In light of Robin Hanson’s latest blah blah blah…

-

Do you really think that a prediction market journalist (or blogger) should disclose his/her event derivative holdings (as you do systematically)?…

Just a question that crossed my mind.

I was in favor of it before I read Hanson. Now, I have doubts… (Hanson, be damned. )

-

-

Good news: They have a new columnist who tries to apply the prediction market approach. He even uses my two favorites tags, “prediction markets” and “prediction exchanges“. Good —although, the last one should be singular, not plural, in this context. (The “betting exchange” tag is discarded. Not smart.)

-

Bad news:

#1. The writer hides behind a pseudonym, “The Predictor”. I don’t think it’s a good idea. With all the crap on the Web, readers are looking for serious articles from known experts. Of all 122 feeds I read daily, only one is published by an anonymous blogger —and that’s a small blog. And I can’t think of any serious, popular blog edited by someone who hides behind a pseudonym. People want to know who wrote what. They want to be able to check the bloggers’ background and disclosure.

[And "The Probabilist" would have been a better pseudonym, anyway. ]

-

#2. The author profile does not convey an image of professionalism:

The Predictor sits in a darkened room 20 hours a day, eats cheese and rice cakes, drinks only low fat lime soda, wears glasses, has long hair, and, above all else, is Betfair’s man on everything prediction markets. He is going to take us through the highs and lows of prediction markets past, present and even future. Be sure to keep an eye out for a fascinating insight into prediction markets and exchanges, what they can do, what they have done and why this can be useful to you.

Is this BetFair man educated in the fields of statistics, probabilities and economics? Has he a long experience in applying the prediction market approach? Is he an experienced prediction market blogger? That would be more interesting to know than to hear that their writer “eats cheese” and “has long hair”.

-

#3. The clueless BetFair blog writer resorts to magic to explain historical pricing:

The polls themselves are often a key source of information that move the markets around, but on this occasion the markets were moving in direct opposition to what polling information was telling punters. The 8,000+ people trading on the market were getting it right to the most frightening accuracy.

This is a typical misunderstanding of how prediction markets work. Any trade is in fact informed/misinformed by either passion or some advanced indicator(s). In a cheap prediction market event study like this one, the writer (here, a British stringer working one ocean away from the topic at hand) suppresses evidence of how some advanced indicator(s) (e.g., some leaked polls) impacted trading, and declares that the BetFair magic was at work. That’s not journalism —that’s charlatanism. It’s mediocre. If the prices moved, it’s necessarily that some information hit the US wires. The fact that this British pseudo-journalist did not discover it does not mean that any hint didn’t exist. The prediction markets are simply information aggregation mechanisms (IAMs) —intelligence in, intelligence out; garbage in; garbage out. BetFair does help traders reverse their psychological arrow of time —remembering the future instead of the past. There’s no magic in prediction markets. There is causality —the prediction market journalist should discover it.

-

TAKEAWAY: As a popular blogger who loves to link out to other resources on prediction markets, I regret to say that I won’t read this BetFair crap anymore and won’t link to it in the future. I bet that the serious people genuinely interested in prediction markets will find other blogs to satisfy their legitimate curiosity about James Surowiecki’s wisdom of crowds.

How come BetFair, which is a serious, legal and ethical prediction exchange, powered by so many great people in Hammersmith, can go so low? It’s unbelievable. There is a striking dichotomy between the excellence of their information technology team and the crassiness of the people who speak out for them.

—

The latest installment (which does not featured a single prediction market chart) is written by Justin Wolfers and Eric Zitzwewitz. A pillar of prediction market journalism is to show dynamics charts of prediction markets. Just describing with words a rise or a free-fall does not do the trick. Here are some charts that the two Wall Street Journal stringers don’t want you to see:

—

2008 US Presidential Elections

2008 US Presidential Election Winner

2008 Democratic Nominee

2008 Republican Nominee

—

BetFair

Next US President

Winning Party

Democratic Candidate

Republican Candidate

—

Next US President Will Be Democratic.

© NewsFutures

—

It’s obvious to anyone (but to Justin Wolfers, Eric Zitzwewitz, and their technophobic editor at the WSJ) that prediction market event study would be more usable with charts legended with news markers. (Maybe that will give Nigel Eccles an idea for his next startup. )

So, let’s pass the truisms…

• If Pepsi withdrew from the Cola Wars, then sales of Coke would go up; 7-Up isn’t quite as similar of a product, and so would benefit a bit less. So too in politics.
• After the rich theater of the early primaries, the lead-up to Super Tuesday now looks pretty conventional, with each race expected to involve two candidates, and no third-party surprises.

And let’s try to spot some prediction market intelligence in their report. I don’t see any. It’s a boring historical report from two people whose specialty is everything (economics, prediction markets, betting, gambling, finance, sports) but politics. The ABC of reporting is to trade your passion. Political prediction market journalism deserves more that two university professors who are out of their turf and are using a clunky content management system to tout a bot-driven play-money prediction sub-exchange.

Justin Wolfers and Eric Zitzwewitz are great prediction market experts, and their write-up in the Wall Street Journal is gentle, but that’s not enough because the competition is formidable. There is a glut of amateur and professional political analysts out there, who live and breathe by politics, and the top one percent of them outputs great, pertinent stuff that draws millions of pageviews. In such a cluttered field, you have to differentiate a lot by innovating a lot. Nothing in Justin Wolfers and Eric Zitzewitz’s political prediction market event study stands out. Instead of carving a niche, they are littering redundant banalities.

That said, they are true pioneers in prediction market journalism. They will make improvement, over the coming years, hopefully.