Ask Fear & Greed: Why are betting markets more accurate than polls?

Published Nov 26, 2024, 1:00 AM

Listener Michael asks: Is there a difference in the sort of metrics or data betting markets and pollsters use in arriving at their prediction? In the US elections betting markets were confidently pointing towards one direction and the polls seemed to indicate a close race.

Join Sean Aylmer & Michael Thompson as they answer listener questions.

If you have your own question for Ask Fear & Greed, get in touch via our website, LinkedIn, Instagram or Facebook!

Welcome to Ask Fear and Greed, where we take your questions and do our best to answer them. I'm Michael Thompson, and good afternoon, Sean Aylmer.

Good afternoon, Michael.

Now, Sean, today's question has come from Michael, not me, another Michael. There is more than one of us in the world, and he sent it via LinkedIn. Yeah, and it's a good one, he says. Hello, Sean and Michael. I hope you're having a good day. Thank you, Michael. We are, He says. Is there a difference in the sort of metrics or data that betting markets and polsters use in arriving at their prediction? The question is premised on the US elections, where betting markets were confidently pointing toward one direction and the polls seem to indicate a close race. It's a good question, isn't it.

Oh? It is. I mean, think about how each side, the polsters, and the bookmakers set the odds. So we take the bookmakers first, whether it's a on track book here, the tab sports spread, whoever. They calculate the odds of something happening, let's say one in three chances. They then have to add a profit margin for themselves, so they don't want to always end up square. So maybe it's one in two and a half is the odds they give when it's actually one in three. One in two and a half is actually five to two in betting parlance. The second So that's the first part. The second part is weight of money. If a stack of money comes for one side, as it did in the US presidential election, where there were some really big early bets on Donald Trump, then the bookies or the market makers change the odds because they actually don't want to lose too much money. So if book it go from offering one in three chance to one in two and a half chance and then all this money comes in for one, you know, whatever's being bet on, maybe they go down to from one and two and a half chance to one and two chants and as results, the payout for accordingly. Now, the idea is that the bookie or the market maker wants to dissuade some people from making the bet, and it's sort of a protection protection mechanism for them. Okay, Now, bookies don't always win in the short term though, you know, to Accina sometimes when they talk about high payout ratios, like sometimes yes, casinos over six month period, they just you know, they end up paying out more than normal. That hurts earnings, but in the long term they win because the odds are on their side. The fact that different bookmakers or market makers offer different ad odds demonstrates that it's not a perfect market. So, you know, lots of data, but there's a human element in this and what they perceive, what they know, what they feel. So that's kind of how bookmakers do it. So I mean a lot of fear. I mean there's a method, but there's a lot of feel for it. Polsters, however, it's a bit the same. So gallup remember any about gall and its goal is to find and all pulses goal is to find a subset of people who best represent the likely voter electorate. So GALB goes out ask people, I mean thousands of people a bunch of questions. They give a score to the answers, and so the questions are things like, have you given much thought to the election? So let's use the US election as as an example. Six months ago, they would have gotten people that have given much thought to the election? Do you know where people in your neighborhood? Again to vote? How do you vote previously, what do your plan to vote? Stuff like that, and it's a bit of a dark art, but they kind of put scores on all that and they work out what the likely voter electorate is going to vote for, and that normally the best ones like Gallup put years of experience into that. I means, certainly it's a dark art, and demographics change and voting tensions change, but in many ways, poles are kind of more data driven, so it's qualitative responses, and I think betting is more I mean, there's a way to money, but there's also like how the book maker perceives things and stuff like that.

So I was doing some reading as well, and it does feel like there is not a definitive answer to this one like that from what I could see. What I was seeing was that the polls are largely based pretty much on sentiment in terms of whereas betting markets take into account all available information everywhere. Essentially because there are fewer restrictions on there are no questions being asked that that kind of require you to kind of narrow down your answer into a very specific kind of am I going to vote this way or this way? Kind of based on what has been said, blah blah blah blah blah. It is just basically, based on all the available information, which one do you think is most likely to win? Put your money behind that one. And so again it's Michael. I don't know whether he's going to be kind of satisfied with this, because he's probably come to us looking for a definitive, non waffle answer and I don't think we've delivered on either of those.

Now, nah. I mean, it's the funny thing why. I mean, the polls were mostly wrong. Trump one easily in the US election. The betting markets were right, but by the time the election was running, they were very close Carmla Harrison Donald Trump in terms of the market. Yeah, so the betting markets were kind of wrong in the last few days. They were right to begin with, but then wrong in the last few days.

I don't know whether that proves or disproves anything that we've actually just basically just says that, you know what, But know whether anyone really knows the answer to this one, I'm not really sure. So look, you know what, Michael, We've done our best. Will give us a B plus for this effort? Yeah, absolutely, Yeah, I think we well unsure, I think, Michael, we've gote that B plus And thank you very much Michael for the question. If you have your own question that you would like us to answer, and it can be as loose or as kind of non definitive as this one, send it on through via the website Fearangreed dot com dot au or LinkedIn or Instagram, Facebook, basically anyway get it through to us, will pop it on the list, Oh, Michael Thompson, and that ask Fear and Greed