Warming: Works only 50% of the time. (The other 50%, the results come from the current Google Search.)

http://209.85.225.103/

-

BetFair was created in 1999 and started off in 2000. Since that time, 2 major things arrived on the world scene. Number one, we have seen the emergence of the prediction market approach. Number two, the Web has taken our lives, and Google has become the dominant Internet search engine. Here are how these 2 major trends are affecting BetFair negatively.

1. Decimal Odds (a.k.a. Digital Odds). - The prediction market approach means that we attack the public with the news and their associated probabilistic predictions, expressed in percentages, where high prices mean high probabilities of happening. BetFair, at the contrary, approach the public with a betting universe and an arcane vocabulary (“backing” and “laying”) where low prices mean high probabilities of happening. That is totally counter intuitive.
2. Non-Indexable Prediction Market Webpages. - Like it or not, Google is now the world’s #1 media. We “google” anything, first thing in the morning. None of the BetFair prediction market webpages can be indexed by Google and the other Internet search engines. That means that BetFair is missing out, in my estimation, on hundreds of thousands of Google visitors each year. Those Google visitors will favor other prediction exchanges (e.g., HubDub) whose prediction market webpages are indexed naturally by the Internet search engines.

-

The British, who drive on the wrong side of the road, don’t have the 2 most important keys of the future.

-

…his own majesty and my very own prediction markets guru David Pennock; yes, he is such a geek but modest.

-

David Pennock

-

July 6th, 2008

Commodity Futures Trading Commission
Three Lafayette Centre
1155 21st St. NW
Washington D.C. 20581

Attention:
Office of the Secretariat- secretary@cftc.gov

Reference:
Concept Release on the Appropriate Regulatory Treatment of Event Contracts
73 FR 25669

-

Just a technical note, before I give you my thoughts. In the following, I call “prediction market” the specific market where one particular event derivative is traded. (For instance, the “Barack Obama will be elected US President in November 2008″ prediction market.) And I call “prediction exchange” the general marketplace where many prediction markets (on political elections and other events) are traded. (Hence, I call HedgeStreet a “prediction exchange”).

Please, allow me to give you my thoughts on the subject of real-money prediction exchanges:

-

ABOUT THE INFORMATION AGGREGATION MECHANISM, FORECASTING, THE LIQUIDITY OF THE SOCIALLY VALUABLE PREDICTION MARKETS, THE DEVELOPMENT OF A US-BASED PREDICTION MARKET INDUSTRY, AND THE PROTECTION OF RETAIL TRADERS

-

The information aggregation mechanism functions well only if there are enough traders. Probabilistic predictions (which are of interest of the economists cited in the CFTC’s concept release) are generated only when there is enough liquidity, that is, when many traders come speculating on an event derivative market (e.g., on the topic of the next political election). Just because forecasters are interested in a topic and want to generate a market-based probabilistic prediction does not mean that traders will flock en masse. Market-generating forecasting is an offspring of the trading activity; if you have too little liquidity, you don’t have any trustworthy probabilistic prediction.

The socially valuable prediction markets should meet 3 criteria:
- their contracts should be very well drafted, so that the probabilistic predictions generated would be useful to society;
- a sufficient number of traders should like the topic;
- there should exist advanced, primary indicators which traders can follow to get early information (e.g., polls, among other sources of information, in the case of prediction markets on political elections).

Here’s a counter example. Yahoo! Research scientist David Pennock (one of the most active and well regarded researchers in this field) has created a set of prediction markets regarding the percentage share of web searches made in the US in 2008, for each Internet search engine (Google, Yahoo!, etc.) That would be extremely valuable, on the paper. Unfortunately, those sets of prediction markets have attracted only a fistful of traders:
http://www.intrade.com/aav2/trading/tradingHTML.jsp?evID=78364&eventSelect=78364&updateList=true&showExpired=false
Hence, no trustworthy probabilistic predictions were generated.

The CFTC should take with a grain of salt the 2008 petition organized by the American Enterprise Institute
http://www.reg-markets.org/publications/abstract.php?pid=1276
that states that “not-for-profit research institutions” and “government agencies” should be allowed to run US-based, real-money prediction exchanges, for the good of society. Just because an organization is smart and fascinated by the prediction markets does not mean that its executives and managers will be capable of drawing traders. Obviously, prediction exchanges should be run by trading specialists and event derivative professionals, and properly regulated. No good will be done by the CFTC if amateurs are allowed to run un-regulated, real-money prediction exchanges.

I see 2 important keys for the development of socially valuable prediction markets.

a) The socially valuable prediction markets (which are not very popular, other than the ones on political elections) should be organized by the generalist prediction exchanges that draw traders en masse because they offer prediction markets on very popular topics.

Sports is a popular topic. If the CFTC go to the website of TradeSports http://www.tradesports.com/ , they will see that TradeSports links, on its frontpage, to the InTrade prediction markets at http://www.intrade.com/ and, thus, send the TradeSports traders to the InTrade prediction markets, which is obviously good for InTrade’s liquidity in general, and especially good for InTrade’s socially valuable prediction markets. In the same manner, the prediction markets on political elections organized by BetFair UK http://www.betfair.com/ are located within their central prediction exchange that is mainly devoted to sports.

The hard fact is that the most popular topic among individual traders (the retail customers of the prediction exchanges) is sports. As long as US laws and regulations won’t allow US-based, real-money prediction exchanges to organize prediction markets on the topic of sports, many US event derivative traders will give their business to offshore, real-money prediction exchanges who accept to take money from US residents (as it is the case with TradeSports-InTrade Ireland).

I understand, though, that the CFTC is working under a jurisdiction that presently outlaws prediction markets on sports.

b) The executives of the popular, real-money prediction exchanges should be willing to create socially valuable prediction markets by collaborating with outside researchers who specialize in certain verticals.

As of today, InTrade is the only real-money prediction exchanges that fill these 2 criteria —a) and b). InTrade’s executives and managers have deployed a considerable effort to create and run an impressive number of socially valuable prediction markets.

BetFair UK have chosen not to develop socially valuable prediction markets, alas —other than those on UK politics, which are well developed and of high social utility. And HedgeStreet does not have yet the CFTC’s stamp of approval to run markets of event derivatives non-financial topics, since that’s the purpose of the May 2008′s concept release.

The economists Justin Wolfers, Eric Zitzewitz, Robin Hanson, Koleman Strumpf and David Pennock (among others) have collaborated with InTrade Ireland to frame interesting questions. Obviously, the research institutions which those economic scientists are affiliated with (e.g., universities, colleges, business schools) have no business running real-money prediction markets.

If the “not-for-profit research institutions” and “government agencies” want to develop socially prediction markets, then they should do it in cooperation with established, popular, regulated, real-money prediction exchanges, who know what they are doing.

(In passing, I fully support Tom W. Bell’s point made in the 5th paragraph of his petition. The CFTC should not favor the not-for-profit prediction exchanges at the expense of the for-profit prediction exchanges. Tom W. Bell’s comment to the CFTC has not yet appeared on the CFTC website, as I type this. http://agoraphilia.blogspot.com/2008/07/lets-tell-cftc-where-to-go.html )

As I said, I follow the prediction market industry since 2003, and the 2 most common mistakes I see made by the people proposing brand-new socially valuable prediction markets are that:
- they forget that the event derivative traders should have fun;
- they forget that, for each prediction market, there should exist advanced, primary indicators that traders should rely on to inform their trades.

I want to tell the CFTC that most people who talk about creating brand-new socially valuable prediction markets are totally unaware of these 2 basic rules.

In the beginning of this comment, I said that prediction markets are forecasting tools (and, hence, decision-support tools) if, and only if, there is sufficient liquidity. I also noticed that the world’s most liquid socially valuable prediction markets are offered by 2 exchanges (TradeSports-InTrade and BetFair) who use popular prediction markets (on sports, the fact is) to support the marketing of less popular, socially valuable prediction markets. (After making that argument, I acknowledged that the CFTC currently works for a legal environment that prohibits prediction markets on sports.)

My point here is to emphasize the uber importance of liquidity on socially valuable prediction markets. In my view, the best situation is when a big, generalist, real-money prediction exchange organizes socially valuable prediction markets and helps them to thrive. Only InTrade Ireland has done that, so far. My suggestion to the CFTC would be to create a legal environment such that their liquidity could be “repatriated” to the US, on a “InTrade USA” real-money prediction exchange.

A related issue is that the CFTC should be concerned about HedgeStreet’s financial health. After its third round of funding, HedgeStreet raised a total of $24.9 million.

http://www.hedgestreet.com/abouthedgestreet/pressreleases/pressrelease_21.html

Lately, HedgeStreet was aquired by an offshore investor for $6 million.

http://www.hedgestreet.com/abouthedgestreet/pressreleases/pressrelease_32.html

Obviously, there has been destruction of wealth, here.

The CFTC did a great job in 2004 when it approved HedgeStreet’s application as a Designated Contract Maker (DCM). The CFTC should now finish the job by creating a legal environment favoring the profitability of HedgeStreet and of other non-intermediated DCMs (e.g., InTrade USA, or BetFair USA, or TradeFair USA) —which I hope will be started up in the future in the US.

What I am afraid with the May 2008′s concept release on “event markets” is that the CFTC does not look into the real issues: the liquidity of socially valuable prediction markets, and the profitability of US-based companies operating real-money prediction exchanges (non-intermediated DCMs).

I’m afraid that all the solutions consisting in “exemptions” and “no-action” letters are false solutions that do not address the real issues.

Finally, for the issue regarding the protection of retail traders, I suggest that the CFTC looks into the worst scandal that occurred in the field of prediction markets —the “North Korea Missile prediction market” scandal. I am sad to say that InTrade Ireland acted in the worst way possible, and, thus, have indelibly tarnished their reputation, alas.

http://www.midasoracle.org/predictions/nkm-scandal/

-

Thanks for listening,

Chris F. Masse
Panorama B, Green Side
305, avenue Saint Philippe
Les Templiers, Sophia–Antipolis
06410 Biot, Alpes-Maritimes
France, European Union

–
–

-

RELATED POSTS:

- Chris Masse’s first comment to the CFTC on “event markets” (prediction markets)

- What the CFTC is asking.

-

-

Now listed on my LINKS page under “Prediction Market Search Engines”.

-

Please, suggest other links to be included on that LINKS page… or on the BEST page… or on the PREDICTIONS page.

Thanks. Will make up to you.

-

Psstt… The multi-exchange search engine (created by those flash-in-the-pan PMIA bozos) has been decommissioned, I see.

-

BetFair CEO David Yu:

We’re delighted to welcome Bo Cowgill in HammerSmith. Bo had a fantastic career at Google, working, in the end, under Google’s Chief Economist Hal Varian. Throughout his years at Google, Bo acquired precious knowledge about the inner working of search engine technology, which will be helpful to BetFair-TradeFair for fighting annoying externalities: Internet rumors and bombastic bloggers.

Bo Cowgill (Internet Meister at BetFair):

Being a number one myself, I’m delighted to work now for the world’s number one prediction exchange, after my work at the world’s number one search engine. I can only sympathize for BetFair. I have been myself the butt of many Internet jokes (associating me with a “Grandizer” I’ve never met); numerous inaccuracies have been published over the years on blogs about my person and my work at Google; my EXCELLENT Google paper has been totally ignored by the prediction market bloggers (PDF file); and my intellectual fight with prediction market ubber-expert Robin Hanson has been totally misrepresentated (the truth is that I won the discussion in the end). At BetFair, I’ll work hard to crash bloggers and forums’ rankings in Google’s web search results. Count on me, big time.

Nick Garner (SEO expert at BetFair):

I’m happy a world-renowned Google expert is joining our SEO team in HammerSmith. I have so far optimized the BetFair sites for Alta Vista (that’s the first search engine listed in my A-to-Z list of search engines), and I am looking forward to reach the “G” letter very soon with Bo’s help.

Mark Davies (Spin Doctor In Chief at BetFair):

Enhancing our SEO team is an integral part of our strategy to take over the Internet from the bombastic bloggers and the unmoderated forums, and this move sits well alongside other recent BetFair initiatives (the BetFair blog, the sports and politics zones, the dynamic compound charts, etc.). Very soon, the Internet will resemble a giant BetFair megaphone, and that shall make the world a happier place. [Sorry to be so short, for once, but I have to leave you, real quick. I must arrive at the horse track VIP gallery before Princess Margaret.]

-

Previously: Google is THE problem, but BetFair should discover the solution.

-

-

-

Here’s their section on politics:

-

Here’s their section on finance:

-

-

My thoughts:

1. Their “about” page still does not list the names of their editor(s) and publisher(s). It would give them credibility.
2. My understanding is that their blog is run by their PR department. That’s a problem. By essence, it means that readers will trust it less than the independent publications out there. Never in the history of the publishing business did a P.R. department of a commercial company managed to run a popular, credible publication —offline or online. It can’t be. Spin doctors want to control information, while journalists want to free it. It’s 2 different worlds, totally incompatible. It’ll never work.
3. They are mixing news articles written by outside journalists or bloggers with P.R. stuff written by people working at BetFair or TradeFair. Again, it’s not a good idea if you want to be trusted by the readers. The two kinds of postings should be published on 2 different blogs —the news blog should be kept completely separate from the corporate blog.
4. A close look at many of their entries show that the BetFair blog is not a quality publication. Their news stories mostly replicate, with less talent, what you can read elsewhere on the Web or in newspapers. There is no innovation.
5. Their postings consist entirely of texts —with very few charts, and almost no internal/external weblinks.
6. I find the BetFair blog very UK centric.
7. Sometimes, they would post for the search engine spiders, and not for the readers. The result is really bizarre, and leaves a funny taste in the mouth.
8. Their blog suffers from many technical problems. The most annoying issue is that their site feed does not reflect entirely their content. I’m a subscriber of their feed within Google Reader. I searched for “Robb”, “politics”, “Clinton”, and “TradeFair”, and I was not able to get the items related to the posts I saw on their website. They should fix this issue ASAP. (I blogged about this technical issue, some weeks ago. To no avail.)
9. Last time I reviewed the BetFair blog, I yelled against the fact that their main writer on prediction markets went with an anonymous byline, “The Predictor” (whose only qualification was that he has “long hair” and that he “eats cheese”). They backpedaled, and now their new prediction market writer goes with the “Mike Robb” byline. It’s much better. (Michael Robb is one of the BetFair employees working at their P.R. department, if I’m correct. His author profile page says he is their “resident political expert”. “Resident” means that he is a BetFair employee, in HammerSmith, as I take it. As whether he is really a “political expert”, as billed, I’ll leave it to you.)
10. From what I know, Michael Robb seems to be a decent and smart ‘chap’ (as they say in England). I wish him good luck. (Here’s his latest output.)
11. They restrict their prediction market approach to politics —that’s a huge mistake, which shows their total misunderstanding of the prediction market approach. They should apply it to sports, too.
12. They are completely clueless as to what form should take prediction market journalism. That’s grave. I’ll blog about that, later on. It’s a long story of prediction market mechanics, usage, intended audience, and strategy. I need to explain all this on an independent blog post, in the near future.
13. Professor Leighton Vaughan-Williams is billed as the “BetFair prof”. It’s a bad idea. It debases Academia. Scholars are paid to develop and propagate our global civilization’s knowledge, not to act as servile P.R. agents for commercial companies. Furthermore, a close look at his prediction market explainer shows that Leighton Vaughan-Williams lacks understanding of the concept of probability as applied to the market-generated predictions. And my readers remember that Leighton Vaughan-Williams has stubbornly refused to disclose the origin of the trading data he used to tout BetFair’s alleged extraordinary predictive power. (The suspicion is that he used InTrade’s data.) All the prediction market scientists I know (Justin Wolfers, Eric Ziztwitz, Robin Hanson, Koleman Stumpf, Paul Tetlock, David Pennock, Lance Fortnow, etc.) do disclose information easily. All this indicates to me that Leighton Vaughan-Williams is not to be trusted. (And scroll down to see the comment on his dubious post, and you’ll see that my negative opinion is shared by at least one informed BetFair trader.)
    
14. Finally, I was disagreeably surprised to see the presence of TradeFair in this BetFair blog. TradeFair was billed as independent from BetFair. If they are really independent, then they should publish on the TradeFair blog —not on the BetFair blog. Makes no sense at all. Makes the TradeFair independence looks like a P.R. trick at launch, as opposed to a reality. The P.R. department is what is worse at BetFair. The BetFair P.R. agents are arrogant ignorants, who don’t understand prediction markets, who don’t comprehend academic papers on prediction markets, who don’t listen, who don’t connect, who publish low-quality outputs that show how deep is their misunderstanding of the prediction markets, and hence, who generate bad publicity for BetFair in the prediction market circles (where InTrade is king). Yet, David Jack decided to put TradeFair in their incompetent hands. Immense error. And when you read each of their 3 binaries explainers, you realize that they are just typical (and mediocre) P.R. pieces, which could have well be published on the TradeFair trading site, in the “help” or “faq” section.

Even though InTrade-TradeSports and BetFair-TradeFair are not directly competitors, any web-literate internationalist interested in prediction markets can compare these 2 prediction exchanges and see almost instantly that the Irish are perfectly Americanized (with the caveat that they lack ethics, alas), whereas the HammerSmith people (in spite of their IT prowess) are just monkeying around with the grace of fatty toads. America hosts the best researchers and analysts in the field of prediction markets. America (not England) has produced the prediction market science (since the end of the nineties). BetFair-TradeFair should cross the Atlantic and get closer to this American bouillon de culture.

BetFair-TradeFair is the world’s #1 prediction exchange, and I value the fact that they decided to be ethical and legal (which contrasts with the pirates of InTrade-TradeSports). However, nobody in HammerSmith gets the prediction market approach (their attempts are superficial), and the BetFair P.R. agents have elevated mediocrity as a performing art. BetFair-TradeFair will have to change in the coming years —otherwise InTrade-TradeSports will remain the king of the prediction markets, forever.

I maintain all my talking about the necessity of good advanced indicators.

However, there is something I should have said to the readers. The prediction markets created in the past by Justin Wolfers, Eric Zitzewitz and Koleman Strumpf all showed very little liquidity for weeks, at inception, but finally started off (even in small ways), eventually. So, the jury is not out yet on David Pennock’s proposals.

Are there any good advanced indicators of future search engine market shares? I was not entirely convinced.

Equally, with Felix Salmon, I’m asking whether there are good advanced indicators of future earthquakes. I’m a bit doubtful.

Which is why I said that any person proposing new prediction markets should be compelled to list the related advanced indicators beforehand.