten me all that much. Like so many tax things these days, there's no shortage of complexity (maybe not so hard for Hungarians).

EU seemed to be showing its woke side with this part:

One main criticism regarding this measure, and several of the other recent family policy measures taken, is that it mostly favours well-off families.

That doesn't strike me as necessarily a bad thing. If richer people have larger families, I suppose there will be a gene-pool imbalance, but sharing the wealth through inheritance will tend to reduce the wealth gap in later generations.

tomo,
At least they were courteous enough to walk out *after* the question rather than during it.

As for the 300 questions for small businesses, it's probably the viral effects of a government monopsony. The big business might have a few clients, but its biggest will almost always be the government, and the government demands "eco" credentials. So the big business just imposes the same on all its suppliers. The cure is, of course, that all the businesses close down, and everything comes from China, no questions asked.

Thought I'd take a look at the ONS new excess deaths scheme. Steps 2 and 3 are the dodgiest ones to me:

2. Fit a statistical model to the number of deaths in previous periods in each age-sex-geography stratum.
3. Use the model to predict the number of deaths in the reference period in each age-sex-geography stratum.

It *might* be realistic to fit models to the demographic strata of the overall population, but fitting a model to the number of deaths means that you expect "trends" in deaths to continue. e.g. you see 11, then 12, then 13, then 14. All good. Sure deaths are growing, but it fits the model perfectly, so growing as expected.

That's taking them at their word for the methodology. Not feeling inclined to scrutinise the mathematical model. As with the climate garbage, cheap computing power makes it convenient for them to use complicated models without proving that they're right, then challenge any doubters to prove that they're wrong.