Content Review: Augur, Veil and Decentralised Prediction Markets
“The best kind of collective decisions are the product of disagreement and contest, not consensus or compromise”
- James Surowiecki, The Wisdom of Crowds
Edit (14/4): Ben from Augur helpfully responded with some clarifications on points made in this post. I have updated/made notes accordingly.
Hi folks! Going to try something a little different today. It’s been a relatively slow week content-wise (maybe I’m just looking in the wrong places) so this provides an opportunity to investigate two (inter-related) projects I’ve been meaning to look at for a while: Augur and Veil.
The purpose here is for me to review range of articles and other relevant materials that have been published relating to the projects in order to summarise how each platform work and help get a feel for the current state of functionality, user issues, upcoming features, etc. Do note that my article summaries are intentionally shorter than the usual level of detail — the complexity of the app is such that anyone planning to use it will need to get deep into the detail, or else risk losing funds.
If this format works, I’ll try to do more of these in future for different platforms. As always, open to comments and suggestions. The weekly article reviews should be back next week.
Now, let’s get into it…
Note: I started writing this article around 10am this morning, at which point I began syncing with the Augur blockchain in the hope of including some examples & visuals in the post. It’s now 9.30pm and its only 74.7% synced. This gives some impression of Veil’s appeal for broader adoption. Note that the Augur app can also be accessed via a web app at augur.net
What is Augur?
At a high level, Augur is a ‘decentralised oracle and prediction market’ built on the Ethereum protocol that allows anyone to create a market on anything with the goal of predicting the outcome of that event. Say, for example, that I wanted to forecast the outcome of a political election, a football match or the price of a financial asset at a particular date, I can create a market and let the push and pull of buyers and sellers find the probability of the outcome.
Each market operates just like any other derivatives market where participants can go ‘long’ (buying shares) or ‘short’ (selling shares). An Augur contract is created when one party purchases shares and another party takes the opposite side of their bet. This P2P element may feel unusual for anyone used to betting on traditional bookmakers’ sites, where you’re betting on an outcome against ‘the house’.
Markets are created using Augur’s native REP token, which are used to govern the market and incentivise fairness and honesty. The person who establishes the market designates a reporter (an external source of truth) to verify the outcomes. Upon expiry of the market, the designated party must report on the true outcome in order to redeem the staked REP and fees.
The forecasts themselves are (currently) placed and settled using Ether. There are three types of markets: binary (i.e. yes/no markets), scalar (range-bound options) and categorical (multiple choice). In both types of markets, all ‘shares’ must equal 1ETH in value. This is a really important point and one that’s taken me a long time to get my head round: it’s the scale of 0–1 ETH that determines the probability of the outcome, i.e. 0 is zero likelihood. 1 is absolutely certain.
In binary markets there is a definitive outcome — either something happens or it doesn’t. You would therefore buy shares based on your perception of the probability of the outcome. E.g. if I bid on a ‘yes’ outcome at the price of 0.7 ETH and you bid on a ‘no’ outcome at the price of 0.3 ETH we can establish a share. Let’s say the market was on the chances of a second Brexit referendum; my 0.7 ETH bid suggests I have a 70% belief in the positive outcome, while your 0.3 ETH bid suggests you have a 70% belief that it WON’T happen.
Scalar markets are where the outcome can fall into a range, e.g. ‘what will be the price of Bitcoin on 31 December 2019?’. Participants therefore buy shares for the range they think, e.g. if the scale was $0–100k — and the current price is $5k — then shorts would pay out for results less than 5k and longs would pay out for more than 5k. In essence, they reward ‘closeness’ to the answer rather than exact outcomes.
In a categorical market there would be multiple discrete outcomes for the user to choose from, e.g. which out of a list of candidates will win an election or which team out of a set of finalists will win a cup competition. In such markets the entire payout goes to holders of the ‘token’ (shares) of the correct outcome. Note that these types of markets exist on Augur, but not on Veil.
Augur’s order book only shows ‘long’ shares, i.e. there is only one order book and those seeking to short the market look at the same market data, but in reverse. Creating an order in one order book, therefore, always creates an opposite order in the other.
The application launched in July 2018 and, according to Predictions.Global, currently has 2,264 live markets (only 43 of which are liquid) and more than $2 million of capital at stake. DappRadar suggests the app has had 66 users in the last 24h (soon to be 67 once my sync is finished) and 183 in the past week. I’ve heard that DappRadar’s figures aren’t reliable.
What is Veil?
Veil is a layer three application built on top of Augur and using the 0x decentralised exchange protocol, which aims to reduce frictions in using the platform. It offers a more intuitive and accessible interface than Augur’s native interface.
While Augur requires all activities (e.g. creating markets, placing forecasts, etc.) to be settled on the Ethereum blockchain, Veil allows for transactions to be batched off-chain and then instantly settled using the 0x protocol. It also overcomes the need for the user to download and continually sync the entire Augur blockchain (i.e. the record of all transactions) allowing the user to get up and running by simply connecting their MetaMask account (still a barrier to non-crypto folks, but definitely an improvement).
It should be noted that liquidity is not pooled between Augur and Veil’s order books (i.e. if you had a long order on Augur and I had a short order on Veil we wouldn’t be matched), and that Veil charges a 1% fee for services.
Veil launched in January 2019. It currently allows markets to be created on financial assets (including crypto), sports and entertainment.
How I became interested / background
Very quickly as context and disclosure, I first became interested in Augur in Oct/Nov 2017 and purchased some tokens (which I still hold) with the intention of using the platform. I made an effort in the summer of 2018 to use the platform. I created three markets and attempted to participate in several other. I found serious challenges both from a UI/UX perspective as well in finding sufficient market depth to actually do anything. With the need to be constantly downloading the latest version of the Augur app and syncing to the blockchain, I eventually lost interest and haven’t used it since. Today I’m downloading v.1.12 of Augur and setting up Veil with the hope of significant improvements in order to use them more frequently.
For the remainder of this post I will be reviewing some articles that discuss the platform/s in greater depth, as well as some broader insights on decentralised prediction markets.
Why would someone use Augur/Veil?
Some common reasons for someone to use Augur might include:
· They’re curious on the outcome of an event and want to use an Augur market to tap into the ‘wisdom of the crowd’ for the likely outcome. The requirement to place bets (‘skin in the game’) provides more likelihood of an accurate forecast than, say, a Twitter poll.
· They are ‘long’ on a particular asset and want to use Augur to hedge their position, e.g. someone mining bitcoin could go short on whether the asset will remain above $5k for the remainder of 2019 to hedge their risk.
· They want to gamble on the outcome of a particular event but cut the middleman (i.e. the bookmaker) out of their bet.
· They live in a region where there are betting restrictions (e.g. the US) and so Augur provides a way to escape regulations.
· They want to create markets / predict outcomes on a niche topic that simply wouldn’t have enough interest or liquidity within their domestic market.
· They want to settle a debate or argument in a public way (see Ari Paul’s tweet in the first article below).
Relevant articles
Below are some top-line summaries of articles I recommend reading. As mentioned in the intro, if you intend to participate in markets on either Augur or Veil I recommend reading these carefully before doing so.
Article #1
Title: A primer on Augur and Veil
Author: Jeremy Paly Batchelder
Organisation/affiliation: CoinList and RelayNode
Date: 10 February 2019
Description:
Jeremy discusses building binary (yes/no) markets on Augur. He breaks down essential terminology, including volume, fees (creator and reporter), open interest, and takes the reader through the process of buying and selling shares on the platform. He then provides a brief overview of Veil and discusses the app’s functionalities.
This is the most practical guide to Augur that I’ve found. It explains the complicated mechanism of participating in markets really well and takes the reader through the process step by step. I would recommend reading this article several times before you attempt to use Augur as it will really save you a few headaches.
Link:
Article #2
Title: A guide to Augur market economics
Author: Paul Fletcher-Hill
Organisation/affiliation: Co-founder, Veil
Date: 18 January 2019
Description:
A nice accompaniment to article #1, this post (from one of Veil’s co-founders) looks specifically at how the two platforms work together. While Jeremy’s post focused on binary markets, Paul looks at both binary and scalar markets and shows how to participate in such markets on the Veil platform.
This article provides some really great examples of how scalar markets work in practice, particularly with regard to how you can get ‘leverage’ on outcomes. Read his example of markets on the price of ZRX tokens and it should all be much clearer.
Link:
Article #3
Title: Augur v2: A tour of the prediction protocol’s first major upgrade
Author: Augur team
Organisation/affiliation: As above
Date: April 2019
Description:
As I mentioned earlier, Augur is currently at version 1.12. The developers will shortly be releasing version 2 of the protocol. This post outlines the updates that will be included in the new code.
These include the introduction of:
· The DAI stablecoin into the ecosystem. This should help to remove some of the volatility associated with the previous version. For example, in V.1 there may be a situation where a market that takes weeks or months to resolve, during which time the price of ETH has changed so substantially to wipe out any profit you’ve made in a correct prediction.
· ‘Invalid’ as a market outcome, i.e. users who see a market that they don’t believe will resolve can vote on this as an option. This can help to resolve disputes where the wording of the question was ambiguous, the event ceased to happen (e.g. a football match was postponed) or a bad actor has intentionally set up a market in order to profit (e.g. has timed a market expiry for after the outcome of the event is already known).
· ‘Use it or lose it’ forking. This is a really important (and controversial!) addition whereby if there is a major dispute over an outcome, all existing holders will be required to participate within 60 days or else lose all their REP. REP is therefore not to be considered as a ‘passive investment’ but instead confers certain responsibilities on all holders.
Link:
Article #4
Title: The Ultimate Guide to Decentralized Prediction Markets
Author: Ben Davidow
Organisation/affiliation: Augur
Date: 13 April 2019
Description:
This post helps to show the potential role Decentralised Prediction Markets (DPMs) could play in society. The author shows that markets can be used to improve the use of knowledge in society (hat tip to F.A. Hayek here). It talks about how DPMs overcome the three ‘Cs’ of centralised markets — closed, controlled and costly, which limit their predictive value. By contrast, DPMs are resilient, reliable and (almost) free, which creates ‘a global, open, borderless liquidity pool that serves as an efficient market for absorbing and aggregating the world’s information.’
Link:
Article #5
Title: Practical applications of decentralized prediction markets
Author: Shingai Thornton
Organisation/affiliation: Crypto enthusiast
Date: 28 February 2019
Description:
This post gives a great overview of the decentralised prediction market space. He talks about Augur, Gnosis and centralised apps built on these protocols such as Veil, Guesser (also on Augur) and Helena (on Gnosis). He discusses the potential use cases, barriers and issues.
Link:
Article #6
Title: On-chain trading with Augur and 0x
Author: Graham Kaemmer
Organisation/affiliation: Co-founder, Veil
Date: 6 February 2019
Description:
Taking a more technical perspective, Graham gives oversight of how Veil works. This includes the mechanism for order matching and how the application makes use of the 0x protocol for off-chain matching.
Link:
Additional resources and tools to check out:
1. Augur’s Github repository. From here you can download the latest release and see what else is being developed.
2. Veil. Use Veil as a means for accessing Augur markets (but, as mentioned earlier, with a separate order book).
3. The official Augur documentation. I read this last summer when trying to use the original software release and didn’t find it practical enough for new users.
4. Predictions Global. An overview of the markets live on Augur.
5. Circle’s research on prediction markets. Great insight into Augur and Gnosis. You can probably skip the section on Stox.
6. Binance research on the design flaws in Augur. Arguably many of these should be ironed out in v2 and/or though use of Veil.
7. A guide on prediction markets. I’ve not actually read this yet, but it looks really cool.
8. Curious Giraffe. Analytics on Augur. I like the name.
9. Reporters’ Chat. Overview of the Augur markets that have moved into ‘reporting’ phase. Those that aren’t resolved by the creator/oracle can be done so by a third party in order to collect the fee.
10. Brave New Coin’s research. Fundamental and technical research on Augur:
11. Crystal Ball Be. More analytics on the Augur platform.
12. What one high stakes trader can teach us about Augur. Ben Davidow shares a story about a trader seeing to place a large bet on the Midterm election results.
Reflections / conclusions
· It’s great that there are finally some informative materials around Augur. When I first tried using in 2018 I couldn’t find anything instructive (the official ‘documentation’ isn’t user friendly), which made getting started much harder.
· The mechanism for creating markets on Augur (buying and selling shares) is really complicated. I think I have a decent grasp on how markets work and even then I struggled to get my head around it. This limits the platform’s broader appeal. Veil seems to have abstracted this to some extent with two buttons (‘Buy Yes’ and ‘Buy No’)
· It takes ages for the Augur blockchain to sync. The average person simply wouldn’t have the patience. Using either the Augur web app or Veil would therefore be a more practical option for most users.
· At a glance (using Predictions.Global), liquidity looks far healthier than when I used it last July/August. For the platform to function as intended the markets still need far greater depth.
· Augur v2 includes some interesting updates. Use of DAI is certainly a major improvement. The ‘use it or lose it’ rule will likely prove controversial.
