4 ms·

It's more complicated. Their collateral ratio has dropped below the liquidation level of 150%. So at best they would only get a fraction of their collateral bac

It's more complicated. Their collateral ratio has dropped below the liquidation level of 150%. So at best they would only get a fraction of their collateral back (any liquidation has a discount and a 13% penalty) - fundamentally they lost because their speculative bet didn't pay off. So yes, they lost more than they should, but describing it like they lost everything because of the liquidation problem alone is misleading. The "attack" was trivial in that there was only one bid at an auction - not enough people liquidating undercollateralized positions, but now so many people are looking at this it's never going to repeat again. It's an obvious risk only in hindsight.

wbl 7y ago
Market makers face this potential problem every day in the real world. They get sophisticated about it or die.
nullc 7y ago
> never going to repeat again So now it just needs a little help from miners to actively exclude all other bidders. My understanding is that in this case there were other bidders but they were drowned out by the winning bidder paying much more gas.
- nootropicat 7y ago
  >So now it just needs a little help from miners to actively exclude all other bidders. A mining cartel that censors transactions is indeed a real risk. Fortunately, ethereum is switching to PoS where even an average person with a smartphone could realistically generate several blocks a day, as opposed to multiple megawatt (or even giga) mining farms, so it's only a temporary issue. >My understanding is that in this case there were other bidders but they were drowned out by the winning bidder paying much more gas. Most likely lack of liquidity and/or gas pricing misconfiguration. Even at an ultra-high 600 gwei (during the peak congestion, the market rate was ~200 gwei) the total fee was less than $10. Simply put: not enough people running liquidation bots. Example: https://etherscan.io/tx/0x239cc6ba8f28b7a3b66cd5e1b558b0c735641d3a44405afe5b28eb2e5ef37665 https://etherscan.io/tx/0x239cc6ba8f28b7a3b66cd5e1b558b0c735...
  - nullc 7y ago
    > average person with a smartphone could realistically generate several blocks a day, I seem to recall the proposed staking minimum being around $200,000... Eth's administrators must have a kink for kidnapping. It's far from clear that "PoS" can result in a system which is both secure and decentralized: https://download.wpsoftware.net/bitcoin/pos.pdf https://download.wpsoftware.net/bitcoin/pos.pdf ... the limited academic work attempting to demonstrate such things have done things like assume that users were using a lossless ordered reliable broadcast medium (which is equivalent to assuming they were communicating over a consensus system). While the history of ethereum has demonstrated that in spite of claims to the contrary in their investment prospectus strong decentralization isn't a feature of the system, there are still many practical challenges even achieving faux-decentralization with PoS. Practically speaking this challenge is demonstrated by the fact that ETH's operators have continually pushed back their promised migration to PoS. Moreover, as was recently demonstrated with "steem" PoS can also easily be abused to rig outcomes just like that above concern with mining. So I think its far from clear that this is a temporary issue. Instead, to me it looks like PoS has turned into a never-arriving panacea being used to excuse all sorts of serious flaws in the ethereum ecosystem in addition to ethereum itself.
    - zeven7 7y ago
      > I seem to recall the proposed staking minimum being around $200,000... I takes 32 ETH to run your own validator node, so at current prices $4,183.
      - nullc 7y ago
        Ah. Indeed, when that was announced the price results in 32 ETH being ~$250k. Why is the same number of ETH an appropriate amount now?
        alphast0rm 7y ago
        ETH's all time high is $1,432.88 [1], so 32 ETH has only ever been worth $45,852.16 max. [1] https://coinmarketcap.com/currencies/ethereum/ https://coinmarketcap.com/currencies/ethereum/
- TheOtherHobbes 7y ago
  Maybe I'm missing something, but I suspect the odds of building a non-gameable system for people whose primary interest is speculative gaming are not going to be good.

wbl 7y ago

Market makers face this potential problem every day in the real world. They get sophisticated about it or die.

nullc 7y ago

> never going to repeat again So now it just needs a little help from miners to actively exclude all other bidders. My understanding is that in this case there were other bidders but they were drowned out by the winning bidder paying much more gas.

nootropicat 7y ago
>So now it just needs a little help from miners to actively exclude all other bidders. A mining cartel that censors transactions is indeed a real risk. Fortunately, ethereum is switching to PoS where even an average person with a smartphone could realistically generate several blocks a day, as opposed to multiple megawatt (or even giga) mining farms, so it's only a temporary issue. >My understanding is that in this case there were other bidders but they were drowned out by the winning bidder paying much more gas. Most likely lack of liquidity and/or gas pricing misconfiguration. Even at an ultra-high 600 gwei (during the peak congestion, the market rate was ~200 gwei) the total fee was less than $10. Simply put: not enough people running liquidation bots. Example: https://etherscan.io/tx/0x239cc6ba8f28b7a3b66cd5e1b558b0c735641d3a44405afe5b28eb2e5ef37665 https://etherscan.io/tx/0x239cc6ba8f28b7a3b66cd5e1b558b0c735...
- nullc 7y ago
  > average person with a smartphone could realistically generate several blocks a day, I seem to recall the proposed staking minimum being around $200,000... Eth's administrators must have a kink for kidnapping. It's far from clear that "PoS" can result in a system which is both secure and decentralized: https://download.wpsoftware.net/bitcoin/pos.pdf https://download.wpsoftware.net/bitcoin/pos.pdf ... the limited academic work attempting to demonstrate such things have done things like assume that users were using a lossless ordered reliable broadcast medium (which is equivalent to assuming they were communicating over a consensus system). While the history of ethereum has demonstrated that in spite of claims to the contrary in their investment prospectus strong decentralization isn't a feature of the system, there are still many practical challenges even achieving faux-decentralization with PoS. Practically speaking this challenge is demonstrated by the fact that ETH's operators have continually pushed back their promised migration to PoS. Moreover, as was recently demonstrated with "steem" PoS can also easily be abused to rig outcomes just like that above concern with mining. So I think its far from clear that this is a temporary issue. Instead, to me it looks like PoS has turned into a never-arriving panacea being used to excuse all sorts of serious flaws in the ethereum ecosystem in addition to ethereum itself.
  - zeven7 7y ago
    > I seem to recall the proposed staking minimum being around $200,000... I takes 32 ETH to run your own validator node, so at current prices $4,183.
    - nullc 7y ago
      Ah. Indeed, when that was announced the price results in 32 ETH being ~$250k. Why is the same number of ETH an appropriate amount now?
      - alphast0rm 7y ago
        ETH's all time high is $1,432.88 [1], so 32 ETH has only ever been worth $45,852.16 max. [1] https://coinmarketcap.com/currencies/ethereum/ https://coinmarketcap.com/currencies/ethereum/
TheOtherHobbes 7y ago
Maybe I'm missing something, but I suspect the odds of building a non-gameable system for people whose primary interest is speculative gaming are not going to be good.