Email Interview: Ken Kittlitz

My responses to a set of questions Chris Masse recently emailed to me:

Chris. F. Masse: Ken Kittlitz, you co-founded the Foresight Exchange (it went by the name &#8220-Idea Futures&#8221- at the time) in 1994. Would you mind telling me two words on your co-founders? Which ones brought the most into the project? Are you still in touch with them? Do you know what they have become?

Ken Kittlitz: David McFadzean got the ball rolling by bringing one of Robin Hanson&#8217-s early prediction market papers to our weekly discussion group. Sean Morgan realized that the WWW, then still in its infancy, would be a great way to create such a market. Mark James, along with Sean, did most of the coding of the initial prototype. Duane Hewitt and myself did most of the work on a paper and presentation that our group presented at a conference the following year.

I&#8217-m still in touch only with David- he&#8217-s currently a software architect at QuIC, a company that creates financial risk analysis/mitigation products.

CFM: What was the spirit of your group at that time (in 1994). Did &#8220-entrepreneurship&#8221- mean something for you, guys? Did you envision a commercial venture, or was it just collegians&#8217- play?

KK: Our weekly discussion group was known as the &#8220-BS Group&#8221- (Biological Simulation, in case you&#8217-re wondering), so I&#8217-d have to admit that &#8220-collegians&#8217- play&#8221- is a fair summary. In 1995, we did try to turn it into a commercial venture, which quickly revealed our lack of business experience. We were all techies of one sort or another, and techies often struggle in the business realm.

CFM: Would you mind telling me two words on GMU professor Robin Hanson? How would you introduce him to some of our readers (I pity them) who have never heard of him?

KK: Robin&#8217-s one of the smartest people I&#8217-ve ever met and, unlike many smart people, not over-specialized. He has deep understanding of a number of fields: artificial intelligence, physics, economics and likely a few others I&#8217-m not aware of. He has a habit of coming up with fascinating, controversial ideas, prediction markets being just one example.

CFM: You co-founded this play-money prediction exchange (Foresight Exchange) in 1994. In 1999/2000, Andrew Black and Edward Wray created and launched BetFair in England. BetFair became one of the most successful British start-ups and its two co-founders are now sitting pretty on a small fortune. In hindsight, don&#8217-t you think that you should have moved to the U.K. and incorporated the Foresight Exchange there, using real money?

KK: In hindsight, I think that I should have done a massively-leveraged short sale of NASDAQ stocks in March, 2000. :-)

The best way forward is always hard to identify, even with tools like prediction markets&#8230-

When we tried to commercialize the original &#8220-Idea Futures&#8221-, starting a real-money market offshore was certainly something we considered &#8212- though at that point, somewhere in the Caribbean seemed the likely venue. Even back then, it seemed likely that prediction markets would be considered a form of gambling, and hence subject to draconian restrictions. The Caribbean can be a nice place to live, but the prospect of never being able to return to North America to visit family and friends was quite a disincentive.

CFM: One thing that strikes me when visiting the Foresight Exchange is that you forbid sports prediction markets, which are very popular on the betting exchanges. Even Bo Cowgill&#8217-s group of Googlers trade on sports, sometimes &#8212-I believe. Sports trading can be fun. Are you a jock hater?

KK: Not really, but the Foresight Exchange was created primarily to focus on science and technology claims. Having it cluttered with a couple of dozen &#8220-tonight&#8217-s game&#8221- claims per day wasn&#8217-t too appealing.

CFM: If I can count, you have more than 12 years of experience in the field of prediction markets. You&#8217-ve seen them all, in all colors and shapes. Do you agree with what Robin Hanson said at the Yahoo! Confab, namely that the DARPA&#8217-s PAM scandal ignited interest in corporate prediction markets? Was the PAM scandal a &#8220-tipping point&#8221-?

KK: No. I think the real tipping point was the publication of James Surowiecki&#8217-s &#8220-The Wisdom of Crowds&#8221-. Those of us interested in prediction markets tend to overestimate the PAM controversy&#8217-s importance- it was a big deal for us, but only an incremental step in the general public&#8217-s awareness of the topic. The interest generated by Surowiecki&#8217-s book showed that prediction markets had &#8220-arrived&#8221- &#8212- they weren&#8217-t just of academic interest, but instead had real-world applicability.

CFM: Note that the DARPA&#8217-s PAM prediction markets was to be public. Which leads to my next question. You and partner David Perry at Consensus Point help Fortune-500 companies setting up and running their own internal prediction markets. Have you ever had the case where one firm opened its corporate prediction markets to contractors and clients?

KK: Some of the firms we deal with are certainly interested in having a fairly wide audience, including customers and contractors, for their markets. I can&#8217-t go into specifics at the moment, however.

CFM: How is Consensus Point doing, so far? Can you draw for us the portrait of the firm that wants to use internal prediction markets? Is it always to forecast sales? Do you sense that the requests come from senior executives or from mid-level prediction markets-enthusiast managers?

KK: Consensus Point is doing very well so far. A lot of inquiries do indeed originate from mid-level managers and researchers, but a fair number also come from the executive level. Sales forecasting is a popular application of the market, but project completion times and commodity price forecasting have also proved to be frequent questions.

CFM: Sorry to ask you this question bluntly. Would TradeSports and Betfair make great competitors of Consensus Point if ever they decided one day to sell prediction market services to organizations?

KK: Quite possibly, but it&#8217-s certainly not a given. Both companies have great trading platforms, but their expertise is in running real-money, public markets. Corporations aren&#8217-t really looking for that sort of domain knowledge when considering how to implement and use a prediction market.

CFM: Would you mind describing in a few words the prediction market services you sell? I guess it&#8217-s web-hosted CDA, but are some firms interested in web-hosted MSR?

KK: We offer both hosted and on-site installations of our software, as well as training, analysis and consulting services. As for MSR versus CDA, see below.

CFM: Speaking of Market Scoring Rules, why did you decide to use this design as the engine for the Washington Stock Exchange? What is its main competitive advantage to CDA? How can MSR best be described: &#8220-betting&#8221- or &#8220-simplified trading&#8221-?

KK: The line between an MSR and a CDA is thinner than you might think! We have a market maker for each stock that provides liquidity by placing bid and ask orders- this is a convenient way of implementing an MSR within a CDA framework. An MSR really helps to start (and keep) the market going, because people always have a price they can buy or sell at. With an unadorned CDA, the bid/ask spread can be enormous, and trading volumes very thin. This alas, is often the case on the Foresight Exchange.

I&#8217-d describe an MSR as allowing for &#8220-simplified trading&#8221- rather than &#8220-betting&#8221-, though I suppose it depends on how much thought the person interacting with it puts in!

CFM: Just curious. When a prediction exchange decides to use MSR, does it have to pay fees or royalties to its inventor, Robin Hanson?

KK: I don&#8217-t believe so, but Robin is in a far better position to answer that question than I am&#8230-

CFM: What is the biggest mistake (if any) you have made since the grand opening of Consensus Point? What did you learn from this big mistake?

KK: No really big mistakes come to mind. Of course, such things are often only obvious in retrospect, so ask me again in a few years.

CFM: What are corporate prediction markets competing against (if any)? Internal polls? Groups of in-house experts? The firm&#8217-s executives? Something else?

KK: Generally, the firm&#8217-s executives. We haven&#8217-t encountered too many cases where firms have been trying to use internal polls as part of their forecasting efforts.

CFM: Are you positive that corporate prediction markets will show something for it? Will the economics literature soon be filled with business cases on how firms can clearly benefit from using internal prediction markets?

KK: Based on my experiences in the field thus far, I&#8217-m confident that prediction markets will prove to compare favorably with the other forecasting methods companies use. This isn&#8217-t to say that they&#8217-ll always yield good information, or be the best thing to use in all situations, but I think they will turn out to be valuable.

Am I positive of this? Not absolutely. But then, I try not to be absolutely positive of anything!

CFM: Now, the question that kills. Tell me frankly. Are corporate prediction markets a &#8220-fad&#8221- or are they just started?

KK: Great question! I think it largely depends on how the prediction market community presents the ideas. There&#8217-s a very real danger that the topic will be over-hyped and, consequently, ultimately dismissed, just as so many other trendy business ideas have been in the past. Today&#8217-s darling is often tomorrow&#8217-s pariah. That would be a shame, since (obviously) I think the markets have a lot of merit.

Note by &#8220-prediction market community&#8221-, I&#8217-m referring not only to those who create and sell prediction markets and associated services, but also people who blog about the topic, create vortals, etc. Not mentioning any names here --) .

CFM: Are prediction markets just one forecasting tool, or do they have a bigger function, in your view?

KK: The pragmatist in me says they&#8217-re just one tool, albeit a great one. The idealist finds something profoundly appealing in their ability to democratize how information is gathered and, ultimately, how decisions are made. The idealist thinks they&#8217-re something more.

Prediction Markets Definitions – REDUX REDUX

No GravatarI would like to comment on the post from the Hatena Diary blog. (By the way, please note that my URL has changed, because I corrected one word in the post title. Sorry for the inconvenience.)

#1. Speculation-oriented prediction markets/exchanges: TradeSports, BetFair.

#2. Hedging-oriented prediction markets/exchanges: HedgeStreet and all the Chicago exchanges that will do binary, European call options.

#3. Forecast-oriented prediction markets/exchanges: Iowa Electronic Markets, AS CLAIMED BY THESE SCHOLARS WHOSE TASK WAS TO CONVINCE THE CFTC TO GRANT THEM A NO-ACTION LETTER. (They would have not gotten it, had they emphasized &#8220-speculation&#8221-. And, of course, &#8220-hedging&#8221- was out of question.) It&#8217-s a &#8220-claim&#8221- that might be discussed, since we&#8217-ve seen that TradeSports-InTrade is a much more powerful predictive tool for the US elections. Ditto for BetFair for U.K. elections.

#4. Decision-oriented prediction markets/exchanges: I would put here the kind of stuff that Robin Hanson is so excited about.

#5. Entertainment-oriented prediction markets/exchanges: Hollywood Stock Exchange, Washington Stock Exchange, Inkling, NewsFutures.

#6. Education-oriented prediction markets/exchanges: The Iowa Electronic Markets fits here, partially, regarding the use that professors around the country make of their markets in classrooms.


– I disagree with Google in #4. Maybe the Google internal prediction markets would fit in #3.

– I disagree with NewsFutures in #3 &#8212-I acknowledge (at least partially) the predictive power of play-money prediction exchanges, of course.


Should we judge markets/exchanges on INTENTIONS or on RESULTS? I don&#8217-t give a damn that TradeSports-InTrade and BetFair were created for speculation– if they have better predictive power than IEM, I&#8217-m fine with them. Ditto for the HSX. I don&#8217-t give the first fig that it was created as an entertainment tool. It&#8217-s the best forecasting tool for the movie business, period.


For the links to the prediction exchanges, see CFM.


Previous Blog Posts:

Prediction Markets DEFINITIONS – not a “taxonomy”

Professor Robin Hanson’s draft definitions is validated by professor Eric Zitzewitz.

Prediction Markets Definitions – REDUX

Prediction Markets Definitions – by Robin Hanson – 2006-11-21


Addendum: Robin Hanson has posted a comment&#8230-

“Oriented” is not clear enough for my tastes. Is this about trader motives? Trader results? Price results? Exchange motives?


My Answer: I meant &#8220-exchange motives&#8221-. [&#8230- See my comments. &#8230-] But now that I think of it, another classification taking account of the &#8220-price results&#8221- makes more sense.

Previous blog posts by Chris F. Masse:

An Email Interview: Alex Kirtland

No Gravatar

(A note from AK: So, in case you haven&#8217-t noticed, Chris has proposed email interviews as a way to get a bit more participation on MidasOracle. I&#8217-m happy to start off by answering his questions to me.)

Chris Masse: What is the best public explanation of prediction markets? As &#8220-stocks&#8221- (Hollywood Stock Exchange, Washington Stock Exchange), as &#8220-event futures&#8221- (InTrade), or as &#8220-event derivatives&#8221- (HedgeStreet)?

Alex Kirtland: Considering that most people have probably never heard of derivatives, that most of the rest don&#8217-t know much about futures, and that almost everyone has heard of the stock market, I&#8217-m going to say that treating prediction market contracts as stocks is probably the best way to go for the general user.

For example, saying &#8220-They&#8217-re like stocks, but different,&#8221- is easier for most people to understand than: &#8220-You&#8217-re buying a futures contract on the likelihood of an event occurring that pays out either 0 or 100 depending on the result.&#8221-

Starting with something familiar, and then introducing complexity, rather than trying to be accurate right off the bat, is usually a better way to go.

That said, if the majority of your users are traders in pork belly futures, then using the stock market as your metaphor to explain prediction markets may just confuse them.

CM: What is the best trading model from a usability perspective: one single class of securities ala TradeSports (where selling means short-selling the &#8220-yes&#8221- contract) or two classes of securities ala Iowa Electronic Markets (where selling means selling the &#8220-no&#8221- contract)?

AK: I don&#8217-t know the answer to that, and I&#8217-ve actually pondered over this for some time. I&#8217-d like to do usability testing/user research to try and figure this out. Just from an academic point of view I think understanding this would be fascinating. But also I think that this has implications beyond prediction markets. Certainly brokerage houses and exchanges might be interested in understanding how to make trading easier for people who may not know how to trade, or are less familiar with trading.

My hunch is that both models work (in fact both models do work), but which one is better is a question of context – who is the user- what is their experience trading- what is the market- is margin involved- if so, how do we communicate that to the user- and so on.

I know there is a large body of research on behavioral economics, which I&#8217-m not as familiar with as I should be, but I don&#8217-t believe anyone has ever researched this specific question.

CM: What is the best pricing model from a usability perspective: a continuous price (HSX) or a 0-100 price (TradeSports)?

AK: It depends on the situation. Sometimes they&#8217-re clearly inappropriate – a linear contract for a binary question, for example. One is not inherently more usable than the other. It&#8217-s more important how it&#8217-s presented to the user.

CM: Should the designer of a new trading screen be innovative or be subordinated to the users&#8217- mental model (if any)?

AK: This is a fascinating question. First understanding a mental model and being innovative are not incompatible things. The mental model is usually a starting point from which innovation can spring forth. This is why you do user research: so you can understand how your users think (or even if they do) about the task you want them to perform.

So, it&#8217-s not so much that innovation (or, better, interface design), is subordinate to the user&#8217-s existing mental model(s), but how do you take advantage of an existing mental model(s) to get the user to more easily do what you want them to do on your site.

Secondly, a lot of readers are probably asking, &#8220-what the hell is a mental model?&#8221- Briefly (and quite vaguely), a mental model is a mental representation of something. For example, most everyone has a mental model of how a restaurant works. You go in, there may be a host, you sit, you order, they bring you food, you eat, you agonize over whether you&#8217-re going to get desert or not, you pay, and then you leave.

There&#8217-s a lot of subtlety to this mental model. Things can change drastically depending on whether you&#8217-re at a diner, a food cart on the street, or a five star restaurant. But the basic process is the same.

It&#8217-s important to note, though, that in and of itself a mental model has nothing to do with the interface of an application. It is usually a hodge podge of heuristics, tasks and sub tasks, and so on, all jumbled together. They don&#8217-t necessarily need to be a true representation of the world, but they need to help the person act in the world.

Referring to the above example, my mental model of a restaurant allows me to go to all sorts of restaurants I&#8217-ve never been to before, have appropriate expectations about what will happen there, and act accordingly.

As an experience designer we&#8217-re not necessarily interested in shaping our interfaces to someone&#8217-s mental model – we don&#8217-t want all interfaces to be exactly like McDonalds – but we do want to be aware of them and take advantage of them &#8230- and not violate them either. We don&#8217-t want to build a restaurant and not serve any food, for example. Once you violate someone&#8217-s mental model of some thing, then they&#8217-ll have no idea what to do next.

Donald Norman&#8217-s book, The Design of Everyday Things, is a good place to learn more about mental models and how they should used when designing an inteface.

CM: Should prediction exchanges set up corporate blog(s) and why? (And if &#8220-no&#8221-, why not???)

AK: Trendio, The Public Gyan, and TradeSports, for example, actually use their blogs quite nicely. These blogs, generally speaking, tell people about contract expirations, changes in margin, new contracts, and so on. They&#8217-re very useful to the people who trade on these sites.

Other prediction exchange corporate blogs are nothing more than self-promotion. That&#8217-s not a bad thing, but it&#8217-s less useful for traders on the site, and more useful for the person promoting the site (or the blogsters covering that site).

Prediction exchange blogs shouldn&#8217-t be treated differently than any other blog: they should publish on a timely basis, and write about something that is of interest to their users. If they can&#8217-t manage that, then they shouldn&#8217-t keep the blog.