The BBC has done a poll about unread books, and found some results that at first might appear surprising:
Some 35% of those who bought or borrowed Vernon God Little, DBC Pierre's story of a US high school massacre, admitted not finishing it.
The figure was 32% for the fourth instalment in the Harry Potter series, while 28% said the same for James Joyce's Ulysses, third on the list.
In reality, though, I think this isn't terribly surprising, because it's a variant of what I think of as the Labrador Retriever Problem. It's got that name because when I went to buy homeowner's insurance before we bought the house, they told me that there are three breeeds of dogs that they won't insure: Pit Bulls, Rottweilers, and Labrador Retrievers.
"That's crazy," you say, "Labs are such sweet dogs." And, in fact, I said that to the agent. Who pointed out that yes, they do tend to be nice dogs, but they're also the most popular dog breed in America.
The chances of any given Labrador Retriever biting someone and leading to an insurance claim are pretty low, but there are so many Labs out there that the insurance companies have had to pay out for thousands of dog bite cases involving Labs. As a result, they will no longer provide bite coverage for people who own Labrador Retrievers.
"Yeah, but what does that have to do with the book thing?" you ask. "After all, Rowling was unfinished by a higher percentage of people than Joyce."
True, but I think the same effect is in play. The Harry Potter book, being a recent pop-culture mega-hit, has probably been bought by a much larger number of people than Ulysses, which means that there are more people out there who have the chance to not finish Harry Potter and the Goblet of Fire than have the chance to not finish Ulysses.
If you provided a thousand Britons with copies of both books, a substantially higher percentage of them would actually finish the Rowling than the Joyce, but that's not what they did. It's not surprising that a higher percentage of people who thought they might want to read Ulysses finished it than those who thought they might want to read Harry Potter and the Goblet of Fire, because the latter group includes, well, just about everyone, including people who don't normally read very much. Those people are unlikely to ever think "Boy, I could really go for some James Joyce right now," but might buy J.K. Rowling just because everybody else is.
On a related note, you might well ask "Why is the fourth Harry Potter book, which is actually pretty good, the one that goes unfinished, as opposed to, say, the fifth book, which was dreadful?" It's because the fourth book is the first eight hundred page Harry Potter book. The previous ones were pretty short, and thus much easier for people to finish, even if they don't read regularly. The fourth book is the first one that required a large investment of time, and is where you'd expect people to start not finishing them. And, of course, people who don't finish the fourth book aren't likely to buy the fifth one...
(I could be wrong about this, of course-- I haven't looked into the methodology of the survey to see if there's some quirk in the way they asked the question that would explain the result. This was my immediate thought on seeing that list, though, and it seems like a good explanation to me.)
(Credit where due: I saw the BBC link via Peg Kerr.)
- Log in to post comments
I started Ulysses three times, but gave up three times as well. I'm afraid I might be skewing the statistics. ;)
Not sure I buy this explanation yet. Check out Clifton's dog bite report HERE. Now this report includes a list (granted by no means complete and may not accurately reflect insurance actionable bites) of serious dog bite incidents related to breed. These numbers are totals, and so already include proportionality effects.
We can see that yes, Pit Bulls and Rotweillers are breeds apart, accounting for far more maimings and deaths than all the other breeds combined. But the Lab is by no means close to them, and is not even a distant third. The Bull Mastiff, the Chow, the German Shepard, the Huskey and wolf hybrids seem to top Labs, even taking into account their vastly inferior popularity.
The first three HP books could be read in one sitting (I know - I did it). The fourth is the first one that requires at least two sittings (I know, it took me two nights: I read the first four during five nights of spring break just when the 4th came out). And once you put the book down it is more likely you will never pick it up again.
Amen to the comment about HP5... dreck, dreck, dreck. At least Rowling redeemed herself somewhat with HP6.
We can see that yes, Pit Bulls and Rotweillers are breeds apart, accounting for far more maimings and deaths than all the other breeds combined. But the Lab is by no means close to them, and is not even a distant third. The Bull Mastiff, the Chow, the German Shepard, the Huskey and wolf hybrids seem to top Labs, even taking into account their vastly inferior popularity.
That's interesting.
As I said, this is the explanation I was given by the insurance agent, who specifically listed Labrador Retrievers as the third breed they won't insure. Maybe people are more likely to file insurance claims against people who own Labs than they are to file claims against people who own wolf hybrids?
I'd think the vast number of people who have Labs and would be paying premiums on them would make up for the relatively small percentage of them that require claims to be paid out. It's the negative outcome probability that should matter for insurance, not the number of negative outcomes.
I liked HP5 better than HP6 personally. HP6 felt like it was just setting things up for 7. Like the middle Matrix.
The explanation for the Labrador Retrievers doesn't sound right to me either. I could imagine that their popularity means that they're more likely to be bought by people who don't really know how to train or take care of them, leading to a higher incidence rate. The facts do seem to get in the way of that reasoning, however.
The chances of any given Labrador Retriever biting someone and leading to an insurance claim are pretty low, but there are so many Labs out there that the insurance companies have had to pay out for thousands of dog bite cases involving Labs. As a result, they will no longer provide bite coverage for people who own Labrador Retrievers.
That's utterly and entirely daft.
That stat seems a little faulty to me too. What would be the difference between the amount of claims from the Lab population, and an equal sized population of multiple breeds with the same average of attacks? Going with this logic, would they deny coverage for the most popular car as well?
I think some higher ups are looking at gross numbers and not paying attention to more relevant statistical info.
Perhaps someone important at the insurance company got bitten by a Labrador-retreiver.
It does seem odd that mere popularity of ownership would be a problem for the insurance company; that they are common would mean that more policies are bought which would proportionally counteract the number of payouts.
And the Harryy Potter vs Ulysses thing is, as you say, something of a self-selection effect. Unless Labrador-retreiver owners tend to be, for example, dickheads, the dogowning equivalent of Volvo drivers, I don't see that ownership of a Labrador-retreiver is the same sort of thing.
It is possible there is a relevant dataset of which I am not privy. Perhaps there are a large number of more modest dog bites not documented in the report I documented above that lead to more modest but numerous insurance claims, and that Labs are drastically overrepresented in this group. I can't imagine how this could be so, but maybe there is. Otherwise, as others have pointed out, it makes no sense to not insure something merely because it's popular, as those not making claims would more than compensate for those who do. It's perceived as opposed to an actual risk.
I can be convinced otherwise by the data though, if it exists.
My understanding of these statistics is that people will buy a "substantial" book if they feel they have about a 2/3 chance of finishing it. There is no result given for Rowling's second book, so I'll assume that it was too short, or perhaps too juvenile to be considered "substantial", or, perhaps, it also has a completion rate of about 2/3. A secondary result is that people are more likely to start a book they will not finish if it is perceived as "a classic" or "good for them", hence the slightly lower completion rate for Joyce.
The even-numbered Harry Potter books are all much better than the odd-numbered ones. I'm hoping this trend does not continue.
The even-numbered Harry Potter books are all much better than the odd-numbered ones.
I don't think I'd agree with that. I thought Harry Potter and the Prisoner of Azkaban was the best of the lot-- the fourth one is pretty good, but it's a little bloated. The second was kind of slight, and the fifth is simply awful, while the first and sixth are ok.
"...The even-numbered Harry Potter books are all much better than the odd-numbered ones."
I could swear I've heard music critics doing their version of this argument in terms of Beethoven's even numbered versus odd-numbered symphonies.
Is adamsj hoping that Harry Potter 7 breaks the parity trend, or fearfully predicting that series ends poorly, or getting his request in for an 8th book of the septology?