Robert Cuffe @robertcuffe Twitter profile

Pinned Tweet

Robert Cuffe

1 month

PM’s fave crime stat (“down by 50%”) leaves out crimes that affect women more than men: groping, flashing, stalking and even rape.

3

22

37

Last Seen Profiles

@XDXDDDDDDA

@billyshears911

@AryanaSala73923

@AlmightyGunnar

@KimboAvax

@Gerard32850880

@JockeyCam

@wonxies

@TonyRoberts2442

@cdrcshekhar

@techryper

@Tata_Rauch

@ikiOridx

@helicoptero2024

@IsabelFloresArt

@dooshaez12856

@DigitalVC2022

@Q0MT6pFmbVqynsM

@vicdelarge

@rinrin20010430

@AdrienneSw25305

@CoreyTurner1996

@jandakembangstw

@kkaebs614

@nao362981089797

@nickSpongebobp1

@Ann91486661

@NaDa15891680

@xaviershewrote

@TrnQuan75751281

@sofiafonsoferre

@DanielleWa35377

@JapanCandyBox

@SILVIA77733779

@bulilow

@liumopodu150617

Robert Cuffe

@robertcuffe

3 years

I know correlation isn’t causation but....

48

291

2K

Robert Cuffe

@robertcuffe

2 years

Largest rich vs. poor gap in inflation since the 00's. For much of the last ten years, inflation has been very close for richest and poorest. Higher fuel and food prices hit poorest hardest - so their inflation has taken off faster in 2022 (blue line, far right of chart).

25

688

1K

Robert Cuffe

@robertcuffe

4 months

Alexa, show me random variation around zero….

17

189

1K

Robert Cuffe

@robertcuffe

4 months

“Crime has fallen by 50%” says @ukhomeoffice @RishiSunak @Conservatives based on the headline official measure of crime in England and Wales. But that excludes harassment, stalking and sexual assault (yes, even rape).

34

315

484

Robert Cuffe

@robertcuffe

4 years

Thanks - rting so my mum can see the clip (if she ever joins Twitter)

Catherine Wyatt

@catherinehwyatt

4 years

You simply won't hear a better explanation of what the number of cases around England actually means -- @robertcuffe on the News Channel earlier

40

491

2K

38

59

462

Robert Cuffe

@robertcuffe

3 years

There are about twice as many unvaccinated adults as there are kids aged 12-15 who could be vaccinated.

40

66

399

Robert Cuffe

@robertcuffe

1 year

It's technically true but not informative to say, as Chancellor Jeremy Hunt did, that the UK saw the biggest growth in the G7 last year. The UK's economy didn't really grow much in 2022. It looks good compared to 2021 because that year included lockdowns.

16

150

363

Robert Cuffe

@robertcuffe

2 years

The ONS coronavirus survey will continue with ~75% of its current sample size and further savings achieved through more online/postal data connection. Sir David Norgrove, outgoing Chair of @UKStatsAuth , announced this speaking at @instituteforgov

10

98

345

Robert Cuffe

@robertcuffe

3 years

How the different vaccines work. Clearly explained by @JamesTGallagher

16

132

309

Robert Cuffe

@robertcuffe

8 months

"Many instances" is pushing it based on the figs the Home Office release. ~1.5% of asylum applications mentioned sexual orientation in the latest figures. About 3/4 of claims that mention orientation are accepted, roughly the same as other reasons for claiming asylum.

Peston

@itvpeston

8 months

"People purport to be gay when they're not actually gay" Home Secretary @SuellaBraverman accuses asylum seekers of lying about their sexuality to gain refugee status in the UK Full interview with @AnushkaAsthana tonight ⬇️ 💻LIVE 9PM @itvpeston 📺11.15PM @ITV #Peston

3K

312

1K

19

89

285

Robert Cuffe

@robertcuffe

4 years

The UK's coronavirus daily death count is being revised downwards. The total is down by just over 10%, but the most recent figures by more than half.

15

107

251

Robert Cuffe

@robertcuffe

3 years

Why is protection from _infection_ SO MUCH better for everyone than protection from illness? @FryRsquared and @rozeggo explain why using table tennis and a fictional zombie apocalypse. v1: no protection

4

104

243

Robert Cuffe

@robertcuffe

3 years

1,052 deaths is lower than last Tuesday (1,449). Don’t be put off by the Tuesday bump. Deaths are coming down. Every day for the last two weeks, the number of deaths reported has been lower than the same day the week before. We haven’t seen a run like that since the summer.

13

25

236

Robert Cuffe

@robertcuffe

3 years

How to balance risks and benefits of vaccination for you. 1-in-a-million is roughly your chance of getting murdered in the next month, or dying in a 250-mile car trip. That's pretty small full stop. And, for most, it's small compared with Covid risks.

AstraZeneca vaccine: How do you weigh up the risks and benefits?

A guide to the risks and benefits of receiving the Oxford-AstraZeneca vaccine.

www.bbc.co.uk

17

89

212

Robert Cuffe

@robertcuffe

3 years

Nearly a quarter of people in their 40s haven't had a jab. Nearly a third of those in their 30s.

26

114

193

Robert Cuffe

@robertcuffe

3 years

What does "very rare" mean? 19 clots in 20,000,000 doses is roughly 1-in-a-million. Similar to your chance of getting murdered next month or dying in 250 miles worth of driving.

14

64

176

Robert Cuffe

@robertcuffe

4 years

Credit to @CebmOxford whose work led to the review and ultimately to these changes. Story here:

10

42

151

Robert Cuffe

@robertcuffe

1 year

Nearly 660k deaths registered in the UK in 2022. The rise in death rates since pre-pandemic times is, by a number of analyses, one of the worst in the last half-century. It returns death rates back to about the levels seen in 2015.

21

73

170

Robert Cuffe

@robertcuffe

2 years

ONS estimate prices rising even faster for poorest fifth*: by more than 16% (vs 11% on average). Food/energy take up a little under half of their spending (vs a third for the average houshold). * poorest by how-much-you-spend rather than how much you bring in.....

5

83

140

Robert Cuffe

@robertcuffe

3 years

Clear chart from ONS: 87% of deaths involving Covid were caused by Covid. 89% of deaths involving pneumonia/flu were not caused by pneumonia/flu.

6

59

134

Robert Cuffe

@robertcuffe

3 years

Good news: long covid symptoms less common than feared. Although a decent proportion of people have symptoms for a month, it falls over time and, by month 3, it’s down to 2.5% pts above the rate for people who never had it.

14

46

126

Robert Cuffe

@robertcuffe

3 years

The main data from the ONS infection survey didn't exaggerate the case for lockdown.

8

30

109

Robert Cuffe

@robertcuffe

3 years

Rumours of higher case rates in vaccinated over 40s (vs unvaccinated) have been greatly exaggerated. We don’t know how many unvaxxed over 40s there are. So we estimate. Rates = cases (known) / people (estimated). Get the people number wrong and you’ll get the rates wrong.

18

31

105

Robert Cuffe

@robertcuffe

2 years

It has been pointed out to me that the typical loan is about 3.5 times income, so this lucky first time buyer has a deposit of nearly £500k. Or is getting a loan closer to 20 times their income.

2

23

96

Robert Cuffe

@robertcuffe

2 years

The UK passed 100k reported Covid deaths on 26 January last year and 125k about six weeks later. The next 25k has taken nearly 10 months: spring lockdown + vaccines really slowed it down.

16

33

89

Robert Cuffe

@robertcuffe

2 years

The Treasury have picked a very unusual example here - the people who benefit the most from the changes in stamp duty. The terraced house in London must cost exactly £625k for them to make this saving - £1 more and they'd be saving a lot less.

15

51

86

Robert Cuffe

@robertcuffe

3 years

We're not currently on track for 7k admissions a day in England. It's a "what if" analysis that looks at what would happen if R hit 1.5 (last week) and stays there and we change nothing.

6

22

78

Robert Cuffe

@robertcuffe

2 years

The rise in the price cap is almost all the rise in wholesale costs (massive blue bit at top of bars). VAT and profits are up too (yellow and purple bits near bottom) but way less: they're a fixed (small) proportion of the total bill.

20

47

70

Robert Cuffe

@robertcuffe

1 year

The Chancellor's chosen stat was that year-on-year comparison rather than how we've grown throughout 2022 or a comparison that looks at whether leading advanced economies have recovered from the pandemic. They're not so flattering.

2

23

66

Robert Cuffe

@robertcuffe

2 years

@ONS say the gap is "the largest since June 2010" and back then, it was actually in favour of lower-income households.

Office for National Statistics (ONS)

@ONS

2 years

We've published experimental data looking at estimates of inflation rates for different types of households in the UK This helps provide insight on the cost of living and how price changes can vary between different groups.

2

23

50

3

30

64

Robert Cuffe

@robertcuffe

4 years

Dear @RoyalStatSoc , if I promise to pretend that I meant it in the sense of "between two gateposts" will you promise not to expel me for crimes against geometry?

4

1

64

Robert Cuffe

@robertcuffe

5 months

Death stats around Eat Out To Help Out Aug 2020: just under 550 Sept 2020: nearly 900 Jan 2021: more than 36,000 (based on deaths involving Covid, according to death certs)

Eat Out to Help Out: What was the impact of the scheme?

How much did the scheme help the economy and affect the spread of Covid?

www.bbc.co.uk

5

30

63

Robert Cuffe

@robertcuffe

3 months

Wages up by nearly 23% since Dec 2019! But cost-of-living up by just over 24% in the same time. And so the value of the average wage is still lower than it was in 2019 - if you measure the cost-of-living properly. h/t @morgan_wild

3

34

63

Robert Cuffe

@robertcuffe

4 years

Doesn't feel high in terms of my-chance-of-getting it. But translates to ~150,000 which is pretty high if you're trying to trace contacts.

9

61

Robert Cuffe

@robertcuffe

4 years

Roughly 1 in 400 people in England have coronavirus - estimate based on survey of households (not hospitals or care homes) by ONS.

11

42

55

Robert Cuffe

@robertcuffe

3 years

Good news from @PHE_uk : 2nd AZ dose reduces cases by nearly 90% These real world data are better than the results from clinical trials, and better than assumed in the most recent models supporting May 12th unlocking.

2

15

52

Robert Cuffe

@robertcuffe

3 years

“We've got higher rates in the younger age groups, but the increases are now happening across all the under 60s," says the assistant director of public health in Bolton. "That suggests to me that we've got transmission within households and that's now part of the situation here.”

2

10

49

Robert Cuffe

@robertcuffe

1 year

@jdportes @ProfTimBale @BenZaranko Ahem. By the excellent @Danict89 on the BBC site yesterday morning.

10

37

50

Robert Cuffe

@robertcuffe

3 years

v2: protection

2

19

47

Robert Cuffe

@robertcuffe

2 years

Confirmation of getting-past-the-peak from @ONS : UK infections falling: down to below 3.5m from 4.3m last week. This survey swabs people at random, so falls aren't due to changes in who gets tested or who can get a test.

5

14

48

Robert Cuffe

@robertcuffe

2 years

How do you square 200,000 new Omicron infections yesterday with <60,000 new cases, total, today?

5

48

Robert Cuffe

@robertcuffe

3 years

@AlistairHaimes Yes - positive results rapid tests done at schools now should be double checked with a PCR test. You still have to isolate from the first positive, but if the PCR comes back negative, you can stop isolating.

13

10

46

Robert Cuffe

@robertcuffe

2 years

UK deaths (for any reason) 8% below the "expected levels" this week - but those expectations have changed. They now include last January's Covid spike.

6

21

47

Robert Cuffe

@robertcuffe

3 years

Do you think of matrices row by row or column by column?

TheBossRoss 🇪🇺 🧶 ❄ 📷 6x💉

@BettinaSRoss1

3 years

This is so confusing.

54

199

1K

7

6

46

Robert Cuffe

@robertcuffe

3 years

48m people with a first dose. About 54m people aged 16 or over in the UK = 6m still to vaccinate. Just over 3m 12-15 year olds in the UK.

3

7

45

Robert Cuffe

@robertcuffe

4 years

Multiplicative growth is hard to grasp. Even for nerds.

Exponential growth bias: The numerical error behind Covid-19

A simple mathematical mistake may explain why many people underestimate the dangers of coronavirus, shunning social distancing, masks and hand-washing.

www.bbc.com

21

14

42

Robert Cuffe

@robertcuffe

3 years

1,449 deaths reported in a day is bad news. But.... It's less than last Tuesday. And Monday's total was less than last Monday's. That's been the case seven days in a row. The numbers are coming down. Not by much (about 10% in a week) but coming down.

2

6

42

Robert Cuffe

@robertcuffe

2 years

Growing gap between dashboard deaths and those caused by Covid. ~140 a day caused by Covid registered in week to 28 Jan in England and Wales ~240 a day reported on the dashboard (also E&W, any death within 28 days of positive test) The gap (~70%) is largest we've seen.

7

16

38

Robert Cuffe

@robertcuffe

1 year

This chart illustrates perfectly why you shouldn’t get too wrapped up in the headline month-on-month GDP growth figure. Ofc the full set of data are informative, but the single headline fig bounces quite a bit around the long term trend. (Which, for the UK, is flat)

3

10

40

Robert Cuffe

@robertcuffe

3 years

About 1 in 500 people in the UK died in the last year and have "Covid-19" on their death certificate. That's more than 2,000 per million people. Handy ballpark number for contextualising rates of adverse events after vaccination.

1

16

39

Robert Cuffe

@robertcuffe

3 years

Total deaths registered in UK in first week of March very close to average levels seen during that week in 2015-19.

3

0

38

Robert Cuffe

@robertcuffe

2 years

Rises everywhere bar London (where picture is a bit muddier). 1/5 Admissions rising everywhere bar London. But "flat" in London masks some hills: big dip and then a rise again in recent days.

7

20

38

Robert Cuffe

@robertcuffe

3 years

Mum, you do have my phone number. You could just call.

1

39

Robert Cuffe

@robertcuffe

1 year

It's hard to put a precise number on how many deaths are caused by delays in urgent care as opposed to, say, the after effects of the pandemic because some of the problems in urgent care are caused by the after effects of the pandemic.

9

13

38

Robert Cuffe

@robertcuffe

3 years

Is anyone else singing “Census Day, Census Day” over and over again to the tune of Jingle Bells?

13

4

39

Robert Cuffe

@robertcuffe

4 months

And, while ‘total’ crime is falling and ‘violent’ crime is falling, sexual assault isn’t.

1

10

37

Robert Cuffe

@robertcuffe

4 years

Now this is what I call a Friday night party.

4

36

Robert Cuffe

@robertcuffe

3 years

The UK is one of the hardest hit nations - but it's too soon to build the league table of second waves.

4

8

37

Robert Cuffe

@robertcuffe

5 years

And the award for best titled statistic of the year goes to....the ONS release on the pub trade focussing on bigger bars: “The Economies of Ale”

1

13

38

Robert Cuffe

@robertcuffe

4 months

...it’s hard to put them into a headline measure of the total “number of crimes”. Victims of sexual assault or domestic abuse might not be able to speak freely in front of family/their abuser. And how many crimes take place in a single case of, say, stalking or domestic abuse?

1

4

36

Robert Cuffe

@robertcuffe

5 months

Oof. Was PM right to say debt is "falling"? Not if it means "forecast to fall as a proportion of GDP in five years' time (but not before then)". @UKStatsAuth say that usage "may have undermined trust in the Government’s use of statistics". Oh, and debt's rising atow.

UK Statistics Authority

@UKStatsAuth

5 months

Today, our Chair has responded to a letter from @sarahjolney1 about public sector debt.

2

15

37

3

18

37

Robert Cuffe

@robertcuffe

2 years

Your energy bill is the bearer of bad news but your supplier's profits are not the problem. Those profits are limited to 2% of the price, barely visible on a chart of the price cap. The real driver of the cap rises is the money that goes to energy generators or gas companies.

3

13

37

Robert Cuffe

@robertcuffe

4 months

@ONS Our write up here:

Key crime stat excludes rape and other sexual offences

A measure routinely highlighted does not include major types of crimes that affect women more than men.

www.bbc.co.uk

1

15

34

Robert Cuffe

@robertcuffe

4 months

@ONS survey 30k people each year to ask about crime. They publish detailed analysis of stalking, domestic abuse and sexual assault (e.g. below) and are researching how best to quantify harassment But...

1

6

35

Robert Cuffe

@robertcuffe

3 years

Case rates levelling in children and teens (maybe even falling). The rise in the parents looks to be stalling too.

1

4

34

Robert Cuffe

@robertcuffe

3 years

Get your positive rapid tests double-checked

7

24

33

Robert Cuffe

@robertcuffe

3 years

All our charts on who's been vaccinated here

0

12

34

Robert Cuffe

@robertcuffe

3 years

Reported numbers of clots after AZ vaccination are still much lower than normal daily numbers of clots.

2

17

34

Robert Cuffe

@robertcuffe

3 years

Money doesn't buy happiness but.....

3

8

34

Robert Cuffe

@robertcuffe

4 years

It's hard to visualise just how less likely school-age children are to die with Covid-19. U15s account for 0.01% of Covid-19 deaths vs nearly 75% from the 75+s. In the middle, 15-44s account for 1%, 45-64 for 10% and 65-74s for the other 15%.

6

21

33

Robert Cuffe

@robertcuffe

4 years

Daily death falls stalling, perhaps even starting to turn. Govt advisers say that we should realistically expect this figure to start creeping up again.

29

11

32

Robert Cuffe

@robertcuffe

1 year

The main story in today's census data: where best to watch the match if you're living in the wrong country? The map below (by @ErwanRivault ) shows people choosing Welsh or English identity in last year's census. Thin pickings for the Welsh in England.

1

11

32

Robert Cuffe

@robertcuffe

4 months

And while we’re at it, violent crime stats tell you _nothing_ (NOTHING!) about sexual violence or harassment. Rape, for example, goes in the “sexual offences” bucket. It doesn’t go in the “violent crimes” bucket. To be fair, neither does robbery – theft with threat of violence).

2

9

32

Robert Cuffe

@robertcuffe

2 years

Freezing tax thresholds means more people will pay more tax. That was the plan when it was announced last March but the effect is even bigger this year because inflation is running so much higher than expected. Figures spotted by @faisalislam , chart by @Danict89

Nicholas Barrett

@nbarrett100

2 years

Rishi Sunak says his "tax plan delivers the biggest net cut to personal taxes in over a quarter of a century". @RobertCuffe , @Danict89 and I tried to find out if people will actually be paying less tax

0

1

3

1

26

33

Robert Cuffe

@robertcuffe

2 years

Two more bits of good news this morning - staff absences due to Covid falling - number of people in hospital primarily for their Covid falling

4

6

33

Robert Cuffe

@robertcuffe

2 years

Also not correct to say that the energy price guarantee will save a typical person in a semi-d £1,150 this winter. That's the estimated saving for the full year.

Robert Cuffe

@robertcuffe

2 years

It is not correct to say that the energy guarantee “is preventing bills going up to £6500 next year”. It’s based on an estimate of peak costs for Apr-Jun that was out of date by the time the guarantee was announced.

1

7

19

1

8

31

Robert Cuffe

@robertcuffe

4 months

And neither is stalking.

1

2

28

Robert Cuffe

@robertcuffe

5 months

The proportion of people who say they wash their towels once or twice a year isn’t far off the proportion who’ll say any old nonsense in a survey.

Millions could be risking health issues because of poor towel hygiene

A survey of 2,200 UK residents reveals men are five times more likely than women to wash their towels once a year or less.

www.dailymail.co.uk

6

7

29

Robert Cuffe

@robertcuffe

3 years

As the number of vaccinated people grows, more and more COVID deaths will be among vaccinated people. Clear analysis by ⁦ @d_spiegel ⁩ and ⁦ @anthonybmasters ⁩

4

5

27

Robert Cuffe

@robertcuffe

4 years

Eight times more pneumonia and flu deaths than Covid-19 deaths doesn't mean that flu is killing more people that coronavirus. Most pneumonia-and-flu deaths are pneumonia. And "involving" is less likely to mean "because of" for pneumonia-and-flu than for coronavirus.

5

13

30

Robert Cuffe

@robertcuffe

6 years

@d_spiegel Thanks to @d_spiegel @stephensenn @statsguyuk for highlighting and discussing. BBC story edited. Any errors are ours not theirs.

2

11

30

Robert Cuffe

@robertcuffe

1 year

Why are interest rates falling for new fixed rate mortgages when the UK's main interest rate is rising?

4

5

30

Robert Cuffe

@robertcuffe

3 years

Biggest rise in registered deaths in the UK since WWII doesn't mean your chances of dying nowadays are the same as they were during rationing.

3

5

30

Robert Cuffe

@robertcuffe

4 months

Caveats around computer abuse and fraud (only added to the crime survey in 2015) are pretty prominent in these claims, and in the ONS release. It has been harder to find caveats about the huge wodge of crimes affecting women that have been left out.

1

4

27

Robert Cuffe

@robertcuffe

4 months

@jdportes Jonathan, the experience differs materially for men and women - violent crime doesn't include sexual offences (rape, groping etc), stalking or harassment Headline figure for total crime leaves them out too. And non-violent domestic abuse.

Robert Cuffe

@robertcuffe

4 months

“Crime has fallen by 50%” says @ukhomeoffice @RishiSunak @Conservatives based on the headline official measure of crime in England and Wales. But that excludes harassment, stalking and sexual assault (yes, even rape).

34

315

484

3

7

27

Robert Cuffe

@robertcuffe

4 months

So @ONS say it’s very hard to get the details you need to add them into the “total” crimes measure. And warn against reading too much into any single fig that tries to summarise long-term trends in all crimes. But that's the one that gets quoted. And it's falling.

1

3

27

Robert Cuffe

@robertcuffe

4 years

“There are no simple answers. ...we will need years to properly assess the effect of the epidemic and the measures taken against it. We’ve now got a league table, but as to why the UK has done so badly, the arguments will go on” from @d_spiegel

6

11

27

Robert Cuffe

@robertcuffe

3 years

Reduction not prevention: three weeks after a first vaccination, there were 74% fewer infections with symptoms. Infections without symptoms were also down but by less (57%). Our writeup here:

3

13

26

Robert Cuffe

@robertcuffe

4 years

More or Less is back. Always worth a listen.

BBC Radio 4 - More or Less, Hawaiian Pizza, Obesity and a Second Wave?

Covid-19 cases are rising in the UK, is it a sign of a second wave of the virus?

www.bbc.co.uk

0

10

26

Robert Cuffe

@robertcuffe

3 years

Last Tuesday: 799 deaths and 10,625 cases reported. This Tuesday: 548 deaths and 8,489 cases reported.

5

3

24

Robert Cuffe

@robertcuffe

3 years

@Kit_Yates_Maths 2, 4, 6, 8, n's growing at a stately rate. 2, 4, 8, 16, exponential growth. that means 8 more couplets at this pace and n exceeds the human race It never caught on as a chant at school.

1

0

26

Robert Cuffe

@robertcuffe

1 year

Big part of counting “excess” deaths is working out what you mean by “expected”. @ONS - and others - looking for feedback on how best to define that in the years after the pandemic.

Sarah Caul

@Sarah__Caul

1 year

Today we released a new blog on what we are doing around excess mortality: (1/3)

13

37

70

2

5

25

Robert Cuffe

@robertcuffe

2 years

Follow-up tests to confirm an infection don’t add much when there’s a lot of coronavirus going around.

2

8

25

Robert Cuffe

@robertcuffe

3 years

Write up by @jim_reed , charts by @England_Rob_

Indian variant: What's causing the spike in Bolton?

The town has seen a sharp spike in Covid cases - driven by the emergence of the Indian variant.

www.bbc.co.uk

8

23

Robert Cuffe

@robertcuffe

4 months

The good news is that domestic abuse has been falling pretty steadily.

1

2

23

Robert Cuffe

@robertcuffe

6 years

1/2 Stato's sidebar. @britishfuture 's report is full of interesting data (). On p3 is The Best Ever demonstration of why open online surveys are unreliable tosh.

1

17

22

Robert Cuffe

@robertcuffe

4 months

Crime analysts and journos and academics all know about this last point. But most of the people I spoke to about the headline measure said something like “hang on, that can’t be….oh yeah…”. It’s this weird feature hiding in plain sight.

1

21

Robert Cuffe

@robertcuffe

3 years

A statistician has lost the right hand glove from each of the last two pairs that he bought. What are the chances that he’ll lose the right hand glove next time around?

50%: coin toss, stats etc

435

100%: because life

458

19

1

23

Robert Cuffe

@robertcuffe

2 years

Huge rises in UK infections estimated by ONS: up from 2.3m to 3.7m in a week. Hint of a slowdown in London, but growth picking up in over 70s.