Assorted links

Comments

5. What price guaranteed Ivy League admission?

Thanks for that link. The Chinese do not screw around.

That such a business exists is an indictment of the entire admission system.

In the 1990s I read about college "consultants" in eastern Europe (Russia or Poland, I forget which) who would make money-back guarantees to parents: if your kid gets into a top university, pay me x amount. If the kid doesn't get in, you owe nothing.

And then the consultants would sit back and do nothing. Even if only 10% of their "clients" got admitted, it was pure profit for the consultants (not counting marketing costs).

So it's easy to imagine how ThinkTank can have a viable business model. The article makes it clear that they are providing some actual consulting help -- and maybe even making the kids into better students, and maybe even better human beings -- so I'm not saying that it's a scam. But one can imagine that the fees they collect may exceed the actual value added, because it's possible to collect fees while producing zero value added.

ThinkTank's vaunted admission prediction model may be nothing more than an ordinary linear regression model, but by providing more accuracy to the estimated admission probabilities it enables the company to fine-tune its prices and the risk profile of its "portfolio" of clients.

A clever scheme, and as I said not necessarily a socially useless one -- but still a bit shady.

#8 Re: "superforecasters" They say they have 20,000 participants, but if they asked 20,000 people to predict coin tosses, it would not be all that surprising to find out that one of them predicted 15 consecutive flips, and there would be perhaps a couple more who got 14/15 and a few more who got 13/15. And the more people you get to participate, the more astounding the performances of some will be.

I've heard there has not been much regression to the mean after an individual or team of individuals is identified as a superforecaster.

One thing that's important to remember wrt prediction: A natural model to use for predictions is coin flips--you make a prediction which you think has a lot of details and insight baked in, but which ultimately is more-or-less a guess.

An alternative model is that we all carry our own little mini-models of the world. At any given time, some subset of those mini-models are pretty good at predicting everything that's going on; later on, those same mini-models are 180 degrees off. The example that comes to mind is Harry Browne and that crowd of financial doomsayers in the early 70s. They correctly foresaw serious inflation in the US for many years, from their own models of the world which were largely built around their political beliefs. And so for those years when we had a lot of inflation, they were pretty good at predicting the future. And then, the world shifted a bit, and their models were all wrong.

I wonder if something like that might happen with predictions more generally. There are some big forces operating at any given time, and if you understand a couple of them passably well, you can give pretty decent predictions locally. But sooner or later, those forces won't be the dominant ones moving the world, and your insights will become much less interesting.

very good comment.

Okay, but immense efforts were made in the 1980s to battle the inflation that some had foreseen in the 1960s and 1970s, so it's hardly surprising that the world moved on to new and different challenges.

We made 250 flips (forecasting questions) since the first batch of super-forecasters were identified. They outpredicted the rest of the crowd by a growing margin. We do think hard about chance and random error. In this case, chance does not explain the results.

One of Tetlock's superforecasters sent me an email explaining his secrets: you have to work hard, be smart, stay immensely informed about obscure foreign affairs, but not get over-excited about them:

http://isteve.blogspot.com/2013/12/tetlocks-good-judgment-project.html

I wouldn't be surprised if a fair amount of competence in this tournament derives from having a sense of just how _long_ it typically takes for stuff to happen. Since the game looks at typically annual time frames so that it can determine winners and losers in a reasonable amount of time, I bet a lot of losers have a tendency to say, "Yeah, that will probably happen" without estimating how long it could take for it to happen.

For example, say there is a question that asks if the coalition government in Britain or Germany or wherever will come undone. In the long run, the answer is surely Yes. But, will it happen within the next year? Powerful people often are pretty talented at kicking the can down the road for another year.

Time scale is very important. At the time of the housing boom, I expected the bubble to crash and a lot of people lose fortunes, but I had no idea when the bubble was going to pop. Certainly not enough to make any money out of it, just enough to know to stay away and hoard cash to buy once it did decline.

Likewise the conflict in Ukraine. Anyone who read Huntington's Clash of Civilizations knew about Ukraine being a "torn country" and that Russia was the "core state" of Orthodox civilization; therefore, they could predict eventually there'd be a crisis. Furthermore, anyone who paid attention to Ukraine's EU association agreement negotiations and Putin's push for an Eurasian Customs Union knew that the two would push Ukraine into a crisis. The Euromaidan could be predicted based on the Orange Revolution, and perhaps even the flight of Yanukovych. However, even if you made it that far, few would predict that it would cause Putin to invade Ukraine to seize Crimea and the Donbas. There were too many unknown details that determined Putin's actions. I'm a pretty cynical guy, and while I was not surprised he invaded, I was shocked he invaded right then.

Knowing the long term outcomes or processes don't necessarily help you determine when things will be done.

Yes, there's a saying that goes something like this: one of the secrets of forecasting is to give them a number or give them a date, but not both.

3. Thiel is a very smart guy, but he underestimates the value of chance, of being at the right (or wrong) place at the right (or wrong) time: the advent of the internet, the critical mass that is Silicon Valley, the surplus of capital seeking a better return, all coming together at once. What I know is that change is the only constant in life.

Correction: Thiel is a very LUCKY guy.

#3. Correction: Peter Thiel disagrees with rayward.

Great article. My favorite part:

"In 2003, as Act II was getting underway, Thiel took care of an awkward task. He told some friends and associates that he was gay. Word spread quickly. “It’s still one of these things that people in our society find incredibly important and interesting about other people,” he says, with what sounds like annoyance."

God bless you, Peter Thiel.

Peter King reported last night that after Michael Sam was cut the NFL frantically called around to see if any teams would have him on their practice squad.

What a world.

Oh, the best part was King said with zero irony "Thus, the NFL and Sam avoided the nightmare scenario." The same nightmare where 1000 or so guys with similar sklls don't get the additional consideration?

Why is Thiel's perspective the rare one?

Suppose that n people will attend games or watch them on TV because Michael Sam is in the league who otherwise would not. What must n be in order for it to be more profitable to the league and to a team to employ Michael Sam than any other free agent defensive end? If he is only on the practice squad, do Sam Profits approach 0?

Repeating now: Why is Thiel's perspective the rare one?

I don't fault Michael Sam for pursuing the Great American Pastime of Milking a Fortuitous Situation for Every Nickel, nor would I blame the NFL if it calculated that boasting an openly gay player might boost the league's bottom line.

All of this, however, is predicated on the bizarre fascination we currently have with being gay. If being gay is the most interesting thing about Michael Sam, he's a boring dude.

Peter Thiel is not a boring dude.

Is this a real question?

Thiel's perspective will be the common one in a generation (in the OECD). The reason it isn't right now is obvious.

A whole generation? Really? Why does gay fascination have so much more staying power than Who Wants to Be a Millionaire?

You do realize you're the only one making such a big deal about how it shouldn't be a big deal that Thiel is gay, right?

Well, me and Peter Thiel.

Also, since the point is "most people think gay is fascinating", I am quite conscious that I am in the minority here, so... Duh?

I have no idea who is talking to who.

My question is why people like reporters say things like nightmare scenario completely seriously about something that happens to 1000 other guys every year.

Btw, as an aside, I think that answer is related to why Time Ragazine's cover picture on Ferguson implies something misleading. I'm not really talking about gay or race.

This is perhaps the gist of Greg Clark's article about immigration and inequality:

"... recent evidence suggests that, in reality, social mobility rates are extremely low. Seven to ten generations are required before the descendants of high and low status families achieve average status. "

That's very roughly how long it took for the French Kings of England to start speaking English.

That is not true social mobility.

Man, Matt Levine is fucking good.

I like to call him "the anti-Matt Taibbi." Good Matt vs Bad Matt.

Great to see Levine getting some recognition. But he still doesn't have as many awards as the clown who wrote the NYT aluminum article in the first place:
http://www.hillmanfoundation.org/thesidney/david-kocieniewski-0/backstory
http://www.pulitzer.org/citation/2012-Explanatory-Reporting

His blog is must-read

5. I thought this would be the answer to another question. My understanding, from discussion with fellow members of the NYC 1%, is that if your child has the theoretical capacity to do the work (i.e., a graduate--anywhere in the class--of a top private or public high school and SATs over 1700), then guaranteed admission to HYP costs in the low seven figures, and admission to the other five Ivies costs in the mid six figures.

Put another way, those are the prices that enable your children to compete evenly with the children of US presidents and senators.

I've heard people bandy about fantastic numbers like this before. But professionally, my firm horse-trades with elite grad schools all the time to get extra people in. The things we trade, however, are nowhere near that valuable. Typically it involves adding people to interview schedules or a marginal hire for a marginal admit. It's possible HBS is a lot easier to get into than Harvard is to get into coming out of high school or that there's some other reason the price is orders of magnitude lower, but I doubt it. I think the high numbers are urban legends.

You can double check my reasoning by looking at the lifetime earnings impact of going to Harvard over some second tier school like Yale on the kind of kid you describe - with high SATs, good GPA, good school. It just isn't very high.

Unfortunately, I don't feel it would be appropriate to name names of personal acquaintances, so I can't provide substantiation of the numbers I have given.

I didn't say the money was well-spent. On the other hand, if your kid is already receiving an inheritance such that they never have to work, what else can you buy them?

Then I think you are saying "Mommy and Daddy gave $1,500,000 to Harvard and junior got in," not "the price for Harvard admission is $1,500,000." The price is more like $20,000. Mommy and Daddy also bought some social status for themselves.

But how much of a status symbol is HYP to the parents?

Let's grant that there is a very mild alpha to bright person going to Harvard over same bright person going to Midwestern state college*. Let's pull a number and assume that there's no functional difference given average time preference between paying $50K to go to HYP and paying

If the kid paid (and certain assumptions vis a vi rational actors), then a "Getting into HYP" thing would cost $50K.

But on the level of the NYC 1%, Daddy cuts the check. And Mommy and Daddy like bragging about having their kid in HYP. Or Mommy and Daddy are members of the 1% in large part because they functionally abandoned their kids/lives to go make money (As a member of the Single 2% at 22** (acc to http://www.whatsmypercent.com/), I resemble that statement), and feel REALLY bad about this, and try to make it up by pushing their kids into Harvard.

In other words, there's $50K of benefit to the kid, and $950K of benefit to the parents, so they charge a million.

* And honestly, the more I learn about this stuff, the more depressed I get about the whole "Social Mobility" concept. It seems like either you've got it or you don't, that "Having it" is inherited (whether that's culturally, nepotistically, or genetically is up for grabs), and that things like "good schools" only make a couple percentage points difference either way.
** In a Bay Area "I make more money in a year than my parents combine for in 5 years and I STILL have roommates because I can't afford to rent my own place" sort of way. CA standards of living are LOW.

1. Hey, my uncle helped develop this! I gotta tell him he got MR'd.
5. Sounds like paying for actual effective parenting by people who don't know how to do it themselves. The cynical involvement in community service and the planning out of resume padding is fairly typical for Bay Area overachievers. College admissions are just a payment milestone.

Is it fair to summarise #9 as "Economics and romance novels are both about people?". Or perhaps we can extend that to "people and the things they want".

#3: there was research on how to make in-vitro beef.....if skin is possible, how far is biofabricated t-bone?

Fancy bothering with beef when you could make lamb. Or bacon. Or duck.

Remember when you guys tied bacon to sticks and made starving Indians chase you around? Harmless merriment to accompany all the eggs, duck and pir in midst of 3 million emaciated dead? Anything beat that road to Imphal, amirite?

If they were so hungry why didn't they just eat the British?

What the heck are you talking about?

Yeah, I don't remember any of that. I know it isnt a defense but I was really drunk.

5. What price guaranteed Ivy League admission?

Is this operation really a guarantee of admission, or is it more of a side bet about the odds of admission given certain input variables? From all the variables that go into the "price", it really sounds more like a bookie operation.

Or to put it another way, the "guarantee" is a way to dramatically increase the price without changing the odds.

#2 - importing lots of poor people increases inequality. What?! I suppose next thing you're going to tell us is social capital and family wealth and incomes are eroding.

5. Isn't about education or knowledge or success or happiness. It's about joining a club, the elite alumni of prestigious universities whose efforts are devoted to increasing that prestige. People with money, the Kennedys, for instance, can buy their dim bulb kids a place in the national elite beside a supposed genius class. The Chinese are just more obsessive about it.

Wow, that Ma article is naïve. Comments were a hoot, though.

As some here have noted, Ma's just a middleman for a really big check.

#2 is misleading. The fact that poor, uneducated people don't magically rise to be rich in another country is expected. The difference is always at the margins. Poor Mexicans will become less poor in the US. So as far as the immigrants are concerned, immigration is still "a dream". Otherwise they would just go back...

For the host country the situation is more complex. I think that it depends on the local economy composition and the saturation of the demand for such workers. However, I think there are many adaptive strategies that emerge that are often overlooked. Companies will lower requirements if the supply of non-educated workers is high enough. This can increase the incentives to open new companies. Welfare is unfortunately the dark side of this, since it provides all the wrong incentives to immigrants and it can completely unbalance the equation.

Comments for this post are closed