Category Archives: Statistics

Stroppy 2024!

Standard

As we bid farewell to the Emmys the spotlight turns towards the Stroppys – or rather the Stroppy, there being only one such award for meaningless twaddle each year. After much thought and hand-wringing the winner is… drum roll… anonymous. Yes, that’s unfortunate, but I don’t know the name of the author of this gem, extracted from a report written for a certain international financial institution based in Washington DC:

James Joyce (portrait)

“A weigh station that is also a public check post shall undertake checks in its capacity as a weigh station unless it is established to the satisfaction of the authorised officer that the payload on the truck or combination being checked originated from the nearby urban agglomeration in which case the check shall be made in its capacity as public check post, and if the truck or combination is found to have committed an infringement requiring immobilisation then that infringement shall be waived and the truck or combination allowed to travel directly back to its loading point for rectification without penalty or permanent record of noncompliance.”

Marcel Proust (caricature)

At only 106 words this sentence can hardly be considered Joycian or even Proustian. But it makes up for that in its opacity, convolution, clumsiness and… well, let’s just sum it up as twaddledom.

I am indebted, not for the first time, to my long-term friend and sometimes colleague Ron Allan, who nominated this passage.

And here’s an unrelated thought bubble… I just read that the average human exhales 255kg of CO2 per year. That’s equivalent to an old-fashioned ounce per hour, so it seems plausible. 

Boeing 747: First roll-out (30 September 1968)

Back to the individual level, the average person exhales about 20t of CO2 over a lifetime. If this number is correct, our species is directly putting 2 billion tonnes of CO2 into the atmosphere annually.  This is as much as 800 million petrol-powered cars.  Or 90,000 Boeing 747s.

Perhaps a childless person who voluntarily undergoes sterilisation should be issued with carbon credits.  What do you think?

By the way, do you happen to know (or be) a film producer in need of a potential blockbusting script for a TV series? If so, please let me know.

Wimbledon

Standard
A street in Wimbledon

There are place whose names conjure up emotions that are out of all proportion to geographical reality.  One thinks immediately of Paris, Samarkand, Xanadu… and Wimbledon.

I have a 326-year-old map of Midlesex [sic] hanging on my office wall. There are a few place names south of the Thames. Roehampton is there, spelt ‘Rowhampton’, but that suburb’s neighbour Wimbledon is absent. Now it rates inclusion in the London A-Z, but this random streetscape (courtesy of Google Maps) hints at nothing extraordinary. And yet…

Last night Mrs SG and I watched spellbound as Ashleigh Barty, carrier of Australia’s sporting hopes, battled with Karolina Plišková to win the Women’s Singles championship. Her victory was followed by royal pomp and graciousness, an interview, a quantity of photographs that in the pre-digital age would have dented the world’s supply of silver, and holding aloft the trophy plate until her arms must have ached.

Ashleigh Barty

And all the time one was aware of the dark green colour scheme of the stands, the worn grass, the military efficiency of the ballboys/girls… it could have been nowhere else but on the hallowed sod that mortals call Wimbledon.

POSTSCRIPT

Mrs SG and I had our second Covid-19 vaccinations this week: AstaZeneca both times, no side-effects.  We were told that our chances of dying from a consequential blood clot are, respectively, 1.9 and 1.8 per million. As a mental exercise I estimated the lifetime odds of dying in a road accident:

Road accident deaths in Australia = 1,580 in 2020
Equivalent to 61.5 per million inhabitants per year
Average life expectancy = 83.5 years
So lifetime odds = 61.5 x 83.5 = 5,135 per million, or 0.5%

I know… lies, damn lies and statistics.  But for me it puts things in perspective.  And of course the chance of dying if you catch the Covid-19 virus is around 2% and the chance of dying of something is 100%. But not for Ashleigh. She has joined the Sporting Immortals. For such as she cremation is but a hiccup.

COVID-19: How Many Will Die?

Standard

It seems perverse of me not to have pontificated or at least propounded about the biggest news story since Harry and Meghan took off – beyond blogging about bog paper. I’m talking about Corona Viral Disease No.19, aka COVID-19.

No-one knows how bad things will get or how soon a vaccine will become available, but I have amused myself by running some plausible numbers. Let’s say that one-third of humanity is infected; that’s within the range that we’ve heard from experts. Then let’s assume that 1% of those unlucky people die. That’s below the rates that are being talked about; but those are based only on the known cases of infection, which are almost certainly the minority of actual cases.

There are about 7.5 billion people alive today, so my assumptions would mean a death toll of 7,500 million x 1/3 x 1% = 25 million. That’s a lot of people, but it’s equivalent to:

  • No more than 50% of the number who died of the Spanish Flu a century ago, when the global population was only 1.8 billion.
  • Less than 50% of the usual number of deaths in a year.
  • About four months of humanity’s natural growth rate (births minus deaths).

Moreover, mortality is going to occur much more than proportionately among the old and the sick, many of whom would die soon anyway. So when future students of demography examine a graph of human population growth they will notice a deceleration in 2020. It may pique their interest enough to glance at a footnote that mentions COVID-19.

NIS Again: SA’s Interconnector

Standard

I coined the phrase ‘Numeric Incompetence Syndrome’ a while back, and yesterday an article in my local newspaper delivered another glaring example. In summary…

South Australia’s connection to the national electrical grid is to be enhanced with a new 900km interconnector to New South Wales and Victoria. The capital cost is expected to be $1.53 billion. The article states: “To cover that, households would pay $9 a year in SA and $5 in NSW.”

Assuming an average household size of 2.7 persons (as in 2016 nationally), there are about 0.65 million households in SA and 3.11 million in NSW. So the total annual amount recovered from households would be ($9 x 0.65M) + ($5 x 3.11M) = $21.4 million. Even allowing for future population growth, this comes nowhere near “covering” an investment of $1.53 billion: to amortise such a sum over a 20-year life at a discount rate of 5%pa would cost $123 million per year, before considering any maintenance costs. So that’s error No.1.

The article goes on to say, “[ElectraNet] estimates the project would deliver overall benefits of $924 million over 20 years…” but adds that “the Australian Energy Regulator … has downsized the project’s 20-year benefit to $269 million.”

Who, in their right mind, would invest $1.53 billion in something that will deliver benefits of only $924 million over 20 years?! How can any sub-editor not see that this cannot be true?! Perhaps the word “net” was omitted, but surely “overall” was inserted to make clear that the writer means gross benefits.

My stroppiness is going off the scale. Journalism is not just about regurgitating people’s press releases; it has to involve some critical thought, some fact-checking, some exercise of common sense for heaven’s sake!

I have emailed the Editor of the newspaper with a link to this post and an invitation to respond and/or to publish a correction.

Surprising Numbers

Standard

I’ve always been fascinated by results of calculations that seem disproportionate to the point of incredibility, but are correct. I get stroppy with people who are neither fascinated by such results nor moved to question and check them. Here are four that came to my attention recently:

  • I peeled an orange and weighed the peel. It was 31% of the unpeeled fruit, so a price of $2.99/kg turns out to be $4.33/kg of the edible part.
  • I read an article whose author decried the terrible devaluation of the US dollar over the past century: “A dollar in 1919 is now worth 5 cents!” That implies an average rate of inflation of 3.04%pa. A lot of central bankers would be very happy with that.
  • Imagine a globe with a diameter of 30cm (1 foot in the old money) representing Planet Earth. The depth of the atmosphere would be 0.3mm – the thickness of a child’s finger nail.
  • Research in the USA in 1997 found that average IQ had increased by 20 points since 1932. Eleven years later, similar research in the UK found a 14 point improvement since 1947. This is called the Flynn Effect. Other research suggests a pretty steady increase of about 3 points per decade – but only up to the late 1990s, when the trend appears to have petered out and reversed.
    [When originally posted, on 19 January, I mistyped “1970s” instead of “1990s”. Sorry.]

Suicide

Standard

There’s been another report on suicide in Australia. People have expressed shock that the highest suicide rate is for men over 85 years of age. I say to those people, “Well, which group defined by gender and age do you want to have the highest suicide rate? Young men? Old women? Children?!”

Personally I think it makes perfect sense that old men are the most prone to suicide. Men are valued – by themselves and by others – for their strength, virility and self-reliance. They tend to define themselves in terms of their income-earning capacity. As they age, all these reasons to be valued fade away.

Men are also known to have fewer and weaker social connections outside their workplaces, and to recover more slowly than women after the loss of a spouse.

According to the latest Australian death statistics, with a bit of adjustment by me for population growth, 70% of males make it to 70; once there, half will make it to 90. I am not shocked if a few blokes in that bracket are no longer enjoying the party and want to leave early.

Numeric Incompetence Syndrome (NIS)

Standard

Here is a direct quotation from my local newspaper:

“Aboriginal women are at least 32 times more likely to be hospitalised because of an assault by their partner than non-indigenous women.

“They are at least three times as likely to have experienced violence in the past year than non-indigenous women.

“All this despite comprising a far smaller proportion of the population.”

The writer (Lauren Novak) is drawing attention to important statistics and goes on to suggest ameliorative measures. But she spoils it for me in that third paragraph. Having correctly compared rates of hospitalisation and exposure to violence, she then demonstrates a failure to understand what a rate is.

I don’t want to pick on Ms Novak, who is a much-awarded professional journalist whom I have never met. This is just one example of what I have now dubbed Numeric Incompetence Syndrome (NIS). It seems to afflict journalists and sub-editors disproportionately, but perhaps that’s because their condition is on public display more often than other sufferers’.

I’m thinking about an annual award for the most egregious manifestation of NIS: a Nissy to sit alongside the well-established and eagerly-awaited Stroppy**. What do you think?

Since writing this I have seen another howler in my local newspaper (one of Rupert Murdoch’s, by the way): a little matter of a displaced decimal point in a graphic summary of the South Australian State Budget, showing annual revenue and expenditure to be A$1.9 billion instead of A$19 billion. Now that’s the sort of thing a sub-editor should pick up, don’t you think?

** The Stroppy Git Award for Meaningless Twaddle, awarded every January, for which nominations are always open.

Motherly Love

Standard

I’ve just read an astonishing piece of news. According to my local newspaper, “New research shows that 77% of Australian mums are putting their own desires on hold to chase those of their children.” This bizarre finding arises from exhaustive analysis by PureProfile, paid for by Go People.

One hopes that Go People will follow this up by funding another round of ground-breaking research into dads’ suppression of their desires for the sake of their children.

Then, if they have any money left, they could hire someone to work out why this might be so. For example – and I’m thinking aloud here, having no research funding to do a proper job – could it be that without the instinct to devote oneself to providing and caring for one’s children none of us would be here?

Interest Rates

Standard

Monetarism was in the early stages of its ascendancy when I was studying economics. I remember one of my lecturers describing Milton Friedman, doyen of the Chicago School, as “that amiable fascist.” Now we are living with the fall-out of Governments’ delegating economic management to their central banks and using fiscal policy as a means of appeasing noisy interest groups and winning elections.

I came across this graph yesterday and felt I should share it:

ustreasury10yieldgraph

If this is how interest rates have to behave – even dipping into negative territory currently in Denmark, Japan, Sweden and Switzerland – to balance inflation and economic growth, we should all be on the lookout for flying pigs and white rabbits with pocket watches.

Suicide

Standard

People in rich and peaceful countries don’t have enough to worry about, so they look around for irritations and talk them up into Big Issues. This is happening now in Australia with male suicides.

Now, please don’t misunderstand me. I know there are people who kill themselves during fits if depression or in response to tragic events, from which they could have recovered to lead happy lives, and I wish they’d had second thoughts. But there are also people who rationally consider their prospects and their options and decide that they’d rather be somewhere else. They’re not enjoying the party so they want to leave. For those people, suicide is rational and even admirable. Their leaving the party means that the net sum of happiness in the world is increased.

But what makes me stroppy is that the people who are promoting this as a Big Issue in Australia speak as though suicide among Australian men has suddenly become a crisis that has to be moved up the political agenda. Notice my italics. The implication is that Australia is the suicide capital of the world and men are disproportionately affected.

Now consider the following facts, drawn from Wikipedia with a WHO citation and with age standardisation:

  • In a ranking of 171 countries, Guyana is ranked No.1 with 44.2 suicides per 100,000 people in 2012. Australia is at No.63 with 10.6.
  • In every country except two (Pakistan and Iraq) the male suicide rate exceeds the female, typically by a factor of about 3.
  • If we compare Australia with the other countries of the developed Anglosphere (our usual benchmark) it falls pretty much in the middle – see the extracted table below, which includes countries that are tied with those of the Anglosphere for a bit of extra colour.
  • The 17 countries at the bottom of the ranking – ie with the lowest suicide rates – are all located in or around the Caribbean Sea or have predominantly Muslim populations. This has nothing to do with the main theme of my post, but I think it’s a fascinating fact and someone should be doing a PhD thesis on it.

suicidestats

Where it says ‘more info’ you can click and see more if you go to the referenced Wikipedia page.

So how the hell did the UK manage to tie with Swaziland at No.105?! Perhaps most of the unhappy Poms have emigrated.