Professor Christopher Wlezien on his research on how election prediction markets compare to polls.

Posted on September 2, 2009 by Chris F. Masse

Professor Christopher Wlezien on his research on how election prediction markets compare to polls:

Markets vs. Polls as Predictors: An Historical Assessment of US Presidential Elections – (PDF file) – by Robert S. Erikson and Christopher Wlezien

About Chris F. Masse

Founder and President of Midas Oracle
This entry was posted in All Best Posts Ever, Analysis (Accuracy & Precision), Analysis (Data), Analysis (Meta), Collective Forecasting, Exchanges & Markets, Market Prices & Probabilities and tagged , , , , , , . Bookmark the permalink.

16 Responses to Professor Christopher Wlezien on his research on how election prediction markets compare to polls.

  1. teaNoranges says:
    September 2, 2009 at 4:33 PM

    There are so many incompetent people in this field.

  2. Panos Ipeirotis says:
    September 2, 2009 at 5:38 PM

    And even more anonymous trolls…

  3. Chris F. Masse says:
    September 3, 2009 at 1:41 AM

    That is why I pressured Ben Shannon to disclose his real identity. There is a better dialog with known people.

    P.S.: I am also a bit critical of the prof’s approach.

  4. Panos Ipeirotis says:
    September 3, 2009 at 4:17 PM

    Yes, knowing the real identity and qualifications is important. If the “There are so many incompetent people in this field.” comment was coming from Robert Engle, I would have taken the comment much more seriously than a comment coming from… teaNoranges. Especially when there is no qualification on why s/he considers Christopher Wlezien to be incompetent.

  5. Panos Ipeirotis says:
    September 3, 2009 at 5:43 PM

    When they make a point anonymity is fine (e.g., teaNoranges makes a point when discussing our paper) . Still, it is difficult to take them seriously. Especially when they have not been around for long and they have no reputation.

    Just think: would you buy an iPod from Amazon.com, or from a website named sellfromthebackofmytruck.com? It is no different here. Reputation matters.

    If you want to extend this to prediction markets: Assume that all players start with the same budget. Players that have been successful in the past, accumulate more wealth. This gives them the ability to weight more heavily in future markets, compared to players that have not succeeded in the past, or compared to players that just started. In prediction markets, reputation = money, and money come from successful past predictions.

  6. Chris F. Masse says:
    September 3, 2009 at 4:46 PM

    Sometimes, smart and educated people want to be anonymous. I don’t know why, but I respect it, if that’s the condition for them to speak to us in the first place.

  7. Panos Ipeirotis says:
    September 3, 2009 at 6:23 PM

    “You assume that an anonymous person is a nobody. Turns out that is not the case, here.”

    I do not believe he is nobody. As I said, his points make sense, when he was criticizing my paper. So, he understands what he is saying. Actually I enjoy our heated discussion. It helps me being more clear in my writing. A little bit of passion never hurts when the other party is intelligent.

    I find tacky the comment that he left here, though. Why? Because he left no justification. And I had no prior to judge whether he is a troll or someone who knows what he is talking about.

    Btw, is there a way to see the history of comments for any particular commenter? I used my admin rights to search the history of comments and find some old comments of teaNoranges. Unlike a streak of 1-line bad comments left lately, he left some nice ones in the past.

  8. Chris F. Masse says:
    September 3, 2009 at 6:07 PM

    Panos,

    I don’t follow you. A little bit of anonymity is useful to get the conversation going on with some people.

    90% of Midas Oracle commenters are not anonymous. 10% are, and if that’s the condition for them to speak here, let’s respect their anonymity.

    (On other blogs, 90% of commenters are anonymous.)

    You keep insisting that Tea&Oranges has no reputation. Actually, he has a reputation on his/her own on Midas Oracle. He has made some smart comments in the past.

    Also, just because somebody is anonymous does not mean that that person is a nobody in the real world. Obviously, Tea&Oranges has a background (and bias) in prediction markets.

    Let’s keep the conversion alive. If a little bit of anonymity helps, well, let’s accept it.

    I try to have people to disclose their full name, but I also respect when they won’t.

  9. Chris F. Masse says:
    September 3, 2009 at 6:10 PM

    QUOTE
    If the “There are so many incompetent people in this field.” comment was coming from Robert Engle
    UNQUOTE

    You assume that an anonymous person is a nobody. Turns out that is not the case, here.

  10. Medemi says:
    September 4, 2009 at 2:43 AM

    Everyone has their reasons for being “anonymous”.
    Mine is simple, everyone on the web that I talk to or have talked to knows me by this name. Only a few people know my real name. It all started with the betfair forum a couple of years ago. It would in fact show a lack of integrity should I choose to use my real name. :-D

  11. Chris F. Masse says:
    September 4, 2009 at 2:13 AM

    “I find tacky the comment that he left here, though. Why? Because he left no justification.”

    Yes.

    To see old comments, I would go inside in the administration area, go in the comments section, and search for the name of that commenter. (Or, externally, I would perform a search site via Google.)

    One can use plugins to dig into comments. I experienced those, however the big problem is that comment searching eats up a lot of SQL resources. (I am still on cheap web hosting.)

  12. Panos Ipeirotis says:
    September 4, 2009 at 12:30 PM

    I have no problem with pseudonymity in general. (i.e., an handle/id with no PII which remains constant for the online life of a person). I oppose pure anonymity, or changing the pseudonymous id.

    I just cannot accept (pseudo)-anonymous accusations. If you have something negative to say, be brave and use your own real life reputation to hurt someone else’s real life reputation.

  13. Chris F. Masse says:
    September 4, 2009 at 1:11 PM

    Yes, that sounds right.

    But leaving some space for anonymity allows people to speak out —otherwise they would stay mute.

    For me the right mix is 80% real people with full name and 20% of people hiding behind a pseudonym.

  14. Medemi says:
    September 4, 2009 at 2:29 PM

    True Chris.

    I couldn’t give you my real name if I wanted to. And it’s not betfair because they know who I am. It’s an even bigger (Japanese) firm who is not happy about my presence, somewhere, in the US. They’ve been stalking these 2 small battery start-up companies in the US who are using the same battery chemistry, walking a fine line between what’s legal and what’s illegal. They’ve been here, digging in my past, my friend’s past even.
    I’ve exposed them, on a number of message boards, and then they suddenly became quiet. You can see why they don’t like me much.
    It’s a true story, and since most of you are from the US, I may call upon your help one day.

  15. Chris F. Masse says:
    September 4, 2009 at 3:39 PM

    Medemi,

    Be careful what you say on forums.

    Best wishes to you.

  16. Medemi says:
    September 4, 2009 at 4:43 PM

    I know, I haven’t made any factual statements, it’s nothing like that.
    Psychological warfare would best describe it, and they hang themselves.

    Thanks for the warning.

Leave a Reply