No Gravatar

Combinatorial Prediction Markets – by Robin Hanson

Video + Slides

Slides from Hanson&#8217-s site – PPT file

Folks, this is great stuff. I may blog about it, again, later on &#8212-if I find time.

You can blog about it on Midas Oracle, if you wish. Or comment on it, just below. Do register yourself.

Subsidizing real-money prediction markets and real-money conditional prediction markets

Should Google subsidize the Lunar X Prize contract on InTrade?

John Salvatier,

Our good friend Bo Cowgill might have already re-created those prediction markets on Google&#8217-s internal prediction exchange at a marginal cost of zero US dollar. No need for him to &#8220-subsidize&#8221- external prediction markets.

[As an appendix, I precise that I am in favor of opening the enterprise prediction markets to external traders, for some questions.]

Subsidizing prediction markets is an old Robin Hanson idea that carries quite a heavy price tag.

Conditional prediction markets is a great idea on the paper. Many people (e.g., Mike Linksvayer) like the idea. However, here is what the uncritical Robin Hanson fanboys blogging on Overcoming Whatever won&#8217-t tell you:

  • The first problem is that nobody trades those things.
  • The second problem is that subsidizing those conditional prediction markets costs an arm and a leg.
  • The third problem is that no major news media outlet has ever quoted the prediction market prices / probabilities generated by those conditional prediction markets.

Peter McCluskey could have rent a French mistress (or a French gigolo) for a full year with all the money he is spending on Robin Hanson&#8217-s idea. Or vaccinated the whole African continent against Malaria. See Peter&#8217-s comment, at the middle of the webpage, here.

Philanthropy and prediction markets are not mixing well &#8212-yet.

Prediction & Decision Markets – Robin Hanson Edition

No Gravatar


Prediction &amp- Decision Markets – (PPT file) – by Robin Hanson – 2008-04-17

And, that one, for your curiosity&#8230- really fascinating ( :-D ):

Evolutionary Game Theory of Interstellar Colonization – (PPT file) – by Robin Hanson – 2008-05-26

Previous blog posts by Chris F. Masse:

  • Kudos to BetFair’s e-mail marketing team?
  • Conditional prediction markets about oil price and SegWay sales… Like the idea, Robin Hanson?
  • Justin Wolfers [*] is the most cited prediction market economist
  • The Orb @ Texas Tech University

The Marketing Of The Reading Of The Public Prediction Markets = What Robin Hanson has deep trouble with, and what the prediction exchanges (e.g., InTrade-TradeSports, BetFair-TradeFair) havent fully computed yet

No Gravatar

Robin Hanson on &#8220-silly&#8221- research topics:

[M]ost people think futarchy (government by [prediction] markets) is silly, even though most think it has a decent chance of performing well […].

Decision markets and decision-aid markets are 2 great concepts pushed by Robin Hanson, the world&#8217-s #1 researcher in the field of prediction markets. But they are just inventions, not innovations. What is important is to find out which population segment or which class of business executives find this stuff productive and helpful.

In that perspective, his presidential prediction markets at InTrade are good ideas, and the liquidity there (helped by an AMM) is decent enough. But they are just betting supports, right now. I haven&#8217-t seen any opinion leaders taking them as a trusted source of information, which is the damn goal. We will see whether that comes true in the future.

If Robin Hanson were really serious in finding a killer app for his concept of decision-aid markets, he would of course come up with conditional prediction markets in the realm of sports, which is the most popular topic in the real-money prediction markets. Alas, I often have the impression that the academics in the field of prediction markets have profound disdain for sports prediction markets.

Robin Hanson on seeking decision advice:

[…] We rarely seek out advice, and when we do it is usually on much smaller decisions. […] One reason we avoid getting advice is that it lowers our status relative to those who give advice. Of course this is also makes asking for advice a good way to flatter and supplicate. Not sure if this explains the puzzle though. But all this doesn&#8217-t seem to bode well for fielding decision markets on the biggest organizational decisions.

Allow me to digress from there. I think that the reading from the prediction markets is like an advice &#8212-in that you have to accept the market message as an authority. If you are an expert with direct access to primary sources of information, I don&#8217-t think you&#8217-d rely on the message from the public prediction markets (which are information aggregation laggards). The big mistake from Robin Hanson and the others has been to sell the public prediction markets as tools for the decision makers. That could happen, but marginally, I believe. Experts and decision makers will firstly want to rely on their primary sources of information and on their analysis.

I think that the population segment which is the more likely to appreciate the consumption of market-generated probabilities would be composed of people who want a chopper view of world events. Prediction market journalism should satisfy this dashboard need.

[Please note that the thoughts expressed above refer to the public prediction markets (as stated in the post title –think BetFair-TradeFair, InTrade-TradeSports, Betdaq, HubDub, NewsFutures, and Hollywood Stock Exchange) —not the enterprise prediction markets, which is a horse of another color.]

Robin Hanson on decision-aid markets:

I don&#8217-t recall ever turning down a chance to consult on prediction markets for a Fortune-500 company. If you know of an opportunity that I&#8217-m missing, do let me know.

Doc, are there more Fortune-500 executives and managers attending a conference on extra-terrestrials or a conference on finance? :-D

Merger Markets on Microsoft-Yahoo

No Gravatar

HP began to explore prediction markets in 1996, but did not even consider applying them to the 2002 HP-Compaq merger. Similarly, Yahoo and Microsoft are two of the companies mentioned most often as being involved in prediction markets (along with their main competitor Google), but I&#8217-ll bet none are considering the by-far-most-valuable markets they could create, on their just-announced proposed merger.

Decision markets could say whether this merger is good for shareholders, by estimating the combined stock price given a merger, and given no merger. Similarly, decision markets could say whether this merger is good for these firms&#8217- customers, by estimating the price and/or quantity of web ads given a merger, and given no merger. This might help convince regulators to approve the merger.

My main doubt here is whether ad price and quantity are good enough measures of the merger&#8217-s social benefits – what other outcomes could such markets estimate, to speak more clearly? And this is a very clear demonstration that these companies are just not serious about finding the highest value applications of prediction markets.

Cross-posted from Overcoming Bias.

Predictocracy = Market Mechanisms for Public and Private Decision Making

No GravatarRobin Hanson:

[&#8230-] The main problem with using [Michael] Abramowicz&#8217-s book as a &#8220-technical manual&#8221-, however, is that he&#8217-s never actually seen, much less touched, most of the blocks he describes. His conclusions are not supported or tested by math models, computer simulations, lab experiments, field trials, nor a track record of successful past proposals – it is all based on his untested intuitions. And he doesn&#8217-t seem inclined to do any such testing himself – he hopes his book will inspire others to do that. There is of course a spectrum of rigor in how solidly one can support a claim. Most business decisions are based on far less rigor that elite academics often demand, and there is surely a place for &#8220-brainstorming&#8221- speculation. Compared with most academics, I admit I have often been more than toward the speculation end of the spectrum, though I have tried to test my speculations via math models, lab experiments, field trials, and have arguably collected a modest track record of success. [&#8230-]

Michael Abramowicz&#8217-s response:

[&#8230-] The incentives provided by two of my technical proposals (the decentralized subsidy approach and the nobody-loses prediction market) are sufficiently straightforward to me that math seems superfluous to me, though I agree that field tests comparing these with alternatives would be useful. Two of the proposals (the text-authoring market and the market web) could certainly benefit from experimentation, but the software needed to implement them would be considerably more complicated than what is needed for existing prediction markets. [&#8230-]

Robin Hanson&#8217-s second take.

Michael Abramowicz&#8217-s post.

Robin Hanson on futarchy vs. predictocracy.

Michael Abramowicz.

Robin Hanson.

I will blog about this book, once I have read it, in the near future.

Predictocracy = Market Mechanisms for Public and Private Decision Making

All the book is online, at the web address above. You can also buy it at bookshops, or at Amazon.


Read the previous blog posts by Chris F. Masse:

  • Many people twitter on prediction markets.
  • Folks, when you have something important to say, write up a full post, not a comment.
  • Prediction Market Journalism
  • TechCrunch is 221 times bigger than Midas Oracle.
  • Earthquake measuring 9.0 or more on Richter scale to occur anywhere on or before December 31, 2008
  • Why Midas Oracle (and not TV news shows or print newspapers) will dominate the future.
  • The Six Degrees Of Separation

The Future of the Prediction Markets

No GravatarEven a prediction market fanboy feeds on the polls &#8212-first and foremost. Steve Dubner, the journalist and co-author of Freakonomics, is, along with his two blog colleagues (Steve Levitt and Justin Wolfers), a strong supporter of the prediction markets. They all have blogged enthusiastically on prediction markets, since the inception of their blog. Steve Levitt calls the InTrade-TradeSports people in Ireland his &#8220-friends&#8220-. Wow. The Freakonomics blog even has a &#8220-Prediction Markets&#8221- blogroll (i.e., a list of external weblinks to the best resources on prediction markets) &#8212-where, of course :-D , Midas Oracle is the cornerstone. :-D To sum it all up, Steve Dubner and his colleagues are true believers in the predictive power of the prediction markets. Good.

Until you analyze this Steve Dubner&#8217-s Freudian lapsus:

[&#8230-] (Fascinating aside: according to a recent Times poll cited in the article linked above, the Florida G.O.P. race is as of now a virtual deadlock between four candidates: Huckabee, Giuliani, McCain, and Romney. This will almost certainly shift as a result of interceding activity, but still, what a spectacle!)

Which forecasting tool does Steve Dubner use to get a sense of the political race du jour? Not the prediction markets&#8230- but the polls.

That speaks volume on the nature of the prediction markets, as forecasting tools.

  1. The polls and the surveys are the primary purveyors of crucial political information, which the political analysts (and&#8230- Steve Dubner :-D ) use to write their reports.
  2. The political prediction markets feed on polls, aggregate them (and other disparate pieces of information), and delivers &#8220-the consensus opinion in a much finer and dynamic way than all the amorphous media buzz&#8220-. They are secondary forecasting tools. They are taken seriously only by the free-market believers (like us) &#8230- but, as the Steve Dubner&#8217-s quote shows, even the prediction market true believers check the polls first.

I think that:

  1. Because of its nature, the prediction market prism, which quantifies the impact of the news, will never be the dominant forecasting tool.
  2. Prediction market journalism will remain on the fringe. It should be developed to serve a targeted audience &#8212-the free-market believers, the busy people, and the event derivative traders.
  3. Conditional prediction markets (a.k.a. decision-aid markets) will never be taken seriously by the decision makers and the public. Robin Hanson&#8217-s tool is very smart, though. Applications should be found within the community of free-market believers, rather.
  4. Enterprise prediction markets (a la Google, Inkling Markets, Consensus Point, etc.) are very interesting because they reflect inside information that can&#8217-t be conveyed by the corporations&#8217- internal media.
  5. The prediction market approach (embodied by InTrade) will always be weaker than the betting exchange approach (embodied by BetFair). I still believe, though, that each prediction exchange should employ both &#8212-which is not the case, right now.


ADDENDUM: In fairnesses to Steve Dubner, it should be noted that, in the same post, he put up a blunt challenge to polling methodology:

[&#8230-] As for the discrepancy between the two polling questions, take note: that’s the difference between a fill-in-the-blank polling question (i.e., “Which is the most important problem …”) versus a leading polling question (”Do you view crime as a ‘very serious problem’?”). The next time you read a poll and think it may be hinky, ask yourself what question the pollsters actually asked. [&#8230-]

Previous blog posts by Chris F. Masse:

  • Red Herring’s list of the top 100 North-American high-tech startups includes Inkling Markets —but not NewsFutures, Consensus Point, or Xpree.
  • Professor Koleman Strumpf explains the prediction markets to the countryland people.
  • Professor Koleman Strumpf tells CNN that a prediction market, by essence, can’t predict an upset.
  • Time magazine interview the 2 BetFair-Tradefair co-founders, and not a single time do they pronounce the magic words, “prediction markets”.
  • One Deep Throat told me that this VC firm might have been connected with the Irish prediction exchange, at inception.
  • BetFair Rapid = BetFair’s standalone, local, PC-based, order-entry software for prediction markets
  • Michael Moore tells the Democratic people to go Barack Obama in Pennsylvania (a two-tier state), but the polls and the prediction markets tell us that that won’t do the trick.

Decision markets are markets where speculators set prices that estimate the consequences of a decision.

No Gravatar

&#8230- writes Robin Hanson (who is into prediction markets since the late 80s, and who is the smartest scholar on this topic).

I would call that &#8220-decision-aid markets&#8220-, then. And I would have the term &#8220-decision markets&#8221- define the strongest form of this tool, envisioned originally by Robin Hanson (PDF file) &#8212-that is, when the execution of the market-generated decision is compulsory.