How to Read Academic Articles

My students in Introduction to International Relations at the University of Rochester are now preparing for their finals. As always, most–if not all–of them are behind in their readings. I think a major reason students fall behind here is because they do not know how to read academic articles properly. Then again, it’s not their fault. For some reason, no one ever directly teaches this vital skill to undergrads. Hell, I don’t believe I knew what I was doing until after I graduated. Oops.

Anyway, here is your triage for finals week:

  1. Read the title. (Seriously.)
  2. Read the abstract.
  3. Read the introduction.
  4. Read the conclusion.
  5. Skim the rest. (If you have time or really want to learn the topic, read the rest.)

Let’s go through the rationale of each step of the process. I’ll be referring to my paper on optimal pitching strategies to make the point clearer. Also, it’s scientific fact that baseball makes everything better.

Step 1: Read the Title
This sounds silly, but it really works. Many tests ask students to identify the author of particular quotes from assigned readings. If you just read the title to all of the assigned readings, you can at least connect buzzwords from the quotes to the titles.

For example, the title of the linked article is Breaking Balls with a Runner on Third: A Game Theoretical Analysis of Optimal Behavior. At this point, we have not read the paper at all. But we still know a lot about what is going on. The subject is breaking balls (curveballs or sliders) with a runner on third base. The author (me) will use game theory to tell the reader something. Thus, if the quote on the test says anything about breaking balls or game theory jargon, we will know who the author of the quote is despite having done no real work.

Step 2: Read the Abstract
I didn’t know what an abstract was until senior year. This is not a good thing. If you are actually going to read an article, the abstract is the single most important paragraph. Read it.

I think the temptation to skip the abstract stems from its location in articles. They appear at the top of the first page or on the title page all by themselves. They are very skipable. But it’s fundamentally stupid to overlook them. Let’s read mine:

Whenever a pitcher throws a breaking ball with a runner on third base, he risks allowing the ball go past the catcher and allowing the runner to score. In this article, I analyze the optimal behavior of the players in such a scenario. With decent control and competent catchers, the pitcher throws breaking balls at the same rate he would if the bases were empty. Knowing that a wild breaking ball will occasionally score from third without a swing, the batter anticipates more fastballs. Yet, because the batter throws pitches at the same rate, the batter’s ability to guess the correct pitch is the same in both instances.

You wanted a summary of the article? Behold, I just gave it to you! Moreover, I did it in ten lines. At this point, you know exactly what I am discussing and what I will prove in the article. If you memorize this information, you can provide relevant discussion of the reading. Usually, undergrads do not need to do anymore. You also cannot see whether my argument holds any water, but that does not become important until you are a graduate student. (In case there was any doubt, my argument is rock solid thankyouverymuch.)

Step 3: Read the Introduction
The introduction is usually the abstract but with a little more information. Sometimes you find motivating examples (not the case in the linked article) and other times you will see a summary of the organization of the article, which gives you a better idea of how the author progresses with his argument (also not the case here). Regardless, the information-to-words ratio is still very high, so you cannot skip the introduction.

Step 4: Read the Conclusion
Children learn to read through works of fiction. Fiction requires the reader to go page-by-page through the novel. This is not the case with non-fiction. Yet most people will plow through an academic article just as they would The Hunger Games. That’s a mistake.

Like the introduction, the conclusion maintains a high information-to-words ratio. Conclusions usually summarize the contents of the article in a couple of paragraphs before providing ideas on how to expand on the research. Normally, “ideas on how to expand on the research” takes the form of admitting to weaknesses in the argument and guesses on how to shore up those problems. For example, I admit that my analysis is pure theory and there is no data to demonstrate that people actually behave in the way I expect them to. So if you are looking for ways to criticize the work, the author basically hands you ammunition on a silver platter.

Of course, if you read the article page-by-page, you probably got bored well before you reached the conclusion and thus never read it. Skipping from the introduction directly to the conclusion avoids this issue.

Step 5: Skim the Rest
Seriously, you aren’t going to get much out of everything in between the introduction and conclusion that you would not have gotten out of what you have already read. Skim everything else, only stopping if you find something interesting. (Some examples/case studies and graphs/charts are good reasons to pause.) Unless you find the research fundamentally interesting or you need to write a thesis on the subject, the nitty-gritty just is not worth your while.

Fin.
The brilliance of this method is that it you know exactly what you should do given the amount of time you have to study. Waited until the last half hour to start studying? Bust out the syllabus and read over the article titles. Have a couple hours? Go through the abstracts. Half a day? Now you can get the introductions in. A full day? Do the conclusions as well. One week? Now you can fit in everything. (Or just read the intros and conclusions. Inevitably, this turns into six days of Call of Duty and one day of studying, but whatever.)

Game Theory of Baseball Talk

Today, I’m giving a talk at the University of Rochester about game theory in baseball. You can see the slides by clicking here. Topics include throwing breaking balls with a runner on third base, optimal defensive positioning against a bunt, and how to catch a ball at a game. Enjoy.

The A La Carte Age of Textbooks

Apple’s textbook announcement from a couple months ago turned out to be a big bust. The new era of textbooks will not be about making the biggest, fanciest, or flashiest work around. This is the era of customizability.

Look at the textbooks on your bookshelf. They likely cost you $50 or more. But, honestly, how many of them have you read cover to cover? Probably not very many, if any at all. For the most part, these thousand page behemoths are useful for three or four chapters. Yet, when you purchase the book, you also must buy another twenty irrelevant chapters. And that sucks.

Now that we have a digital platform to work with, it is time to change the philosophy of textbooks. Yes, we can all keep publishing our thousand page behemoths. But we should also publish every chapter individually. Thus, students who purchase our textbooks buy exactly what they need and nothing more.

This model is impossible for the mass-produced paper model. A twenty chapter textbook would turn into 21 versions—one for each chapter and an all-inclusive edition. Publishers simply cannot anticipate consumer demands and produce the appropriate quantity of each version. So we are stuck with one pricy version.

Digital publishing does not face such constraints. For example, I have written three chapters of my Game Theory 101 textbook. The first chapter, Game Theory 101: The Basics is available all by its lonesome. (It’s also been Amazon’s best selling game theory book every month since July 2011—the month of its release.) So is the second chapter: Game Theory 101: Extensive Form. And the third chapter: Game Theory 101: Advanced Strategic Form Games. Students can purchase exactly what they need and nothing more. Or, they can purchase everything for a discounted price. The important thing is that the decision is in their hands, not the publisher’s.

If anything, Apple’s textbook announcement seemed to go in the opposite direction. iBooks still lives in a world that thinks ISBNs are necessary. I’m not sure what purpose an ISBN serves in 2012 when I can simply Google a title and find the book that way.

Actually, I take it back. I know exactly what purpose ISBNs serve: bureaucracy! A single ISBN costs $125. You can purchase 10 for $250, or $25 a piece. Regardless, ISBNs make a la carte textbooks unnecessarily expensive, as the publisher has to purchase a unique ISBN for each individual version. So Apple is going to be sitting this revolution out until it kills its ridiculous ISBN policy.

Death to the ISBN! Viva la a la carte!

Game Theory Tells Us ______

I hate it when I read a sentence that says “Game theory tells us _______.” For example, “game theory tells us prisoners should confess to crimes.” If you ever see something to that effect, know that the author doesn’t really understand what game theory is about.

The example is a reference to the prisoner’s dilemma. In that model, the prisoners only want to minimize jail time and defect accordingly, which leads them to an outcome that is strictly worse for both players than had they cooperated with each other.

But game theory does not generally tell us that the prisoners should rat each other out. Whenever we solve a model, we make some assumptions, formulate those assumptions into equations, do some game theory, and produce conclusions. Game theory is nothing more than overly glorified math. (Sorry to burst your bubble.) The nice part is that game theory maps assumptions to logically valid conclusions. So, really, game theory only tells us how to solve the game.

The assumptions do all of the dirty work. In the prisoner’s dilemma example, the critical assumption is that players only want to minimize jail time. Alternatively, they might want to minimize jail time and maintain their friendship. In my textbook, I formulate this as a stag hunt. The players want to keep quiet if and only if the other keeps quiet as well, while they players want to confess if and only if the other confesses as well. Consequently, in this game, the prisoners might stay silent. Changing the assumptions changes the prediction.

Bottom line: the assumptions tell us everything.

Here & Now Interview

Here is a link to an interview on Here & Now today, in which Robin Young and I discuss the game theory in Words with Friends:

http://hereandnow.wbur.org/2012/02/09/words-with-friends

The book in question is here.

A Piece from NPR’s All Tech Considered

NPR’s All Tech Considered called me a while ago to ask about the strategy behind with Words with Friends. Here’s the end product:

Winning Words With Friends: It’s All about Strategy

The article includes a bunch of links to my YouTube strategy videos. I ultimately compiled those strategies and more into this book.

New Working Paper on Bunting

If you are interested in the game theory of baseball, you should head over to my working papers page. I just uploaded a new article on optimal hitting strategy during a no hitter. As it turns out, a batter can never bunt with a no hitter in progress and still maximize his team’s probability of victory. The article has a full proof of the claim.

Pricing Your Book Intelligently

If you read through the self-publishing message boards, price point is a popular topic. Outside of writing process, selecting the right price might be the most important decision an author faces. I see a lot of different theories about what should determine price. Most of what people say is wrong. You only need to think about one thing when selecting your price:

Pick the price that maximizes your profit.

That’s it. That’s all. Just pick the price that maximizes the size of your monthly check. Picking anything else is stupid.

This is a tall order. It’s not immediately clear which price will generate the most money for you. Amazon offers 35% royalties on books between $0.99 and $2.98. Anything between $2.99 and $9.99 nets 70% (minus a usually negligible data transfer fee). So a $0.99 book needs to sell between five and six as many books as a $2.99 book to keep pace. Whether the higher demand from the cheaper price justifies selling at $0.99 remains to be seen and probably varies from book to book and genre to genre. An author might have to experiment with different price points before settling on the right one for him.

An author must also weigh the value of sacrificing extra profit from a single book to generate additional profit for rest of the author’s library. For example, the author of a trilogy might want to sell the first book at $0.99 and the other two books at $2.99. The cheap introduction brings more readers into the series, which in turn will lead to additional sales of the more expensive selections. I run a similar strategy with my game theory textbook, selling the first chapter at $0.99 and a larger version at $2.99.

Some people’s pricing strategy baffles me, however. Some people refuse to price books at lower prices because they do not feel it adequately reflects the amount of work the author put into the book. Absolute nonsense! The time you spend writing a book is a completely sunk cost once you are done. Placing a book at $4.99 because you think that is the value of the book is beyond ridiculous. The market determines the appropriate price. If selling at $0.99 or $2.99 brings in more profit, your resistance to lowering your price only makes it more difficult for you to pay your rent. It also leads to fewer people buying into your brand and therefore fewer sales of your other books. These authors need to get over themselves.

Just pick the price that maximizes your profit. It’s that easy…at least in theory.

The Power of a 1-Star Review

I am a little worried.

My larger textbook—Game Theory 101: The Basics and Extensive Form—has been doing quite well. This month, it has averaged a little more than ten sales per day. I know that might not sound a lot, but it projects to more than 3500 sales over the course of a year, or $5600. That is a lot of money for a graduate student living off of a small stipend. It’s also a lot for an academic book, most of which will get about 200 sales over their lifetime.

However, my book received its first review yesterday. It was not so good:

I regret this purchase, even at only $2.99. This is way below “101” level. This is about a 6th grade level…[i]f you are considering this product, you are better off reading the Wikipedia articles on Game Theory – it will be free and much more productive.

Needless to say, he gave me 1 star. Now, whenever someone searches for game theory on Amazon, they will see my book is worth one measly star.

How does that impact a seller? Despite averaging more than ten copies per day this month, I have not sold a single one since that time. Yikes.

The part of the review that I put above—the book is very basic. (There is more to the review that I believe is incorrect, but I excluded it from above to keep this conversation focused.) But that’s the point. I want the book to be simple. Game theory should be simple. There should be no magic or hand waving. The number one complain I receive from students taking game theory is that they have no idea how to solve games despite attending every lecture, section, and office hour they could. That’s why I created my video series, which eventually evolved into the textbook.

But his suggestion to read Wikipedia instead is ridiculous. Go read the Wikipedia page on how to solve for mixed strategy Nash equilibrium. Then try to solve the game in this video. Best of luck to you.

My personal feelings aside, consider an objective metric of the quality of a book: the rate of return. Amazon gives you a seven day window to return a Kindle book that does not meet your expectations. My small textbook has a rating of 5 stars (from a single, trollish review). In the last two months, it has a return rate of 0.69%, meaning a little more than 1 in 150 of these books gets returned. (For perspective, I’ve heard anything less than 5% is good, so I’m well below that mark here.)

The bigger textbook? The one with a 1-star review? Zero returns in the last two months. None. Buyers, it appears, like the larger textbook better. Yet it appears from the reviews that the opposite is true.

Edit: Looking back, I noticed I only calculated this from US sales. There was a single return of the bigger textbook from Amazon UK. So the return rate wasn’t exactly zero for the world, but it was darn close.

I’m curious why Amazon does not show the return rate for each book. Writing a review is time consuming, so you might get one review for every couple thousand books you sell. Consequently, the reviews don’t really tell you what the average person thinks of the book. Returning a book, on the other hand, is not costly, since it puts money back in the customer’s pocket. It is a better metric of a book’s value to its customers.

Anyway, hopefully my sales will return. In the meantime, if you have read Game Theory 101: The Basics and Extensive Form, please write a review. =) Whether you think it is five stars or one, at least the review will better reflect what everyone thinks as a whole.

Interview on Mind Your Decisions

Presh Talwalkar recently interviewed me on Mind Your Decisions. You can go directly to it by clicking here. If you are curious about how I got into this field in the first place or my thoughts on the way game theory is taught, this interview is for you.