← More or Less
More or Less
How extraordinary is Ye Shiwen?
00:27:42
There was controversy this week after Ye Shiwen, a young Chinese swimmer, won the 400...
Loading summary
Transcript97 lines
- [00:00]
A
Hello and welcome to More or less. If Radio 4 is the Olympic Games, we're the beach volleyball. This week, we'll be asking whether the news that homelessness has risen by 25% means that homelessness has risen by 25%. We'll be looking for the most dangerous place in crime drama. And TV's Yan Wong will be here as promised, to answer one of your numbersome questions. What's it all about? Well, here's a clue. Can you identify this musical introduction? What about this one? And this? We'll give you the answers later in the program. But first, the Olympics. And there was controversy this week after Yixiuan, a young Chinese swimmer, won the 400 meter individual medley in fine style. But then the head of the U.S. swimming Coaches association, who incidentally isn't part of the U.S. team, called that performance disturbing. And the swimmer was forced to deny that she'd been taking. Well, Wesley Stevenson's here. Wes, let's just remind ourselves what happened in that race.
- [01:05]
B
So at first, no one was dominating. So through the butterfly and the backstroke
- [01:10]
A
stages, then the third stage, the breaststroke, things started to take shape. The American swimmer Elizabeth Beisel started to pull away.
- [01:17]
B
And no one had really mentioned Yeshua at this point.
- [01:20]
A
Yeah, the commentators had said that Beisel had slammed the door on the right.
- [01:23]
B
Slammed the door. She was that far, far ahead.
- [01:25]
A
And then. Two lengths to go.
- [01:27]
B
Two lengths to go. As the competitors started swimming freestyle, it looked like this happened. It was extraordinary.
- [01:39]
C
Yes.
- [01:39]
B
Yuan just roared past Elizabeth Beissel and she won easily.
- [01:42]
A
She broke the world record. So well done. Yes, you won. Or is there, as this particular US coach has hinted, something more sinister going on? Now, where's Yeshuan? Said that the criticism's just a case of sour grape. So let's look at the charges against her. At first, she beat her personal best by five seconds.
- [02:01]
B
Yeah, she did. But you've got to remember Ye Shiwan is very young, so you'd expect some dramatic improvements. The Australian swimming legend Ian Thorpe said this week that he beat his personal best by five seconds when he was a teenager. And also if you take Ye's fellow swimmer in the 400 meters individual medley, that's the Australian Stephanie Rice. She beat her personal best in 2008 by six seconds when she set the world record in the Aussie Olympic trials. And no one shouted drug cheat at her. Ye's still growing. I mean, she's 12 centimeters taller than she was two years ago when she was 14. And so you'd expect her times to improve.
- [02:35]
A
Now, the Second issue is that she smashed the world record. She beat it by over a second.
- [02:39]
D
Yeah.
- [02:40]
B
Let's take Stephanie Rice again. When she beat her world record, she beat it by a wider margin back in 2008. So, I mean, it's impressive, but it's not that remarkable.
- [02:49]
A
Now, the final element in the case against Yeshuan, over the final length, this last 50 metres, she was swimming faster than Ryan Lochte, who won the equivalent men's race.
- [03:00]
B
Yeah. This comparison with Lochte probably isn't that telling. It's not the first time that Lochte's been slower than a woman over the last leg of that type of race. In Beijing in 2008, when he got the bronze, he was slower than the Italian Alessia Felipe by more than half a second, and she only came fifth.
- [03:17]
A
And it's probably worth pointing out he didn't actually swim the whole race slower than Je Shi Wang.
- [03:21]
B
No, no, absolutely not. I mean, he raced 23 seconds faster than Ye overall. He just paced himself very differently. He was well ahead of the race.
- [03:30]
A
He was well ahead by the time he reached the last leg, and he was comfortably in control of the race, whereas Ye had a lot of work to do to catch up with the leader.
- [03:37]
B
Yes. And Lochte also had three races the next day, so he probably had a reason to kind of save a little bit of himself for those. And this is the problem. We've. I spoke to Dr. Ross Tucker from the Sports Science Institute at the University of Cape Town, and he says the comparison doesn't necessarily mean very much.
- [03:54]
D
She swam it as fast as Lochte. Yes, but Lochte didn't swim it as fast as some of the other men in that same race. And so therefore, his performance compared to the best men in that race for that discipline or for that leg was maybe not as impressive. Other people are saying, ah, but hang on. If you're going to look at Yi's final lap, then we must also look at Rebecca Adlington when she won the 800 meter title because she was even faster. So the point is that when you start to look at performance in this way, analyzing performance and trying to prove doping is such a futile task.
- [04:27]
B
And Rebecca Adlington was, of course, the British 400 and 800 meter gold medalist from Beijing and swam faster than both Lochte and Yexiuan in the final leg of her 800 meter race at the World Championships last year.
- [04:40]
A
Just thinking about the way that Yeshiuan paced herself, it's actually an amazing thought. If she was able to improve on the way she allocated her effort, the way she paced herself, she might actually go faster.
- [04:51]
B
Yeah. And this is what Dr. Tucker thinks.
- [04:53]
D
I would say, based on what we know about the end spurt and how many seconds you can gain at the end compared to lose in the race, she's probably worth another second and a half to two, maybe even more. If she got the pacing strategy more even. In other words, if she allocated her energy more across the four strokes, I think she could probably go even quicker.
- [05:14]
A
So we've covered all of these charges, all of these allegations, and actually what we're saying is it was impressive, but there's nothing about it that's really suspicious. Was there anything about the race that raised suspicions?
- [05:24]
B
Well, there are a couple of things that stick out. Now, Ross Tucker had this observation.
- [05:30]
D
Lochte is swimming that final freestyle leg pretty much as you would expect. I worked out that most of the competitors at that level are doing the freestyle about 16 to 20% slower than the equivalent world record for the leg and the same for the females. So in the last world championships, the winner of the individual medley and in fact the top five are doing the final 100 freestyle about 18 to 20% slower than the 100 freestyle world record, whereas Yixiwen was at 12% slower.
- [06:01]
A
Okay, so in that sense, it was an exceptional performance.
- [06:04]
B
Yes. And there's another measure you can look at as well. Reza Noubri is a maths professor at Bloomsbury University in Pennsylvania in the usa. He's applied something called threshold theory to world records in the past. Now, what he does is to take the previous best performances and the amount of time that elapses between them and establishes what records might be possible with a 90% confidence. Now, we gave him the data for the best times in the 400 meters individual medley before Yeshuan's race, and he came up with this estimate.
- [06:34]
D
Yeah. Based on these 10 measurements that you sent me, the lower band should be 4 minutes, 29 seconds and 100th of seconds. She did it in 4 minutes, 28 seconds and 4300th of seconds. So in that sense, you may call that exceptional.
- [06:57]
A
So this race was extraordinary. But then sometimes we do get extraordinary athletes.
- [07:02]
B
Yes. And when Professor Nubri ran the data on the men's 100 metres before Usain Bolt first broke his record, he said that with 90% confidence that the next record wouldn't exceed 9.62 seconds. Of course, now we know that Bolt's record stands at 9.558 seconds. So you're right. We sometimes get extraordinary athletes and this is why Ross Tucker says that while we should ask questions, we should be careful about what we assume from the performance.
- [07:30]
A
I suppose you can understand why people are suspicious.
- [07:32]
B
Yes. I mean, there's a cloud that's hung over Chinese swimming ever since 32 of their swimmers tested positive in the 1990s and only seven months ago another teenage swimmer tested positive for the drug epo. And the reason that people are asking questions is that they failed to in the past when there've been other drug problems. And it's not just true of swimming.
- [07:52]
E
No.
- [07:52]
A
There's athletics for example.
- [07:54]
B
Well, yeah, I mean athletics is a great example, particularly women's athletics. We know that there was a huge problem with drugs in the 1980s and this means there's a problem with a lot of the records set in that era. As Ross Tucker says, it's unlikely they'll ever be broken.
- [08:07]
D
The problem now is that the sprint records in particular, 100, 200, 400, the 800 meter world record is the oldest one on the books. It's 1983, was set by a Czech athlete, Jamila Kratos Vilova. And the reality is that no female is coming close to them. We just don't see 800 meter woman running in the range of 153 anymore. So the female records are largely out of bounds or out of play because the effect of the doping in the 1980s was just to push them beyond the realms of sort of belief.
- [08:42]
B
Florence Griffiths Joyner, who was dogged by allegations of doping up until her premature death at the age of 38. Her 100 meter record is 10.49. The closest anyone has got to that in the 24 years since is 10.64 seconds, which is a huge gap in sprinting terms.
- [08:59]
A
And her 200 meter speed is amazing too.
- [09:03]
B
Yes, it's more than 3/10 of a second faster than the current fastest time set by Alison Felix last month. That was a time 21.69 seconds. The 400 and 800 meter records are equally as remarkable in terms of how far ahead they are of the current athletes performances. I mean, both are held by Eastern Europeans who were competing in the 1980s, although neither tested positive for drugs and both are denied doping allegations. The 1500 meters is interesting too. This record is held by a Chinese athlete who was coached by the controversial Chinese coach, Ma Junren. If you compare the record to the nearest non controversial time, the Chinese runner is five seconds ahead. Other athletes Ma coached were banned in the 1990s after positive drugs tests, although not this particular record holder. Another Chinese athlete from this era, Wang Junzhai, also holds the 3,000 meters and 10,000 meters records. And again, they're a long way ahead of the current athlete's performances too, although again, she's always denied using drugs.
- [10:05]
A
So what we're saying is, in women's athletics, there are a huge swathe of records that just appear to be completely out of reach by contemporary athletes who we hope are clean. There are a lot of questions to raise, but not particularly about Yishuan. Thank you, Wes. Earlier this week, the BBC website brought us the following alarming claim about homelessness.
- [10:27]
B
Homelessness rose 25% in England over the past three years.
- [10:32]
A
Of course, you may be wondering, is that true? And the simple answer is, yes, it is. But here's where it gets interesting. That word homeless, in the words of the great Inigo Montoya, I do not think it means what you think it means. The figure is referring to the statutory homeless in England. These are households who apply to their local council for help with housing and whom the council has a responsibility to help. Suzanne Fitzpatrick, professor of Housing and Social Policy at Heriot Watt University, explains who makes that list and and who doesn't.
- [11:04]
F
Those priority need groups are primarily Families With Dependent Children and also households which contain a vulnerable adult, usually because that adult has some sort of a health problem or is perhaps an elderly person. In England, most single homeless people are not accepted as statutory homeless, even if they apply to the local authority. Unless they can demonstrate some specific vulnerability, then most single people or couples without a child would not be accepted as statutory homeless. The position is entirely different. In Scotland. For example, everybody who is unintentionally homeless is now accepted as statutorily homeless in Scotland and is entitled to rehousing by a local authority. The point is that most people who are found statutory homeless are not literally sleeping on the streets. That is a tiny minority of people. It's usually that you are in some very acute form of housing need which is recognised in legislation and accepted by the local authority to which you apply.
- [12:05]
A
When I hear the word homeless, I automatically think of rough sleepers. And even if I don't, the BBC story gave me a helpful hint by serving up a photograph of two people lying in the open air in sleeping bags. But most rough sleepers aren't covered by this definition of statutory homelessness. We see this quite a lot on more or less an intuitive sounding label that actually describes something quite different from what the uninitiated might have in mind. It's also interesting to look at the Longer term trend. As we've heard, there has been a 25% increase in the number of statutory homeless households households over the last three years. But 2009-2010 was the trough of a long fall in the number of statutarily homeless households. Last year's number is higher than in the last couple of years, but it's lower than at any year in the decade running up to 2008 and 2009. Homelessness peaked in 2003, 2004 at 135,000 households. And that's almost three times the current level and against a benign economic background. So why are people homeless and why are the numbers rising? Again? Intuitively you'd assume that because of the credit crunch, the recession and government austerity, it's because people fall behind in their rent or their mortgage payments, but perhaps not.
- [13:19]
F
The key driver of homelessness is housing affordability. It's the ability of people and households who have to find accommodation not being able to do so within their means. Both because people are becoming economically weaker at the bottom end of the income spectrum, but also because of market is continuing to tighten. So housing affordability is the key driver of homelessness, not actually forced exits from the housing market. It's actually frustrated entry into the housing market that is the most important factor in driving up homelessness. Mortgage arrears and repossessions are not the key driver of increases in statutory homelessness. And they never have been. Not in the last major housing recession either. Of course, some people experiencing mortgage repossession and rent arrears do become homeless, but they are a small minority. It's 4 or 5% of homelessness acceptances are due to those causes. Relationship breakdown and young people being asked to leave by family, friends, parents are far, far more important as immediate causes of statutory homelessness.
- [14:23]
A
According to Professor Fitzpatrick, while the numbers of people who are statutory homeless fluctuate, the proportions of people in that situation, for different reasons, actually stay fairly constant. And repossessions are a pretty small percentage. One more thing, these homelessness numbers aren't just a measure of how many people find themselves in trouble. It could also be a reflection of how local authorities deal with them. They have guidelines to judge who's homeless, but over the years, are they interpreting them differently? Are prevention schemes having an impact on these numbers? It's hard to tell. The situation is worrying. Other estimates for homelessness, for instance, counts of rough sleepers, are also on the rise. And as changes to housing benefits and other cuts in funding come through in the next few years, this rise might not be a blip. You're listening to more or less in association with the Open University now. At the start of the series, TV's Yan Wong volunteered, begged us to be allowed to answer the most interesting and creative questions you could throw at him. And Anna Mouton from South Africa gave us this. I'm always amazed by the number of songs one can recognize on hearing the first, second or two of music. Since music has a limited number of building blocks and there are mathematical rules for how these can be combined to sound musical, is it possible to calculate the total number of potential opening bars? Surely it must be finite. Jan, before we get to the answer, a little test for you. Ok, what were they?
- [15:53]
E
The first one was Thriller, wasn't it? The trouble is my memory's not very good, so I do not remember the other two.
- [16:00]
A
I haven't heard it either. I thought the first one was Thriller. I had no idea what the second one was. The third one sounded like Another One Bites the Dust by Queen. Okay, I've got it now. Thriller. Great Balls of Fire, I think. And Another One Bites the Dust. So, Jan, it is an interesting question. How on earth did you go about answering it?
- [16:23]
E
Well, I think actually there's something going on here with what we just listened to, or where we're not just listening to the notes, we're listening to the sort of quality of the sound. And I think that's what most people actually do. And it sounds like the questioner was asking, really, if you just take the notes and the speed at which the notes are played, and that's a bit easier, you can try and sort of enumerate out the mathematical possibilities. There are a lot of them, really, like, really quite a lot, because it
- [16:48]
A
seemed to me, I guess, I just had in mind the number of keys on a P. Well, you know that there aren't that many. Say you're going to play three or four notes. I mean, obviously you get many thousands of possibilities, but clearly my intuition is off here.
- [17:00]
E
Well, there's a thing called the combinatorial explosion. So the more possible little choices you make, they add up to enormous amounts. So, for example, if you just took the rhythm and you imagine you could, say, fit 32 sort of beats into two bars, then you're talking about about 4 billion rhythms that you can fit in without even throwing notes in. If you throw notes and chords in, you get numbers that are well over the number of atoms in the observable universe.
- [17:28]
A
So just tapping something out with a, you know, with a Pair of drumsticks. How many did you say there were?
- [17:34]
E
4.3 billion.
- [17:36]
A
So 4.3 billion rhythms in two bars?
- [17:38]
E
In two bars. If you can do it reasonably fast.
- [17:40]
A
And then. Well, 32 a second is pretty fast, but fine, we'll let you have that then. Once you start incorporating the possible pitches, then presumably it gets much, much bigger.
- [17:49]
E
Yeah, absolutely. You start getting into more than the stars in the observable universe. And then if you allow chords as well, there's lots of different sorts of chords you can play and the number rapidly gets huge. But the thing is, of course, you don't. Real music doesn't explore all that space.
- [18:04]
A
That was going to be my question, because you can't just thump several different notes on a piano and call that a tune. I mean, actually, the number of different combinations that composers must have in practice explored must be much smaller.
- [18:17]
E
Much smaller. The problem is the listener thought there might be sort of mathematical ways you can narrow that down. I don't think there are any. Not any strict rules. And if anyone did invent one, I'm sure some composer would just try and break it immediately. And so there's sort of rules of thumb, but there isn't any strict thing that allows us to say, this noise is music, this noise is not music. And so it's difficult to know how to cut it down to fewer possibilities.
- [18:37]
A
So that's the kind of the pure maths approach to this, the combinatorial approach. Do we have any empirical evidence? I mean, there are big databases of songs, right?
- [18:44]
E
So, yes, there are. And it's rather nice because the very first bit of a song, it's called the music musical insipit, if you know Latin, there are databases of that because those are sort of copyright free, and there are maybe a couple of hundred thousand of those in various databases around the place. And people have tried sort of mining these because some of the listeners might know websites where you can type in tunes or sing tunes, hum a hum a tune, and it tries to recognize what the tune is.
- [19:10]
A
Okay, so what do we learn from these databases?
- [19:12]
E
So it looks like as long as you take into account both the sort of pitch and the rhythm, then you can identify a large number of tunes. You can narrow it down to just one tune within about eight notes, probably more like six to eight notes. So it's quite quick.
- [19:27]
A
Okay, so the pure maths approach to this, the combinatorial explosion, says there's basically no constraints. They're almost infinite. Number. Not infinite, but almost infinite. Yep. Okay, edit that out, please. Producer There's a very large number, a very, very, very large number. For all practical purposes, infinite, because far more than the number of atoms in the universe. If you take a database of songs, you can distinguish one song out of 10,000 with just 6, 7, 8 notes. Nevertheless, I mean, there are songs, pieces of music that sound an awful lot like each other.
- [19:59]
E
There are Baa, Baa, Black Sheep, Twinkle, Twinkle Little Star have the Same bar.
- [20:03]
A
Your musical taste is remarkable.
- [20:06]
E
I have young children and so it's a bit surprising. There are definitely these constraints on there. You can use those first sections to try and work out how constrained the music is compared to sort of just allowing any old notes to go. And it turns out that really real songs do explore quite a lot of the space. The predictability of just a random set of notes. They're very unpredictable.
- [20:28]
A
Yeah, if you played me four notes and it's totally random, I have no idea what the fifth note's going to be.
- [20:32]
E
The predictability of a song is not that much more predictable from these measures of the initial insipids, the starts of songs. And so it seems like you're exploring reasonably a reasonably large amount of the mathematical sort of space when you're writing songs.
- [20:47]
A
The amount of music out there is almost unlimited. And yet we still have X Factor. Jan Wong, thank you very much.
- [20:54]
E
Thank you.
- [20:55]
A
Yan Wong, never wrong. If you have a question for Jan, please let us know more or less@BBC.co.uk and we'll try and get one more answer out of him before the end of the series. Earlier in the week, housing expert Henry Pryor popped up on the Today programme with.
- [21:12]
E
And if you travel between here in central London, for example, and Manchester, the average house price falls by £3,500 for every mile you drive.
- [21:23]
A
A number of you emailed to ask whether that could possibly be true. We called Henry Prior, who told us he'd compared average house prices in Salford with a particularly expensive part of London, Kensington and Chelsea. We checked his sums and we found the real figure is even harder to believe. It's more than £4,500 per mile. Kensington and Chelsea prices are just over a million pounds on average, and in Salford they're less than £100,000. So that's over £900,000 difference in just over 200 miles. There are a couple of complications here. These averages might be comparing Kensington Mansions with Salford Bedsits, or alternatively, Kensington Bedsits with Salford Mansions. We've not looked at the property mix. We think the story is clear enough, but of course the gradient isn't really £4,500amile. If you start in Kensington and travel eight and a half miles in the Salford direction, you'll find yourself in Brent with average property prices of £320,000. We reckon that the property price gradient between Kensington and Brent is £85,000 per mile, and then from Brent to Salford It's a mere £1,100 per mile. Henry Prior, incidentally, seemed a little unnerved to receive a call from us. He later emailed to say that hearing from More or Less was like spotting a police car in your rear view mirror. It doesn't matter whether you've committed a statistical crime, he said, you still feel guilty. If only we had such a chastening effect on politicians. Our More or Less website is new and improved with podcasts and articles, and it's available at the old address of BBC.co.uk more or less. And of course, you can always get in touch via More or less@BBC.co.uk Speaking of which, the More or Less inbox is always bursting with insightful and informed observations about the great issues of the moment. This isn't one of them.
- [23:17]
C
Would More or Less consider doing a segment on the popular TV mystery series Midsommar Murders? A great number of people die per episode. I've wondered over the history of the series how the number of dead bodies per year in midsummer compares against the total number of murders in the average English county.
- [23:41]
A
That email from loyal listener Dee Yin Lu and Midsummer's Inspector Barnaby does have to contend with a pretty high body count. But how high? And can we learn anything useful from investigating it? Well, probably not, but let's do it anyway. Charlotte MacDonald, over to you.
- [23:56]
C
Well, De Hyun Liu isn't the first person to notice the high murder rate in Midsommar. There are on average around 2.6 murders per episode and around 8 episodes a year. Midsommar is a county, so if we assume that Inspector Barnaby investigates every murder in the county, 21 people get bumped off there every year. Now, being a fictional place, I don't know the exact number of people who live there, but one of the counties used for filming is Oxfordshire, so if we assume its population is the same as that real county, we get to an annual murder rate in midsummer of 32 per million.
- [24:29]
A
Right? And that's high, I assume?
- [24:31]
C
Yes. For example, Thames Valley Constabulary, which polices the real county of Oxfordshire, reports a murder rate of ten in a million. Even London, the most murderous place in Britain, has a rate about half that
- [24:42]
A
of Midsummer that's interesting, though. Midsummer's dangerous, but it's not absurdly dangerous. It's not like the murder rate is a thousand times that of Oxfordshire, which we might have expected. And presumably then there are places in the world that do have a murder rate like Midsummer's.
- [24:55]
C
Yes, I was surprised too. And yes, Midsummer's murder rates are not beyond the bounds of possibility. Chile and Latvia have murder rates about on a par with Midsummer. And in Honduras, the murder rate is a massive 910 per million, almost 30 times the rate in Midsummer. It's a pretty horrifying number. So let's return to the world of fiction for a moment and think about another question. Where's the most murderous place in TV crime drama? It turns out, after frankly, a lot of work, that Midsommar is a peaceful place compared to Cabot Cove in the us, which is the fictional hometown of Murder She Wrote's Jessica Fletcher.
- [25:41]
A
I hope you're going to tell me they are not still churning out Murder She Wrote.
- [25:44]
C
Well, I. You talking about. It's a great program.
- [25:46]
A
Yeah, yeah, come on, keep going.
- [25:48]
C
But while they were, the redoubtable Jessica was solving over five cases a year in Cabot Cove itself. Sometimes she went on the road, but on average, five murders a year took place on her doorstep.
- [25:58]
A
But actually, that's a pretty small caseload compared to Inspector Barnaby.
- [26:02]
C
Well, yes, but cbs, who broadcast the program, told the New York Times once that Cabot Cove's population is three and a half thousand. That would put it ahead of Honduras and way ahead of Midsummer, with a fearsome murder rate of 1,400.
- [26:17]
A
This all puts Honduras into perspective, though. Cabot Cove's murder rate is ridiculous. It's five people in this tiny place and it's actually not that far ahead of Honduras. Charlotte, just one more question. I don't suppose you've worked out the murder rate in late 1970 is TV Los Angeles, have you? I mean, they were enough to keep Quincy, Jonathan and Jennifer Hart, Perry Mace and Jim Rockford and Columbo busy all at the same time.
- [26:57]
C
No.
- [26:57]
A
I see. Before we go, we asked you at the start of the programme to identify these three musical openings. Easy. Perry Como's Magic Moments, Satisfaction by the Rolling Stones, and of course, Frank Sinatra's New York, New York. And we're leaving this week. It's all we have time for. Next week we return to the scene of a terrible economic crime, Trumpton. Until then, goodbye. More or Less was presented by Tim
- [27:32]
F
Harford, the Financial Times undercover economist. The producer was Richard Knight and the editor was Richard Varden.
- [27:39]
C
More or less is made in association
- [27:41]
F
with the Open University.