Bo Cowgill.com

MY FAVORITE ANDROID APPS (SO FAR)

2009-03-11T00:09:00.001-07:00

Since December, Google employees have been carrying T-Mobile G1 devices with the Android operating system installed. Here are some of my favorite applications so far:�

G-Backup: Backs up your call log, text messages and pictures to your Gmail account. $4 one-time charge to download.�
PhonePlus Callback: Typically when I don't answer a call, I like to reassure the caller that I'm planning to get back to them.�The PhonePlus app adds a button to your incoming call screen that sends the person a text message: "Will call you back in a bit!," or a message of your choice. $1 for a high quality version (one-time payment).�
Loopt: The original social location sharing service. Of course, Google has its own�Latitude�service as well, which is also a social location sharing service that works on Android.�
My Tracks: An application you can launch to record your GPS tracks every few seconds. This is useful so that you can share where you hiked, biked, drove or whatever on your vacation or day trip.�Here's a recent motorcycle trip I took around San Francisco.�After you record your journey, you can easily upload the data to Google Maps directly from your phone and send to your friends.�
Zombie, Run!: This game, by my friend and colleague�Peter Dolan, locates you on a map of your town using your phone's GPS (or cell phone triangulation). It then populates the map with "zombies" -- mosters trying to eat you whose locations appear on your map nearby you.�Your job: Escape the zombies moving towards your location on the map. If you don't run fast enough, you get eaten and you lose!�
Google Reader Mobile: Note that this isn't an installable app, just a nice mobile friendly version for Android's browser.�
ChompSMS: An SMS client that looks like the iPhone's. As if the default doesn't already.�
Trap: Jezzball for Android. For those dull moments in the subway where you don't have a connection for your feed reader.�

Granted, some of these should be part of the OS. Maybe they will in future iterations. But one of the great things about an open handset is that developers can plug the gaps while the OS evolves.

OOPS!

2009-03-09T13:48:00.000-07:00

A seriously unfortunate house listing. Check out the Street View for this house by Google Maps.

PREDICTIVE MODEL MARKUP LANGUAGE

2009-03-05T16:53:00.000-08:00

Here's a cool product: The Predictive Model Markup Language (PMML). PMML is an open standard for sharing predictive models.

This allows different statistical packages and data mining systems to talk coherently to each other about a particular model. PMML supports a variety of algorithms -- so whether your model is based on regression, trees or whatever, it can be expressed in PMML.

Here is a package for R and here is a tutorial and background. If you're looking for support, check out Zementis.

FREE ZIP CODE TO CONGRESSIONAL DISTRICT DATA

2009-02-26T13:37:00.000-08:00

If you're doing quantitative research on Congressional districts, its often helpful to get a mapping between zip code and Congressional district. Shouldn't this kind of data be free and easy to find?

It wasn't, at least in my experience. There are a number of commercial providers of this data who seem to be charging way too much.

Fortunately, a few cool new organizations have popped up that can help. Sunlight Labs, a technology branch of the Sunlight Foundation, has an API that will allow you to extract this data. Similarly, Watchdog.net makes this data accessible via an API as well as text files (check out this 1.5GB mapping of zip+4 values to Congressional districts or this smaller file containing zip codes and districts only, without the +4).

In case these files move, the links came from this Watchdog.net support page. While you're at it, why not donate to the Sunlight Foundation?

2009-02-20T17:55:00.001-08:00

RECENTLY I WAS ON A PANEL at the Predictive Analytics World '09 conference in San Francisco about the R Programming Language and the R Project for Statistical Computing.

The topic of the panel was "How Google and Facebook are using R" -- and I was joined by Itamar Rosenn of the Facebook Data team (for some info about what their team does, see this slightly old but still informative talk).

Jim Porzak gave some remarks about the basic features and history of R, and David Smith gave a presentation about Revolution Computing -- a company aiming to do for R what Red Hat did for Linux. Red Hat seems to be the inspiration for a lot of open-source startups these days.

Google uses R a lot -- as far as I can tell, more than any other stats package. We have a fairly active discussion forum on R topics consisting of over 200 people. Since unqualified praise is boring, I offered some critiques of the language on the panel -- I was thinking that members of the audience might be R contributors who might be considering future projects. These shouldn't obscure the fact that the R Foundation has done a brilliant job of creating a highly flexible and valuable piece of open source technology.

Michael Driscoll moderated the panel and had a writeup here. If I get some time this week, I'll write up my full remarks.

2009-01-29T23:05:00.000-08:00

HAL VARIAN in McKinsey Quarterly.

2009-01-27T15:19:00.001-08:00

YOU KNOW ITS A RECESSION WHEN: More people are searching Google for coupons than Britney Spears.

The switch came in early 2008. Does this have something to do with Britney's declining popularity? I don't think so. Her query share numbers have been around the same since 2006.

Hat tip: Hal Varian.

2009-01-22T09:24:00.000-08:00

NEW PREDICTION MARKET PAPER by Neil Malhotra and Eric Snowberg here. See review from the Social Science Statistics Blog here:

The most clever insight in the paper is that you can combine data from different prediction markets to estimate an interesting conditional probability -- the probability that a primary candidate will win the general election conditional on winning the nomination. (If p(G) is the probability of winning the general election and p(N) is the probability of winning the nomination (both of which are evident in prediction market contract prices), p(G|N) -- the probability of winning the general election if nominated -- can be calculated as p(G)/p(N).) In the first part of the paper, the authors focus on how individual primaries in the 2008 election affected this conditional probability for each candidate. This is interesting because classic theories in political science posit that primary elections force candidates to take positions that satisfy their partisans but hurt their general election prospects by making it harder for them to appeal to the electoral middle. If that is the case, then ceteris paribus one would expect that the conditional election probabilities would have gone down for Obama and Clinton each time it looked like the primary season would become more drawn out -- which is what happened as results of several of the primaries rolled in.

As it turns out, p(G|N) didn't move much in most primaries; if anything, it went up when the primary season seemed likely to extend longer (e.g. for Obama in New Hampshire). Perhaps this was because of the much talked about positive countervailing factors -- i.e. the extended primary season actually sharpened each candidate's electoral machines and increased their free media exposure. Of course, Malhotra and Snowberg have no way of knowing whether the binding effect of primaries exists and was almost perfectly counterbalanced by these positive factors, or whether none of these factors really mattered very much. [...]

The second part of the paper explicitly considers the problem of assessing how "surprised" the prediction markets were in particular primaries (without explaining why this was not an issue in the first part), and employs a pretty ad hoc means of upweighting effect estimates for the relatively unsurprising contests. Some kind of correction makes sense but it seemed to me that the correction was so important in producing their results that it should be explained more fully in further revisions of the paper.

2009-01-22T09:20:00.000-08:00

CHECK OUT THIS NEW PAPER by Hal Varian discussing some of the microeconomic features of online ad auctions. Its a followup to this previous 2006 paper on position auctions and also cites David Pennock.

2009-01-17T12:50:00.000-08:00

ONE OF THE MORE MYSTERIOUS LINES I'VE EVER READ IN AN ACADEMIC PAPER: Here:

We have no reason to thank the University of Chicago Press.

Okay!

2009-01-14T11:30:00.000-08:00

YELPDROID.COM: One of the cooler, lesser known features of Google Gears lets users tell a website your location (without having to enter it manually). This enables website publishers to give you location targeted content. For example: Imagine navigating Yelp without having to enter your address.

I'm not sure why Yelp hasn't implemented this, but fortunately my colleague Rob On moved without them. Check out Yelpdroid.com from a machine that has Google Gears installed (this includes all Android phones).

Its simple: Just type what you're looking for, and Yelpdroid will send your query to Yelp with your location appended. Note: You'll need to give Yelpdroid permission to view your location so it can send it to Yelp. Yelpdroid and Yelp don't otherwise read or store your Gears-provided location.

How does it do this? If you're on a mobile, it uses GPS or cell tower triangulation. If you're on a device without GPS or cell towers (such as a desktop or laptop), it uses your IP address. This works astonishingly well at deriving your approximate location.

You also need to authorize each website to view your location data (otherwise all kinds of companies would know potentially private information about where you are). See more about this feature here.

2009-01-13T19:56:00.001-08:00

HERE'S A BETTER COPY OF GOOGLE AND YAHOO'S LETTER TO THE CFTC ABOUT PREDICTION MARKETS.

This is identical to the one on the CFTC's website (linked from this page of responses to the CFTC's request for comment on prediction markes) -- except that it can be more easily copied and pasted.

2009-01-13T18:53:00.001-08:00

THE CFA INSTITUTE HAS written to the CFTC about its forthcoming ruling about prediction markets.

As part of its efforts to repond to the Concept Release, the Centre recently surveyed the membership of CFA Instutute. Of the nearly 1,100 respondents to this survey, 712, or 65%, indicated that they believe event contracts provide valuable information that can improve the decisions of investors, commodity producers, manufacturers, and others. On the other hand, the respondents were more ambivalent about both the usefulness and appropriateness of these contracts for investment purposes and the need for regulators to assume direct oversight of these contracts and markets.

For example, in response to a question about the usefulness, just 52% of respondents agreed with the suggestion that event contracts "can help investors, commodity producers, manufacturers, and others hedge specific risks." This includes 40% who agreed only that these contracts were useful for hedging purposes, and 12% who said there [sic] were useful for hedging and "suitable for customer investment accounts." Only 2% said they were suitable for customer investment accounts, but not for hedging.

Likewise, only 52% of respondents believe derivatives market regulators should have direct oversight of event contracts. Of these, 29% said direct oversight should come with permitting these contracts to trade on established futures and options exchanges, while the remaining 23% said this direct oversight should include a restriction on these contracts to trading on specialty prediction markets. By comparison, 9% said regulators should have responsibility only to review and approve specialty prediction markets, and 18% said they should have no authority over prediction markets or event contracts at all [Footnote: The remaning 20% had no opinion].

2009-01-09T12:48:00.000-08:00

CASS SUNSTEIN appears likely to lead the Office of Information and Regulatory Affairs in the White House.

The Wall Street Journal says the post "oversees regulations throughout the government," "wields outsize power" and "job will be crucial as the new administration overhauls financial-services regulations, attempts to pass universal health care and tries to forge a new approach to controlling emissions of greenhouse gases."

Here's an article I wrote with Cass in 2006 in The New Republic, and here's an open letter to regulators that he signed in 2008 about prediction markets.

In this letter, he said "The Commodity Futures Trading Commission (CFTC), the federal regulatory agency that oversees futures market activity, should establish safe-harbor rules for selected small-stakes markets."

This is a position similar the one Google and Yahoo! have taken.

On that note, our coalition will shortly have more members.

2008-12-30T05:11:00.001-08:00

NEW PREDICTION MARKET PAPER by David Lee and Enrico Moretti of Princeton and Berkeley, respectively.

2008-12-22T00:55:00.001-08:00

INTERESTING 2004 PAPER from Steve Levitt about why financial markets and gambling markets -- despite many common features -- have evolved different structures.

2008-10-27T10:38:00.000-07:00

IRISH BOOKMAKER PADDYPOWER has already paid out on Obama.

One wonders what they'll do if McCain somehow squeaks out a win!

2008-10-20T07:21:00.002-07:00

WISDOM OF CROWDS AUTHOR JAMES SUROWIECKI now has a blog.

2008-10-20T03:41:00.001-07:00

VERY INTERESTING POST on the Consensus Point blog about the prediction markets at Best Buy.

2008-07-10T16:23:00.000-07:00

A COALITION OF REGULATED EXCHANGES including NYSE Arca, International Securities Exchange, Chicago Board Options Exchange and the American Stock Exchange have written a letter to the CFTC about event contracts whose purpose seems to be protecting the SEC's turf.

BLOOMBERG ON GOOGLE INTERNAL TRADING

2008-07-10T11:32:00.000-07:00

I THOUGHT CLAYTON HARRISON did a great job in this article in Bloomberg on Google's prediction markets and the research around them.

2008-07-09T15:45:00.000-07:00

THE NOTE is a great daily political newsletter. Unfortunately, ABC News is a bit too old media to realize that the note should have its own full-text RSS feed. To get The Note by RSS, you've got to sign up for a feed containing all of ABCNews political stories -- a fraction of which are The Note.

While we're waiting on ABC to come into the modern era, I've created a feed using Yahoo! Pipes that scrubs out all of the non-Note stuff out of ABCNews.com's political feed. Enjoy. And check out Yahoo! Pipes, which is a very cool and intuitive application.

"COWGILL BATTERING MINORS"

2008-07-05T15:06:00.000-07:00

This must be the worst headline AP has ever written. The actual story is about how my cousin Colin is having a great first season in professional baseball. He was signed by the Arizona Diamondbacks earlier this year.

Also, congrats to my other cousin Jean Ellen who was valedictorian from Dartmouth in June.

YOU TELL 'EM!

2008-07-02T14:39:00.000-07:00

"The regulation of [election and policy event contracts] follows from the history of enlightened, flexible innovation exemplified by the CFTC." -- Jason Ruspini.

Also check out the encouraging statement from the International Swaps and Derivatives Association, whose input is likely to be taken seriously because of its member financial organizations.

Link via Chris Masse. The big question I want to know: Will the CME or CBOT write in? What about anyone from the enterprise prediction market space?

STEVEN PINKER'S STUFF OF THOUGHT

2008-06-24T22:08:00.001-07:00

Worthy read.