Cryptocurrency to predict (and influence?) the future.
If you’re into this stuff already (as in, you clicked in here to read about an ethereum-based altcoin that most people have never heard of) the idea of forecasting the future shouldn’t phase you. The draw of cryptocurrency itself is more about what’s possible than what actually exists, so enthusiasts usually have some thoughts, good or bad, about where the world is heading. Gnosis just wants to give you a better view.
To some extent, the past can be used to approximate the future. If you asked 5,000 people yesterday whether or not they like some brand of chewing gum and 63% said yes, it’s a safe bet that if you ask them the same question tomorrow, you’ll get similar results. Problem is, it takes a significant amount of time and money to ask people questions.
Sometimes the value of a prediction can justify that expense. Even something as mundane as chewing gum habits can have value if you’re a candy manufacturer evaluating the possibility of bumping up your gum production line. If a prediction can help inform major business decisions, it’s worth that time and money.
If chewing gum is important, political elections are probably more important. Election results have lasting effects that are felt far and wide. It’s well worth the time and money then, to generate predictions of election outcomes. There are many, many organizations dedicated to using scientific polling to do just that. And plenty of research has gone into analyzing, testing and refining polling techniques. Still, it’s not the most reliable forecasting technique we have available.
See, there’s something interesting about us humans. Not a rule necessarily, but a tendency at the very least. When money is involved, we pay attention. Not a flattering observation, but true. If you‘ve got money riding on the outcome of the next presidential election there’s a strong possibility that you’ll do a little more research. In fact, there’s a strong chance you’ll make a more accurate prediction of the results.
Betting on political races is nothing new. From the mid 19th century until the early 1940’s there was even a formal market in place in the United States for election betting, presidential and otherwise.
Through this market, if you were confident in a certain political candidate, you could purchase a share in their victory. These ‘prediction markets’ were viewed almost as an extension of Wall Street. Odds were posted daily in prominent newspapers and trading volume often rivaled stock trading. The going rate for a particular candidate was strongly indicative of that candidate’s probability of success. This market was eventually pressured out of operation, but during it’s existence it maintained a more successful record for predicting election outcomes than modern polling techniques.
It’s inferred that the accuracy of the market was driven by it’s inherent financial incentive. The New York Times puts it very well in an article published on October 10, 1924 (I found this NYT quote in this research paper from UNC, pg. 9):
“Wall Street is always the place to which inside information comes on an election canvas … [and] it is a Wall Street habit, when risking a large amount of money, not to allow sentiment or partisanship to swerve judgments … any attempt to force odds in a direction unwarranted by the facts will always instantly attract money to the opposite side.”
The efficacy of this political prediction market was largely forgotten by history. But then the internet came along. Online transactions could reduce latency, increase transparency, and offer the flexibility required to get around the localized regulations that seem to have driven historical large-scale prediction markets to a close.
Developed as a research project through the University of Iowa in the late 80's, the Iowa Electronic Market (IEM), a digital reincarnation of the old election prediction market, was one of the first examples of an online prediction market. A study in 2008 of IEM’s success rate covering 5 presidential elections, found that predictions were more accurate than polling 74% of the time.
Although the concept of the prediction market is well suited to politics, it is by no means limited to political predictions. Prediction markets may be applied to the forecasting of all kinds of events. They can be used to inform business decisions, trade on box office success, monitor climate change, facilitate sports betting, and the list goes on.
Online prediction markets are still forced to operate through some form of central authority. In order for them to function, someone needs to be responsible for monitoring results and overseeing transactions. That’s not a bad thing necessarily, but it is a vulnerability.
Today, one of the draws of cryptocurrency is the ability to leverage the authority of a widespread network to eliminate certain traditional authorities. So the question arises, what if we outsource the heavy lifting of a prediction market (the transactions, the verifiable authority, etc.) to the crowd? By providing the right incentive, it should be possible to create a network that processes prediction transactions and verifies their corresponding real-world occurrences automatically, transparently, and in real-time. The evolution of cryptographic technology then, has the potential to take the concept of the prediction market to a new level.
How exactly does a prediction market work?
Gary works in accounting. He loves burritos. He eats them at least once a week. We want to form a dynamic prediction of whether or not today will be burrito day.
Go around the office and ask your coworkers to pony up. Minimum bet is a dollar. For each dollar wagered, you issue two tickets. One ticket signals a vote in favor of burrito day. The other ticket is a vote against burrito day. Let everyone know that when the results are in, winning vouchers may be redeemed for one dollar. Losing vouchers will be valueless.
This is a pretty boring experiment if you just take money from everyone and then give it back later. To make it interesting you need to stimulate some trading.
Since Gary only eats one burrito per week the odds aren’t great that today will be the day. Most of your coworkers are going to be looking to unload their burrito day tickets. They can sell their tickets if they find the right buyer. If you’re willing to accept a little more risk in hopes of a larger payout, this is a good time to pick up some burrito day tickets on the cheap, or maybe sell some non-burrito day tickets at a premium.
You started this mess, so you get to act as the middleman, collecting, matching, and processing their orders. As the day goes by, information is aggregated and sentiment shifts. With a financial incentive to predict the correct outcome, your market’s participants will be motivated to look for ways to get ahead in this market.
Here there is a built-in correlation between the price of each outcome and the probability that the outcome will occur. If, for instance, the going rate for a pro-burrito ticket is 32 cents, you know that roughly 32% of your market is willing to entertain the possibility that Gary will eat a burrito today. You are forecasting a 32% chance of burritos. Any information that your coworkers come across, say an open menu on Gary’s desktop, or a comment regarding lunch options, is likely to quickly be reflected in share pricing.
Where does cryptocurrency fit in?
There are multiple organizations working towards some form of a decentralized prediction market. Most notably, Gnosis(GNO) and Augur(REP). Neither is the clear winner at this point by any means. I focus on Gnosis here, but I provide a brief description of Augur below with some suggestions on where you can learn more.
Gnosis is working on a framework for hosting and interfacing with decentralized prediction markets. You would use the Gnosis framework to create a market and link it to an event. Through this market, users may purchase something akin to a share in the event.
For each share, you contribute what Gnosis refers to as a ‘collateral token’. In return, you receive a set of ‘outcome tokens’, one for each possible outcome of the event. Those who contribute collateral tokens are able to trade outcome tokens, limited only by market liquidity really, up until the moment that the corresponding event has occurred. To launch a prediction, you simply register an event and activate the market.
Now, one of the key difficulties of smart contracts implementations is incorporating reliable data into the system. It’s difficult to verify data, even something as simple as the current time, across a widespread network. So one of the challenges of an automated prediction market is providing reliable verification of event outcomes.
In some cases, Gnosis intends to allow references to data from external platforms to settle the outcome of an event. It’s safe to say we’ll see some services being developed in the future to provide verifiable data to smart contracts tools. But, for cases where a verifiable source cannot be found, Gnosis proposes a solution referred to as the “Ultimate Oracle”.
Sounds magical, but it’s a pretty simple protocol. Basically, wagers on an event outcome would be collected during a short period of time after an event has occurred. To dispute an outcome, you would simply place money on an alternate outcome. If an outcome is a clear frontrunner over a specific period of time, the Ultimate Oracle will settle the outcome and distribute funds accordingly. This would allow users to leverage the platform itself to verify event, and in turn, this would likely generate some useful data for other applications.
The Platform Itself
Gnosis describes their platform as consisting of three separate layers. The first layer, the core layer, will provide the system’s central smart contracts. The second layer, the services layer, will provide some basic packaged services for developers. The third layer is represented by applications built on the Gnosis platform.
But the system is still in it’s early stages. A timeline representing current and projected achievements is available on the Gnosis website. The timeline predicts release of a trading interface in the first quarter of 2018. You can fool around with a rough beta interface currently. Keep an eye on their github if you’re interested in continued development.
Gnosis will incorporate the use of 2 ERC20 compliant tokens, Gnosis (GNO) and Gnosis Wizard (WIZ). The GNO token has a fixed supply of 10 million GNO tokens. They plan to charge for the use of certain services that operate on the core platform. Fees will be payable in ETH, BTC, or WIZ.
WIZ is generated by holders of GNO. There will be a mechanism in place, similar to a Proof of Stake setup, by which GNO may be tied up in a time-locked contract for a set period of time, in return for a preset value of WIZ.
Each Wiz token will be redeemable for 1$ worth of fees on the Gnosis platform. Additionally, the rate at which WIZ tokens are awarded is proportional to the use of WIZ tokens on the platform. This should effectively peg the value of the WIZ token at or very near 1 USD. It’s important that the WIZ token maintain low volatility in order for it to serve as a vehicle of collateral on the platform.
Sales of the Gnosis token originally brought in over 12 million dollars worth of ether. The amount of GNO sold in the Gnosis ICO represents only 5 percent of the tokens created for the project, implying a total market value of nearly $300 million. That’s a lot of money for a project that‘s still this early in the making. In contrast, Augur raised $5 million for 80% of their tokens during their October, 2015 ICO.
It’s not really fair to talk about Gnosis without mentioning Augur. They’re very similar. A full evaluation of Augur is beyond what I’m willing to call the scope of this article right now. As always, the best place to learn is the whitepaper anyhow. But they also have one of the best promo videos I’ve seen. You should also check out their website.
It’s not uncommon to ignore the wider ramifications of cryptocurrency. Many still see it as a financial vehicle or a technological oddity. At the heart of these concepts, and their value, is the suspicion, maybe even hope, that the systems that emerge from this evolution may have potential beyond financial gain.
The weight of this concept really comes home when you realize the role that cryptocurrency may play in future markets. When you’re talking about decentralized, programatic currency, the value of creating dynamic monetary systems is most obvious. If I can hold something of value and orchestrate it’s behavior in a specific and reliable fashion, the applications are endless.
Maybe a little less obvious, but prediction markets look very much like the way to inform that behavior, to provide the data that makes money move. If you can predict the future, you can plan for it, maybe even influence it.
It’s one thing to envision the ability to control the flow of value, but the ability for the flow of value to control itself based on real-time data is more disruptive still. What’s more, the real-time data in this case describes public sentiment.
All in all, the coolest thing about these platforms, both Augur and Gnosis, is just that they represent the kind of idea that you just really want to see exist. What happens when the wisdom of the crowd is given economic power?
Martin Köppelmann and Stefan George are college buddies that initially took an interest in developing a decentralized prediction market.
Joseph Lubin is an Ethereum co-founder and early partner.
Matt Liston is a co-founder for the Augur platform who seems to have jumped over to the Gnosis team.