Incentive to contribute unevenly distributed information

Seedling Prediction Market: Conversation with John Delaney, CEO, Intrade and Russ Thomson, Meritology

Any time that I can find anyone who can bat the ball around, I continue to wonder about the idea of establishing a prediction market to seed early stage innovation in Ontario and Canada.

With their permission, I am publishing below an exchange that I had with John Delaney, CEO of prediction market platform Intrade and Russ Cameron Thomson the founder of Meritology who blogs at Value Made Visible.

I know that this will be very difficult for anyone to wade through unless someone is really interested in the possibilities but at this stage I can not justify the time to edit.

I am still trying to determine if the basic idea works or is theoretically flawed.


From: michael at socialcapitalvalueadd dot com
To: russell thomas at meritology dot com
Subject: An Innovation-Seedling Prediction Market (in conversation with John Delaney)
Date: Tue, 13 Apr 2010 23:30:13 -0400

Thanks Russell for leading me down the learning curve.  I will check out the readings.

Let me share some additional thoughts (indicated by MC>>) below to continue the conversation with you and/or others …

FR: russell thomas
Date: Tue, 13 Apr 2010 11:11:51 -0700


I think it’s a promising idea, but runs the risk of being a “beauty contest”, because the award of seed capital is predicated on having the highest value in the prediction market, which is could easily become self-fulfilling. This would be a serious theoretical flaw.

MC>> Mitigating the “beauty contest” risk is one of the motives for using the pm in this context.  Small market venture capital suffers from “beauty” blinders.  They are closed social networks with few perspectives on “beauty” (where beauty is a metaphor for high potential businesses).  The idea is to create a pm that broadens the input on what is “beautiful”.  Perhaps we design & promote the market to ensure that the majority of the players are internationally based – Hong Kong, New York, Silicon Valley, Boston, London, etc.

I also wonder how your prediction market is different from a minority equity share or option in each company, from an incentive point of view.

MC>> The pm would not be linked to equity in the start up company.  Early stage companies do not have enough liquidity to open up the networks that can participate.  As a meta-market, the pm could be open to the social networks of companies (who have insight into the qualities of the company) as well as international participants who may not know the company as well, but have better information about the prospects for next rounds of finance, potential partners/channels, global competition and demand where it counts (the US, Europe, Asia – big markets).

Hmm ..  let’s say the option unit price is set at a level that would discourage and/or make obvious gaming.  If the unit price was $1, wouldn’t it be pretty obvious if a few related players were buying to boost a company? Once that was reported wouldn’t the rest of the market turn against this prospect?  The $500K is not a windfall.  There will still be rigorous investment terms and demanding investors to satisfy.  If everyone knows that $500K will be invested in the highest potential company every month, why not just improve the company until it is the best in the pool or enter another company?

I would also be concerned about the information available regarding each start-up.  If it’s just the “elevator pitch” plus founder biographies, then the market would favor “star power” over less flashy teams that might be more viable and sold.  To counter this, it would be very desirable to have information provided by the social network connected directly to each start-up, rather than just the start-up itself.  Part of this social network should be early customers, partners, and suppliers.

MC>> Again, being attracted to “star power” is another problem with shallow venture capital pools in small markets that we would like to overcome with a pm.  Better, broader, more informed perspective on who has “star power” and why, is exactly what we are after. The idea is to bring global perspective to balance local connections to scarce seed stage capital.  While there would be standard “listing” requirements, the companies would be free to become as transparent about their activities as they would like and to develop expertise in developing globally respected businesses that they communicate effectively in global terms, rather than the language of local government grants or a few early stage investors who have money due to vested interests.

Finally, venture capital-oriented start-ups must evolve and learn as they progress through the venture process.  Some VCs say that the #1 success factor is the ability to rapidly learn and adapt, even if that means discarding the original product or market plans.  Thus, your prediction market is really trying to evaluate each start-up’s ability to learn and adapt.  This is very hard to evaluate solely as outcomes (i.e. years to profitability, years to IPO or liquidity event).

MC>> I agree that early stage companies must be agile, evolve and learn, even if that means dramatic changes to the business plan.  I will check out the links that you have shared below.  I think the pm would account for this by placing higher value on the people behind the companies (founders, employees, partners, customers, etc) and favouring companies who have higher “margins of error” available to them.  By this, I do not only mean higher margin products, I also mean companies that have validated fast growing, global markets and some sort of uncompetitive advantage.

Instead, you might do much better to focus your prediction market on metrics relative to a process or lifecycle model.  I recommend the Bell-Mason diagnostic, as described in the book High Tech Ventures. <>    This will put *everyone’s* attention on what matters most: successfully progressing through each growth stage.  The prediction market could have questions like this:

*       Successfully completes Bell-Mason Stage 2 in 12 months or less, *and* $50K or less

This has the great advantage that it represents near-term accomplishment, not far-off outcomes.  They can also be assessed objectively by experienced executives or VCs.  (The Bell-Mason diagnostic stages and evaluation rules would need to be modified for “e-Ventures”, e.g. social networks platforms, etc.)

In fact, you could set up a combinatorial prediction market for each of the 12 dimensions of the Bell-Mason diagnostic. (For more on combinatorial prediction markets, see: .)  This could be very powerful from an information discovery viewpoint, because it would reveal conditional dependency between various factors at each stage of development.   (This raises a very interesting valuation challenge, which would take you into options theory because the incremental value of each stage is the option of participating in future stages.  Your PM doesn’t need to explicitly include such models, but maybe PM participants would!)

MC>> I took a quick look at the slides.  Thanks!  I will also check out the other link.  I hope others will also feel free to suggest additional readings.

Lastly, I’m not such a fan of “winner take all” contests.  You mention that one winner will receive $500K.  I think it would be better to have a diversified payoff schedule, including some awards to “Honorable Mention” companies and also special categories – solo entrepreneur, college student team, and so on.  For the “winner”, you could have a matching fund where angel investors could participate as a pool.  After all, being the winner should make the start-up attractive to regular angel investors.

MC>> The “winner” in each instance gets a $500K investment subject to standard due diligence and terms.  But the market would be permanent so any listed company would stay on the market, eligible for the next month.  Part of the rational is to illustrate for solo entrepreneurs, college student teams and the like the kind of conditions they must meet to be attractive seed investments.  Angel investors would be welcome to compete with the Fund.  They would have to get organised to get the best deals.  The pm would impact their perspectives – hopefully broadening their notions of “beauty” and “star power”.

Hope this helps,

MC>> It does help.  I hope that this conversation continues.



From: Michael Cayley
Sent: Tuesday, April 13, 2010 9:14 AM
To: John Delaney
Subject: An Innovation-Seedling Prediction Market (in conversation with John Delaney)

Thanks John – I am taking your advice and sharing our thread with the PM google group.

First & foremost, I am wondering if you or anyone else can see any theoretical flaw in the basic idea? (see details below)

Do you & everyone else in the group agree that the idea could work if implemented properly?

Is getting “unequivocal” settlement of the market mostly a function of achieving critical mass?  Can anyone point me to reading or give me a sense of “how many of the right participants” we would need in this market to eventually be confident in funding at least one start up per month?

If we require incentives to achieve the critical mass how would we estimate how much?  Are we talking about something feasible?

At this point, I am most interested in comments/insights that either validate or invalidate the basic idea.

I expect that doing it right will beg many questions and thoughtful solutions.


Subject: The Al Jazeera Network Dedicates a Full Show to Prediction Markets
Date: Tue, 13 Apr 2010 12:29:10 +0100
From: john delaney at intrade dot com
To: michael at socialcapitalvalueadd dot com

Hi Michael,

Thanks for posting the show on your blog.

Here are a few quick comments.

Your questions are very interesting and I know other service providers and prediction market aficionados will have additional comments.

If there was some asset to incentivize people to predict on a set of markets it would certainly help participation. Perhaps some portion of the seed stage capital could be dedicated to motivating those who are predicting it would be very helpful. I believe that the contributors to the fund would love to get significant crowd wisdom on the investment prospects.

However settlement of these markets could take quiet a long time until the result became unequivocal.

Perhaps a way to approach the best questions to ask is to have those contributing the seed fund to define them. Perhaps the questions would then include.

1.      The company to be operating profitable within X years
2.      The company to be generating sales of Y in A years
3.      The company to have a monetization event of $Z or more within B years.

Feel free to post your mail and my reply on the PM group.

Best regards,


From: Michael Cayley
Sent: Monday, April 12, 2010 4:39 PM
To: John Delaney
Subject: The Al Jazeera Network Dedicates a Full Show to Prediction Markets

Thank you for sharing that John.

I have embedded the video in my blog <>  where I have introduced the idea of using a prediction market to spread a globally oriented innovation meme in Ontario and/or Canada <> .

I would be grateful if you or a member of your team might share some of your thinking of the possibilities of employing a prediction market in the way that I have described.

Let’s say that a fund is put together and that will award $500K in seed stage capital to a promising start up on a constant, reliable basis – perhaps every month, perhaps every 3 months … in any event, the funding date will be a known quantity.

Seed stage companies can then “list” themselves … I.e. share information about their prospects as a company.  A prediction market would be used to access the information available through the listing and all other sources and determine the probability of a company “achieving a $10-million venture capital round” or “achieving sales of $100-million” or “obtaining $500K in seed financing from the fund” or “going public on the NYSE” or …

I know the question is critical but don’t know what question would be appropriate.

On the “funding day” the company that is doing best on the prediction market will get the $500K seed funding, subject to final due diligence and standard investment terms.

The companies would be based in Ontario or Canada but the prediction market would be global and the backers of the fund would actively promote it to investors (venture capital, institutional, etc.) all over the world to assure critical mass and that the prediction market outcomes would be a function of local and global information.

Any thoughts?


Dr. Robin Hanson, Chief Scientist, Consensus Point & John Delaney, CEO of Intrade with Riz Khan on Al Jazeera

Very informative discussion with these leaders of prediction market thinking.

I continue to wonder if a prediction market could be used to help spread a globally oriented innovation meme in Ontario and/or Canada.