3 ms·

> The top 10% of miners control 90% and just 0.1% (about 50 miners) control close to 50% of mining capacity This is incorrect. Replace "miners" with "mining po

by 3np 5y ago

> The top 10% of miners control 90% and just 0.1% (about 50 miners) control close to 50% of mining capacity This is incorrect. Replace "miners" with "mining pools" and they're closer to the mark. What's happened is that the block construction (done by the pool operator) and the PoW (done by miners) are mostly decoupled. Some miners will run everything themselves, but they're not represented in those numbers. (There are some things that can change here in stratum2, the new protocol used for coordinating mining pool s, but whatever) Sure, pool operators can abuse their power. But as the paper shows, hash power is liquid; miners do move between pools for various reasons. While there is a bit of a systemic risk there, it's not necessarily as bad as it might seem. IMO the biggest risk here is censorship; and we saw how responsive miners are to things like that earlier this year with the whole "OFAC-compliant" debacle.

Kbelicius 5y ago
> This is incorrect. Replace "miners" with "mining pools" and they're closer to the mark. It is correct. They are talking about miners not mining pools.
- 3np 5y ago
  Thanks for pointing that out and unfortunate that I misrepresented it, granted I didn’t read the article properly. As they note themselves it is a best-effort estimation based on publicly available data. Without digging into the data and methodology further, I’m assuming that they are probably at least in the right order of magnitude. One thing I’m noting is that they exclude ~70% of miners and assign ~30% as exchange miners (note that those two figures only coincidentally add up to 100%). One possible methodology error I can see (and again, further digging would have to be done to verify) is that exchanges often pool their user wallets together; it could be that for some exchange(s), they mistakenly lump together several users as one. Since their claimed concentration is very high, it should be feasible to look at the top 50-100 to cross-check. Would be awesome to get more research like this. Overall it’s an insightful article and the authors show great understanding.

by 3np 5y ago

Kbelicius 5y ago

> This is incorrect. Replace "miners" with "mining pools" and they're closer to the mark. It is correct. They are talking about miners not mining pools.

3np 5y ago
Thanks for pointing that out and unfortunate that I misrepresented it, granted I didn’t read the article properly. As they note themselves it is a best-effort estimation based on publicly available data. Without digging into the data and methodology further, I’m assuming that they are probably at least in the right order of magnitude. One thing I’m noting is that they exclude ~70% of miners and assign ~30% as exchange miners (note that those two figures only coincidentally add up to 100%). One possible methodology error I can see (and again, further digging would have to be done to verify) is that exchanges often pool their user wallets together; it could be that for some exchange(s), they mistakenly lump together several users as one. Since their claimed concentration is very high, it should be feasible to look at the top 50-100 to cross-check. Would be awesome to get more research like this. Overall it’s an insightful article and the authors show great understanding.