In the middle of May I speculated--reasoning from first principles only and with no inside information--about how prediction markets might help the book industry pick winners:
"For some types of books (and authors) prediction markets might be appropriate."
That sparked a note from a reader, tipping me off that:
"Confidentially, something's about to launch very soon in this area."
Sure enough, MediaPredict launched a week later, eliciting this highly positive piece in the 5/21 NYTimes: "Publisher to Let the Public Have a Vote on Book Projects".
Media Predict is soliciting book proposals from agents and the public, and posting pages of them on the site. Traders, who are given $5,000 in fantasy cash, can buy shares based on their guess about whether a particular book proposal is likely to get a deal, or whether Touchstone Books, an imprint of Simon & Schuster, will select it as a finalist in a contest called Project Publish. If either happens within a four-month period, the value of the shares go to $100 apiece; if not, the share price falls to zero.
The devil as they say, is in the details. Traders are voting not on how popular a work could become with the book-buying public but with the binary, intermediated and likely recursive question of whether the editors at Simon and Schuster will select it.
In other words, S&S editors are going to be breathing their own exhaust--now with more butt-covering camouflage. (Which reminds me a little of this.) It's the kind of conversation I've seen some couples get into:
Man (off-handedly and with a smile): What would you like for dinner, honey?
Woman: Oh, whatever you'd prefer is just fine with me!
Man (thinking burgers and beers with the guys): I'd prefer what pleases you, dear.
Woman: Well I don't know. What were you thinking?
Man: I was thinking I'd like to know what you're thinking.
And 'round and 'round and 'round it goes. That's an imperfect and terribly stereotyped analogy, and no, it's not my happy marriage. Well, OK sometimes maybe it is. But let's move on, shall we? :)
The more abstract version of that analysis is this: The system S&S has set up with MediaPredict has no outside, arbitrating reference point--such as how many copies a particular book would actually sell.
Hey, now there's an idea: ask the market directly about its spending enthusiasm for a given title. (Print-on-demand technology, not to mention digital rights management, render completely moot the objection that only a few select books can be produced. The hurdle revolves around marketing and a legacy business model of annointing a few winners.)
Unfortunately, S&S's Touchstone Books does not appear to take that important dis-intermediating step. I doubt it's MediaPredict's fault. I suspect (without any proof except my experience with how Fortune 500 executives think and in particular how publishing executives think) that the use of prediction markets may have been de-clawed in this case to avoid dis-enfranchising S&S editors. Think about it:
The PM predicts Bestseller status for a book the editors decide to pass up. It gets picked up by another publisher, fulfilling the PM's expectations and making bundles of money for an S&S competitor. At that point, S&S editors not only have egg on their face, but the S&S board (and shareholders) are emboldened to ask: Tell us why exactly we need you if all you're going to do is to second-guess and stand in the way of this smart little prediction market thingy we've set up over here to perform the same function?
Justin Wolfers (respected PM guru at Wharton) notes the same basic flaw, though you'd have to read to the end of the article to find his critique.
[Wolfers] said that if Simon & Schuster relies on the
traders' judgments to select a book, it could skew the bets themselves.
"If they say we find it really persuasive that everyone bet on book
A, they're just looking at a book that everyone bet that everyone else
bet that everyone else thinks is the best book," Mr. Wolfers said. "So
you don't end up with the wisdom of crowds, but the infinite reflection
of crowds looking at crowds."
This is at least the second time I've noticed a major media outlet burying what should have been the lead on a prediction markets story--and with it, Wolfers' critique. If I were him, I'd be a little peeved. Last fall, the Washington Post carried a piece ("The Top Pickers vs. The Pack") that completely mashed two irreconcilable ideas, as I noted in "WaPo Misguided on 'Experts' vs. The Swarm":
The article takes the academically and empirically
well-proven notion of collective intelligence (aka the wisdom of
crowds), and presumptively twists it into the more comforting and
familiar (if statistically invalid) notion that "father-knows-best"--only we haven't done a good enough job of finding the right "father" up until now.
I don't know why the media's treatment of prediction markets is missing the mark, but they need to get smart faster. PMs are not going away. Not taking seriously the critiques of Wolfers and others isn't helping. The NYTimes digs itself further into a hole by taking a flawed comparison at face value:
Media Predict is modeled after other so-called prediction markets like the Hollywood Stock Exchange, which allows traders to bet on the four-week North American box office receipts of movies, or the Iowa Electronic Markets, which allow people to bet on election results...
The glaring difference: the HSX and the IEM ask traders to vote on highly distributed end-result questions. As I've observed many times (e.g., here, here, here and here) prediction markets--while very cool indeed, and applicable to many more things than those to which they are being applied today--are just plain lousy at calling things like Papal elections or the nomination of judges where information is closely held by a select group. I suspect it's even worse when that group has sponsored the market and each is trying to predict the actions and preferences of the other.
If the same principle were to be applied to American Idol, voters would weigh in with their opinions not on which singer(s) they liked best, but on which one(s) they thought Simon Cowell was going to like. That's not entirely without its entertainment value, but it adds little to what Mr. Cowell was going to say anyway. Simon (& Schuster) could take a lesson from what the other Simon (Cowell) lacks: humility. The crowd is often right. If they're paying, they're even more right.
As for MediaPredict itself, I'm reserving judgment. At this stage of market development, more new players trying more approaches, covering more questions and attracting more corporate attention is a good thing--and entirely expected. As the MP folks write on their blog: "Love us or hate us, we’re here…" That may be a little extreme. How about: I love that you're here.
Recent Comments