You are currently browsing the monthly archive for October 2011.
My street is a Halloween Mecca. People flock from neighboring blocks to a section of my street and to the street just North of us. (Ours is an East-West street as are most of the residential streets in the area.) And I have noticed that in other neighborhoods in the area and in other places I have lived there is usually a local, focal Halloween hub where most of the action is.
And on those blocks where most of the action is the residents expect that they will get most of the action. They stock more candy, they lavishly decorate their yards, and they host haunted houses. They even serve beer. (To the parents)
I think I have figured out why we coordinated on my street.
In a perfectly symmetric neighborhood lattice, trick-or-treating is more or less a random walk. With a town full of randomly walking trick-or-treaters every location sees on average the same amount of traffic. Inevitably, one location will randomly receive an unusually large amount of traffic, those residents will come to expect it next year, decorate their street, and reinforce the trend. Then it becomes the focal point.
In this perfectly uniform grid, any location is equally likely to become that focal point. That is the benchmark model.
But neighborhoods aren’t symmetric. One particular asymmetry in my neighborhood explains why it was more likely that my street became the focal point. Two streets to the South is a major traffic lane that breaks up the residential lattice. In terms of our Halloween random walk, that street is a reflecting barrier. People on the street just to the South of us will all be reflected to our street. In addition we will receive the usual fraction of the traffic from streets to the North. So, even before any coordination takes hold our street will see more than the average density of trick-or-treeters. For that reason we have a greater chance of becoming the focal point. And we did.
Traffic could cause roads: greater traffic leads to greater expenditure on roads. Or roads could cause traffic: greater supply of roads leads to more driving and hence traffic. Which way is it?
This is studied by Duranton and Turner in a forthcoming paper in AER. They use an instrumental variables approach to identify causation. A 1947 plan envisaged a highway network connecting existing population centers. Importantly for this study, tha plan was based on existing population centers not forecast traffic demand. Hence, the impact of the greater availability of roads on traffic can be studied (while controlling for factors such as population). The authors find:
For interstate highways in metropolitan areas we find that VKT (vehicle kilometers traveled)
increases one for one with interstate highways, confirming the ‘fundamental law of highway congestion’
Provision of public transit also simply leads to the people taking public transport being replaced by drivers on the road. Therefore:
These findings suggest that both road capacity expansions and extensions to public transit are not
appropriate policies with which to combat traffic congestion. This leaves congestion pricing as the main
candidate tool to curb traffic congestion.
- You can’t tickle yourself but someone can hold your hand and tickle you with it. (Try it)
- Syncopated rhythms trigger an automatic response where you bob your head or tap your feet because your brain demands that the missing beats be counted. In a strong sense, the music causes you to dance to it.
- Even though it would take you several minutes to list all 50 States, you know right away that there is no state that begins with E.
- If you are eloping, it is easier to back out at the last minute because there aren’t hordes of guests coming from all over expecting a wedding. Therefore marriages that begin with elopement will tend to last longer.
I leave it to Paul Krugman to advice Greece. I’ll stick to Greeks. In a couple of weeks from now, while everyone is at the rally complaining about the austerity plan, the government will shut down all banks. They will get out of the euro. On Monday, Greek ATMs will issue euronotes with a picture of Plato photoshopped onto them. They will be called drachmas. Plato’s picture will cause the value of the note to fall to a fraction of the value of the euro. So, in the next couple of weeks, following the simplest prescription of Hirschman’s Exit, Voice and Loyalty, Greeks should move all their bank accounts abroad. Of course, the country that gave us Plato, Aristotle, Pythagoras and, my personal favorite, Thucydides, does not need my shallow advice. Via the Daily Mail:
Fat-cat Greeks have secretly shifted more than €228billion euros out of their country’s crisis-hit banks and into accounts in Switzerland, according to a report.
The big money is fleeing the country as rich Greeks fear the possible re-introduction of their old currency, the drachma, would instantly halve the value of their euros if they are left in Greek banks.
Netflix has increases prices. What should Redbox, the kiosk DVD rental firm do? Thye could cut or maintain prices and gain share. Or they could raise prices, lose some sales and gain margin. Redbox has decided to raise prices:
The new rental rate will be $1.20 per day, instead of the current $1 daily rate. Redbox prices will remained unchanged for Blu-ray discs at $1.50 per day and video games at $2 per day.
Why don’t companies do this more often? One explanation:
[I]t spooked investors, especially because Redbox appears to be picking up customers still stewing over the higher prices at Netflix. Coinstar’s shares plunged 10 percent in Thursday’s extended trading.
To outsiders, sales are more observable than profits/unit, because variable costs are kept private by firms but sales figures are publicized. The market can see lower sales but has a harder time calculating the profit implications – lower sales could mean higher profits. Or stock market analysts are crazy short termists.
This one is just not fair:
But that’s a statistics question. Here’s the game theory question.
What percentage of students in the class will answer A) to this question?
A) Less than 50%
B) 50%
C) Greater than 50%
This article surveys the frontiers of toilet-reading science. Few downsides, some upsides.
No writer owned the arena of toilet reading more than Henry Miller. He read truly great books on the lavatory, and maintained that some, Ulysses for instance, could not be fully appreciated elsewhere. The environment was one that enriched substantial works – extracted their flavour, as he put it – while lesser books and magazines suffered. He singled out Atlantic Monthly.
It’s pretty old, but worth reading given his new book.
Trivers has been teaching himself things and then growing bored with them his whole life. In 1956, when he was 13 and living in Berlin (his father was posted there by the State Department), he taught himself all of calculus in about three months. Around the same time, and with more modest success, Trivers-a skinny child picked on by bullies-tried to learn how to box, doing push-ups and covertly reading Joe Louis’s ”How to Box” in the school library.
Akubra Cadabra: Tobias Schmidt.
As reported on the Planet Money blog:
It sounds ridiculous today. But not so long ago, the prospect of a debt-free U.S. was seen as a real possibility with the potential to upset the global financial system. We recently obtained the report through a Freedom of Information Act Request. You can read the whole thing here. (It’s a PDF.)
The problem? No debt means no T-bonds. Without T-bonds what happens to all the financial instruments linked to T-bond yields? The white paper was co-authored by my old Berkeley mate Jason Seligman.
All Cambridge (U.K.) undergrads have (had?) to struggle through a chapter by chapter reading of Keynes’ General Theory of Employment, Interest and Money. Some come out of this confirmed Keynesians and even Marxists and then go on to work in the City of London. This rich irony comes at the cost of some confusion for hordes of the intellectual elite as the book is extremely hard to read. It turns out that Keynes offered a lucid synopsis of his theory in the QJE in a reply to his critics. My only quibble is that I wish he had used the odd equation here or there – he speaks in equations but does not dare spell them out presumably for fear of losing his reader. But here is a spectacular passage on uncertainty vs risk:
Why does it matter? Because it affects demand for money and hence the interest rate:
The whole article is full of amazing insights. i’s are not dotted or t’s crossed. Many papers remain to be written.
(Hat Tip: Nabil Al-Najjar)
An auctioneer is never tempted to employ a shill bidder.
To be sure, he might want to make the winning bidder pay a higher price and using a shill bidder is one way to make that happen. For example, in an English auction the seller could shill bid until the price reaches a point where all but one bidders have dropped out. That price is the highest revenue he would have earned without shill bidding, and by shilling a little bit longer before finally dropping out, the seller could try to extract something more.
Of course, this comes at some risk for the seller because there is a chance that the high bidder will drop out before the shill bidder does and then the seller misses out on a sale. Nevertheless, a shill bidder pays off on average if the seller thinks that this small-probability loss is outweighed by the large-probability gain.
Nevertheless, the seller would never be tempted to do this.
The reason is that he could achieve exactly the same thing using reserve price. Before the auction even begins he can ask himself what he would want to do if the price rose to that level. If he decided that he would want to use a shill bidder to raise the price even further then he could bring about exactly the same effect by setting his reserve price at the desired level.
That is, a shill bidder is just a reserve price in disguise.
(ps, you don’t have to get very fancy to see why this is wrong.)
Here is the gist of the new mortgage plan:
The changes, which will take effect over several months, will let people qualify for new loans no matter how far the values of their homes have fallen, so long as they have made at least six consecutive monthly payments. The plan also will reduce borrowers’ fees, for example, by dispensing with the need for an appraisal in many cases and by automatically transferring mortgage insurance to the new loan.
Fannie Mae and Freddie Mac generally require refinancing lenders to assume responsibility for any problems with the original loan because in making the new loan they are relying in part on that original documentation. That has made lenders reluctant to refinance loans for which they are not already responsible. That provision will now be waived, in exchange for a fee.
But the program still applies only to loans that Fannie and Freddie acquired before May 31, 2009. It does not reduce the amount that borrowers owe. And only borrowers with less than 20 percent equity in their homes are eligible; those with more equity must seek a refinancing through the standard and more expensive channels, although the government is considering making some of the same changes, like reducing fees, for those borrowers.
One of my favorite colleagues, Debbie Lucas, who has now moved to MIT Sloan, co-authored a study of the impact of a large scale mortgage refinancing plan while she was at the CBO:
Relative to the status quo, the specific program analyzed here is estimated to cause an additional 2.9 million mortgages to be refinanced,
resulting in 111,000 fewer defaults on those loans and estimated savings for the GSEs and FHA of $3.9 billion on their credit guarantee exposure, measured on a fair-value basis. Offsetting those savings, federal investors in MBSs, including the Federal Reserve, the GSEs, and the Treasury, would experience an estimated fair-value loss of $4.5 billion. Therefore, on a fair-value basis, the specific program analyzed here would have an estimated cost to the federal government of $0.6 billion….Because the estimated gains and losses are small relative to the size of the housing market, the mortgage market, and the overall economy, the effects on those markets and the economy would be small as well.
The plan Debbie studied is larger than the one being implemented right now. Something more radical is necessary to have a big impact. It might have to allow borrowers to write down their principal. Who has the guts to propose a bailout of homeowners? Ironically, Glenn Hubbard, advisor to Mitt Romney.
Economics is a unique discipline in that the technical ideas have to be explainable to regular people. Of course, the ideas in economics are not as technical as phyiscs, but there is essentially nothing at stake in being able to explain Maxwell’s Demon to a lay person, although that is certainly a talent.
And most disciplines that must be explained to the great un-unwashed are not technical enough for that to be any challenge.
Hand-in-hand with that distinctive feature is the fact that economic ideas are about things that regular people already have opinions on (usually strong ones) [how's that for un-dangling a preposition!] To be able to have a useful dialog with them requires that you understand their opinions and most importantly why they have them.
Because strong opinions don’t come from nowhere. There is always some logic behind them.
So to be a good economist you must be familiar with, and see the logic behind, wrong ideas. This requires you to be sufficiently dumb. Because really really smart people would never have entertained those ideas and they will find them completely foreign and not worthy of consideration alongside the correct logic.
On the other hand you do have to be sufficiently smart to know why the logic is incomplete. But that by itself is usually not so demanding.
What is demanding is to be sufficiently dumb and sufficiently smart to be able to do both, AND also to be able to explain to BOTH the regular people and the smart people the other side.
Finally, I would add that being sufficiently dumb is crucial for finding good research projects. It has to do with the elusive “non-obvious” ideas. in practice “non-obvious” means “not obvious to the regular guy.” To spot those projects you need to have a replica of the same mental infrastructure that the regular guy is equipped with.
I do most of my reading through Google Reader, and when I get an idea for the blog I post it to Google Buzz. The small number of followers I have there will sometimes comment and/or vote for ideas that pop up there. When its time to write something for the blog, I go back there for ideas.
Buzz will soon be retired by Google and Reader is going to be crippled. That’s going to affect me. For one thing, I won’t have access to Google Reader feeds from many of my favorite curators, most notably Courtney Conklin Knapp, the soy sauce of the internet. And I need a new place to pre-test my ideas. (I used to think that I would use Twitter for that, but my Twitter identity has become overrun with tweets like “I went back in time so I could be the first person to write about the paradoxes of time travel.”)
Google says that the retired services are to be replaced by Google+ so I am switching to Google+. In fact I have been using it for a while now, jotting down some ideas and getting feedback before posting them here. Google+ is more of a “social” social network than Buzz so instead of just dumping links and incomprehensible notes-to-self, I am writing little rough drafts. It works well. Writing doesn’t come easy for me but for some reason when I know that what I am writing is verifiably a rough draft, I loosen up a bit and it comes easier. The feedback is great too.
(One potential downside is that I write something wrong, people on G+ point out that its wrong and I am too embarrassed to post it here. You learn a lot from wrong ideas so that would be a loss.)
So if you are on G+ I hope to see you there, and I would be happy to get your feedback.
By the way, know any good blogs? It seems like a large number that I subscribe to on Reader have gone dark so I am looking for some new ones. “What’s Hot” in Google Reader is already gone!
Social Choice Theorists are going to see an experiment in action as San Francisco votes for a new mayor using rank-order voting. Here is how it works:
Each voter lists up to three candidates in ranked order: First, second and third choice.
If one candidate gets more than 50 percent of the first-place votes in the first round of counting, he’s the winner and there’s no need to look at the second and third choices.
But if no one has a majority, the candidate with the fewest number of votes is eliminated from the future count and his second-choice votes are distributed to the remaining candidates.
If still no one cracks the 50 percent mark, then the candidate with the second-lowest vote total is eliminated and his second-place votes are distributed. If the voters’ second choice already was eliminated, it’s the third-choice vote that goes back into the pool.
This continues until one candidate has a majority of the remaining votes. Last November, it took 20 rounds before Malia Cohen finally was elected as supervisor from San Francisco’s District 10.
There are two strategic issues. First, there must be an incentive for strategic voting via Gibbard-Satterthwaite/Arrow. Hence, sincere voting and strategic voting will differ. Second, the candidate positions and in fact the issue of who enters as a candidate is a key factor in the rationale for switching to rank order voting in the first place. Some voters must hope that third party candidates can now enter and have a chance of winning. Others must hope that more centrist policies are adopted by the candidates in the hope of being voters’ second or third choice.
I assume there are many formal theory papers in political science on this but am not familiar with them…anyone have any ideas?
Advertisers want information about your tastes and habits so they can decide how much they are willing to pay to advertise at you. That information is stored by your web browser on your hard drive. Did you know that every time you access a web page you freely hand over that information to a broker who immediately sells it to advertisers who then immediately use it to bid for access to your eyeballs?
Here’s how it works. Internet vendors, say Amazon, pass information to you about your transactions and your browser stores them in the form of cookies. Later on, advertisers are alerted when you are accessing a web page and they compete in an auction for the ad space on that page. At that moment, unless you have disabled the passing of cookies, your browser is sending to potential advertisers all of the cookies stored on your hard drive that might contain relevant information about you.
However, many of the really valuable cookies are encrypted by the web site that put them there. For example, if Amazon encrypts its cookies then even though your browser gives them away for free, they are of no use to advertisers.
That is, unless the advertisers purchase the key with which to decrypt your cookies. And indeed Amazon will make money from your data by selling its keys to advertisers. It could sell them directly but it will probably prefer to sell them through an exchange where advertisers come to buy cookies by the jar.
The interesting thing about the market for cookies is that you are the owner of the asset and yet all of the returns are going to somebody else. And its not because your asset and mine are perfect substitutes. You are the monopolistic provider of information about you and when you arrive at a website it is you the advertisers are bidding for.
How long will it be before you get a seat at the exchange? Nothing stops you from putting a second layer of encryption over Amazon’s cookies and demanding that advertisers pay you for the key. Nothing stops me from paying you for exclusive ownership of your keys, cornering the market-in-you, and recouping the monopoly profit.
File under “Feel Free To Profit From This Idea.“
(I learned about the market for cookies from Susan Athey’s new paper and a post-seminar dinner conversation with her and Peter Klibanoff, Ricky Vohra, and Kane Sweeney.)
(Picture: Scale Up Machine Fail from www.f1me.net)
Via BoingBoing, why is the Indiana Election Commission putting cubes of styrofoam in their mailers?
The Styrofoam cube enclosed in this envelope is being included by the sender to meet a United States Postal Service regulation. This regulation requires a first class letter or flat using the Delivery or Signature Confirmation service to become a parcel and that it “is in a box or, if not in a box, is more than 3/4 of an inch thick at its thickest point.” The cube has no other purpose and may be disposed of upon opening this correspondence.
The price of a 6.5oz bottle of Coca Cola stayed fixed at 5c from 1886 to 1959. Daniel Levy and Andrew Young document this fact and ask why this might have been the case in a period that saw two world wars and the Great Depression. They offer two technological explanations for price rigidity:
First, we demonstrate that an installed base of vending machines with nickel-only capability, and the
evolution of the technology that could accommodate multiple type coins and change making,
imposed an important constraint on the ability of the Coca-Cola Company to adjust the Coke’s
price. Second, at the 5¢ price per serving, the smallest price increase compatible with the
consumer still using a single coin was a 100 percent jump to 10¢. A monetary transaction
technology for smaller price adjustment while keeping consumer “inconvenience costs” low in
terms of the number of coins needed for purchasing a bottle of Coca-Cola, was not available.
How can you get around these technological constraints?
You could lobby your friends:
Woodruff [Coca Cola CEO] submitted a request in 1953 to the newly elected President Dwight Eisenhower (his hunting companion and friend) himself, to get the U.S. Department of Treasury mint a new 7 1/2-cent coin.
Eisenhower forwarded the request to the Treasury Department officials who did not like the idea.
Or ingeniously in an early example of mechanism design, you could randomize bottle delivery while retaining the nickel technology in the vending machine:
“Instead of offering one ‘Coke’ for 6¢ the coin cooler offers 8 ‘Cokes’ for 45¢,
which is only 5.625¢ (5 5/8¢) per bottle. [The] coin cooler [delivers] either an
empty bottle or no bottle at all for one nickel in every nine deposited. This
absence of ‘Coke’ is called an official blank. Please be warned that, if you fail to
deposit nine nickels, at worst you will strike the blank and have to deposit
another nickel for your ‘Coke.’ At best you will miss the blank (8 times out of 9)
and your ‘Coke’ will cost only a nickel, but as stated, on the average ‘Coke’ sells
for 5.625¢ per bottle—the only price at which it is offered”
The plan might actually have been tried out in Chicago and Canada!
(Hat Tip: Tilman Klumpp and Xuejuan Su)
Assume that people like to have access to a community of people with similar habits, tastes, demographics, etc. A “community” is just a group of some minimal absolute size. Then the denser the population the more likely you will find enough people to form such a community.
But this effect is larger for people whose tastes, habits, and demographics are more idisyncratic than for people in the majority. Garden-variety people will find a community of garden-variety people just about anywhere they go. By contrast, if types of people are randomly distributed across locations, the density of cities makes it more likely that a community can be assembled there.
But that means that types won’t be just randomly distributed across locations. The unique types are willing to pay more to live in cities than the garden-variety types.
From the great blog Mind Hacks:
Because of this, the new study looked at volleyball where the players are separated by a net and play from different sides of the court. Additionally, players rotate position after every rally, meaning its more difficult to ‘clamp down’ on players from the opposing team if they seem to be doing well.
The research first established the belief in the ‘hot hand’ was common in volleyball players, coaches and fans, and then looked to see if scoring patterns support it – to see if scoring a point made a player more likely to score another.
It turns out that over half the players in Germany’s first-division volleyball league show the ‘hot hand’ effect – streaks of inspiration were common and points were not scored in an independent ‘coin toss’ manner.
Quoting The Angry Professor:
I need a gmail filter that sends a custom vacation message to certain people. The filter needs to scan the message for the latest date mentioned and determine the geographical location of the sender. It must next add four weeks to that latest date. The longitude of the sender’s geographic location must be advanced by 180o and the latitude multiplied by -1. After triangulating the dry land nearest to the rotated coordinates, the filter must finally send the following “vacation” message:
Thank you for your email. I am [on dry land nearest the point exactly half-way around the globe from you]. I will not have email contact until [computed date], but I will try to respond to your message as soon as I return.
Via Vinnie Bergl, here is a post which examines pitch sequences in Major League Baseball, looking for serial correlation in the pitch quality, i.e. fastball, changeup, curve, etc. The motivating puzzle is the typical baseball lore that. e.g. the changeup “sets up” the fastball. If that were true then the batter knows he is going to face a fastball next and this reduces the pitcher’s advantage. If the pitcher benefits from being unpredictable then there should be no serial correlation. The linked post gives a cursory look at the data which shows in fact the opposite of the conventional lore: changeups are followed by changeups.
There is a problem however with the simple analysis which groups together all pitch sequences from all pitchers. Not every pitcher throws a changeup. Conditional on the first pitch being a changeup, the probability increases that the next pitch will be a changeup simply because we learn from the first pitch that we are looking at a pitcher who has a changeup in his arsenal. To correct for this the analysis would have to be carried out at the individual level.
Should we expect serial independence? If the game was perfectly stationary, yes. But suppose that after throwing the first curveball the pitcher gets a better feel for the pitch and is temporarily better at throwing a curveball. If pitches were serially independent, then the batter would not update his beliefs about the next pitch, the curveball would have just as much surprise but now slightly more raw effectiveness. That would mean that the pitcher will certainly throw a curveball again.
That’s a contradiction so there cannot be serial independence. To find the new equilibrium we need to remember that as long as the pitcher is randomizing his pitch sequence, he must be indifferent among all pitches he throws with positive probability. So we need to offset the temporary advantage of a curveball this is achieved by the batter looking for a curveball. That can only happen in equilibrium if the pitcher is indeed more likely to throw a curveball.
Thus, positive serial correlation is to be expected. Now this ignores the batter’s temporary advantage in spotting the curveball. It may be that the surprise power of a breaking pitch is reduced when the batter gets an earlier read on the rotation. After seeing the first curveball he may know what to look for next and this may in fact make a subsequent curveball less effective, ceteris paribus. This model would then imply negative serial correlation: other pitches are temporarily more effective than the curveball so the batter should be expecting something else.
That would bring us back to the conventional account. But note that the route to “setting up the fastball” was not that it makes the fastball more effective in absolute terms, but that it makes it more effective in relative terms because the curveball has become temporarily less effective.
The latter hypothesis could be tested by the following comparison. Look at curveballs that end the at bat but not the inning. The next batter will not have had the advantage of seeing the curveball up close but the pitcher still has the advantage of having thrown one. We should see positive serial correlation here, that is the first pitch to the new batter should be more likely (than average) to be a curveball. If in the data we see negative correlation overall but positive correlation in this scenario then it is evidence of the batter-experience effect.
(Update: the Fangraphs blog has re-done the analysis at the individual level and it looks like the positive correlation survives. One might still worry about batter-specific fixed effects. Maybe certain batters are more vulnerable to the junk pitches and so the first junk pitch signals that we are looking at a confrontation with such a batter.)
Bruce Riedel who ran President Obama’s AfPak review now favors containment over engagement with Pakistan as I mentioned in Part 1. He thinks military aid should be reduced sharply and substituted with reduced tariffs and the like. How is the Pakistani army going to respond? In the past, scientist A Q Khan sold nuclear secrets to countries that are hostile to the United States. Was the Pakistani Army complicit in the nuclear trade? That information is not in the public domain.
What is to stop the Pakistani Army employing or reverting to nuclear trade under containment? Again, thinking of the United States as the Principal and the Pakistani Army as the Agent, the principal has to come up with some way to give incentives to the agent. With commitment, the obvious instrument is a threat, e.g. the United States will attack/invade Pakistan if they observe a sign of nuclear trade. If the threat is credible, the Pakistani Army will not engage in nuclear trade and the United States will not have to carry out its threat. But the threatened action is costly and it is impossible to commit to carry out the threat in the scenario where it is meant to be implemented. Hence, there would be a credibility issue even if Pakistan were not nuclear. The fact they are nuclear makes the threat incredible. Indeed, it is hard to see any “stick” that can credibly be used to give incentives.
That leaves “carrots”. The United States can make a transfer to the Pakistani Army if and only if there is no sign they are engaging in nuclear trade. It is credible to take the carrot away if there is a sign of trade. If there is no sign of nuclear trade, it is credible to make the transfer because the Pakistani Army has a credible threat of start nuclear trade if a transfer is not forthcoming – after all they get surplus from nuclear trade. If the transfer is large enough and signals of nuclear trade are more likely if nuclear trade actually occurs, the Pakistani Army will play along with its part in the equilibrium too.
This logic is familiar from the efficiency wage model of Shapiro and Stiglitz – if the principal cannot use sticks, she must use carrots and ends up giving surplus to the agent. But here I only want to point out an irony: under the old regime of engagement, the United States paid the Pakistani Army to take a costly effort and hunt down terrorists; in the (possible) new regime of containment, the Unites States pays the Pakistani Army not to go take the profitable action of engaging in nuclear trade. Either way, a transfer is made.
The “Vivace” (vivacious) is a burger topped with bacon, salted spinach, marinated onions and mayonaise with mustard seeds.
The “Adagio” (slowly, like the musical term) is also a hamburger, topped with sweet-and-sour eggplant strips, sliced tomatoes and salted ricotta, all between a bun covered in sliced almonds.
There’s also a tiramisu.
via Arthur Robson:
While appeals often unmask shaky evidence, this was different. This time, a mathematical formula was thrown out of court. The footwear expert made what the judge believed were poor calculations about the likelihood of the match, compounded by a bad explanation of how he reached his opinion. The conviction was quashed.
And the judge ruled that Bayes’ law for conditional probabilities could not be used in court. Statisticians, Mathematicians, and prosecutors are worried that justice will suffer as a result. The statistical evidence centered around the likelihood of a coincidental match of shoeprint with shoes owned by the Defendant.
In the shoeprint murder case, for example, it meant figuring out the chance that the print at the crime scene came from the same pair of Nike trainers as those found at the suspect’s house, given how common those kinds of shoes are, the size of the shoe, how the sole had been worn down and any damage to it. Between 1996 and 2006, for example, Nike distributed 786,000 pairs of trainers. This might suggest a match doesn’t mean very much. But if you take into account that there are 1,200 different sole patterns of Nike trainers and around 42 million pairs of sports shoes sold every year, a matching pair becomes more significant.
Now if I can prove to jurors that there was one shoe in the basement and another shoe upstairs, then probably I can legitimately claim to have proven that the total number of shoes is two because the laws of arithmetic should be binding on the jurors deductions. And if there is a chance that a juror comes to some different conclusion then it would make sense for an expert witness, or the judge even, tell the juror that he is making a mistake. Indeed a courtroom demonstration could prove the juror wrong.
But do the “laws” of probability have the same status? If I can prove to the juror that his prior should attach probability p to A and probability q to [A and B], and if the evidence proves that A is true, should he then be required to attach probability q/p to B? Suppose for example that a juror disagreed with this conclusion. Could he be proven wrong? A courtroom demonstration could show something about relative frequencies, but the juror could dispute that these have anything to do with probabilities.
It appears though that the judge’s ruling in this case was not on the basis of bayesian/frequentist philosophy, but rather about the validity of a Bayesian prescription when the prior itself is subjective.
The judge complained that he couldn’t say exactly how many of one particular type of Nike trainer there are in the country. National sales figures for sports shoes are just rough estimates.
And so he decided that Bayes’ theorem shouldn’t again be used unless the underlying statistics are “firm”. The decision could affect drug traces and fibre-matching from clothes, as well as footwear evidence, although not DNA.
This is a reasonable judgment even if the court upholds Bayesian logic per se. Because the prior probability of a second pair of matching shoes can be deduced from the sales figures only under some assumptions about the distribution of shoes with various tread patterns. The expert witnesses probably assumed that the accused and a hypothetical third-party murderer were randomly assigned tread patterns on their Nikes and that these assignments were independent. But if the two live in the same town and shop at the same shoe store and if that store sold shoes with the same tread pattern, then that assumption would significantly understate the probability of a match.
Via Tyler Cowen, a quote from Daniel Kahneman on why a sandwich made by someone else tastes better.
When you make your own sandwich, you anticipate its taste as you’re working on it. And when you think of a particular food for a while, you become less hungry for it later. Researchers at Carnegie Mellon University, for example, found that imagining eating M&Ms makes you eat fewer of them. It’s a kind of specific satiation, just as most people find room for dessert when they couldn’t have another bite of their steak. The sandwich that another person prepares is not “preconsumed” in the same way.
Put aside the selection effect that conditional on a person making a sandwich for you, it is likely that they are a better sandwich maker than you. Even in a randomized sandwich trial the effect would be there but I have a different theory why:
A large part of tasting is actually smelling. You can verify this by, for example, eating an onion with your nose plugged. Our sense of smell tends to filter out persistent smells after being exposed to them for awhile so that we cannot smell them anymore. This means that when you are cooking in the kitchen, surrounded by the aromas of your food, you are quickly de-sensitised to them. Then when you sit down to eat, it is like tasting without smelling.
When your spouse has done the cooking you were likely in another room, isolated from the aromas. When you walk into the kitchen to eat, you get to smell and taste the food at the same time. That’s why it tastes better to you. The same idea applies to leftovers. It takes much less time to reheat leftovers than it took to prepare the food in the first place so you retain sensitivity to more of the aromas when it comes time to eat.
Bruce Riedel who ran President Obama’s AfPak review now favors containment over engagement:
It is time to move to a policy of containment, which would mean a more hostile relationship. But it should be a focused hostility, aimed not at hurting Pakistan’s people but at holding its army and intelligence branches accountable. When we learn that an officer from Pakistan’s Inter-Services Intelligence, or ISI, is aiding terrorism, whether in Afghanistan or India, we should put him on wanted lists, sanction him at the United Nations and, if he is dangerous enough, track him down. Putting sanctions on organizations in Pakistan has not worked in the past, but sanctioning individuals has — as the nuclear proliferator Abdul Qadeer Khan could attest.
It is useful to think of the US-Pakistan game as a principal-agent relationship. The US (principal) would like to “pay for performance” and make a transfer if and only if the Pakistani army (agent) capture terrorists and quash the Taliban. Performing this task is costly for Pakistan for many reasons. For one, they use the terrorists as proxies in their fight against India. But if the US values elimination of terrorists enough, there is a transfer or sequence of transfers that are large enough to persuade Pakistan to work hard on America’s behalf. For the transfer scheme to work, the US has to be able to commit to pay. If Pakistan is too successful, then the US has no incentive continue paying them. Knowing this, Pakistan does not want to work too hard on America’s behalf. Do enough work to keep the money rolling in but not enough to kill off the goose laying golden eggs.
This delicate balancing act can tip one way or another with random events. After one huge such event, the capture of Osama Bin Laden, the relationship has gone sour. Perhaps, we are in a new phase, that is the gist of Riedel’s column. But how should this be managed? I need to think about part 2….
