ICONbet has received a lot of support from the ICON Foundation and currently provides most of the daily transactions on the struggling ICON blockchain. Based on our previous dealings with ICON we were naturally suspicious of this DApp. The website for the DApp includes a link to a whitepaper that attempts to show that the game is fair by using a Secure Hash Algorithm (SHA3-256) of the block timestamp.
The whitepaper ends with the following text “The ICONbet team is globally distributed. Given the ever-changing and uncertain regulatory environment for blockchain-based games, tokens, and DAOs, the team has decided to remain anonymous“. It’s safe to say that the team behind the DApp are part of the ICON team and have access to the centralised nodes used by the ICON blockchain.
The ICONbet DApp follows most of the rules of roulette and as it’s supposed to be operated by a Decentralised Autonomous Organisation (DAO) the game is known as DAOlette. One major difference between DAOlette and real roulette is the restriction on placing bets on 0 (shown as a bomb on the wheel).
Many players have complained of the wheel coming up on a bomb more often than statistics would suggest and this manifests itself in one of two ways. Either 3 bombs in 3 spins of the wheel or 3 in quick succession of around 20 spins. Despite heavy promotion on Twitter and paid ads on DApp sites there are only a couple of dozen players.
So is ICONbet a scam or not?
We started by using the website to place bets on the wheel spin and noticed the same alarming frequency of bombs reported by other players. Maybe the website was biased towards certain users or specific ICX addresses. As the DApp is simply a smart contract on the ICON blockchain we built our own bot to “spin the wheel”. The bot has several advantages over the website like avoidance of malware, much higher betting frequency and the ability to analyse the results.
The ICON blockchain uses 24 centralised nodes to create blocks and each node creates 10 blocks before the duty is passed to the next node. The time to produce a block increases slightly when the block producer changes so this is a good time for a node to drop an extra bomb on an unsuspecting player. Our bot was designed to skip the first block produced by each node to avoid these bombs.
With centralised nodes producing the blocks the assertion that the game must be fair as it’s based on SHA3-256 of the block time is simply wrong. Even our basic bot can demonstrate that taken over an average of 10,000 wheel spins it only takes around 400 microseconds to change any winning spin block time into a losing spin block time (or vice versa). The ICON blockchain has a block time of approximately 2 seconds or 2000000 microseconds. Therefore increasing or decreasing the block time by 400 microseconds is an adjustment of just 0.02%. Block times on the network vary much more than this so the node operators can easily adjust the times to suit the wheel spin outcome they desire.
As shown above we made it to 3rd place in the ICONbet wager war on the 21st September. This was basically achieved due to the high frequency wheel spins created by our bot compared with using the ICONbet’s clunky website. For most people, us included, a top 3 position comes at a cost. Based on the claimed house hedge of 1.5% you can expect to lose over 1,000 ICX during the day. This is compensated by 15 to 25% of the wager war fund but this is only 10% of the house excess and some days there is no excess, despite over 300,000 ICX being wagered. The top 3 position also earned us about 110,000 TAP tokens.
TAP tokens are the DAOs native token but it heavily favours the anonymous ICONbet development team so they don’t have much intrinsic value. Not surprisingly the ICON blockchain shows that ICONbet financiers also play DAOlette. LuckyWhale and Rok23 BXMT currently hold at least 6.5M and 4.9M TAP respectively across 3 known blockchain addresses. As LuckyWhale took top spot on the day we came third we checked their bets for the day and the following day. On the day they came first from 445 wheels spins they were hit with just 8 bombs compared with at least 21 that most players would receive. More remarkable was the betting pattern over the 2 days. LuckyWhale placed dozens of large bets (1000 to 4000 ICX per wheel spin) and they were almost all winners compared with bombs against 10 to 100 ICX bets.
Statistics show that to earn 6.5M TAP you need to lose a huge amount of ICX if the “wheel” is not biased. One of LuckyWhale’s wallets can be found here but the more important wallet is here. It shows the first wallet transferring 22,634 ICX over a 2 week period to the second wallet and then sent off to a cryptocurrency exchange. So during the time a lot of the 6.5M TAP was earned LuckyWhale was also regularly earning ICX by “beating” the house hedge.
So even if you have had a bad run with “nuclear” bombs compared with regular bombs you can console yourself in the knowledge that some punters are regularly beating the odds. Check out all the bets placed by Wager War winner for 9 October 2019, Mart0593. Does a strategy of placing apparently random large bets works as a strategy at DAOlette? Probably not as indicated by a punter which we have nicknamed ForeverLosing. The odds of a losing spin are exactly the same whatever amount of ICX you wager, 1 ICX, 10 ICX or 4000 ICX it’s all the same. After more than 300 wheel spins the punter was down about 4% compared with the claimed house hedge of just 1.5%. So ForeverLosing adopted the random big bet strategy and increased the regular bet size of 2 to 10 ICX to a whopping 1500 ICX. Big bet yes but with a claimed 81% chance of winning it should be fairly low risk. Boom, first spin at the new mega bet size and it’s a losing bet! That took the unlucky punter to cumulative losses of almost 21%.
ForeverLosing’s online dataset has over 1,100 wheel spins, cumulative loss of almost 14% and still about 4% excluding the single mega bet. That particular punter should have contacted us before placing his mega bet as it broke a few of the golden rules of ICONbet that you pick up after tens of thousands of wheels spins. Please contact us if you haven’t worked out the obvious pitfalls of this DApp. All the ICONbet gambling data that we analyse is available on ICON’s tracker but it’s not very user friendly so our bot does all the hard work for you. Use our contact page if you want a permanent record of your DAOlette wheel spins or you want a copy of the golden rules. We ask for a small donation for running our bot over your data but we even accept the ICONbet TAP token that you no longer want.
- ICON Foundation supports ICONbet
- All the ICONbet developers are anonymous
- It takes under 400 microseconds for an ICON node to change a winning spin to a losing spin
- Apparent wheel bias
- Not allowed to bet on bombs which makes manipulation by nodes easier
- Players not sticking around to use the ICONbet DAO
We will leave it to the reader to decide if they think ICONbet is a scam. We welcome comments from P-REP candidates, ICX holders and anyone interested in the ICONbet DApp. If you found this information, or any of our articles useful, please contribute to the BLOC8 blockchain evangelical cause of cleaning up crypto. If you would like copies of the data or further details please use our contact page.