Polls, nuns and life partners

Summary

Podcast Summary: More or Less — "Polls, Nuns and Life Partners" (May 1, 2015)
Host: Tim Harford (BBC Radio 4)
Guest Experts: David Cowling, Ben Page, Deborah Mattinson, Sister Gemma Simmons, Matt Parker
Episode Theme:
This episode of More or Less explores the accuracy and interpretation of statistics and numbers featured in UK news, political debates, and daily life. Key topics include election polling–particularly in marginal constituencies, the validity of "reading age" statistics, a reported resurgence of nuns in the UK, the health implications of jogging, and the mathematics behind finding an optimal life partner.

---

Key Discussion Points & Insights

1. Polling in Marginal Constituencies and the Limits of Uniform Swing

• Setting the Scene in "Trumpetonshire"
  Tim Harford humorously investigates voting intentions in a fictional marginal constituency ("Trumpetonshire") to illustrate the difficulties pollsters and the public face predicting election results at the local level.
  • National polls don't translate directly to seats due to the UK's "first past the post" system.
  • The most common method, "uniform swing," can be imprecise, especially in unique local contexts ([03:01]).
• Uniform Swing Example
  • 2010 results: Conservatives 40%, Labour 34%, Lib Dems 16%, UKIP 6%, Greens 4% ([02:45]).
  • Applying national polling changes produces a neck-and-neck forecast for 2015, with both Conservatives and Labour estimated at 38% ([03:01]).
  • "All these polls come with a margin of error of around 3%...we simply don't know it." — Tim Harford ([03:48])
• Expert Opinions on 2015 Election Outcome
  • David Cowling (BBC):
    “Scotland wrecks everything...I'm genuinely perplexed.” ([05:09])
    • Predicts a swing toward Labour could tip marginals, but acknowledges unpredictability given the rise of parties like UKIP.
  • Ben Page (Ipsos Mori):
    “In this election, all sorts of weird things could happen...UKIP could get 10 or 15% of the vote ... and end up with only one MP.” ([05:48])
    • Highlights disparities between national vote share and actual MPs returned.
  • Deborah Mattinson (BritainThinks):
    "Anybody that's predicting anything is bonkers...if you're asking who I think is most likely to be Prime Minister...I think you have to say Ed Miliband...he's got more friends on the left." ([06:50])

Notable Segment:

• David Cowling on Trumpetonshire's prospects:
  "If Labour can mobilise its support amongst the biscuit workers in Chigley, and if UKIP penetrate behind the lace curtains of Tory Trumpton, there's a real chance this seat will change hands." ([07:26])

---

2. Debunking the "Reading Age of 11" Statistic

• BBC Moneybox Claim:
  Katie Morley (Daily Telegraph) says "the average person in the UK has a reading age of 11." ([08:23])
• Statistical Reality:
  • The OECD - cited as Morley's source - does not measure "reading age." Their literacy studies compare adult skills internationally, not against a child’s reading level ([09:00], [09:18]).
  • In a 2013 report:
    • England's adults ranked 22nd out of 24 countries on literacy.
    • The claim of an "average reading age of 11" is not directly supported by any OECD metric ([10:03]).
• Key Contextual Quote:
  "It's a measurement we see banded about a lot, but it's not a measurement used by the OECD." — Tim Harford ([10:03])
• Takeaway:
  The pension advice website referenced might genuinely challenge many readers, but claims about reading age are commonly misunderstood or misreported.

---

3. The "Nun Boom": Understanding Trends Behind Statistical Headlines

• Media Headlines:
  Reports claim the number of new nuns in England and Wales has trebled in five years ([11:41]).
• Actual Numbers:
  • 2009: 15 women began nun training; 2014: 45 ([12:09]).
  • This is an increase from a very low base: "You could fit all last year’s new nuns on a double-decker bus." — Tim Harford ([12:22])
• Proportional Perspective:
  • Up from 0.006% to 0.02% of adult women ([12:31])
• Immigration as a Factor?
  Sister Gemma Simmons clarifies that while some new nuns are immigrants, most are not coming directly to the UK with the aim of joining a religious order ([13:56], [14:03]).
• Historical Context:
  The total number of nuns in Great Britain has been declining for decades; about 6,000 remain ([12:54]).
• Pope Francis Effect?
  The "uptick" predates Pope Francis, so it's unlikely this is a major cause ([15:02]).

---

4. Can You Jog Yourself to Death? Examining the Danish Study

• Alarmist Headlines:
  News claimed "too much jogging as bad as no exercise at all" ([15:32]).
• Statistical Flaw Highlighted:
  • Among 36 “strenuous joggers” over 12 years, only 2 deaths.
  • Too small a sample size to draw meaningful conclusions ([16:28]–[16:36]).
• Author's Admission:
  • Dr. Peter Schnorr (study author) concedes:
    "You're completely right. We should have said we suspect that it is so, but we can't say for sure. Everybody makes some mistakes in papers." ([17:18])
  • The study did not provide reliable evidence that strenuous jogging was hazardous ([17:27]).
  • "It shouldn't have been misunderstood, because if you go into the statistics there, you will see confidence limits that are really bad...This is too thin." ([17:49])
• Key Insight:
  Sensational headlines often oversimplify or misrepresent the actual strength of statistical evidence.

---

5. The Mathematics of Optimal Stopping: The “Stop-and-Choose” Dating Dilemma

• Listener Correction:
  Dr. Carey Bunks corrects a previous episode’s math about how many people to “reject” before making a final dating choice ([19:32]).
  • Matt Parker had advocated rejecting the square root of the pool (e.g., first 10 of 100).
  • But Dr. Bunks points out the mathematically optimal strategy is about 1/e (~36%) of your options ([19:53]).
• Explained:
  • The ‘secretary problem’ suggests you should reject and sample the first 36 out of 100, then pick the next who is better than any prior ([21:01]).
• Matt Parker’s Defense:
  • The 1/e rule works for maximizing the probability of choosing the absolute best (“all or nothing”).
  • The square-root rule gives the best average outcome if “pretty good” is also acceptable ([21:51]).
• Quote:
  "If your goal is to get the best person possible in your range...use the 1/e method. But if you're OK with the second best, and want the best on average, use the square root." — Matt Parker ([21:51]–[23:00])

---

Notable Quotes & Memorable Moments

• On Polling Complexity:
  "Scotland wrecks everything, I think, in terms of trying to make a prediction, and now I'm genuinely perplexed." — David Cowling ([05:09])

• On Misreported Reading Age:
  "It's a measurement we see banded about a lot, but it's not a measurement used by the OECD." — Tim Harford ([10:03])

• On Small Number Statistics:
  "Only 36 strenuous runners were studied. That's even fewer than the number of new nuns in Great Britain." — Tim Harford ([16:36])

• On Nun Headlines:
  "You could fit all last year’s new nuns on a double-decker bus...you’d still have room for luggage." — Tim Harford ([12:22])

• On Correcting Mathematical Myths:
  "Dr. Bunk is absolutely correct. That is well known. That is the famous solution...published in 1961. However...the method I described is the solution to...I want to get the best person on average." — Matt Parker ([21:51])

---

Timestamps for Key Segments

• Polls in Marginals & Uniform Swing: [00:00] – [08:01]
• Reading Age Myth: [08:23] – [11:24]
• The "Nun Boom": [11:24] – [15:12]
• Jogging Study Debunked: [15:30] – [18:38]
• Maths of Optimal Dating / Secretary Problem: [18:54] – [23:07]

---

Conclusion

This episode of More or Less pulls the curtain back on the misuses of numbers, probabilities, and statistical rules that pepper UK news and culture. From the inherent imprecision of polling in unfashionable marginals and the true meaning behind reading age claims, to the reality behind nun "booms," overhyped jogging scare-studies, and the math of love, Tim Harford and his guests deliver humor, clarity, and thoughtful correction. The episode stands as a timely reminder: numbers, deftly wielded, can clarify our world—or obscure it just as easily.

Transcript

[Tim Harford]

Thank you for downloading this programme from BBC Radio 4. I'm Tim Harford. Hello and welcome to More or Less, the programme that looks for the truth behind the numbers in news, in life and yes, in politics. This week we'll be asking whether the nation really is reading no better than an 11 year old. And we'll return to the mathematics behind both optimal dating and jogging yourself to death. But first, with less than a week to go before polling day, More Or Less is visiting a key constituency that's been unjustly ignored by the mainstream media. Yes, Trumpetonshire. It's a key marginal constituency currently held by the Conservatives, boasting traditional Labour supporting areas near the Cressley biscuit factory in Chigley, alongside the staunchly Conservative market town of Trumpton. Loyal listeners may recall that I spent some time as treasurer of Trumpetonshire County Council myself, but I'm not involved in the parliamentary battle. I'm just curious as to who will be elected as Trumpetonshire's mp. Time for a quick ring around to take Trumptownshire's political temperature. Hello, Mrs. Minton here. Oh, hello, this is Tim Harford calling. I'm just phoning to find out who people are going to vote for next week. Do you do this for a living? Oh, no, no, I'm not a pollster, I'm a radio presenter. But Trumpetonshire is a key marginal and I'm just trying to work out if it's likely to stay Conservative. I hear that you voted Tory last time.

[Deborah Mattinson]

That's right.

[Tim Harford]

Do you think you'll vote for them again? Oh, I don't know. Oh, well, thank you, Mrs. Minton. Yes, goodbye. Let's try another voter. Hello, Clamp here. Oh, hello, Mr. Clamp. I'm just ringing round trying to get an idea of how people are going to vote next week. Oh, I see. Can I ask you who you're going to vote for? No, no luck. Whatever next. Oh, dear, this isn't very successful. We'll have to try something else. Unfortunately, there aren't many opinion polls that are done on a constituency level this election. Lord Ashcroft has been spending a lot of money polling individual seats, but for some reason he hasn't paid for one in Trumpetonshire. Most polls you hear reported in the news are national polls. They've taken a sample of the whole country and worked out the proportion of votes that will be cast across the country. But that doesn't translate directly into the number of seats because of the first past the post voting system we have in this country. So how do we work out who's going to win in Trumpetonshire? You may recall that in the 2010 election, the constituency of Trumpetonshire got the following results.

[Matt Parker]

Here are the 2010 shares of the vote in the Trumpingtonshire constituency. The Conservatives on 40%, Labour 34, the Liberal Democrats 16%, Ukip down at 6% and the Greens just behind them on 4%.

[Tim Harford]

The way most experts estimate it is by applying national polling data, a simple method is called the uniform swing. Let's look at the Conservatives. In 2010. They got a 36% share of the vote across the whole of the UK. As we record this programme, the opinion polls put the conservatives on 34% of the national vote. That's a loss of 2 percentage points point, and we can apply that to Trumpetonshire. That suggests the Conservative vote share in Trumpetonshire might fall from 40% by 2 points to 38%. Labour got 29% of the national vote in 2010 and opinion polls put them 4 percentage points up on that at 33%. So if we apply a 4 percentage point gain in Trumpetonshire, Labour would get 38%, exactly the same as the forecast for the Conservatives. It's clearly a tight battle here in Trumpet. Bear in mind, too, that all these polls come with a margin of error of around 3%. It might be that either the Conservatives or Labour are comfortably ahead and we simply don't know it. So this already looks like a tough seat to call. But is the uniform swing method really a sensible way to predict how Trumpetshire will vote? Among polling aficionados, there's a lively debate as to whether the uniform swing method is still useful. Some analysts rely on it, others are sceptical. There are always local considerations. In Trumpetonshire, the sitting MP has been in office for 20 years. It's always been a tight marginal, but somehow he's always scraped through all the years of New Labour hegemony. He managed to cling onto this seat. So maybe it's less marginal than it seems. But another factor is the rise of ukip. Camberwick Green is full of fruit pickers from Eastern Europe these days, and with their message about immigration and EU membership, UKIP has been very visible here in Trumpetonshire. As for the UK election as a whole, we spoke to a handful of polling experts to give us their predictions. This is David Cowling, editor at the BBC's Political Research Unit, the BBC's top expert on opinion polls.

[David Cowling]

Golly gosh. Well, I think before the explosion in Scotland, if I had to Bet the mortgage on it. I would have thought that there was the better chance that Labour would be the largest single party. But Scotland wrecks everything, I think, in terms of trying to make a prediction, and now I'm genuinely perplexed.

[Ben Page]

The average error on our final poll in the 2010 general election campaign was about 1.1% for each of the parties. And of course, the margin of error on each of these polls is around 3 or 4%. So we were quite happy with that.

[Tim Harford]

This is pollster Ben Page. He's the chief executive of Ipsos Mori, a polling organisation.

[Ben Page]

In this election, all sorts of weird things could happen. UKIP could get 10 or 15% of the vote. Let's wait and see. And end up with only one mp. The Lib Dems, having tried to get rid of First Past the Post, could be saved by it. Get 7% of the vote and still get at least 25 MPs. So all sorts of differences between votes and the final seats in the House of Commons. And again, depending on the day of the week, the greens are getting 7 or 8% of the vote. But again, probably only going to have Caroline Lucas as MP in Brighton if she holds on there. Very unlikely they would win in any other seat.

[Tim Harford]

While Ben Page takes a statistical approach, there are alternatives.

[Deborah Mattinson]

So we identified five different marginal seats, five seats that we thought would tell a very different story. They've been selected to reflect the constituency as a whole.

[Tim Harford]

This is Deborah Mattinson, founding director of Britain thinks her research company has been running focus groups gauging opinion during the campaign. We asked her if she could predict the outcome of the election.

[Deborah Mattinson]

No, it's too close to call. Anybody that's predicting anything is bonkers. I think, actually, if you're asking me who I think is most likely to be Prime Minister as things stand, I think you have to say Ed Miliband. Because if, as I would expect, you know, it's too close to call between the two parties and you end up, for the sake of argument, with a very, very close result. I think that that, in the end, means that Ed Miliband will be the Prime Minister because he has more options available to him, he's got more friends on the left.

[Tim Harford]

But of course, what we really want to know is which candidate's going to win in Trumptonshire.

[David Cowling]

Well, we couldn't have chosen a more interesting seat. I'm reminded of Disraeli's famous quote when he was campaigning there in 1868's general election. As goes Trumpetonshire, so goes the heart of England. Looking at the seat, it takes a 3% swing for labour to win there. Now, all the Polling suggests that IBDEM 2010 share will be down and some will go to Labour and that Ukip's share will be up to the discomfort the Conservatives. So if Labour can mobilise its support amongst the biscuit workers in Chigley, and if Ukip can penetrate behind the lace curtains of Tory Trumpton, there's a real chance this seat will change hands.

[Tim Harford]

David Cowling, of course you can find out on next week's program what happened. And we'll be looking back on some of the other predictions with David, Ben and Deborah. Lots of owl eared listeners emailed us at more or lessbc.co.uk, in shock at this claim which was broadcast recently on Radio 4's Personal Finance programme Moneybox.

[Katie Morley]

That gentleman seemed probably a lot more sophisticated than the average person in the uk. And we need to remember the average person in the UK has a reading age of 11.

[Tim Harford]

That was Katie Morley, the Daily Telegraph's senior personal finance reporter. She was on Moneybox to talk about the Government's new pensions website. It's meant to explain to people the options they have in getting hold of their pension. But Katie reckons it's too complicated and too difficult to use, hence her comment about reading ages. But was she right? Can the average person in the UK really only read to the same level as an 11 year old?

[Katie Morley]

So I have to confess that wasn't something that I'd planned to say. It was one of those stats that was sort of, you know, bumbling around at the back of my head and it sort of came out.

[Tim Harford]

It is a live radio program, we have to acknowledge.

[Katie Morley]

Yes, it was live and it basically came from the OECD report that I'd read a while ago.

[Tim Harford]

What does it mean to say that an adult has a reading age of 10 or 12 or 15? Is that actually something that the OECD said?

[Katie Morley]

That is something that they said that I interpreted it as. Anyway, I'm no education expert, but to me what that would mean is that materials that would be too complex for a child of that age to read, people would struggle with. So, for example, this pension wise website contains a lot of financial jargon, a lot of sort of words in there that I think any school child would really struggle with and, you know, any sort of average adult would struggle with as well. I mean, to be honest with you, I think I would struggle with the pension wise website. That's why that popped into my head.

[Tim Harford]

Katie Morley of the Daily Telegraph. And Katie, of course, is not the only reporter to talk about reading age. It's a measurement we see banded about a lot, but it's not a measurement used by the oecd. We spoke to one of its education experts, William Thorne, who told us that reading age just isn't a helpful way for the OECD to talk about literacy levels because the organization covers dozens of countries and a broad range of ages. But OECD data do allow us to compare countries. And adults in England didn't fare well in the most recent study. In 2013, the OECD studied 24 countries and the adults in England ranked 22nd, with Northern Ireland finishing only one place higher. The study didn't cover Scotland or Wales, who came top? Japan. So Katie got her numbers muddled. But she may have a point. If the new pensions website wants to be widely understood, it's going to have to keep things simple. As the general election looms, the nation's newspapers are scrutinizing what they see as the crucial issue facing the country today. A sudden surge in the number of nuns. Sister act how nuns are staging a surprise comeback in the uk Having none of it.

[Sister Gemma Simmons]

Growing numbers of British women take up the habit.

[Tim Harford]

Women becoming nuns hits 25 year high. Our nun correspondent, Hannah Moore joins me now. And can we agree right now, Hannah, to avoid terrible puns?

[Hannah Moore]

I think that's what everyone wants, Tim, although I am the best reporter, bar none. So this story is based on data from the Catholic Church showing that the number of new nuns in England and Wales has trebled in five years.

[Tim Harford]

And that might sound surprising in the context of a more secular society. There are fewer Christians in the UK and the nuns seem to be bucking the trend. So, Hannah, there are three times as many new nuns. Three times what?

[Hannah Moore]

Three times? Not very much, Tim. In 2009, 15 women entered training to become Catholic nuns. That's known technically as formation. Last year, 45 women did.

[Tim Harford]

So you could fit all last year's new nuns on a double decker bus. You can supply your own punchline here. I suppose if you have one, you'd still have room for luggage.

[Hannah Moore]

Yes. The proportion of adult women deciding to enter training and learn to be a nun each year has soared from 0.006% to almost 0.02%.

[Tim Harford]

So not only are those numbers difficult to say, they're also very, very small. Give us the context. What's the historical trend?

[Hannah Moore]

Overall, the number of new nuns was higher in the 1980s than it is today. But the basic story is that nuns have been dying or leaving their orders in greater numbers than new nuns are joining up. The total number of nuns in Great Britain has fallen by a in the past two decades and it's now 6,000.

[Tim Harford]

So this might seem like the end of a way of life, but still, why has the number of new nuns ticked up? Is it just a blip or does it have a cause?

[Hannah Moore]

One possible explanation is immigration. Catholicism is big. In Poland, for example, the National Office for Vocation, which is part of the Roman Catholic Church and which collects the numbers we've been discussing, says that this is part of the story.

[Tim Harford]

So immigrants from Poland coming to the UK to become British nuns?

[Hannah Moore]

Perhaps not.

[Sister Gemma Simmons]

I think one has to be careful. This isn't about people just coming over to England in order to join a religious order, because religious orders are subject to the Home Office regulations about immigration and visas, just like everybody else.

[Hannah Moore]

This is Sister Gemma Simmons. She's the director of the Religious Life Institute at Heythrop College at the University of London.

[Sister Gemma Simmons]

What we do sometimes get is people from, say, other parts of the European Union who've come over here, who've settled, and then from there think that they want to join a religion order and opt to join the order in this country as opposed to their country of origin. I belong to a network where people who are interested in religious life from all around the world can write in and say, your order looks like the profile of an order. I'd like to explore a bit more. And I suppose I get about 20, 25 applications a year from people from Nigeria, Indonesia, you know, all over Africa and Asia. I always answer them, but would normally say, look, you would be subject to visa restrictions. We cannot simply take people who are only able to stay here for a short time. Would you like me to put you in touch with a religious order in your own country?

[Tim Harford]

So is this the Pope Francis effect? You know, fashionable new Pope from Argentina,

[Hannah Moore]

first ever Pope from outside Europe, possibly in the uk. Although you could argue that the tiny uptick in new trainee nuns was already happening.

[Tim Harford]

Hannah Moore, thank you very much. You're listening to More or Less. Our email address is more or lessbc.co.uk. earlier this year, a study was published and it made headlines around the world, striking fear into the beating hearts of hardcore runners.

[David Cowling]

Too much jogging.

[Tim Harford]

As bad as no exercise at all. The researchers studied 1,000 healthy runners and almost 4,000 healthy non runners in Denmark over the course of 12 years. The results published in the prestigious Journal of the American College of Cardiology. And the headline conclusion was that light and moderate joggers, on average, live longer than people who have sedentary lifestyles. That result wasn't a surprise. But what was a surprise was the finding that strenuous runners do not live longer than couch potatoes. Strenuous runners were, broadly speaking, people who ran for four hours a week at a pace of seven miles an hour. We covered this story in our last series and at the time I spoke to Dr. Alicia White, who analysed the study for the National Health Service website. How many people in the strenuous jogger category actually died?

[Hannah Moore]

Two.

[Tim Harford]

Two?

[Hannah Moore]

Yes, two.

[Tim Harford]

That doesn't sound like a lot.

[Matt Parker]

No, it's not.

[Tim Harford]

Only 36 strenuous runners were studied. That's even fewer than the number of new nuns in Great Britain. So although the overall study was big, Alicia reckoned that the number of strenuous runners was just too small to draw conclusions. What's more, we don't know whether the two people who died had died from a heart attack, cancer or being knocked down by a bus. Well, now we've managed to get in touch with the author of the study, Peter Schnorr, a clinical cardiologist and founder of the Copenhagen Heart Study, which produced the data for this research. And you know what? He agrees. He concedes that they didn't have the evidence to say that strenuous jogging was bad for you.

[David Cowling]

You're completely right. We should have said we suspect that it is so, but we can't say for sure. Everybody makes some mistakes in papers.

[Tim Harford]

A complete and frank admission that the evidence wasn't strong enough to support his research paper's conclusion. And Peter Schnorr says experienced readers of research papers would have realised that. He says it was obvious from the statistical analysis that you couldn't have confidence in the findings that strenuous joggers had the same average life expectancy as those with a sedentary lifestyle.

[David Cowling]

It shouldn't have been misunderstood, because if you go into the statistics there, you will see confidence limits that are really bad. And if you normally read papers, you could say, ah, this is not good statistically. This is too thin.

[Tim Harford]

But most journalists around the world weren't going to be delving into the statistical tables. They were going to take their lead from the researchers conclusions. So does Peter Schnorr regret all those shock, horror headlines that resulted? Does he worry he might have put people off running?

[David Cowling]

I don't think so, because when you have these deaths, you always have deaths in marathon. And so on, if these people had been investigated because they're always a little ischemic heart disease and so on, it could have been prevented.

[Tim Harford]

You're listening to More or Less with me, Tim Harford. Now, from one correction to another. Loyal listeners may remember Matt Parker, author of Things to Make and do in the Fourth Dimension, dishing out mathematical advice on our program ahead of Valentine's Day this year.

[Matt Parker]

When you're looking for a life partner, as you're interviewing people, or I believe dating is what it's commonly called as you're going through the candidates. If you meet a candidate and you assess their qualities and then you reject them later on, it's very hard to go back and say, look, I've changed my mind. It turns out you were the best I'm ever going to do. But on the flip side, if you settle down with someone straight away, you have no idea what the general quality of possible candidates was. So the question is, how many people do you reject to get a sense of the standard? And at what point do you stop rejecting and you start choosing? And that is called the optimal stopping problem.

[Dr. Carey Bunks]

Dear More or less, with regards to your program on the maths of dating, I would like to point out that your guest, Matt Parker, has made a serious error in his mathematical approach to finding the best life partner.

[Tim Harford]

Oh, dear. This is loyal listener Dr. Carey Bunks, who wrote to more or [e-mail] to express his concern.

[Dr. Carey Bunks]

Matt suggested that if you could date a total of 100 people, you should sample and reject the square root of that number. That's to say, the first ten, and then subsequently choose the first one that was better than any in that first initial group. Unfortunately, his advice is both erroneous and highly suboptimal. In fact, to produce the best result, the sample size should not be the square root, rather the inverse of the natural number E. That's to say 36%. So for 100 potential dates, the optimal approach is to sample and reject not the first 10, but rather the first 36. I believe Matt would agree that this is not only a lot more fun, but also a superior result. Regards, Dr. Cary Bunks.

[Tim Harford]

There's no answer to that, except to haul Matt Parker back into the studio to explain himself. Matt, you are back here. What do you have to say for yourself?

[Matt Parker]

So you've had a complaint that my description of what is called the secretary problem is apparently suboptimal, and it's called

[Tim Harford]

the secretary problem because it was originally about choosing secretaries. Our loyal listener has complained. Dr. Bunks Dr. Bunk, debunking you, has complained that actually you had suggested interviewing and or dating the first 10 out of 100, the square root. Correct. Before then choosing the best. The first person who is better than any of that 10.

[Matt Parker]

Yes.

[Tim Harford]

And in fact it is well known that you should be interviewing and or dating 1 over E. What is E?

[Matt Parker]

Yeah, well, don't get me wrong, a big fan of E. It's one of these numbers like PI in mathematics, one over E is about 36, 37%.

[Tim Harford]

Which if you had potentially 100 dates, interview the first, I think 36, 36, 37. And then you pick the first person who comes along who's better than the first 36 or 37. So what if you. What's your excuse?

[Matt Parker]

Dr. Bunk is absolutely correct. That is well known. That is the famous solution to the problem where you interview one E of your candidates, then pick the next best one. And that was originally published in 1961. So the Secretary problem was being thrown around in the 1950s when hiring secretaries was at an all time high. However, since then research has continued. The method that Dr. Bunk described, the 1E, that is the method you use if your goal is to get the best person possible in your range of options. But you don't care what you get if you don't get them, so it's all or nothing. The method I described is the solution to. If you tweak the problem slightly to be I want to get the best person on average. And so you go, look, I'd love the best person, but if I don't get the best person, I'm pretty happy with the second best as long as I get the best person on average. And so Instead of assigning 100% to the best person and zero to everyone else, you assign them weighted values on how good they are. And so Dr. Bunk is completely correct for that particular situation. If we were living half a century ago, that would be the best solution. But things have moved on.

[Tim Harford]

Well, Matt Parker, thank you very much. You are nothing if not thorough and nothing if not recent in your use of research.

[Ben Page]

Thank you.

[Tim Harford]

Thank you to Matt Parker. His book is Things to Make and do in the Fourth Dimension. And thank you also to our loyal listener, Carey Bunks. That's all we have time for, but if you want to tell us about any statistical curiosities, please get in touch via More or [e-mail]. our website is BBC.co.uk more or less. And next week the program will be broad live on Friday afternoon when we'll be trying to make sense of the election results. Until then, goodbye. And just a reminder that you can download more [e-mail] Radio 4.