Whether you have already played on “traditional,” centralized prediction markets or have never heard of them before, there are good reasons to try out the new decentralized platforms. Bodhi Network, a new decentralized prediction market platform, launched live on the mainnet, where similar projects, Augur and Gnosis have had over 3 years without a usable product. In this article, we are going to explain what prediction markets are, how the new decentralized ones are different from the traditional ones, and why they might have huge potential.
We would like to thank the team at Bodhi Network for their contributions to the design and implementation of the research and to the analysis of the result.
What are prediction markets?
Simply put, prediction markets are markets where participants play against the outcomes of events. The market prices indicate what the participants, or the crowd, think about the probability of a certain outcome. The main purpose of such markets is to aggregate the popular beliefs about the potential outcome of a certain event.
[bctt tweet=”The main purpose of a prediction market is to aggregate the popular beliefs about the potential outcome of a certain event”]
Prediction market existed long before scientific polling came into being. It is believed that betting on the outcome of political events was done on Wall Street as early as 1884. The first electronic prediction market – the one relatively similar to those traditional markets that exist now – first came about in 1988. It was Iowa Electronic Markets that is still functioning. It was launched by the University of Iowa, primarily for scientific purposes. The point is, online prediction markets using real money are prohibited in the US as they have been regarded as illegal gambling since 2006 when Congress passed the Unlawful Internet Gambling Enforcement Act. However, launching such markets for scientific purposes can be considered legal. Such a prediction market, PredictIt, was created for scientific purposes in 2014 by the Victoria University of Wellington.
So, how do traditional prediction markets function? Market contracts are created for certain events – for instance, presidential elections. To function properly in the US, the biggest prediction markets have to secure a no-action letter from the Commodity Futures Trading Commission, and that commission decides on the regulations for the prediction market to function. For instance, they might ask to put a cap on the maximum number of traders or on the thematic categories of bets (e.g. political decisions are allowed while betting on sports events is not).
A prediction market contract trades between 0 and $1. When the contract is resolved, you get $1 for each winning vote you made, and nothing for each losing vote.
The current price of a “vote” for certain outcomes reflects the probability of this outcome as believed by the market participants. For instance, if the current price of a vote (or share) is set at 30 cents, it means that the general belief is that there is a 30% chance that this outcome is correct. Or here is a bet on PredictiIt about whether or not Putin will be the president of Russia at the end of 2018:
One share for “yes” costs 94 cents, and one share for “no” costs nothing. It means that 94% of market participants believe that Putin will remain Russian president by the end of 2018. You can buy votes for the outcome that you believe is correct. When the market is closed, for each winning share you bought you will get $1, and nothing for each losing share. The mechanism is simple.
The reason why prediction markets are considered accurate is that money serves as proof-of-stake, so people are more likely to give informed opinions on various topics than they would have done in a simple poll.
[bctt tweet=”Prediction markets are considered accurate because money serves as proof-of-stake”]
How decentralized prediction markets are different
The very idea behind the decentralized prediction market is similar to the “traditional” ones such as IEM or PredictiIt. The main difference is that they are built on a decentralized ledger, meaning that no single authority controls them. This makes these networks more flexible and secure, and it gives them a number of other advantages that will be covered below.
Currently, there are four major decentralized prediction markets: Augur, Stox, Gnosis, and Bodhi Network. The first three rely on Ethereum and the market contracts are administered as Ethereum smart contracts, while the latter is based on the QTUM network. The first three platforms issue their own ERC-20 compatible tokens for trading, and Bodhi tokens are QRC-20 compatible as it runs on QTUM.
Now, let’s move on to the advantages of decentralized platforms over traditional ones.
Why decentralized prediction markets could be better than traditional ones
Based on Bodhi’s whitepaper, let’s try to understand through an infographic how a decentralized prediction market could be better:
The data from decentralized prediction markets can be used commercially
Traditional prediction market data can only be used for scientific purposes, so only a few selected academic institutions have access to it. However, decentralized markets’ data can be used externally, including for commercial purposes. That means that in the future users might get some money from their bets not only through winning on the market, but also from the companies that “order” the opening of contracts for certain events. Sounds attractive, right? And this opportunity can be attractive not only to users but to the companies as well. Here’s why:
Companies are already using internal prediction markets to make the “wisdom of the crowd” help them with the decisions about new products, employment, marketing policies and so on. Some big companies, such as Google, run their own internal prediction markets. The others, including General Electric, Qualcomm, Motorola, and Best Buy, receive services from Consensus Point, a company that specializes in running internal commercial prediction markets to help companies make business decisions. Consensus Point’s solutions prove that prediction markets are highly accurate – the accuracy rate is 90%. The average project savings from their solutions are 58%. Impressive! And they manage to do it with just more than 300 thousand participants.
[bctt tweet=”Companies are already using internal prediction markets to help them with the decisions about new products, employment, or marketing policies”]
Now, imagine how successful similar prediction techniques would be if they relied on a broader and more global public? And that is exactly what decentralized prediction markets can offer. Companies can be able to rely on the replies of their customers, not on their internal assessment, when making decisions. And they will be eager to pay for that while the users will be able to get their share of these money as new prediction markets become decentralized. Sounds good, but for now that is just speculation.
Why you can win more money on decentralized markets than on traditional ones
The calculation of winnings on Predictit and similar traditional prediction markets is pretty straightforward: for each winning share you get $1, and for each losing share you get nothing.
So if you decided to take part in the Predictit’s bet back in 2016 on the US presidential election and bought $100 worth of shares for Trump on September 26th of that year when the price on Trump’s shares was at 35 cents, you would get a total of $285 dollars by the time the bet was closed. So your winnings would be $185. Not bad, right? But what if we told you that you could win a lot more if the betting was only administered through a decentralized prediction market?
The point is that the winnings calculation mechanism is different from market to market. Some decentralized prediction markets, such as Gnosis and Augur, use the mechanism similar to that of traditional markets like Predictit, which is still pretty straightforward. But Bodhi Network, another decentralized prediction market, uses a totally novel calculation method. In this calculation the total number of traded shares within a bet matters – the more shares that are traded, the more money you can win. This means that you’ll be able to earn a lot more money if you participate in a popular bet than on other markets.
On Bodhi Network you place bets not in US dollars but in QTUM. Your total winnings would also be dependent on the price fluctuations of QTUM.
Let’s consider you want to wager $100, you would have to have QTUM tokens worth that much to wager.
Your winning bets = $100 worth of QTUM
Case 1 – Suppose there are $5,000 (equal amount in QTUM) in total bets (meaning all the participants of a bet spent $5000 on bets in total). Total winnings of all participants – $1000 (including your $100), and total losing – $4000.
In this case, on a decentralized platform, 1% of total losing ($40) goes to the result setter, and the remainder is split among the winner. As you put on stake $100, and total winnings are $1000, you have 10% of the winning side. Then you will get 10% of the losings.
You win $396.
Case 2 – Once again, there are $5000 in total bets. But now total winnings and losing are $2500 each. You are again among the winners. 1% (or, in this case, $25) goes to the result setter, the remaining $2475 is split among winners. This time you are 4% of the winning side ($100/$1000), so you win $99.
Case 3 – This time out of $5000 in total bets $4000 are winnings and $1000 are losing. Then 1% (or $10) of losing goes to the result setter, and the remaining $990 is split among winners. As you are 2.5% ($100/$4000) of the winning side, you get $24.75.
But how much do you even get into your pocket?
On PredictIt there are the following fees:
- 10% fee on earnings
- 5% on withdrawal
And Bodhi takes 0% in fees (you pay for the QTUM gas fees only), which makes it better for users in any case. This is a pretty good reason to try out Bodhi Network, isn’t it?
The current industry landscape of decentralized prediction markets
Platform: Ethereum. Augur runs on Ethereum, and the transactions are done in Ether. Its own cryptocurrency, REP, is used only by oracles and reporters (see below). Augur runs on totally automated Ethereum contracts, therefore the creators of the platforms cannot influence the bets in any way.
Fees: None. Augur won’t charge any transaction fees.
Mechanism: Augur bets will be created by oracles who have to stake a certain amount of ETH or REP when creating a bet. This amount is returned to them when the market is resolved. This mechanism incentivizes oracles to correctly resolve the markets. Besides oracles, there are also reporters. They help in resolving the markets – for instance, in the cases when the resolution is disputable or when a single resolution from a single oracle is deemed not trustworthy enough. Reporters also stake REP or ETH. They receive a portion of the ETH lost by the losers in the bet. The more reporters stake and the more actively and correctly they resolve the bets, the more ETH they receive.
Platform: Ethereum. Augur runs on Ethereum, the transactions are made in GNO and OWL, two Gnosis-specific tokens. GNO can be paid for in ETH, and can be converted into OWL via Gnosis smart contracts. OWL is used to pay the transaction fees on the platform, and it is also planned to be used for the withdrawal of money from the platform.
Fees: The contract creators, or oracles, only pay ETH gas fees when creating a contract. Traders pay 0.5% fees for every transaction in OWL.
Mechanism: Similar to Augur, bets will be created by oracles. However, the mechanism of market resolution on Gnosis is a bit different. For some bets there will be centralized oracles – a single oracle will decide the outcome. It can be used for the bets where the result is well-known such as the current year’s Nobel prize winners. For the other bets, decentralized oracles will be used. As Gnosis said, “The common scheme is that the truth is a schelling point. Different decentralized oracle mechanisms usually set up various coordination games where participants are voting in an economically incentivized way. If they vote along with the majority, they gain some value, and if they vote against the majority, they lose value.”
Fees: As claimed in the Stox whitepaper, transaction fees are expected to be around $0.02-$0.2 per transaction.
Mechanism: Bets are created and resolved by the oracles who are incentivized by the fact that they receive a share of transaction fees. Stox claims that their platform will also benefit from cooperation with business providers such as invest.com – a platform for online investments launched by the founders of Stox. As of 2016, it had $3 million in revenue and 50 million users.
Platform: QTUM. Unlike other prediction markets, Bodhi runs not on Ethereum but on QTUM, which combines the Bitcoin Core infrastructure with Ethereum Virtual Machine. Bodhi is one of the first and largest DApps on QTUM. More importantly, QTUM is mobile-compatible, so Bodhi is likely to be one of the first decentralized prediction markets to get a functioning mobile DApp. QTUM is used to participate in bets and BOT, Bodhi’s own token, is staked to create events or vote to correct the reported results of events.
Fees: No fees, QTUM gas fees only.
Mechanism: Bodhi has a “replaceable oracle,” meaning that anyone can be set as the initial oracle to decide the result of the event. If the result is set properly, this initial oracle gets 1% of the total prediction amount. For instance, if there were $100,000 in total losing bets on the event, the initial oracle would get $1,000 for setting the correct result.
The initial result can be challenged by other people. To do so, they would need to stake tokens against that result. If the votes are not unanimous (if there is a challenge), arbitration goes to the next round, and fresh additional BOT tokens must be staked to back their result. Anyone in the public can also stake purely on the validity of the result, regardless of whether they participated in the bet itself or not.
If there is still not a unanimous decision, users must keep staking additional BOT tokens to back their result, convince others to help them, and generally do what they can to achieve a unanimous agreement on their result. In short, the final result has to be backed unanimously by the public.
As we have discussed in the article, current prediction market platforms are quite useful as their results tend to be more accurate than those of polls and other prediction methods. However, they are prone to bias and their operations are very limited and focused on the academic side of things. The new decentralized platforms could be a game-changer in this industry. They won’t be prone to the regulatory attacks and they will incentivize people to participate in the market by giving them a chance to earn more money with their bets.
Right now, there are four major decentralized prediction markets. On the one hand, the diversity of market solutions might harm the development of the industry as potential demand for the platforms will be dissolved among these platforms. On the other hand, these platforms will compete with each other and will strive to make their solutions the best on the market, which in the end will only be beneficial to users.
[bctt tweet=”You can make some extra money just for your opinions on certain issues, which is certainly a good reason to try prediction markets right now!”]
In the future, decentralized prediction markets might get really big, and it makes sense to get a taste of them when the industry is still in its infancy. In the end, you can make some extra money just for your opinions on certain issues, which is certainly a good reason to try prediction markets right now!