I really like the idea behind MAGI: a CPU-mine-able coin that seeks to reduce the ecological impacts of the coin. It does this by attempting to limit the network hash rate by scaling back block rewards as the hash rate increases. This is the first coin that I mined, in large part thanks to Novaspirit's article (https://www.novaspirit.com/2017/10/19/crypto-mining-sbc/). But I've become disillusioned with the coin since then.
The creators argue that their scheme will lead to a more decentralized network as participation in mining is emphasized over raw hashing power [MAGI White Paper: https://arxiv.org/abs/1409.7948].
It is anticipated that the new block rewarding system leads to a more decentralized cryptocurrency. - Joe Lao
Unfortunately, there are a number of flaws with MAGI.
- The Tragedy of the Commons: Perhaps the least problematic of the flaws, but illustrative a magical point of view required to believe in MAGI. This is an old economics problem, the common example of which is cows grazing in a town's commons (shared pastures). Individual can extract short term gains by grazing as many cattle as they can there to the overall detriment of the commons, which get turned into a mud pit thanks to all of those cows. Cow herders seem a long way from crypto miners, but the idea is the same: people will not act for the collective good (keep block rewards high by reducing network hash-rate) when short term gains are at stake (are you going to turn your mining rigs off so that others can extract more?).
- The security of blockchains depend on the high hash rates. Very basically, blockchains depend on distributed voting: the network as a whole has to come to an agreement about what the current state of the ledger is. I may lie in my vote and say that I have a million cow-bucks when I only have ten, but if the majority doesn't agree, then I can't get away with it (this is an over-simplified example: the original bitcoin whitepaper directly addresses an actor changing previous transactions [https://bitcoin.org/bitcoin.pdf]). Hashing power is the equivalent of votes, so if I can muster the hashing power, then I get to write (probablistically) the blockchain, and if I say I have a million cow-bucks, it is so. The lower the network hashrate, the more vulnerable the blockchain is to attacks because it lowers the barrier for one entity to gain a majority of the votes.
- There are specific attacks that are at least in part exploiting the low hash rate. If you've seen any of the pools for MAGI lately, you'll see an endless string of oprhans, occasionally encapped by huge payouts. It appears that a flaw in the implementation is being exploited to force higher payouts (40XMG instead of the typical < 1XMG).
The payout flaw, however, can only be taken advantage of because the bad actor is able to gain a majority of the network hashrate.
Any system will have flaws in the implementation, and perhaps MAGI will be able to fix this one. But unless they fix the flaw in the coin's \emph{design}, then attacks will still be feasable. Network hashrate isn't the enemy: it's necessary to security. Is there a way for MAGI to achieve it's initial goals while maintaining security? Is that even possible? Until that problem is solved, MAGI is a fatally flawed coin and attacks will continue.