Prediction Markets
Editors Note: This is the first post from TheTodd, a long-time friend of mine who has 2 degrees from the U of I. Todd is an authority on many things, and I have always enjoyed his unique perspective. I hope you will join me in welcoming TheTodd to the agora.
~by TheTodd
Compared to most of you, I know very little about politics. I don’t read much news about any of the US presidential candidates, and I don’t have my finger on the pulse of the American people. Despite my lack of knowledge, by occasionally looking at a few numbers and graphs, I have a good idea about how likely each candidate is to win and which candidates are gaining and losing ground.
How is this possible? It’s quite simple, really. People bet on the election, and I follow the odds they’re betting at and how those odds change with time.
There is a well known theory in finance called the Efficient Markets Hypothesis. Wikipedia defines it thusly: “The efficient market hypothesis (EMH) asserts that financial markets are ‘informationally efficient’, or that prices on traded assets, e.g., stocks, bonds, or property, already reflect all known information and therefore are unbiased in the sense that they reflect the collective beliefs of all investors about future prospects.” I know, I know. This market isn’t totally efficient. Please read on.
In a past life, I spent a great deal of time observing the markets for betting on sports. Those markets were not totally efficient, but they were usually quite close, and - with a few exceptions - as the size of the market grew, the market became more efficient. If you’re one of the 5 people interested in the Southwest Wyoming Christian State vs. Northwest Baptist Montana basketball game, you might be able to get an edge with careful analysis. After your $100 bet on that game, the odds would significantly change.
However, if you’re betting on a well-studied NFL game, then you can safely assume that the odds you’re betting at are very close to the true odds, because if the odds were inaccurate, some people with a lot of money and excellent analysis skills will throw money at the game until the odds move enough to eliminate any edge. The best professional gamblers tend to focus on these larger markets because they can bet tens or hundreds of thousands of dollars in them. Their time is wasted trying to find a way to make a few bucks on the SW Wyoming game.
I believe that even bigger markets (like the stock and currency markets) are even more efficient, but again, I don’t think they’re completely efficient. What’s the point of all of this efficiency talk? Well, the betting market for the 2008 presidential election is large enough that I consider it to be “reasonably efficient”. By that, I mean that when I look at the odds, I feel like I’m looking at numbers that are probably just a few percentage points away from the current
true odds in most cases.
Here is the market’s current estimation of the odds to win each nomination:
Republican:
Giuliani 30%
Romney 25%
McCain 18%
Huckabee 14%
Paul 8%
Thompson 4%
Other < 1%
Democratic:
Clinton 62%
Obama 30%
Edwards 5%
Gore 2%
Other < 1%
Some trends:
McCain: Started the year as the clear GOP favorite (above 50%) but steadily dropped until falling below 20% in May. May saw a small resurgence (up to close to 30%), but a steady drop through the second half of May and all of June put him at around 5%, where he remained until a small gain in November and a very sharp gain over the last 4 days. Huckabee: Started the year at around 4%, and spent April-July below 2%. August-early October was spent at around 3%. In mid-October, everything started going right. By the start of November, he was around 7%. By the start of December, he was past 12%. He got close to 20% earlier this month, but he’s fallen back a bit in the last few days.
Paul: Below 1% until June. June and July were spent at 3%. In August, he rose past 5%. October pushed him past 7%. He hit a peak of 9% in early November, dropped sharply (to around 5%) by early December, but he’s back at 8% now.
Thompson: Started the year at around 4%. Gradually rose to around 20% by mid-April. Dropped back to around 15% in late April and May before another steady rise to close to 30% in July. Since July, it’s been a slow, steady fall, though.
Clinton: Started the year at around 47%. Bounced around in between 42-53% until beginning a steady rise in July. This rise peaked at close to 75% in October, but she’s dropped since then.
Obama: Started the year at around 22%. Remained there until a sudden rise to around 30% in March. Remained there until another rise (to close to 40%) in July. Late July-mid October was a steady fall down to around 12%, but he’s been back on the rise since early November.
Edwards: Started the year at about 16%. Hit 18% shortly thereafter, but steadily fell until hitting 5% in July. Rose back to around 7% in August and September but fell back to 4% in October, which is about where he’s remained since.
Gore: Started the year around 8%. Fluctuated rapidly between 6% and 12% until May, where he remained steady at 10% for about a month and a half. In mid-June, he fell until hitting 5% in early August. He then started to rise again, hitting a peak of about 12% in early October before sharply falling in mid-October to below 5%.
Comment by Augur on 22 December 2007 at 5:32 pm:
Thanks for the great post Todd, I have a few questions and observations
Questions:
1) Can I actually make money betting against Gore to win the nomination? it seems completely insane that some people still think he is going to run.
2) Can you purchase on these markets at any time? Like if there is a political event, say a sex scandal, how quickly do these markets react?
3) What kind of volume are we talking? Any rough estimate how many people and how much cash is being gambled in these markets?
4) Have you seen any numbers on the likelihood of Bloomberg entering the race as a third party candidate or the odds of him being successful if he does enter the race?
Observations:
1) I think the best buy is Romney because most people aren’t considering all of the following factors:
a) primary (order of the states) favors Romney
b) compressed primary schedule helps Romney and Rudy and hurts Huckabee b/c Huck has no money, and unlike past primaries, after Iowa he wont have time to raise more
2) Edwards is probably about right, I would be tempted to say that he has a better chance than the market suggests b/c Hillary and Obama are singularly focusing on each other, while Edwards is showing slow gains, plus he lived in Iowa forever, but the reality is probably that even if he wins Iowa, he doesn’t have the money to be competitive through the rest of the primaries.
Comment by thetodd on 22 December 2007 at 7:25 pm:
My answers:
1) While researching this question, I discovered that Intrade (the site these odds came from) changed their fee structure. The new fee structure seems to cause some strange behavior in situations where the odds are extreme (like 2% or 98%). In particular, there is incentive to try to trade out of a position instead of hanging on to it until the result is decided. I suspect many people are trying to trade out of their positions before the Gore market dies completely. If you bet against Gore and hang on to your bet until it’s decided, your profit after the fees will be very small. It’d probably be about $1 profit for each $200 bet. The new fee structure appears to cause some irrational behavior in the Gore market, but I don’t think it has much effect on the other markets at this time.
2) You can bet at any time, as long as there’s someone on the other end willing to take the bet. If the people making the offers on the other end get the news first, they will withdraw their offers or change the prices. In sports markets, unless you had inside information, chances are you would be beaten to the punch by people like me who watched the markets like a hawk. I’m not sure how many people (if any) are paying such close attention to the smaller political markets.
I have not tried to make a general study of how political markets react to news, but I do have one anecdote. In October, Gore made a public statement where he reiterated that he had no plans to run. His odds dropped significantly in the 12 hours BEFORE he made this statement.
3) It’s hard to say how much money is being bet on the election overall, since the betting takes place on many different sites, most of which don’t reveal any information about betting volume. However, for political events, intrade.com is the most active site. At Intrade, it appears that volume on the GOP nomination market has been about $7 million so far, and the volume on the Democratic market has been about $10 million so far. I don’t know how many people are participating, though. In my post, I was trying to convey the idea that the volume in this market was smaller than for a large sporting event like a typical NFL game, but it’s a bigger market than you’ll find for minor sports events.
4) There is a (fairly small) market for betting on which party wins. Current odds in that market are 61% Democrat, 37% Republican, 2% other. There is also a market for betting on the general election winner. Bloomberg is at 0.7% there. This market is still smaller than the primary markets, though. This 0.7% is probably more reliable than the Gore 2%, since there is not much incentive to close trades at irrational prices in this market at this time.
Comment by Augur on 22 December 2007 at 10:04 pm:
Thanks for the answers Todd, interesting example about the Gore announcement.
Comment by John Bambenek on 22 December 2007 at 10:07 pm:
The problem with EMH is that it overlooks the fact that humans, even a large group of educated people looking at the same group of data, are not rational creatures. They aren’t completely irrational, but they certainly aren’t rational either.
The economy isn’t entirely driven by fundamentals, certainly not day-to-day, it is driven by emotion, rumor, group-mentalities, etc. And add on top of that the people who do trade on stocks aren’t necessarily representative of everyone and that group has its inherent biases.
Comment by thetodd on 22 December 2007 at 10:44 pm:
I think John hit upon most of the reasons most markets aren’t perfectly efficient.
Another one I’d like to add is that we haven’t perfected the process of valuation. The best of us are very good at determining the proper value of some assets, but we can still improve.
When I first encountered EMH in the late 90s, it appealed to me very much. However, the dot com boom/bust and my subsequent sports betting experience caused me to replace it with the “sort of efficient markets hypothesis” :)
Comment by Brian on 23 December 2007 at 12:22 pm:
I think another big problem with EMH specifically as applied to predicting the outcome of the upcoming primaries is that expert analysis is largely worthless, so it doesn’t matter much even when multiplied by a large number of people. Primaries always have a tendency to go completely screwy right before the voting actually starts, in ways that pundits and analysts were never predicting. The 2004 Democratic primary is a good example: Dean’s jumped in the polls largely out of nowhere, then plummeted largely out of nowhere, leaving a space for Kerry and Edwards to take first and second and Gephardt (thought to be an Iowa favorite) a dreadful fourth. And all of those changes happened in a very compressed period of time at the end of a very long and fairly static campaign.
This year is even MORE unpredictable, given the heightened importance of Iowa and New Hampshire and the fact that there is no way to accurately poll in Iowa due to the nature of the caucus there and the difficulty in determining who is likely to actually vote there. For the Democrats, because the top 3 are pretty much in a dead heat in Iowa polls, that means NOBODY has ANY CLUE who will win there, and who wins there will have a big impact on what happens in subsequent states. The GOP race is, impossibly, even more chaotic.
All of which isn’t to say that these prediction markets aren’t at all efficient, just that they might be quite a few notches below some of the examples of the more efficient markets you talked about.
Comment by thetodd on 23 December 2007 at 3:14 pm:
Brian, your post raises a question I’ve been wondering about for some time now.
In the other markets I have some experience with (especially in the sports market), the people that are commonly acknowledged as experts - often popular journalists, ex-athletes, and ex-coaches - aren’t really experts at prediction. They’re experts at other things, like persuasive writing, strategy, playing and preparing for games.
In a football game earlier this year, the topic of gambling game up, and one of the announcers put it well. He said something like “If I really could pick the winners, I wouldn’t be up here in the booth.”
The prediction experts - the ones making money betting on the games - generally aren’t public figures, and they generally don’t publish any of their analysis.
I wonder if this is true in the political markets as well, or if the “experts and pundits” are driving a lot of the activity.
Comment by tet on 23 December 2007 at 6:50 pm:
I’ve been seeing more and more neurological research, Todd, on how the human brain functions during the decision-making process.
People are most certainly able to make rational, informed decisions. I’d be willing to bet you right now that somewhere on the University of Illinois campus tonight, I could find someone who could logically devise the series of decisions that would give the people of Ethiopia the highest liklihood of eating well for the next decade.
Unfortunately for decisions inday-to-day life, however, this is not what usually happens. The research into where the brain fires when a decision is made seems to indicate that there is a high degree of correlation between the degree of involvement of the individual making the decision and the liklihood that the decision will be made by emotional methods, rather than logical ones.
In other words, if a decision involves you getting your foot chopped off, I can make a perfectly logical one. If it is my foot instead, the decision is almost 100% likely to be made using emotional brain hardware.
The “jump” to the emotional side can be easily triggered in most people. Try it yourself: The next time you are truly stymied by a decision and the logical, informed choices seem perfectly balanced assign each decision an number, then either roll a die (or flip a coin for a yes/no decision.)
In the split second before the die stops rolling or the coin is revealed, you will know beyond a shadow of a doubt what the proper decision is–it’s the result that you wanted the die to pick during that second.
Immediately upon this decision being made, you will feel your decision-making machinery revving up in your head feeding you justifications for the decision. Once the coin is revealed, you will feel a sense of loss if it’s the wrong choice or an adrenaline rush of joy if it agrees with you.
Now, how does this meta-wiring of human beings impact what we’re talking about here?
If someone is going to be making money by predicting the results of an election, the lower the amount of money possessed by the person betting, the greater the impact on the person and the lower the liklihood that the prediction is going to be made on logic.
This would then seem to indicate that if you divide the amount of the bet by the monetary worth of the bettor, you could get a value that would indicate the likely logical value of the prediction.
This might even work with economics. Certainly markets don’t really behave like the individuals in them are logical or informed. There are bubbles a lot of the time where investors irrationally latch onto things like tulips, click-through sites or fixer-up real estate.
Any of this make sense?
Tom
Comment by thetodd on 23 December 2007 at 7:20 pm:
I just noticed that part of my post went missing at some point. I also posted trends for Giuliani and Romney. Here they are.
Giuliani: Started the year around 15%. Rose past 40% in March, but
slowly fell, dropping below 25% in June. Rose back above 40% in
October, hitting a peak of 45%, but he’s been falling since then.
Romney: Fluctuated between 15-25% for most of the year. In mid-October
his fortunes began to rise, hitting a peak over 30% in mid-November.
He dropped quickly after that, falling below 20%, but since the 12th,
he’s been on the rise again.
Tet, your comment was very interesting, and I’ll look at it more closely later, but I don’t have time at the moment.
Comment by thetodd on 24 December 2007 at 12:46 pm:
Tet,
I liked reading your post because I’m very interested in understanding how and why humans behave irrationally sometimes. In particular, I’ve spent a lot of time reading material on heuristics and biases and behavioral finance. Your explanation of the “flip a coin to decide” phenomenon was still new to me, though!
I think you’re on the right track with your theory about investment size/net worth, but I would like to propose some refinements:
1) We should account for the diminishing marginal utility of each additional dollar. Someone with a net worth of $100,000 would probably suffer more losing $50,000 than someone with $10 billion would suffer losing $5 billion.
2) There are formulas (like the “kelly criterion”) that dictate the proper bet size for net worth growth maximization if you know how large your advantage is on your bet. It is also irrational to make too small of a bet. This is a problem I’ve had. I got too scared to recognize, properly evaluate, and maximally exploit some advantageous situations. I am not sure how to account for this in our rationality formula.
3) You can only invest so much in any market before the prices change (because nobody is left who is willing to sell to you at the old price). So, in some cases, the bets of high net worth individuals will be smaller than they would’ve liked. Similarly, there is also sometimes a minimum investment size, so low net worth individuals are often faced with the choice of either not betting at all or betting a bit more than they would’ve liked.
And of course, if we have more information than just a long list of bet sizes and net worths, we can also do a far better job at estimating the bet’s rationality. Sports books can usually quickly identify who the winners are, because they have other information, like your win/loss results, the timing of your bets, and a history of the size of your bets. With all of this extra information, it’s amazing how much one can infer.
Comment by kofi the one dumb comment can disinterest a generation on 27 December 2007 at 12:17 am:
I’ve come to the party late and so I’m jotting thoughts, starting with the original article, as I go:
1) Your lead-in presumes betting on the presidential elcetion. In what venue are people betting? Sure, there are ‘markets’ online that pretend to track these chances, but their volume is so low its laughable to say they are more efficient or accurate than my buddy’s weekly football pool. What is the source of your proposed efficient market? These pools? Telephone polls? Internet polls?
2) Maybe I missed it (and since you indicate early in the year maybe you have a poll to back up your claim) but I don’t recall Johnny ever riding so high as 50%. Not when he was running in 2000. I’m sure he had favorables in that range during the past seven years but not that I’ve seen regarding this election cycle. Cite please?
3) I’ve read up through the Clinton paragraph now and slightly confused. Are we talking about national polls? Did Hill really pea k at 75% nationally for the blue primary in Oct? Three out of four democrats wanted her to represent them as president and now less than one in two do? I don’t believe it. I would have read something about such a dramatic shift in the newspaper or seen it on my nightly news.
4) Is Edwards really in the single digits nationally? I keep seeing his numbers in Iowa and keep thing there’s a chance a pretty boy lie-yer will run my country. Reading your numbers I guess I am wrong.
5) I became to bored to read the comments when I inadvertently punished my eyes by reading Augur’s ignornat comments about bullshit event markets.
Comment by thetodd on 27 December 2007 at 12:29 am:
Kofi,
1) See the second comment, section 3.
2,3,4) These are not poll results. These are estimated nomination win probabilities. 5% for Edwards means that he has about a 5% chance to actually win the nomination, not that 5% of people will vote for him in the primaries.
These are the odds you could’ve bet at throughout the year. A $3 bet on Hillary when odds were 75% would’ve returned a profit of $1 if she goes on to win. A $1 bet on Edwards at 5% right now returns a $19 profit if he goes on to win.
Comment by kofi the unless now that its so close and up for grabs nobody wants to commit too hard to their..... on 27 December 2007 at 12:41 am:
it happened… my name became too long… in conclusion: kofi the unless now that its so close and up for grabs nobody wants to commit too hard to their…… views lest they come out on the opposite end of their nominee?
Having read through the comments I’ll add a bit more. Intrade? As in TradeSports? As in its nearly 2008 - haven’t you heard of this stuff? This was a big deal a few years ago when articles were being written about the virtues of ‘group-think’ or ‘mass-think’ or whatever they were calling it(today it’d probably be dubbed iThink or some nonsense) and how it beat all analytical models. Woopty-frickin-doo it doesn’t take much to be them ‘experts.’ Whether or not this is true is open to debate, and the inherent flaw of sites such as intrade are the low volume. But Christ? Seriously? A week before NH and this is what Urbana has to offer? A little reverse peristalsis based on inaccurate trend lines based on irrelevant national polling? You want to talk about something, let’s talk about some state polls. Let’s talk about whether or not our polling data is accurate. Let’s talk about how a win here or a second place fishion there might impact the subsequent votes. This blog has been circle jerking about the next president since its inception. Now’s the time to spout. Fuck the event market nonsense. Let’s hear some real.
Comment by kofi the i dont take serious people with 'the' in front of their name; unless they are 'theKoffi' on 27 December 2007 at 12:50 am:
TheTodd:
1) Thank you for referring me to other comments. I got their eventually, but without your guidance I may have been lost forever. We can’t say for sure, but it’s possible. With a life vest like you hugging me so closely, I’ll know never to fear the dark cold waters.
2) Thank you for the math lesson. Whether or not I needed a refresher isn’t important. What is important is the faith you pour into bullshit, inaccurate, questionable low volume event markets. Is it the exiled Nigerian princes day job while he waits for a generous westerner to share their bank account information and launder life times of wealth or is it a legitimate entrepreneurial activity by the mafia as they expand from brooklyn into cyber-brooklyn? We’ll never know. But you were there first. And you talked about it 5183rd.
I’d mock you about all this but I assume every good teacher just wants to reach one student - much like you reached me with your math skills - and every good journalist just wants to inform one person - much like you taught Augur that he can earn six figures a year without knowing about something that has been written about 5182 times before. Chalk up this day as a success, sir TheTodd.
Comment by kofi the i dont take serious people with 'the' in front of their name; unless they are 'theKoffi' on 27 December 2007 at 12:50 am:
TheTodd:
1) Thank you for referring me to other comments. I got their eventually, but without your guidance I may have been lost forever. We can’t say for sure, but it’s possible. With a life vest like you hugging me so closely, I’ll know never to fear the dark cold waters.
2) Thank you for the math lesson. Whether or not I needed a refresher isn’t important. What is important is the faith you pour into bullshit, inaccurate, questionable low volume event markets. Is it the exiled Nigerian princes day job while he waits for a generous westerner to share their bank account information and launder life times of wealth or is it a legitimate entrepreneurial activity by the mafia as they expand from brooklyn into cyber-brooklyn? We’ll never know. But you were there first. And you talked about it 5183rd.
I’d mock you about all this but I assume every good teacher just wants to reach one student - much like you reached me with your math skills - and every good journalist just wants to inform one person - much like you taught Augur that he can earn six figures a year without knowing about something that has been written about 5182 times before. Chalk up this day as a success, sir TheTodd.
Comment by kofi the i dont take serious people with 'the' in front of their name; unless they are 'theKoffi' on 27 December 2007 at 12:50 am:
TheTodd:
1) Thank you for referring me to other comments. I got their eventually, but without your guidance I may have been lost forever. We can’t say for sure, but it’s possible. With a life vest like you hugging me so closely, I’ll know never to fear the dark cold waters.
2) Thank you for the math lesson. Whether or not I needed a refresher isn’t important. What is important is the faith you pour into bullshit, inaccurate, questionable low volume event markets. Is it the exiled Nigerian princes day job while he waits for a generous westerner to share their bank account information and launder life times of wealth or is it a legitimate entrepreneurial activity by the mafia as they expand from brooklyn into cyber-brooklyn? We’ll never know. But you were there first. And you talked about it 5183rd.
I’d mock you about all this but I assume every good teacher just wants to reach one student - much like you reached me with your math skills - and every good journalist just wants to inform one person - much like you taught Augur that he can earn six figures a year without knowing about something that has been written about 5182 times before. Chalk up this day as a success, sir TheTodd.
Comment by thetodd on 27 December 2007 at 1:42 am:
Kofi,
I am fairly new here, so I’m not sure if that’s your normal writing style, but you sound quite angry.
If you felt my answers or original post were condescending, I apologize. I didn’t intend that. If you were just bored reading about something you already knew, I’m sorry, but, since I don’t know you, I have no idea what you do and don’t already know.
“As in its nearly 2008 - haven’t you heard of this stuff?” - In the original post, I note that I pay little attention to politics or news. The post was about political markets because politics is a popular subject on this blog and sports isn’t. My related background is in sports betting. I was completely unaware that political markets were in the news years ago. Despite that, it seems that some readers here also weren’t aware.
“Whether or not this is true is open to debate, and the inherent flaw of sites such as intrade are the low volume” - This was true when you first heard about these markets. The volume is not so low anymore. Those volume numbers I posted in the second comment significantly understate the actual volume of money being bet, since there are many betting sites. Also, each betting site is not its own market, independent of the other sites. If the odds at one site differ significantly from the odds at any other site, software will identify an arbitrage opportunity, someone will make some risk-free money, and the odds will move toward uniformity.
While I have no data about political markets, I have collected substantial data on sports markets with similar volume that confirm their relative efficiency beyond all doubt. Faith doesn’t have anything to do with it, unless you count faith in the axioms that are used to prove oft-used theorems and formulas in statistics. In sports markets of comparable size to the primary markets, large edges are nearly impossible to find. There are small edges to be had, though.
You could certainly argue that there are differences between the political and sports markets. I hope someone here tries to do that, because I’m sure some of you can come up with some thoughts on this matter that I haven’t had yet.
“A little reverse peristalsis based on inaccurate trend lines based on irrelevant national polling” - Why do you think the betting action is driven by national poll results? If I was going to bet in this market, I would use much more information than just some national polls. I strongly doubt most of the bettors - especially the larger bettors - are that dumb.
“I assume every good teacher just wants to reach one student” - Yes, I expected my post would contain some information that was new to some of you and not new to others. Ideally, I want to write a post containing information that’s both new and interesting to all of you, but I can’t do that because I have no idea what you already know or what you find interesting. All I can do is guess based on the previous posts and comments I’ve read.
Comment by kofi the you all should quit your day jobs and start worshipping at the altar of kofi on 27 December 2007 at 12:17 pm:
I’m just kinda cynical and sarcastic all the time. I’m not angry at you, TheTodd. I’m angry at the world.
“As in its nearly 2008 - haven’t you heard of this stuff?” was directed at Augur. I like to pick on him. I like to pick on everyone. Except Hanno. I don’t like to pick on him, but I do it anyway because I feel obligated.
I also refuse to believe that the volume on these event exchanges has grown to any meaningful level. Without being shown some sort of tangible evidence that the daily volume is in the tens or hundreds of millions of ’shares’ or whatever they trade exchanged, I just don’t buy it.
Even if such volume did exist, I don’t buy into the mass think nonsense. I don’t believe its legitimate. It was a passing fad five or six years ago and hasn’t been brought up much since. Must be because its so wildly successful.
Anyway, TheTodd, don’t take it personally. Your post just needed a dissenting comment lest we all quit our day jobs and begin worshipping at the altar of Intrade.
Comment by Hanno on 27 December 2007 at 12:58 pm:
Todd, Kofi is our resident misanthrope. Pay his ire little mind. He seems to get a rise out of bloviating. Though a few of us wonder if we actually know him…his last comment makes me wonder even more.
Comment by thetodd on 27 December 2007 at 1:05 pm:
Fair enough. Volume is indeed not tens or hundreds of millions of shares.
I also agree with you that mass think is not great at prediction. Actually, in the sports betting market, “mass think” is abysmal at prediction. This may sound hard to believe, but the average sports bettor does worse than someone picking at random.
The efficiency is generated by a small number of very skilled people and groups with large amounts of money.
Do these very skilled people also exist in the realm of politics? I don’t know.
The reason I suggested the politics market has such people is that the amount of money to be made is large enough (since a similar amount of money attracts the smart sports bettors)
However, there are also reasons to suggest that such smart people might not exist in the politics market. For instance:
1) You can turn your money over more quickly in the sports market. If you can get an 0.25% advantage in basketball, then you can get an average 0.25% return on your money almost every day. If you get a similar edge in politics, you have to wait for the event to end or hope someone will buy your shares.
2) Political markets are newer and might not be as well-understood yet. Computer analysis of sports for prediction purposes has been going on since the late 70s and has been improved and refined since then. Political markets are only a few years old.