Emile Servan-Schreiber of NewsFutures:
January 29th, 2008 at 7:25 am
Another important difference with NewsFutures (where people have been “trading news” in 2000) is that hubdub doesn’t give away any prizes to performers. That, perhaps, is a direct consequence of the false good idea of letting people create their own markets. This not only creates many opportunities for fraud (if, for instance, the creator of the market also controls the outcome), but it also encourages incoherent outcome definitions, unverifiable outcomes, and duplicate or junkyard markets. Same problems that Inkling’s public markets suffer from.
Also, the mere idea of a Web 2.0 makeover of prediction markets is laughable. To paraphrase a good ol’ song from the 90’s, prediction markets were web 2.0 before web 2.0 was cool.
Emile Servan-Schreiber’-s criticism is pertinent. However, his conclusion (’-no’- to self-management of event derivatives) is too radical. Without Inkling Markets, we wouldn’-t have had Michael Giberson, who loves experimenting with play-money prediction markets. Somebody will come up, one day, with the right technology (e.g., a reputation system) patching the flaws that EJSS addresses. One day, in the future, we will be able to enjoy both worlds, because they will have merged into one: the libertarian prediction exchanges and the disciplined prediction exchanges.
My good doctor Emile, remember JFK, who pushed his country to do things “-not only because they are easy, but because they are hard“-, …-and succeeded. 
Just because event derivative management by traders is problematic does not mean that we should give up right now. Kudos to Inkling Markets and HubDub for trying, and acknowledging criticism from veterans.
—-
UPDATE: Emile Servan-Schreiber comments…-
No one has a monopoly on user-driven content. Every exchange out there lets people propose their own ideas for markets that might be of interest to themselves and others. For instance, on NewsFutures, a lot of the general forum discussion is back-and-forth between the users and the admins about which markets to create next. Where NF differs from FX, Inkling and Hubdub is that the NF admins (which, by the way, are recruited from the user-base itself) have the final control on wording as well as settlement. That’-s what guarantees the coherence of the exchange, which in turn means we are able to offer prizes, whereas the likes of FX, Inkling and Hubdub likely cannot because they give too much control away to unknown entities to guarantee the fairness of the contest.
I like NewsFutures, and I get all that. But I’-m saying that, nowadays, on the Web, people want DIY tools. That’-s why HubDub and Inkling Markets are appealing to them. They don’-t have to discuss “-back-and-forth”-. They create the event derivative they want. Straight from the producer to the trader.
Read the previous blog posts by Chris. F. Masse:
- RIGHT-CLICK THIS IMAGE, AND FILL IN THIS SURVEY, PLEASE.
- Papers on Prediction Markets
- The Journal of Prediction Markets
- The 45-degree Line
- Implied Probability of an Outcome –BetFair Edition
- Justin Wolfers on Rudy Giuliani = not convincing… yet
- The Florida primaries thru the prism of the InTrade prediction markets
Re: “This not only creates many opportunities for fraud (if, for instance, the creator of the market also controls the outcome), but it also encourages incoherent outcome definitions, unverifiable outcomes, and duplicate or junkyard markets. Same problems that Inkling’s public markets suffer from.”
.
In my experience, purely from the point of view of a user of Inkling-public-market (IPMs), I haven’t seen any evidence of fraud of the form EJSS suggests (and Inkling at least tries to prevent such self-dealing).
.
On the other hand, “incoherent outcome definitions, unverifiable outcomes, and duplicate or junkyard markets” are all problems on the IPMs. Still, in my view:
.
– Traders can avoid trading in poorly defined or hard to verify markets, limiting the harm from those problems. Still, these are probably the biggest problem for traders, because it is sometimes hard for non-specialist traders to see the holes in a product definition.
.
– Duplicate markets can limit liquidity by dividing traders into groups and they clutter up the exchange. But a single trader active in both markets eliminates the first problem, and clutter problem is relatively minor.
.
– Junkyard markets? Meaning either (a) abandoned by the market manager — probably the second biggest burden on the IPM — or (b) markets no one cares about but the manager, which adds to clutter but otherwise not costly.
.
So, I’m still in the camp of thinking it is a truly good idea to allow users to create markets on play-money exchanges. (Commenting in general, and not on Hubdub specifically.)
OK. Still, Inkling Markets should have done more to establish a reputation system from day one.
Inkling and HubDub have taken two completely different models:
.
Inkling:
– User creates market, admins have to approve
– User cannot trade in their own market
– User must cash out the market as they see fit (many issues here, including abandoned markets and incorrectly-cashed-out markets)
.
HubDub:
– User creates market and it’s added automatically
– Users can trade in their own market
– HubDub takes responsibility for closing out the market, but users must state what criteria should be used in order to close it out
– Flagging capabilities for bad markets. When enough people have flagged a market it is voided and all bets spun back to the users
.
In principle I think user-generated-markets are good since you never know the kinds of things people will be interested in trading. That said, opening markets up to these means that you’ll likely be wading through a lot of c**p. That happened to Inkling, though they seem to have gotten a handle on it. There will likely be a lot of problems on HubDub for a while until a bit of a community develops to do the flagging/voiding.
.
My two cents…
No one has a monopoly on user-driven content. Every exchange out there lets people propose their own ideas for markets that might be of interest to themselves and others. For instance, on NewsFutures, a lot of the general forum discussion is back-and-forth between the users and the admins about which markets to create next. Where NF differs from FX, Inkling and Hubdub is that the NF admins (which, by the way, are recruited from the user-base itself) have the final control on wording as well as settlement. That’s what guarantees the coherence of the exchange, which in turn means we are able to offer prizes, whereas the likes of FX, Inkling and Hubdub likely cannot because they give too much control away to unknown entities to guarantee the fairness of the contest.
I like NewsFutures, and I get all that. But I’m saying that, nowadays, on the Web, people want DIY tools. That’s why HubDub and Inkling Markets are appealing to them. They don’t have to discuss “back-and-forth”. They create the event derivative they want. Straight from the producer to the trader.
Emile, why don’t you certify a class of event derivative managers? Take Michael Giberson. He is a PhD economist. He knows more than you do about economics. He read all the papers on prediction markets. (He even worked close to Robin Hanson, wow.) If you certified this guy, then he would create his prediction markets on NewsFutures instead of creating them on Inkling Markets, and you would get all the marketing benefits of having a PM blogging economist linking to your exchange.
.
You could establish a list of criteria (PhD in economics, etc) to certify some people, and integrate them as administrators either in the NewsFutures exchange or in a parallel exchange that you could create on the side.
That’s an excellent idea. If Michael would like to create/manage his markets on NewsFutures or Bet2Give, or both, I’d happily give him admin access. What say you, Michael?
Emile, would you extend your invitation to *any* PhD in economics who has a good knowledge of the prediction market field?
I agree with Chris that there is value in more or less immediate/direct user created markets, even though it results in some of the problems that Emile identifies.
There is little “coherence” to the Inkling exchange in the way Emile means – while the underlying technology is standard across the markets, at the content level there is a large diversity. Some complicated markets are well defined and some simple markets are poorly set up. Market coverage of standard topics like politics and sports is somewhat scattershot.
As I put it once before, I like it in part because it gives the place a bazaar like feeling. No one calls a bazaar coherent – coherence is maybe an attribute of a department store or mall.
Unlike Chris, however, I don’t necessarily think NewsFutures needs to become more like Inkling, or Inkling needs to become more like NewsFutures. There is room for many approaches, particularly as we are early in the development of online prediction markets.
Either Inkling Markets and HubDub should do CDA or NewsFutures should let people manage event derivatives from A to Z.
.
Hey, Giberson Incorporated, will you take on EJSS’s proposal?
I don’t think anyone needs a Ph.D. in economics or any other field to be a great market manager. For instance, check out the profile of NF’s chief public markets admin: http://us.newsfutures.com/home/people.html#don.
I’m ready to extend market-admin access on NF to anyone who is passionate enough about PMs and some specific issues to create some cool markets and spend the time needed to administer them seriously. Obviously, this can’t be an automatic process, and I’d do the vetting myself.
By creating/managing markets on NF you’d gain instant access to one of the largest, most experienced, most international crowd of prediction traders. Ad-revenue sharing deals are possible too.
1. Emile, if you want this information to be published on a Midas Oracle post, let me know (or do it yourself, or publish that on NewsFutures and I’ll re-publish here, whatever you want).
.
2. Speaking for me (not for Mike G. or anyone else), I wonder whether some new markets may be of interest to the Midas Oracle readers (and of course the existing NewsFutures traders). Not necessarily about the prediction market industry.
.
Ideally, it should fit into my “advanced indicators” framework…
http://www.midasoracle.org/200…..n-markets/
.
If one day I have a special idea, I’ll tell you.
.
It should be a socially valuable prediction market, but also a market that provides some fun… Let me think…
.
[EJSS managed to trap me. I have to prove I can be creatively useful, now. Horse of another color than just blogging.]
Emile, I’m certainly interested in taking you up on your offer.
.
But I agree that you want cool markets and someone to dedicate the time necessary to properly administer the markets. It will take some thinking about market ideas and a little thinking about scheduling on my end.
.
I’ll email you off-line and we can figure something out.