FAQ – Frequently Asked Questions

FAQ – Frequently Asked Questions

Reputation [REP] tokens have been distributed.
Here’s how to view and access REP tokens you purchased.

Be mindful of your mobile phone (in)security.
It isn’t safe to use SMS for Two-Factor Authentication (2FA).
Check out this important warning from Kraken, and be careful out there.

Guides

Guide to Creating Markets on Augur

Glossaries

Advanced Terminology
Appeal Bond A bond posted by anyone to have an Open Market re-adjudicated by every reporter. If the consensus remains the same when the market is resolved, the Appeal Bond is paid to reporters to supplement trading fees. If the final consensus is different than the one appealed when the market resolves, the Appeal Bond is doubled in value at the expense of reporters that reported in opposition to the consensus and sent to the Appeal Bond poster.
Blockchain Fork When a chain of transactions splits into two seperate time-lines due to irreconcilable disagreement. When a Fork Bond is posted towards an appealed market this conensus disagreement creates a time-line in which the original consensus is valid and another where it is not. This duality should eventually collapse as the invalid fork rapidly loses all value. This collapse of activity and value is monitored by the Augur system to autonomously select the honest fork for Open Markets.
Early Resolution Bond A bond paid by anyone to request a market be resolved before its expiration date. It is returned to the poster if the next consensus from reporters is anything except “Market is not ready for early expiration,” which results in the bond being paid to the reporters.
Fork Bond A bond posted by anyone to have the reputation history fork in order to contest the consensus that resulted from an appealed markets re-adjudication. In the first fork of reputation, reputation is adjusted as if the previous consensus is correct. The fork bond is distributed respectively to all reporters that committed to this consensus. In the second fork of reputation, the fork bond is returned to the bond poster. In addition to the standard 20% penalty, all reporters that committed to the previous consensus pay a penalty of X\%\ of their reputation holdings where X is the percent of all reputation spent on the Fork Bond.

 

All Terminology
API A toolkit that allows third party systems to interact with Augur’s functionality and data in an automated manner. Augur’s API is open to all.
Appeal Bond A bond posted by anyone to have an Open Market re-adjudicated by every reporter. If the consensus remains the same when the market is resolved, the Appeal Bond is paid to reporters to supplement trading fees. If the final consensus is different than the one appealed when the market resolves, the Appeal Bond is doubled in value at the expense of reporters that reported in opposition to the consensus and sent to the Appeal Bond poster.
Appeal Period The last 24 hours of the Reporting Period. When the Appeal Period begins the Augur platform runs the reported data through analysis and determines the current consensus for each expired market. During the appeal period the consensus outcome can be contested by anyone willing to post a contesting bond. Each time a consensus is appealed, a new consensus is obtained from a more reliable and more expensive source. The first appeal increases the number of reporters to every single reporter. The second appeal ignores reporters and instead determines consensus using market forces. If a contesting bond is not paid to a market during this 24-hour period the consensus most recently determined becomes the final consensus.
Arrow–Debreu model An economic model that there is a correct schedule of prices for everything that would allow supply and demand to converge across all goods and services. In the context of a prediction market, it’s posits that there’s a correct probability for every potential outcome across all subject matter and provides a guide (and expectation) for the pricing of securities in a derivative market and the pricing of probabilities in a prediction market.
Augur An open, decentralized prediction market platform
Automated Market Maker A program guided by a formula designed to emulate and “jumpstart” organic market activity. See Liquidity-Sensitive Logarithmic Market Scoring Rule.
Binary Markets A prediction market with only 2 possible outcomes, for instance “yes” or “no.”
Bitcoin A cryptocurrency, currently the largest by total network value and daily transaction volume
Blockchain A distributed network that establishes consensus on network transactions, generates a tamper-proof history of such transactions and engages in other actions based on a decentralized consensus of all peers.
Categorical Markets A prediction market with a list of more than 2 outcomes, for instance “Hillary Clinton,” “Donald Trump” or “Someone else” will win the 2016 US presidential election.
Clustering Grouping sets of objects by a common factor or factors – an early method of establishing decentralized reporter consensus in a previous version of Augur’s oracle system
Collective Intelligence see “Wisdom of the Crowd”
Combinatorial Markets Markets in which outcomes are influenced by one another and are not mutually exclusive.
Commit And Reveal A cryptographic measure that prevents the revelation of a decision until all parties on the system have committed to their decisions
Commit Period The first 30 days of a Reporting Period. During this time reporters are required to log in and submit encrypted reports about the markets they are assigned.
Commodities Futures Trading Commission The government body charged with regulating derivatives markets in the United States, including centralized prediction markets
Consensus The result of reporters voting on a market outcome. If not significantly and successfully disputed, it is accepted as truth and determines the distribution of all funds remaining in a market.
Cryptocurrency Digital money limited in supply by mathematical proof of ownership and behavior incentives for those that authorize transactions.
Decentralized Managed in a distributed manner that is not subject to the risk a potential single point of failure
Decentralized Autonomous Organization (Dao) An organization whose processes are defined and managed to a significant extent by algorithms – smart contracts – operating on a decentralized basis
Early Resolution Bond A bond paid by anyone to request a market be resolved before its expiration date. It is returned to the poster if the next consensus from reporters is anything except “Market is not ready for early expiration,” which results in the bond being paid to the reporters.
Ether The foundational cryptocurrency of the Ethereum network, accepted as payment for computational power and data storage.
Ethereum A blockchain-powered distributed application platform that executes smart contracts which make up Augur’s back-end. Augur interacts with it to purchase provably accurate computational power and data storage.
Expired Market A market that has passed its expiration date. Expired Markets are scheduled to be resolved by reporters.
Expropriation Bond When reputation is forked, one of the forks represents the fork being justified. On this fork of reputation, 20% of reputation is taken from each of the reporters that committed to the consensus that stood before the fork was initiated. When the market is resolved this reputation is given to to the reporters that committed to the final consensus.
Fork The creation of a new variant of an existing blockchain caused by an irreconcilable disagreement and the subsequent split of the transaction history into two separate time-lines. When a Fork Bond is posted towards an appealed market this disagreement over consensus creates a time-line in which the original consensus is valid and another where the claimed consensus is valid. This duality soon collapses as the invalid fork rapidly loses all value. This collapse of activity and value is monitored by the Augur system to autonomously select the honest fork for Open Markets.
Fork Bond A bond paid by anyone to have the reputation currency fork in order to contest the consensus that resulted from an appealed markets re-adjudication. In the first fork of reputation, reputation is adjusted as if the previous consensus is correct. The fork bond is distributed respectively to all reporters that committed to this consensus. In the second fork of reputation, the fork bond is returned to the bond poster. All reporters that committed to the previous consensus pay a penalty of X% of their reputation holdings where X is the percent of all reputation spent on the Fork Bond.
Indeterminate Lacking conclusive information to decide the question, one of the determinations an Augur reporter can make — for instance, besides “yes” or “no” in the case of binary questions.
Initial Liquidity Requirement The starting value of currency assets a market obtains that allows it to begin trading shares and pay for reporters labor.
Intrade A centralized prediction market that was once the most popular in the US before it failed due to regulatory, legal and internal management problems
Javascript A programming language used for Augur’s front-end interface and for middleware it uses to communicate Ethereum network functionality to/from the front end
Limit Orders An order to buy or sell a security at a specific price or better.
Liquidity-Sensitive Logarithmic Market Scoring Rule (LS-LMSR) The Logarithmic Market Scoring Rule refined to account for changes in market volume, making it a more responsive and precise emulator of organic conditions at the expense of relaxing the loss limit of the LMSR
Logarithmic Market Scoring Rule (LMSR) A formula for automated pricing of prediction market outcomes that directs an automated market maker to offer at least four useful properties critical to well functioning prediction markets: 1) set prices for a set of mutually exclusive outcomes form a realistic  set of odds, 2) be availability as a “buyer” or “seller” of last resort in the absence of one or the other, 3) have prices be updated in a realistic manner approximating how “organic” markets typically work and 4) restrict the market creator’s potential losses to a fixed limit
Market A system where people buy and sell products, services or securities from each other for mutual gain
Market Fee The maximum fee a market can charge, set by the market creator
Market Creator Anyone who asks a question about the future by creating a prediction market that defines possible outcomes and providing initial liquidity to that market
Market Scoring Rule In the context of a prediction market, a framework for determining how a price changes in response to a series of actions by users
Merkle roots Condensed, encoded, cryptographic summaries of transactions. They provide a shortcut — essential information on the transaction history of a cryptocurrency network  without having to download the entire blockchain.
Open Market A market that is past its open date but has not yet reached its expiration date.
Play Money Market A financial exchange involving tokens with no value, with competition for reputation typically being the incentive to participate and amass a larger token balance.
Prediction Market A financial exchange for speculating on potential outcomes
Principal Component Analysis (PCA) a statistical method for isolating the biggest sources of variation in a system
Proof Of Stake One method by which cryptocurrency networks establish consensus and harden network security involving the weighting of each node’s clout by the size of their coin balance
Proof Of Work One method by which cryptocurrency networks establish consensus and harden network security involving computer work on mathematical puzzles
Pythereum A programming language of ethereum
Reporter Anyone that has a Reputation Token. They resolve Augur’s prediction markets that have reached their expiration dates by voting on which predicted outcomes occurred. Reporters lose reputation tokens if they report false results and are paid for their labor with a portion of market trading fees. Reporters lose up to 20% of their reputation tokens if their reports conflict with the final consensus.
Reporting Period A 60-day cycle during which markets progress towards a final consensus using reports from reporters and appeals from the community. The reporting period is cycles through a 30 day commit period then a 29 day reveal period, then Appeal Period (24 hours).
Reputation Token (Rep) A tradeable asset numbering 11 million tokens whose possession makes its owners the reporters of Augur’s platform.  Each token entitles its owner 1/22-millionth of all market trading fees in return for faithful reporting on market outcomes.
Reveal Period The 29-day period after the Reporting Period. At least once during this period every reporter must briefly log in to decrypt the reports they submitted.
Resolved Market A market that has received a consensus from reporters without that consensus being appealed.
Scalar Markets Markets in which the outcome can be any number between the limits set by the market creator, for instance “The Standard & Poors 500 Index will hit ____ by January 1, 2017”
Scaling “Futureproofing” a system to handle more activity over time
Schelling Point The point at which (or decision upon which) a consensus is likely to congregate (settle)
Serpent A programming language of ethereum
Sidechain Technology in development that aims to allow different cryptocurrency networks to communicate with each other
Sidecoin A means of replicating bitcoin’s current distribution onto a new blockchain
Singular Value Decomposition (SVD) see Principal Component Analysis
Smart Contracts Computer code on the Ethereum network that represents fixed sets of rules governing what is to be done or not done with specific cryptocurrency balances under specific conditions.
Stablecoin A cryptocurrency designed to ensure purchasing power equivalent to a specific fiat currency – for example, Dai.
Sybil Attack Attacks on a system, like a reputation or feedback system, using the creation of multiple false identities created to undermine the system or otherwise weaken its integrity
Trader Anyone who buys or sells shares in a prediction market. Traders make profit by selecting markets they are more knowledgeable about than the average trader. They trade shares in these markets based on their expectations about event probabilities.
Trading Fee The fee a trader pays for a transaction. Because fees are applied to predicted profit, trading fees are lower than the advertised market fee when trading options with odds close to 0% or 100%. Trading Fees are split 50/25/25 to Reporters/Market Creators/Liquidity Providers.
Trustless Not requiring trust in a third party. Note that the meaning of this term is different in cryptocurrency circles than in the mainstream.
Two Factor Authorization (2FA) The requirement of a two-step process to access an account involving entry of at least one password and information from another device owned by the account holder
Wallet An application for storing, receiving and sending your cryptocurrency balance and accessing other available cryptocurrency network functions.
Wisdom of the Crowd The notion that certain frameworks, like a prediction market, can act as a magnet that pools the biases, awareness, knowledge and insight of many diverse people with different perspectives in such a way that the overall total is “greater than the sum of its parts”
Validity Bond An anti-spam security deposit Market Creators must pay to create markets. It is refunded when the market is resolved with any outcome other than “Indeterminate/Spam” or “Unethical.”

 

For All Users
Augur An open, decentralized, autonomous prediction market platform.
Consensus The result of reporters voting on a market outcome. If not significantly and successfully disputed, it is accepted as truth and determines the distribution of all funds remaining in a market.
Cryptocurrency Digital money limited in supply by mathematical proof of ownership and behavior incentives for those that authorize transactions.
Decentralized Autonomous Organization (Dao) An organization whose processes are defined and managed to a significant extent by algorithms – smart contracts – operating on a decentralized basis
Ether The foundational cryptocurrency of the Ethereum network, accepted as payment for computational power and data storage.
Ethereum A blockchain-powered distributed application platform that executes smart contracts which make up Augur’s back-end. Augur interacts with it to purchase provably accurate computational power and data storage.
Expired Market A market that has passed its expiration date. Expired Markets are scheduled to be resolved by reporters.
Liquidity Provider / Maker Anyone that offers to buy or sell shares at prices not currently available. They are only given their portion of the fees if their offer finds a buyer or seller.
Maker Fee The trading fee for placing stop and limit orders. They are typically set lower than Taker Fees.
Market Fee The maximum fee a market can charge, set by the market creator
Open Market A market that is past its open date but has not yet reached its expiration date.
Prediction Market A financial exchange for speculating on potential outcomes
Resolved Market A market that has received a consensus from reporters without that consensus being appealed. Once any market is resolved, the consensus becomes permanent.
Taker Fee The trading fee for placing orders that are immediately executed.
Trading Fee The fee a trader pays for a transaction. Because fees are applied to predicted profit, trading fees are lower than the advertised market fee when trading options with odds close to 0% or 100%. Trading Fees are split 50/25/25 to Reporters/Market Creators/Liquidity Providers.
Wallet An application for monitoring, receiving and sending your cryptocurrency balance and accessing other cryptocurrency network functions.
Wisdom of the Crowd The notion that certain frameworks, like a prediction market, can act as magnets that pool the biases, awareness, knowledge and insights of many different people with different perspectives in such a way that the overall total is “greater than the sum of its parts” in its situational awareness.
For Market Creators
Binary Markets A prediction market with only 2 possible outcomes, for instance “yes” or “no.”
Categorical Markets A prediction market with a list of more than 2 outcomes, for instance “Hillary Clinton,” “Donald Trump” or “Someone else” will win the 2016 US presidential election.
Combinatorial Markets Markets in which outcomes are influenced by one another and are not mutually exclusive.
Scalar Markets Markets in which the outcome can be any number between the limits set by the market creator, for instance “The Standard & Poors 500 Index will hit ____ by January 1, 2017”
Market Creator Anyone who asks a question about the future by creating a prediction market that defines possible outcomes and providing initial liquidity to that market
Validity Bond An anti-spam security deposit Market Creators must pay to create markets. It is refunded when the market is resolved with any outcome other than “Indeterminate/Spam” or “Unethical.”

 

For Reporters
Appeal Period      The last 24 hours of the Reporting Period. When the Appeal Period begins the Augur platform runs the reported data through analysis and determines the current consensus for each expired market. During the appeal period the consensus outcome can be contested by anyone willing to post a contesting bond. Each time a consensus is appealed, a new consensus is obtained from a more reliable and more expensive source. The first appeal increases the number of reporters to every single reporter. The second appeal ignores reporters and instead determines consensus using market forces. If a contesting bond is not paid to a market during this 24-hour period the consensus most recently determined becomes the final consensus.
Commit Period The first 30 days of a Reporting Period. During this time reporters are required to log in and submit encrypted reports about the markets they are assigned.
Indeterminate An additional outcome option given to all reporters to choose if none of the outcomes defined by the market creator are verifiable.
Reporter Anyone that has a Reputation Token. They resolve Augur’s prediction markets that have reached their expiration dates by voting on which predicted outcomes occurred. Reporters lose reputation tokens if they report false results and are paid for their labor with a portion of market trading fees. Reporters lose up to 20% of their reputation tokens if their reports conflict with the final consensus.
Reporting Period A 60-day cycle during which markets progress towards a final consensus using reports from reporters and appeals from the community. The reporting period is cycles through a 30 day commit period then a 29 day reveal period, then Appeal Period (24 hours).
Reputation Token (REP) A tradeable asset numbering 11 million tokens whose possession makes its owners the reporters of Augur’s platform.  Each token entitles its owner 1/22-millionth of all market trading fees in return for faithful reporting on market outcomes.
Reveal Period The 29-day period after the Reporting Period. At least once during this period every reporter must briefly log in to decrypt the reports they submitted.

 

For Traders
Limit Order A request to automatically buy or sell shares as long as the price is equal to or better than the limit you specify.

For example: “I predict this event has at least an 80% chance of happening, please buy any shares selling at less than 80%” OR “I predict that this event has less than a 40% chance of happening, so if anyone is buying shares for more than 40%, please sell my shares to them.”

Stop Orders An order type which includes a specific trigger price which delays the entry of the order until the specific trigger price is met or exceeded. Once the trigger price has been reached by a trade, your order will be executed.

For example you’re thinking: “If the odds fall below this much, I will no longer think my prediction is valid, so please sell my shares” OR “If the odds go above this much, I will be confident enough in my prediction to buy these shares”

Trader Anyone who buys or sells shares in a prediction market. Traders make profit by selecting markets they are more knowledgeable about than the average trader. They trade shares in these markets based on their expectations about event probabilities.

Note: all orders are executed in “price / time” priority.

REP (Reputation) Tokens

$REP – Where can I purchase?

REP is currently traded on a number of third party exchanges such as Poloniex, Kraken, Bittrex, Gatecoin, and ShapeShift. REP can be purchased with a number of different crypto currencies and fiat, and is a standard ERC-20 Ethereum token.

Can I participate in Augur without REP?

Yes, Augurs current settlement currency is in ETH, although we have plans to support any ERC20 token on Ethereum. REP is only for reporting on the outcomes of events.

How did Augur create a fair crowdsale without knowing the value of a REP?

The crowdsale operates as a kind of auction where the price of REP tokens is ultimately determined by the market.

The bigger the percentage your purchase makes up of the total crowdsale, the bigger the percentage of all outstanding REP you will get.

Let’s set aside the 20% of all REP reserved for Augur’s team, advisers and the Forecast Foundation and consider the following example:

Imagine a pie with 11,000,000 pieces to be eaten.

Now imagine that Alfonzo pays $1 (or the BTC equivalent) to the Forecast Foundation at the very beginning of the software license/pie sale.

Because Alfonzo is the only person to pay the Foundation thus far, he now lays claim to all 11,000,000 pieces of pie and/or software licenses.

However, shortly thereafter, a second person, Bo decides to pay $10 to the sale as well. Consequently, Alfonzo now lays claim to only 1,000,000 of the licenses/pieces of pie. Bo receives the other 10,000,000.

However, if a third person, Sarah, purchases $1,000, then the distribution of pie/software licenses is redistributed yet again. Of course, Bo and Alfonzo could always purchase more if they feel like they would like to have a greater impact on maintaining the Augur software through reporting. In this sense, the crowdsale works as a kind of auction.

At the end of the software license sale period, reputation will be distributed based on the amount purchased by each address participating and the point at which participants contributed (to reflect their discount %, if any)

The reason we did it this way is to let the market set a value for each rep instead of us. We capped the number of rep and let value fluctuate instead of having an unlimited number of rep but us choosing the unit price. It was more like an auction than a traditional sale.

How does Reputation (REP) work?

 

Reputation can be thought of as a sort of “score” attached to an individual’s public and private address and is both divisible and exchangeable like bitcoin. However, that is where the similarities to cryptocurrency end.

REP has a fixed supply of 11 million and can not be mined.

Those who hold REP are expected, in regular installments, to report on the outcome of a random selection of closed events/predictions in the system. Reporters have a time limit in which to do this. We expect each reporting session to initially be quite quick, but may take as much as an hour as Augur grows in popularity.

Reporting should be a fairly intuitive process, as most markets will have had time before the voting period to auto-resolve themselves (as rational actors will be incentivized to sell their losing shares before a market reaches zero value, leading most markets to have 99-to-1 odds at closing time.)  If markets have not resolved themselves, this may require Reputation holders to do some Googling; but in such cases it is more likely the decision in question will have been poorly worded, undeterminable, or unethical, and thus reporters should mark the decision as such.

If Reputation holders fail to report on the outcome of events assigned to them during the reporting period, or report dishonestly, our system redistributes the lazy or dishonest Reputation holders’ REP to those who have reported both regularly and honestly.

They can lose up to 20% of rep if they don’t report each reporting period. If everyone else reports
accurately the “lazy” person will lose 20%. If there are some other lazy people, and some people who outright lied, you’ll lose some, but a bit less than 20%.

For more on detail on reputation, please visit this blog post

 

How much REP do I lose for failing to report, and how do I reduce this penalty?

Reputation can either be active or dormant. All reputation starts off dormant. When you decide to activate your REP, this is your indication you will show up for work. If your REP is active and you “don’t show up for work” for over a month, you will be penalized 20% of your reputation holdings, and your account will automatically be moved to dormant.

When REP is dormant, it is not penalized, but also doesn’t get trading fees and cannot report. REP is a token that comes with both responsibility and reward; it is not a currency.

I didn’t make the October 3 deadline for providing a valid ETH address. What can I do?

Everything about Augur’s deployment and processes is automated and decentralized — handled by smart contracts — including the initial distribution itself.

As part of the system, all unclaimed REP (unclaimed because the system had no valid address to send REP to by the set deadline) is retained by the “Genesis” account that distributed all REP. You can find that account, its balance and transaction history here (it shows more than 6,000 unclaimed REP after distribution of the other 10,993,000+ REP):
https://etherscan.io/token/REP?a=0x0000000000000000000000000000000000000000

It’s part the Ethereum system now — neither we nor anyone else can access it.

You are free to check back on that link as often as you’d like – you’ll find that the balance will stay there until the system goes fully live. At that point, the system will treat it like any other REP that fails to report in line with the consensus: it’ll be subject to rounds of incremental redistribution every reporting round from those who failed to report on market out comes with their REP to those who did report in line with the consensus.

For those this happened to, we’re sorry this happened to you. We made multiple efforts over more than a month to alert everyone to the deadlines in addition to e-mails we directed specifically to people who had invalid or nonexistent addresses.

Augur’s entire system relies heavily on REP holders who are aware of and carry out their responsibilities. It relies on rewards for those who do and penalties for those who don’t – and for both rewards and penalties to be determined by autonomous code beyond individual human control.

The more REP (Reputation) tokens I have, the more I get in trading fees?

Yes, the greater the percentage of all REP tokens you possess, the greater the percentage of trading fees you get. For example, doubling your reputation holdings will double the fees you receive relative to other users. On the Augur platform, 50% of all trading fees get distributed to active REP token holders with each payment proportional to their holdings.

What are recommended Ether wallet options for storing my REP tokens?

We don’t have official recommendations, but widely mentioned immediately available options (some still in the process of supporting REP) include:
1) uPort
2) Jaxx
3) Mist
4) myetherwallet.com
5) Ledger Nano S

What happened to the crowdsale page? How do I access my REP?

The crowdsale page was taken down some time ago as a security precaution. Check the link at the top of this FAQ page for instructions on accessing your REP.

You’ll first need your ETH payout address — if you need a reminder of what that address is, please check your e-mail inbox and computer files from the time of the crowdsale. It might contain references to this address (crowdsale participants who bought with BTC were given an e-mail and download option to get their private key file at the time the address was generated).

If you bought with ETH, by default, we sent the REP automatically back to that account, so check your ETH transaction records.

General Overview

Are Augur’s prediction markets limited to Yes or No?

No, while “binary” markets dominate the front end of our demo, the back end of the system already has the capability to support scalar markets (the S&P will drop to 1500) and categorical (US presidential election will be won by Clinton, Bush, Trump, Sanders, etc.) and we’ll likely expand to other types of prediction markets

Can traders sell their shares as soon as an event happens or do they have to wait for the resolution to be finalized by reporters?

Traders are always allowed to buy and sell from the moment a market is created until it is resolved and closed by reporters. Almost all traders will sell immediately after the event happens. This is because after everyone knows the outcome, the value of the winning shares will be over 99% of their full value. Sometimes the outcome is known before the event date, and this process can start happening even earlier. What process? Who is buying these shares?

Typically once an event happens, arbitrageurs will show up and start buying all the shares off of normal traders. These are people that buy shares worth >95% and wait until they are worth the full 100% after resolution. The traders are happy because they get virtually all of their winnings immediately and can use them right away to trade on other markets. The arbitrageurs are happy because they get a virtually risk free place to invest their money for profit. The value of the winning shares will continue to get closer and closer to 100% until the profit margins become too small for arbitrage investors.

If you sell before the outcome is decided by referees, you do have to pay another trading fee, but since trading fees will be quadratic (aka bigger near even odds and smaller at the edges) it’ll be a tiny fraction of the normal fee.

When traders buy and sell shares it may seem like they are predicting an event, but because the blockchain is not aware of the outside world, they are technically predicting which outcome reporters will select when they resolve the market. Reporters are honest because of the incentive architecture that applies to them. In effect, reporters will report an outcome that always matches the event, such that betting on an event is the same as betting on how reporters will report on that event.

This is because reporters play a ‘game’ with only 2 profitable strategies. The first strategy is to tell the truth, and not just any truth, one that the majority of traders believe. The second strategy is fraud. For fraud to be successful, attackers must donate enough money to honest reporters to double their wealth, a cost equal to the market cap of REP. Additionally they must wait through 4-6 months of appeals to create a fork filled with fake markets containing more money than exists in all of Augur combined. Then, once they trick Augur into thinking their fork is the one customers prefer, they have to hope that even after 4-6 months of all traders being warned their money will be stolen if they don’t withdraw their funds, that there is enough money left to recover the hundreds of millions of dollars they just lost. The reliability of reporters is backed by this incredibly high cost of fraud. It is this low risk reliability that arbitragers rely on to give traders good deals on their winning shares as soon as the event happens.

How is the create market bond and fee calculated?

The equation for calculating the market bond can be found here.

The minimum creation fee is 0.025 ETH. This can change depending on the maker / taker fees set for your market.

The bond is calculated via:

( Creation Fee * ( Total Events Past 24 Hours ) / ( Total Events in Reporting Period ) ) / 2

The idea behind this is that if you attempt to spam create markets, the total events will go up, making it more expensive to create a market. But it is now dynamic based on the amount of events created in the past 24 hours in relation to the total events in the current reporting cycle.

What about legal and regulatory issues involving prediction markets?

Our views are detailed at the following links:
http://augur.strikingly.com/blog/regulation-and-augur

What is going on with reporting during the beta?

While in Beta, Augur runs on the Ethereum testnet and our own private chain. Due to this, reporting cycles are only 2 days long. This is for testing purposes when we test full reporting cycles. Your only allowed to submit reports during the first half of this, so you only have one day to report during the beta. We also frequently reset our private chain, resetting your account and transactions.

What’s the difference between events and markets?

An “event” is the thing that reporters report on. A “market” is the thing that traders trade on. The only distinction between them (from a user’s perspective) is that you can create multiple markets (with different fees) for a single event. *Usually* if you want to create a market you actually want to create the event and the market simultaneously, and ignore the distinction between them.

Events: An event is some real world event with either a binary (yes/no), scalar (some number, i.e. 124.27), or categorical (Bush, Hillary, Trump, or Bernie, or none of the above) outcome. Reporters report on events. An event creator would pay a bond, in the case that their event is a spam event / gets reported as “indeterminate” the bond would be split amongst the reporters. An event creator doesn’t stand to make any profit — most of the time, market creators will simply create an event before creating their market. However, if there’s already say a “Will Hillary Clinton win the US Presidential election in 2016?” event, a market maker can use that instead of creating their own event.

Markets: This is where the predicting actually takes place. A market can contain one event. A market has a few key properties: 1) initial liquidity provided by the market creator (who pays a small fee + initial liquidity — however, this is optional), and the trading fee (also set by the creator). A market creator sets the initial properties.

When is Augur launching?

Our current timeline is looking like a late summer 2017 launch. We are in process of beginning security audits, which will be followed by a single bug bounty market along the lines of “Will this market be hacked?”. Once completed, Augur will launch live on the main Ethereum network.

Will Augur allow for limit and stop orders?

Yes, limit orders are built into Augur but stop orders are not.

Stop orders are difficult to implement because ethereum software is only running when it is paid to do so. This means if a user wants their stop order to be executed when they are offline, they have to pay other users to check if the order is ready, and to activate the stop order on their behalf. To make matters worse, activating the stop order involves paying ethereum to review every single stop order to make sure the user is not lying. Checking every stop order is so expensive that it becomes impractical above ~50 stop orders. It would also be unfair because each customer placing a stop order would pay a bigger fee than the last one. Ultimately, adding stop orders would require the poster of the stop order to pay high fees into a holding account that later refunds who ever fronted the cost of doing the expensive transaction. Currently stop orders are impractical, but will be revisited as advances are made in the field of trustless decentralized computation.

Will other contracts be able to query Augur outcomes?

Yes, Augur is actually 2 products that can be used alone. One is a prediction market, the other is an oracle. The results from the oracle service are publicly available, and anyone can pay what will likely amount to several dollars to have an official statement regarding the objective truth about a subject.

System Mechanics

How does Augur prevent spam attacks from overloading reporters with work?

Several things interfere with the ability to overload reporters:

The amount of reporters assigned to each market is dependent on the amount of money remaining in the market at expiration. Spam markets will likely have little or no volume and will only be assigned the minimum of 3 reporters.

Markets come with a market creation fee. This fee is priced such that even if every single market was a spam market with zero volume, reporters will still obtain a decent minimum wage despite the lack of trading fees. Every cost that Augur imposes on users or employees are mathematically calculated to be as inexpensive as possible without harming the network. This practice includes the the market creation fee.

Market creation requires a validity bond. Validity bonds are priced using metrics that represent the accuracy of reporters, along with their supply and demand. Validity bonds are returned to the market creator if reporters agree the market has a determinate outcome. They serve to prevent spam and provide a buffer between the potentially volatile demand for reporters compared to the relative inelasticity of supply. Every 2 months the average amount of markets created per day is measured. This is used to benchmark expected market creation rate. Validity bonds are priced using peak-load pricing. This means if a day comes where markets are being created at double the normal rate, then validity bonds will be doubled. Spamming effectively results in reporters getting overtime pay.

Validity bonds are also affected by the performance of reporters. If more than a few percent of markets are not reported on, or have their decision appealed, the validity bond begins to rapidly increase.

Is there a total number of REP to be issued?

 

Yes, 8.8 million (80%) of all 11 million tokens are being sold to the public as part of the crowdsale The remaining 2.2 million tokens will be distributed to the team/advisers (16%) and to the Forecast Foundation (4%), a non-profit which will be formally responsible for managing the maintenance, enhancement and promotion of the platform.

Why are trading fees less than advertised when odds are close to certain?

Applying flat fees to the total transaction volume would result in scenarios where trading is pointless. For example, if a market was trading at 98% odds and had a 3% fee, trading these shares would only result in losses and trading would stop. We would like trading to continue with little friction at all odds and for the fees to be merit based instead of flat. To do this fees are applied to potential profit instead of volume. The basis of this is ‘chance of winning’ * ‘potential profit’ and the resulting quadratic formula is TransactionFee = 4 * Odds * (1-Odds), where odds is the odds of the share you are buying.

The fee is maximized at 50% odds.
The fee is 90% of what is advertised when odds are 66 or 34.
The fee is 70% of advertised when odds are 77 or 23.
The fee is 50% of advertised when odds are 85 or 15.
The fee is 20% of what is advertised when odds are 95 or 5.
Visually represented with X = Odds of purchased share and Y = fraction of maximum fee: http://www.wolframalpha.com/input/?i=x+*+(1-x)+*+4+from+0+to+1

Why are trading fees split 50/25/25 for reporters, market creators, and makers?

Short explanation:
We decided that market creators no longer have to provide initial liquidity, the primary cost of market creation. The benefits are increased specialization of roles and removal of incentive to bypass liquidity requirements. To reduce income to match reduced costs, market creators were reduced from 50% fees to 25% fees. The remaining 25% is used to subsidize makers. (traders that provide liquidity to markets)

Full Explanation:
Due to the complex nature of the Augur incentive architecture, the exact effects of any given fee split are difficult to predict. An objective mathematical determination of the perfect fee distribution for a given set of measurable goals borders on impossible. However, the general effects of changes can be predicted:

Reducing fees sent to reporters would reduce the value of REP, reduce the number of reporters, and proportionally reduce the cost of a successful attack on Augur.

Reducing fees sent to market creators would reduce the number of markets and increase trading fees.

Reducing fees sent to liquidity providers (makers) would reduce the size of the order book, reduce accuracy of predictions, reduce transaction volume, increase spreads, and increase the difficulty of a new market finding customers willing to risk providing liquidity.

We have a few goals in mind for a fee distribution model.
(i) It should target simple fractions in order to be easy to mathematically work with and understand.
(ii) When optimal market creator behavior is mathematically predicted, it should result in typical trading fees that are drastically cheaper than centralized competition, yet not so low the fee becomes irrelevant. We are targeting to have typical transactions seeing fees between 0.1% and 4%.
(iii) We would like to respect economic research that has demonstrated that allowing zero and even negative fees to liquidity providers not only pays itself off in the form of more customers, but also increases the size of the order book. It additionally reduces or eliminates the price difference between buying and selling known as the spread. Paying traders to create conditional orders instead of immediately filled orders will result in greater accuracy, more customers, and more transaction volume.
(iv) We would like to respect those that bought reputation under the expectation of getting 50% of fees.

Item four requires us to maintain 50% fees to reporters. That leaves 50% of the fees remaining. Initially, Augur planned to combine the role of liquidity provider and market creator into a single role that got 50% of fees. However, this does not allow specialization. People who are best at creating markets are people that are good at predicting the amount of demand from customers. People who are best at providing liquidity are those who are best at predicting the outcome of the market and have lots of money to throw around. In hindsight, combining these 2 skills into a single role was unwise. It excluded poor people from making markets while discouraging anyone from making a market unless they were skilled simultaneously at both predicting demand and predicting outcomes.

In order to separate the skills of predicting demand and predicting outcomes, we removed the requirement for market creators to provide a large amount of starting money to the market. For market creators this cash infusion requirement represented the majority of the cost of creating a market. With market creation reduced to a fraction of the expenses, we cut their income in half leaving market creators with a 25% share of the fees. The 25% that was freed up is now paid in subsidies to anyone that successfully completes a conditional order, for example a limit order, on the order book. These liquidity providers, also known as makers, assume additional risk over other traders while at the same time providing value to the community in the form of increasing prediction accuracy and trading volume. After internal debate and economic research, we came to an agreement that the best use for this spare 25% would be to subsidize liquidity providers.

Also note, nothing about Augur’s structure is written in stone – as we’ve demonstrated. We adapt and incorporate the best ideas out there (just check out our github history), and we will be relentless in doing what’s needed to solidify our first-mover advantage, maintain our decentralized nature and achieve widespread mainstream adoption.

Why is there a delay between when a report is generated and when it is released?

It’s an anti-collusion measure, commit and reveal: https://en.wikipedia.org/wiki/Commitment_scheme

How does a time delay help the commitment scheme? Couldn’t a shorter loop still work with that mechanism?
The advantage of resolving in a very short time such as 24 hr is much smaller than it initially seems. Once an event date happens the odds will be nearly 100% for the winners. If traders withdraw their winnings after the event date but before the resolution, they only stand to lose a tiny fraction of a percent by withdrawing. The only traders that will not immediately withdraw are those that care about that tiny fraction of a percent enough to wait for resolution. This tiny minority of traders having to wait longer for their fraction of a percent are the only people harmed by a longer resolution time.

On the other hand, longer resolution time comes with strong benefits. Giving reporters longer deadlines to complete their assigned labor increases the number of reporters that choose to work, thereby increasing security. Having 30 days to complete 30 days of work at times of your choosing is far more practical than requiring every reporter to show up for a minuscule amount of work every single day, and risking a 20% loss of their assets if they “no call no show” (reporters can flag themselves as dormant to virtually remove the failing to report penalty). In addition to increasing the number of reporters that complete their assignments, a long reporting period gives the community time to notice an attack and withdraw money before the attackers can steal it. With a 2 month reporting period, it would take an attacker 4-6 months of obvious fraud to advance past each tier of security. This allows traders plenty of time to withdraw funds before they can be stolen.

Will creating multiple accounts give me more influence or reputation?

No. REP tokens have to be bought. They cant be gained by just creating new accounts, and the system determines voting clout (and the percentage of all trading revenues you get) based on how many REP tokens you have, not how many REP accounts you have – that’s irrelevant to the system. “Sybil attacks” using this method will not work.

Will market forces lower trading fees to be practically zero?

Market creators decide the fee rate for each market they create. Setting the fee higher results in more income per trade, and if a clone of your market exists with lower fees, will also result in losing customers to competition.

So, what is to stop a race to the bottom where market creators engage in a price war to zero fees?

First, markets require a creation fee that covers the labor cost of the minimum 3 reporters. This creation fee removes the possibility of reporters working for no pay to resolve a market that did not get any customers. Second, market creation requires a validity bond. Validity bonds are priced using metrics that represent the accuracy of reporters, along with their supply and demand. Validity bonds are returned to the market creator if reporters agree the market has a determinate outcome. They serve to prevent spam and provide a buffer between the potentially volatile demand for reporters compared to the relative inelasticity of supply.

Both the creation fee and the validity bond represent costs to the market creator. These costs result in a minimum fee a market creator must set in order to break even. The more volume a market is expected to get, the lower the fee can be set without losing money.

However, these costs are still very cheap, and this means there is a low barrier to entry for competition. This low barrier of entry would result in price wars and mass duplication of the most popular markets. Because market creators are allowed to adjust their fee after creating the market, it would also incentivize market creators to use predatory pricing to obtain de facto monopoly status. To prevent the spamming of temporarily sub-competitive markets, the validity bond is affected by the current fee. The lower the fee is set, the higher the validity bond requirement. The validity bond is priced as an inverse of the fee such that if a market creator wishes to cut their fee in half, they must first double the size of the validity bond. A fee of 0% is impossible because as the fee approaches zero, the validity bond requirement approaches infinity. The result is that the barrier of entry for market creation is no longer equal for every market. The barrier is greatly lowered for markets that expect few customers and have high fees. Likewise, markets that expect to be popular and therefore have low fees will require a large amount of capital to create.

 

Technical

Can I use Augur with my local wallet instead of the hosted one built into the website?

Yes, all features needed to use Augur will be built into the website itself, and that includes the wallet where your money is stored. However, we have also made the opposite thing possible. Augur has a program that can be installed such that you do not have to use a browser at all, and can keep your passwords and money on your own computer with your own local wallet.

Can I use the ethereum account that was created by the crowdsale as a normal ethereum account?

Yes, *but* in order to send that on to the sale contract and get credit for it you’d have to import the new address into geth and send with geth. In other words, it’s much easier to just use geth to begin than to use the generated one.

How will Augur handle updates to its code after launch?

On ethereum software code can not be changed or updated, however data stored in contracts can. How do ethereum developers update code while that code is currently in use and immutable?

Augur follows the standard practice of having a contract that maintains a master directory of contracts in its data. Any time one Augur contract wants to interact with another, it checks the master directory to know what address to interact with. This master directory allows individual components of Augur to be hot swapped.

The master directory contains an approved administrator as an ethereum address. All requests to swap out contracts in Augur must come from this address or be rejected.

Initially this administrator address is simply an account held by the lead programmer. However, we plan to switch this address out with a contract that will add the requirement of approvals. Once this contract is in place, users can make requests to have a specific Augur contract swapped out with an updated version. This request is submitted to the approval contract. The approval contract then accepts votes from holders of reputation tokens. If the update proposal is approved, the approval contract submits the change request to the master directory, which in turn updates its data.

Because REP holders have absolute authority over this approval process, potential conflicts of interests arise. REP holders would vote in approval of any proposal that directed to them more money and power. This would eventually result in REP holders exhibiting monopoly-like behavior. They would vote to exploit the users as much as possible without collapsing the business. We are creating Augur to be autonomous. This means it is designed to resist attempts to tamper with its original intended purpose (providing accurate forecasts of the future along with objective determinations of truth to humanity as cheaply as possible). In order for Augur to maintain integrity, it will not allow many functions to be updated. Any functions or databases that present of a conflict of interest by encouraging human corruption will be compartmentalized into contracts that can not be changed without widespread community support for an improved clone of Augur to be created.

 

Is the June 17 DAO hack relevant to Augur? Does it have a similar vulnerability?

As far as we can tell, no.

Before the hack, they fixed the bug here
https://blog.slock.it/announcing-dao-framework-1-1-35249e2e001#.n1f9147ha

but they didn’t push it live because they thought it only applied to part of the code not used yet and not to splitting, which is active. Unfortunately, the attack worked when splitting too.

Here are the relevant lines of code:
https://github.com/slockit/DAO/blob/a188e9e78ab253e4307238f12de333ca625556bf/DAO.sol#L747
https://github.com/slockit/DAO/blob/a188e9e78ab253e4307238f12de333ca625556bf/DAO.sol#L686 is what calls withdrawRewardFor which has a vulnerability (vuln).

Basically in the live version of the dao line 747 is after line 749, which allows splitDAO to be recursively called without proper consequence.

Another potential fix would’ve been if this:
https://github.com/slockit/DAO/blob/a188e9e78ab253e4307238f12de333ca625556bf/ManagedAccount.sol#L60
[which withdrawRewardFor calls] were a send as opposed to a call, since the call allows the vuln here as well, but a send does not.

You can recursively call splitDAO with the withdrawRewardFor loophole to have this line https://github.com/slockit/DAO/blob/develop/DAO.sol#L662 repeat a bunch before it gets zeroed out allowing the child dao to have a ton of eth it shouldn’t.

With a call transaction the contract you call can execute arbitrary code at the beginning, which means it can recursively call the original function a bunch of times before your balance ever ends up getting decremented. So the call in L60 above is part of what enables this vulnerability [as does the improper incrementation order in 747]

As far as how it applies to Augur, we don’t have this specific vulnerability as far as we know [we use send], and we recently added code to prevent the callstack overflow exploit.

We’ve planned from the start to have rigorous third-party security audits as soon as we’re out of beta and before we go live as a real-money platform.

Is there API documentation?
What are the main components of Augur? What language is it written in?

Ethereum contracts in serpent and a UI and interface to the contracts in JavaScript

What do all of the terms in the private key json file stand for?

 

Where’s the code?

Platform / Personal Security

How can I access my REP if I have my private key / keystore file?

MyEtherWallet.com is the easiest way to access your REP. Keep in mind, you will need to have a small bit of Ether in your address to send the REP (to pay for gas costs). Keep your private key / keystore file PRIVATE and out of any centralized services (email, calendars, dropbox, etc) – as a malicious actor can steal your REP with such info.

How can I use REP with a hardware wallet?

We have a guide for storing your REP on a Ledger Nano S here. A Trezor is similar, as well as a cold storage wallet (printing out your private key as a QR code on a secure computer). Please reach out to us if you want more detailed information on how to safely store your REP.

How is Augur approaching code security after the DAO attack?

We decided it’d be better to make the code comply with NASA and the Jet Propulsion Lab’s coding guidelines, the Motor Industry Software Reliability Association guidelines, and then various recommendations from approximately 15 blog posts people made after the DAO hack before beginning audits.

We want the code to be as easy to audit as possible and be as simple and secure as we can get it before going to previously planned third-party security audits and bug bounties.

The key is getting it as bug free as possible, then having ways to recover from any vulnerabilities easily on contract as opposed to hoping for an ethereum fork. We have plans for this, including incentivizing people to try to hack it.

Will Augur contracts be vulnerable to the interactivity parasite attack?

These aren’t a valid concern at least as laid out by most (if you include one chain within another chain it can be an issue, but that’s a pretty far fetched attack and not what most people mean when they bring this up).

We brought up this attack a long time ago before we realized you could have permissioned function calls on ethereum. If the leech can’t prove it got some payout from x augur market or can’t prove the outcome of x augur market it can’t leech. Of course, anyone can see this outside the chain, but anyone can see the result of a US presidential election outside the chain as well, it’s getting it into the chain w/o a lot of trust that’s the hard part.

“It’s not true. Contracts cannot view other contracts except through deliberately publicly exposed methods, so the interactivity property doesn’t really leave substantial cause for concern.” ~ Vitalik Buterin

https://www.reddit.com/r/ethereum/comments/3fmuvz/the_augur_reputation_token_crowdsale_will_begin/ctqhyqs

Note that currently any blockchain based system can execute a “parasite” attack by incorporating Merkle proofs of the host blockchain into its own chain.

Videos

What advertisement or marketing videos exist?

How A Decentralized Prediction Market Works:

Reputation 101: How Augurs Reputation Tokens Work:

What educational videos are available?

Joey Krug @ SF Bitcoin Devs (August 23rd, 2016)

Joey Krug @ Epicenter Bitcoin (August 13th, 2016)

Joey Krug @ Coinbase (July 22nd, 2016)

Jack Peterson @ Ethereum.nz (Jun 13th, 2016)

Ethereum Enthusiasts @ RISE Barclays NYC: Augur 101: When Ethereum Meets Prediction Markets (Jun 9th, 2016)