Customer Reviews

The Predictioneer's Game: Using the Logic of Brazen Self-Interest to See and Shape the Future

5 star:		(35)
4 star:		(23)
3 star:		(22)
2 star:		(3)
1 star:		(2)

 

 

This review has been edited to correct some misstatements pointed out by the author. I was working from a prepublication version that did not have all the end-notes, nor a reference to the website. Moreover, the author's comment to this review adds some useful material. On the basis of that, I would raise my rating from three stars to three and a half if that were allowed, but my basic opinion has not changed.

This book is likely to teach you some fascinating and useful material, but I can't recommend it wholeheartedly because it may drive you crazy as well. The basic idea is simple. Experts know a lot, but are bad at making predictions about human affairs. Simple models based on quantitative game theory are more accurate, and even when they're incorrect they expand your thinking in useful ways. Moreover, these models allow you to simulate alternatives and generate outcomes as good or better than the best human strategists can achieve.

To evaluate this book, it's useful to separate that claim into two parts. I'm a quant, and therefore I think it's pretty well established that you make better decisions by asking experts what they know and letting a computer trace the logical implications than by following the experts' recommendations. I also accept that simple quantitative models do remarkably good jobs, and are only rarely surpassed by complex qualitative analysis. Anyway, if you don't accept those positions, there's no point in even opening this book. So to me and probably to you if you're still reading, asking experts simple questions with answers on a scale from 0 to 100 and combining the results in a reasonable way, is an excellent approach to most decisions. Call this the basic quant position.

The author goes further than the basic quant position in three respects. First, he makes much stronger claims for the superiority of his approach. Second, he advocates specific game theory analysis that involves complex modeling, as opposed to simple rules such as guessing that the outcome will be somewhere near the average opinion weighted by salience and power. He doesn't justify his methods as practical shortcuts that seem to work, he repeatedly claims that they are backed by science and logic, unlike alternatives. Finally, he goes beyond prediction to use his model to engineer outcomes. This is quant-on-steroids. It made me recall that John Nash, the most important game theorist with respect to this kind of work, maintained he was Emperor of Antarctica.

To paraphrase Dizzy Dean, it ain't megalomania if you can back it up. This is where it starts to get irritating. In the preface, before you get to page 1, he writes, "I've been predicting future events for three decades, often in print before the fact, and mostly getting them right. . . . I have made hundreds, even thousands, of predictions--a great many of them in print, ready to be scrutinized by any naysayer." This guy makes extraordinarily bold claims about his quantitative prediction ability, and he doesn't keep track of his record? He can't even recall within a factor of ten how many predictions he has made? Who decided he was "mostly right?" And why is anyone who wants to know the record a "naysayer?" Even if you did all the work, unearthed all the printed predictions (he only references a few in the book) and found he was zero for 200, he could just say he got thousands of other ones right, ones you didn't find. This sounds more like a Nostradamus defender than the "science" he's always claiming.

To bookend that frustration, by the end of the book, despite frequent promises, he has not revealed his model! He does have a version on-line that allows you to play around with it, and has more details on how it works, but still no clear, top-down description. In the book, you get hints and bits and pieces, but no clear explanation of how he arrives at his predictions. And this is not the only broken promise, there are frequent comments that he will "go into this further" or "provide more details," later; I can't find one example where he follows through. On the other hand, there are a few facts that get repeated far too many times.

Those two things would be enough for a lot of people to conclude he's a fraud. But there's an awful lot of good, clear, insightful analysis in between. He gives examples of political and diplomatic predictions he has made, discussing the inputs and basic form of the analysis. There are accounts of corporate and legal struggles where he maneuvered to an outcome favorable to his clients. He also applies the methods to history, to ask what might have happened. This is all fascinating stuff, and the data and conclusions speak for themselves. They show plausibly that this approach could work, it is practical to implement and it leads to conclusions that are surprising, but can shown to be logical. Without a lot more details these stories don't prove the model works, but they represent a coherent claim that it does.

However, this brings us to another problem. Some of the accounts are not credible. An account of how he maneuvered a fifth-choice candidate into a CEO job requires us to believe the board of directors split into five groups of three that agreed among themselves on the exact preference order for the five candidates, that these generated a cyclic preference order that included all five candidates (a mathematically possible but unlikely result only previously observed in game theory textbooks), that all this information was known with certainty beforehand and that none of the board members was smart enough to consider coalition-building or voting a second choice when it was clear a first choice vote would be wasted. Moreover, the Rube Goldberg scheme that worked seems far less promising than simple politicking to either build a coalition or change a slight preference.

This is the least credible account (unless you include the million dollars he was offered by Libya to engineer the removal of Anwar Sadat from power in Egypt, or the 10% of Zaire dictator Mobutu's external wealth offered to keep him in power), but none of the stories include basic information to allow fact-checking. In some cases the need for confidentiality is clear, but why no names of the government officials who hired him, or the partners at Arthur Andersen? Why is he protecting the agents of Libya and Mobutu? Does the brokerage firm that bought his advice in 1992 still insist on remaining anonymous? (And does it even still exist?) Some discretion is understandable, but this book reminds me of the fictional spies who have all labels removed from their clothing and possessions.

The final irritation is only one account of a missed prediction is given, and it is explained implausibly. The author predicted Hillary Clinton's healthcare plan would pass in 1994. He claims that the outcome was changed by the Rostenkowski scandal. That's hard to believe, since Rostenkowski resigned from his leadership position before the first bill came to Congress. Rostenkowski was not a strong supporter of healthcare reform. There's no doubt that his political skills would have been useful, had he chosen to push the plan, and the scandal did weaken the Democrats in general, but many other things happened that seemed to be at least as important. So either the prediction was dependent on lots of unpredictable events, and therefore should have been given as a probability distribution instead of a point estimate, or Rostenkowski was special, in which case the prediction should have been healthcare will pass or fail based on how Rostenkowski does. And why was the prediction not updated as the scandal worsened?

I know this is a long review, but I've only covered some of the bigger irritations. If you're an easy-going, tolerant sort who wants to learn some important practical and theoretical aspects of prediction, by all means read this book. If not, you might want a blood pressure check before you attempt it.

The Predictoneer's Game is a story about the author and the accuracy of his predictions than a discussion on the application of Game Theory to decision making. While the book contains descriptions of Game Theory, the majority of its pages are dedicated to the author explaining past predictions and how they were destined to be accurate. If you are looking for a book on how to understand and apply game theory for yourself, then you will need to look elsewhere in my opinion.

In many ways, Bruce Bueno De Mesquita's book is similar to those written by Nicholas Taleb. Both authors explain their thoughts and tools in a self-referential style that you will notice throughout the book. In more than one place, this tone can be a bit overbearing. This is more of a book for those that admire game theory rather than people who want to learn how to apply game theory. This does not mean that the book is not enjoyable and worth the read - it is just not what I expected.

The Predictioneer's Game is organized around the authors experience applying game theory to a wide range of situations from peace in the Middle East, North Korea nuclear disarmament to how to purchase an automobile. Each of these case stories is interesting examination of the events and the motivations leading to the actor's decisions.

The book does describe approaches for applying game theory. The appendix contains sample calculations all of which will get people started. These will help get people started in applying game theory to understand the behaviors and motivations of players in a decision, but according to the author there is more to the model than a table of influence, salience and position figures. As an example of this limited support, the author talks about weaknesses in the model, errors you should avoid, but it provides limited examples of how to fix it.

It is an interesting, but challenging read. This book requires a quiet room where you can concentrate on reading and understanding the game theory ideas by reading between the lines.

The stories and case studies are interesting and make this book worth the read. However, I had expected more from one of the founding minds and leading practitioners of game theory. So it's a good book, more from the perspective of reading about the forces behind current and past events than identifying tools that the reader can use to make their own predictions.

Please note, this book introduces the author's model for a general audience. If you seek greater math disclosure read his equally excellent PREDICTING POLITICS.

The author, a political science professor at NYU built his reputation developing Game Theory models that predict foreign affairs outcomes with a 90% accuracy as confirmed by a publicly released CIA study. This is amazing given that Game Theory is not effective at forecasting (Kesten Green/International Journal of Forecasting). The reason why the author's forecasting performance is so good is that he often uses many quantitative methods beyond Game Theory. For instance, his model to predict company fraud is a Logit Regression model. His work on the Cold War seems like a combination of Discriminant Analysis and Monte Carlo Simulation. His Global Warming regulation projection seems also based on Monte Carlo simulation.

When he first developed his model in 1979, he was shocked that it would often contradict his own expert opinion and most often turn out right. The Defense Department made him predict 17 different foreign negotiation issues. His model got all 17 outcomes right!

He states that a computer can readily grasp what no human mind can because his model can capture readily all the layers of information being traded among decision makers. With just 10 players, this number of info exchange is already 10 factorial or 3.6 million exchanges.

In chapter 4, when dealing with North Korea nuclear containment negotiation he explains the underlying basics of his predictive model. He states his model needs only four inputs (converted from English into quantitative scores) to predict an outcome:

1) The relevant players influencing the outcome;
2) The players' position on the issue (No = 0, Yes = 100);
3) How much they care about the outcome (salience; does not care = 0, cares a lot = 100); and
4) How influential they are (with an Influence weight).

To figure the answers to the above questions, you talk to the relevant experts or research the media (The Economist, NY Times, etc...).

Regarding the North Korea situation with 6 defined potential outcomes, the author first generates a baseline forecast as shown in Appendix I. He calculated the weighted average position of all players that came out to 60 corresponding to "slow reduction in nuclear development and capability in exchange for U.S. granting diplomatic recognition." To refine the estimated position of 60, he makes manual adjustments reflecting minute shift in players positions. And, he reruns the numbers until all realistic change in players' positions are accounted for. And, this gives him how a position on an issue evolves over several rounds of negotiations.

He has used his model in a wide range of circumstances including: foreign policy negotiations, company mergers, company fraud, and litigation. And, the related predictions were surprisingly accurate.

The author is quite a historian. His analysis of historical flows from Christopher Columbus to nowadays and his many related model applications are pretty fascinating. He uses the data from the Correlates of War Project, an amazing resource for budding historians.

Later he moves on to contemporary issues to develop forecasts with his students at NYU. The one on Global Warming green house gas regulation is particularly interesting. His resulting model outcome that any regulation implementation will fail makes sense. His optimism that ultimately technology will solve what regulation could not is also commonsensical.

On the other hand, the author states his model did not predict the failure of the Clinton Health Care plan. But, he advances it was solely due to Rostenkowski retiring from Congress. But, the Clinton Plan never gathered adequate support in Congress because Hillary was unwilling to compromise (no way to pass a legislation through Congress).

Sometimes his proposals appear far fetched. His implementation of altering Board members voting system for a new CEO into a Byzantine single elimination system is hard to believe. Why would any Board member agree to such burdensome procedure? Also, his Palestinian vs Israeli tourism revenue sharing proposal is probably Utopian. Nevertheless, those few minute rebuttals do not distract from the overall excellence of the model.

I bought this book because I had seen de Mesquita talk at TED
[...]

I found the TED talk intriguing but annoyingly devoid of the details that would help one understand how the model really works. This was part of a shrewd strategy employed by de Mesquita to make me buy his book. So, de Mesquita 1 - Stefano 0, I'll definitely give him that.

The book is *worse* than the TED talk: nowhere in the book de Mesquita gives you an honest explanation of how the model works. To begin with, it is even unclear what the parameters of the model are. In some sections of the book (Chapter 4, an analysis of North Korea's nuclear strategy) it seems as if there are three parameters for each decision maker involved in a given issue: 1) what is their preferred outcome (expressed numerically on a sliding scale) 2) how influential they are; 3) how much they care about the issue. In other sections (Chapter 11, analysis of climate change international negotiations), it seems the parameters are four: the previous three plus 4) desire for agreement.

de Mesquita's claims are that: i) a diligent subject matter expert can reliably identify the decision makers involved in an issue and assign to each of them a value for these three or four parameters; ii) the model takes this parameter assignment as input and charts the interactions among decision makers identifying *the* outcome with the most power behind it (the outcome preferred by the decision makers that collectively yield the majority of power); iii) the model is stable (i.e. small variance in the assignment of initial parameters does *not* result in drastically different outcomes).

There is one big problem: the book does not contain a proof/explanation of claim ii). You will look in vain for a description of how the interaction among decision makers is modelled. de Mesquita goes through numerous real life examples telling you about certain things happening or not happening after a certain number of negotiation rounds but you are *never* shown how to compute the collectively preferred outcome at step N as a function of the values of parameters 1)-4) at step N-1. Since I bought the book to learn precisely that, I am completely disappointed.

To add to the confusion, there are many charts (e.g. chart 10.9, prediction of power shifts in Iran) in which de Mesquita seems to imply that what his model predicts is the value of parameters 2) and/or 3) for some of the actors. in other words, what had been previously defined to be an *input* to the model is now treated as its *output*.

For the rest, the book is stuffed with completely unsupported claims of the form "as we can see, the model predicts X". No, Bruce. We *cannot* see that because you are not showing any of the details that would be required to make that claim. And it is really mysterious why de Mesquita would not be showing the details: apparently he teaches the model at NYU so its details can hardly be a trade secret.

That said, although the book is a grotesque collection of non-sequiturs, they are highly interesting non-sequiturs. de Mesquita's point of view on many historical or current events is much less developed than it should be for the book to be valuable, but definitely insightful: he seems like the kind of guy who could enliven a dinner party with really good stories worth additional thought and research. My personal favorite is his game-theoretical analysis of the Worms Concordat on the Pope-King procedure for the selection of bishops (Chapter 11).

Finally, one pet-peeve: who the heck designed the charts and who approved them for print? some of them (10.8B, page 198) are simply illegible (go ahead, I dare you). Some seem to have been put together by somebody who had to learn the basic of the Microsoft Office suite in a hurry. For example: was it really so hard to justify the boxes on figure 11.1 (page 206)? who drew the lines of the convex preference sets in fig 2.1 (page 22)? These remarks sound catty but diagnose a problem: nobody who were really serious about professional data analysis would be caught dead producing such cartoonish disasters. To me this is a symptom that this book was the outgrowth of a PowerPoint culture, one in which what's important is to catch the attention of a distracted decision maker with an-intriguing-but-please-not-too-detailed slide. This visual slovenliness dovetails perfectly with the intellectual sloppiness of the rest of the book.

Bottom line: skim it if you find it at the library but don't buy it. There's little to learn from here and none of what is advertised.

If you buy this book looking for an insight into the practical application of game theory and an informative guide on the application of quantitative analysis to political science - prepare to be deeply disappointed.

Bueno De Mesquita is mostly interested in explaining to the reader how extraordinary clever he and his model is - but gives virtually nothing to the reader who is interested in the mechanics of the application of the theory. The book is mostly composed of superficial and highly subjectively presented case studies, with plenty of self-congratulatory prose on how extraordinarily accurate his modeling has been proven to be but with really nothing that you can sink your teeth into as to the details.

The few cases he does go into detail with are so simplistic that they really don't need a computer model applied - just a pencil and the back an envelope.
So, at the end of the book, you are left with the feeling of being cheated - the book promises some thought provoking and original perspectives - but the only viewpoint that I was left with was the one that attending a lecture by this egotistical academic would require a bucket (to vomit into) and an extra large room to accommodate his inflated ego.

Professor Bruce De Mesquita is clearly "a beautiful mind" and appears to have made significant contributions in the realm of poltical forecasting. As an economist, I appreciate and resonate with his line of thinking in the book - and often think more people should consider arguements like his.

I have to say over all, though, I did mind myself disappointed by the book. I felt the writing to be disorganized often and his analysis of the games which he reflected upon to be convoluted at times. He continually gives examples only to withhold names (to protect his clients) but this has the effect of watering down his perspective and simplfying his games to seem unrealistic (for example something about his CEO game seems fishy - I'm not sure I buy that the board could so easily be persuaded to accept the voting format which he proposes). He at multiple times states that a preditioneer must make public predictions and allow themselves to be proved wrong - but even when he claims that he is going to engage in such practice he writes about his predictions (and his students) and then confirms that they were correct by events following the prediction. This gives the appearance that he may have choosen only the predictions that come true. He spends enormous time describing how awesome and successful his model is - and by and large devotes little time to actually decribing/presenting his model in a precise (mathmatical) way.

Overall some interesting ideas based on sound ideology, but the presentation was seriously lacking and thus got in the way of the ideas.

The book begins with good tips for buying a new car, followed by a nice intro about the basics of game theory, particularly for the laymen. Therefore, those readers already familiar with game theory can jump directly to chapter 4. From here on I bear patiently asking myself "where is the beef?"

In the end Mr. De Mesquita does not shows how he actually applies game theory to make his almost always accurate predictions, particularly within the field of political sciences; he will just give a hint, and talk about his models, under which circumstances this approach is useful, how his students use it, and why he his predictions failed (the few times he did). If his algorithms for game theory are proprietary (certainly they must be), then why did he bother to write this book? Ego? Marketing for his services? Quite a deception.

Predictioneer is a very well written book. Game theory is introduced and developed in an understandable discussion of root problem identification, quantifiable stakeholder assignment and the mathematical process to derive an objective prediction input to the decision making process. Interesting contemporary problems are discussed. The reader will understand the `why' of game theory algorithms but not the `how' beyond the most basic and simplistic examples. Like all mathematical models, the devil is in the details and we don't get a peek at what's in the box.

Positives:

Good primer to operational game theory

Good tips on the methods and considerations for identifying and quantifying stakeholder blocs, stakeholding individuals and their ability to influence the outcome, salience or self interest and positional authority.

Interesting storytelling and contexts of game theory in operation.

Negatives:

I couldn't avoid feeling that the book is an advertisement for the author's services in a number of places.

The game theory example of `why Sparta fell so quickly' was a good example of how game theory inherently includes the subjectivity and information of the model's driver. This game theory post mortem failed to consider the uniquely effective contribution of Epaminondas revolutionary strategy of `total war'. Sparta disappeared from history but not for the game theorized reasons. GI/GO lives!

Game theory application still has a long way to go to be the decision engine of choice but I could be wrong ;)

Despite the critique below, I consider this book to be a valuable read with many intriguing ideas presented by the author.

While it's admirable that this author capped his intriguing book by putting his prediction model to the test, when it came time to put the pedal to the metal, Mr. Mesquita hit the brake pedal instead.

In chapter 10 of "The Predictioneer's Game", Mr. Mesquita dared to be embarrassed by putting his new and improved prediction model to the test by predicting various important events in Pakistan such as future distribution of political power. Unfortunately, many of his and his students' predictions failed to materialize.

For starters, there was no military coup between February 2009 and July 2009. Mr. Mesquita's model had predicted a decline in Pakistani military power that would serve as the impetus for such a coup during the period stated, which as of late October, 2009 has yet to occur.

His model also predicted "The new president [Obama] is not likely to do much of anything about the rise of terrorist influence within Pakistan at least through the end of 2009." Wrong again. A January, 2009 Times Online article titled "President Obama orders Pakistan drone attacks" had this quote on the subject: "Missiles fired from suspected US drones killed at least 15 people inside Pakistan today, the first such strikes since Barack Obama became president and a clear sign that the controversial military policy begun by George W Bush has not changed." Mr. Obama, as it turned out, pursued a more vigorous effort of hunting down Taliban and Al Qaeda operatives and senior leaders in Pakistan as evident by the following quote from an October 24, 2009 "The Long War Journal" article: "The US has carried out 45 airstrikes inside Pakistan so far this year [2009]. In all of 2008 [when Bush was in office], 36 strikes were carried out."

Mr. Mesquita stated, "The analysis shows that substantial domestic pressure is likely to push for cuts in American aid to Pakistan." Wrong on that account as well. In an October 21, 2009 Newsweek article titled "About Those Billions", Ms. Katie Paul, the author reported "2009-2014: A new five-year, $7.5 billion assistance package was passed by Congress in September and signed by President Obama in October..."

Lastly, the author's prediction model showed "After June 2008, their (the Pakistani government's) approach along with everyone else's, including the Pakistani military and the U.S. is projected to be little more than rhetorical opposition to the militants with almost no serious commitment to go after them.... The air just pours out of the anti-militant effort." This prediction was supposed to hold through December 2009, which was the end of the forecast period for this project. Unfortunately, this too did not pass the muster. With the Taliban threatening to overrun the country, a serious offensive to push back the insurgents began in earnest in May 2009 in Swat Valley by the Pakistani military aided by U.S. drones, intelligence and surveillance videos (L.A. Times, The Washington Post). On October 24th, ABC News reported the Taliban was on the run from the Pakistani Army in its latest offensive in South Waziristan with 24,000 troops committed to the operation.

None of the glaring failures I have mentioned above from the predictions stated in chapter 10 are to discredit the author and the field of game theory. I am a big believer in the value of this field of study and its application to foreign policy and business cases. However, I am not as convinced as the author about the accuracy and simplicity of "The Predictioneer's Game". The models the author described harbor much subjectivity in such input factors as the assignment of salience and influence of stakeholders. Game theory is as much an art as it is a science.

A quick read of the cover blurbs indicates that the author can predict future political and economic events accurately, like Tiananmen Square or the Gulf War. It seems like a fantastic claim. After reading the book, I cannot say whether the author can do this or not. He seems to have a following, but in this book he does not explain exactly how he figures out what will happen tomorrow. Perhaps the math is too complex for a popular work like this, but he doesn't go into the math at all. Maybe he wants to protect his proprietary algorithms, which are the source of his income. He does make some predictions about future events, which will give him credibility if they turn out the way he predicts. He also talks about past predictions that he has made, but I cannot vouch for the accuracy of those claims, because my only source is the author himself.

I could not find the word Predictioneer in my Webster's second edition. Bruce Bueno De Mesquita has forged a new career path by force of his talent. He uses game theory models, based upon logic and evidence, free of ideology and partisanship, to craft solutions and predict outcomes to real world problems. In this book he gives a high level overview of his methods, tells stories about his past accomplishments, and makes predictions about the future. This is a quick and interesting read, worth your time if your interest is piqued.

Mesquita plays the game of what if with some historical events, from Spain's sponsorship of Columbus to the origins of World War Two. The most fascinating application of his model is on current problems, and his projections for our future. His comments on the political situation in Pakistan and the future of the environmental movement were fascinating. Time will tell whether his claims come true.

The book is well written, engaging and entertaining, without being overly technical or tedious (my initial fear). The author is a good storyteller, and he reveals some interesting details about his past work, without revealing the identity of his customers. I was disappointed that he does not describe the details of his work, the math or the computer programs, in any meaningful way. I felt that he hides the real substance of his insights beyond the reader's grasp.