• Will sports wagering expand to political events? Should it?

    The Los Angeles Times has a news report on Las Vegas bookmakers taking some bets on non-sporting events.

    Let me explain why that’s important and why it should be encouraged.


    As Vegas sports books expand betting options, MVP voting is fair game. Will presidential elections be next? (Los Angeles Times)

    And though bets on nonsporting events — such as the Academy Awards, the presidential election and reality TV shows — are not allowed in Nevada, they are permitted overseas, and interest in such betting provides a push for casinos to expand their portfolios beyond traditional wagering.

    But regulators and casinos remain mindful of tampering or influence where voting, rather than scoring, decides an outcome. They have instituted safety measures, such as halting betting well before voting takes place and capping the amount that can be won — usually about $10,000. Most casinos require identification through a player’s card for bigger bets.

    “We’re not going to take six-figure bets on these,” said Jay Kornegay, Westgate’s vice president of race and sports book operations. –LA Times


    My Comments

    These “Prediction Markets” provide valuable information to the public

    • Everything we know about these betting markets suggests they give us the best, unbiased forecasts of future outcomes
      • To put it concretely, you cannot regularly beat the “line” in pro football bets unless you have valuable, non-public information, such as the extent of a player’s injury.
      • If you actually have such information and can legally bet by using it, your bet will change the odds. That gives the public some real information, in the same way that price movements for milk or chicken at the supermarket give you information about shortages or surpluses.
    • The deeper the betting pool, the more reliable the forecasts
      • Are they more reliable than polls? Yes, because the bettors incorporate information from the polls in their betting.
      • In fact, they incorporate information from every source and have powerful incentives to get more information and to distinguish signal from noise.
      • There are such markets now in England, but Americans are restricted in using them; there are small markets (largely for academic purposes) in the US. Larger, deeper markets would be far more informative.
    • Allowing legal betting on political events such as presidential or Congressional elections would make these markets much, much deeper–and give the public better information
    • The only reason to block these bets (besides general opposition to gambling) is the possibility the events themselves could be corruptly influenced by bettors.
      • The fact that many states have lotteries indicates there is no general opposition (though state legislatures may not want any competition)
      • Corrupt influence is far more likely in sports, such as mobsters bribing fighters to throw a match
      • It can also happen in financial markets, such as a major oil-producing country betting on the futures markets, using inside knowledge that the state itself will expand or cut its production
    • The risks of bettors corruptly changing outcomes they bet on is remote and almost certainly less than in other markets were wagering is permitted.
    • For these reasons, political wagering should be permitted in states that also permit sports wagering

    Some examples where it would be valuable for the public to know the odds:

    1. Will the government of Venezuela be overthrown by Dec. 31, 2017
    2. Will Bashar al-Assad still be in office by December 31, 2017?
    3. Will Boris Johnson be Prime Minister of the United Kingdom on September 1, 2017?
    4. Will Pres. Donald Trump complete his full first term in office?
    5. Will the Democrats win control of the House of Representatives in 2018?

    Lots of people are interested in these events and some would like more accurate forecasts about the outcome. For example, oil producers would like to know more about Venezuela as they plan their own production. By the same token, they may have good sources of information about the situation there so their bets would inform others by changing the odds.

    Knowing the odds in political-prediction markets would be very informative to them and to policymakers.

    On international events, for instance, it would give the National Security Council, CIA, Defense Department, and State Department additional information to incorporate in their policy planning.

    For these prediction markets to work well, the outcomes would have to be well-defined before the events. To bet on the outcome in Venezuela, for instance, I need to know the exact definition of “overthrowing the government.”

    That is obviously more complex than setting a 3.5 point spread for a football game, but it is already done in today’s thinner prediction markets. So, I don’t see any insurmountable technical reasons why political prediction markets cannot be expanded.

    They would certainly be useful–to the public and to policymakers.


    If the general topic interests you, I highly recommend a fascinating (and non-technical) book:

    The Wisdom of Crowds Paperback by James Surowiecki

    Surowiecki is not a cheerleader for all crowdsourcing.

    The book explains, in clear, lucid prose, when it works well and when it does not.


  • All the polling averages and betting markets in one snapshot . . Saturday, September 24

    Polling averages and political betting markets as of Saturday, September 24.



    ♦ What about control of the Senate? Very tight race, with conflicting predictions.

    • Princeton Election Consortium estimates the Republicans maintain control (49 seats to the Democratic side, including Independents who caucus with them).
    • PredictIt market says the Democrats win will control (53% chance).
  • Who will win the Senate?

    The Washington Post makes its initial forecast on the 2016 Senate. They predict the Republicans will hold it. 

    34 seats are up, 24 of them are Republican.

    Aaron Blake, who writes “The Fix” for the Washington Post, sees the race boiling down to four toss-ups:

    ♦ Indiana
    ♦ Nevada
    ♦ New Hampshire
    ♦ Pennsylvania

    If Blake is correct about the other states, then the Democrats have to win 3 of these toss-ups if Hillary wins (because the 50-50 tie would then be broken by VP Tim Kaine).

    If Trump wins, then the Democrats have to win all four toss-ups to flip the Senate.

    Still, it is worth noting that this forecast assumes that all states go the direction they are “leaning.” That’s far from certain, and there are lot more of those iffy contests on the Republican side.senate-predictions-wapo-9-20-16-edited-final





    hat-tip-100px-w-right-margin♦ Hat tip to Mike Bauer for sending me the article.


  • All the polling averages and betting markets in one snapshot . . September 18

    Polling averages and political betting markets as of Sunday, September 18.



    ♦ Which states are toss-ups?

    • Real Clear Politics says: Ohio (18), Florida (29), Iowa (6), Colorado (9), Nevada (6), Arizona (11), Michigan (16), Wisconsin (10), Pennsylvania (20), New Hampshire (4), Virginia (13), North Carolina (15), Georgia (16), Maine CD2 (1)
    • Princeton’s Sam Wang says only  North Carolina, Wisconsin, and Nevada
    • Notice that both RCP and Princeton say the election now hinges on these toss-ups.

    ♦ Larry Sabato and his co-authors offer a very interesting perspective on why Clinton still maintains an Electoral College lead despite the fundamentals that should favor him (Larry Sabato’s Crystal Ball)

    Trump is neither mainstream nor conducting a campaign that is anything close technologically and financially to the Clinton effort.
    In our view, this is why — along with strong partisan polarization — the contest, while close, has had Clinton pretty consistently in the lead: Trump is underperforming the fundamentals and reducing the odds of a GOP win.

    Larry Sabato, Kyle Kondik, and Geoffrey Skelley,
    “The Fundamentals: Where Are We in This Strange Race for President?”

    ♦ What about control of the Senate? The Princeton Election Consortium estimates 50 Senate seats for each party. The PredictIt market gives the Republicans a slight edge. (H/t to David Vun Kannon for reminding me to include the Princeton numbers)



  • Polls tightening but prediction markets still bet heavily on Clinton

    Clinton has a 68% chance of winning, according to the PredictIt! betting market (as of Sept. 7).

    That’s a 3% decline since August 31, reflecting considerable tightening in several polls and a few showing the race effectively tied.

    There’s been a steady drip-drip-drip of back news for Hillary, but informed players in the betting markets don’t think it will sink her, at least not yet.

  • Predicting the Presidential Election . . September 4

    Most prediction markets and polling averages still say Hillary Clinton is highly likely to win and that her electoral margin will be substantial. Trump’s likelihood of winning has increased by several percentage points in the betting markets after another week of disclosures from the FBI investigation.

    Trump partisans point to some recent polls that show him leading. That’s why it is important to temper those evaluations with a look at the polling averages and political-prediction markets, which Zip Dialog presents regularly.

    For Trump supporters, the most encouraging news comes from the Real Clear Politics electoral map, which shows 155 toss-up votes in the electoral college. If he surges in those states, he has a narrow path to victory.

    Predictions 9-4-16 revised2