This is a good point. I don't think betting is always good. And in my case it doesn't apply perfectly, since I never went around advertising my stance on this question (unlike, say Nate Silver and his bet with Morning Joe).

The problem of changing one's mind is interesting, but I'm not sure it's a bigger problem than people just saying stuff that isn't rigorous or thought out.

A mechanism that promotes people actually believing what they say isn't a bad thing.

The reason I like entering into bets is that they force you to admit when you're wrong, and to examine why you were wrong. Far too many pundits have perfected the art of being blithely unaware of all their past errors. Which is surely a bad thing.

I'll add. An outcome bet really serves best as a true BS tax if it's made public at the time, and also tied to a call.

For example, me and Felix made a public bet on whether Greece would leave the Eurozone in 2012 (I won). We made that on a Reuters video, that got published.

In the specific case of this bet, I was eager to take what I thought of was a good bet, which at the time was also outside-consensus. But perhaps there was not benefit to me as a pundit.

Also, to take just one example, I changed my mind on the NYT paywall shortly after entering into a bet about it with John Gapper. I wouldn't necessarily go so far as to say that the bet was the *cause* of me changing my mind, but it certainly didn't make it harder for me to change my mind, as Ben implies.

i actually find this interesting. it's not an elegant solution to the problem of pundits not admitting they were wrong or they've changed their minds, it's at least something.

That said, I suspect the kinds of people who would enter into a public bet like this aren't the ones with the problem in the first place.

i don't have a general ethical concern with pundits doing this. For a reporter it would be far trickier, since it does create an incentive for a particular outcome.

No offense to Joe/Felix etc - but I think the relative distance you are from market-moving reporting the less ethically challenging the gambling thing becomes. Debate would be much different if we were talking about M&A reporters betting on unannounced deals.

This connects to whether the gambler's reporting has any ability to influence outcomes: I'd be fine with a markets reporter betting on the outcome of the Syrian revolution, less so with one betting on whether the market will close up or down that day.

The implicit gambles people take would also be a fun thing to look at - ie. I'd love to know the timing of when economics reporters buy houses etc.

@Tom, I don't understand. Are you saying that M&A reporters have the ability to influence whether a deal will get announced, or that market reporters have the ability to influence whether the market closes up or down on a certain day?

Although I love gambling: Tangentially to Tom's point, this sort of hinges on certain preconditions: a. the existence of a well-defined public personality and the idea of the independent actor. From an editorial standpoint, there's a real brand risk for the lesser known.

And, b. situations with binary outcomes, or creating situations with binary outcomes (unless you're in some kind of parlay disaster). That works for elections or who will leave the EU, but not everything. But maybe that's good--it forces reporting out of equivocating, more focused on reality/outcomes.

@Felix - No, in both cases it's more about creating an extra complication re perceptions of the reporter's thinking / biases. Most conflict of interest stuff is more about dealing with potential perceptions of conflict rather than explicit fraud.

But yes, for market-moving M&A reporting, I think bets would feel wrong to the average reader: if we know the person who got the scoop on Dell going private had made a bet on that happening a week before, would it seem significantly less problematic than if they bought Dell stock a week before?

One problem with reporting is that almost everyone is incentivized create news and crisis where it doesn't really exist, especially those of us who think all the time about traffic.

That strikes me as a gigantic conflict of interest that overwhelms any betting we do.

From a pure editorial standpoint, I'd be better off if the Eurozone crisis went on 10 years, and one country after another almost left.

This is one reason prediction markets are a good idea. Joe has a large binary bilateral bet; if he changes his mind it'll be hard to cut his losses by laying it off. So he's got an incentive, or more realistically a confirmation bias, to skew his coverage that way.

In a prediction market if you change your mind in response to new information you close your trade and cut your losses, so the incentive is to get it right every day rather than double down on a losing bet.

Obviously for markets pundits you don't need prediction markets - you have actual markets - though I suppose many people worry about the ethics of that too.

Btw on Joe's point on traffic incentives: one possibility is that it only "overwhelms any betting we do" because you don't do enough betting.

@Tom, the problem with reporting on Dell going private if you own Dell stock is that you have a financial incentive to report it in a way that might push the stock up further. You know, give lots space to people you're quoting saying that the price is too low, that kind of thing. If you just had a binary bet on whether the deal happens or not, there's no such conflict.

And as Joe quite rightly says, the real conflicts facing journalists are almost never financial anyway. I wrote about that a bit here. blogs.reuters.com

@felixsalmon, damn, that's a long post.

I think we can agree that one worse incentive than betting on outcomes is being paid by the word. Thanks for coming to my branch!

Being paid by the word > being paid by the page view.