Calculate some sort of centralization/decentralization index?

[WiW]

I was thinking about it, especially in regard with this¹ topic, that most of the bitcoin network's security comes from the decentralized nature of the network. If the centralization/decentralization is the most important facet of our network, and the whole network is public, couldn't/shouldn't we calculate a rough index?

I'm no programmer or mathematician, but I know how to google, and I couldn't find such a thing. If we know the global hashrate, and we know how each and every block is awarded, isn't this possible? If all the blocks are awarded to the same peer on the network, you have a decentralization index of 0. If all the blocks are awarded equally to each and every peer of the network, you have a decentralization index of 1. This is assuming block reward distribution = relative hashing power. Call it the CDI or DCI, if you will.


I'm aware that it's a little more complicated than that, especially considering that a botnet master can distribute the blocks awarded to him across endless seemingly unrelated peers, while mining pools are supposedly supported by many unrelated peers however the hashing power of the entire pool is ultimately controlled by a single entity. This is why I'm calling it a "rough" index.

So, like I said, I'm just thinking aloud, but let me know what you think.

¹https://bitcointalk.org/index.php?topic=140588.0

[1713993196]

1713993196

[1713993196]

1713993196

[DannyHamilton]

. . . and we know how each and every block is awarded, isn't this possible? If all the blocks are awarded to the same peer on the network . . .

I'm not aware of any way to know with any certainty at all which peer solved the block.  You can determine which bitcoin address (or addresses) the block subsidy is associated with, so I suppose you can get an upper limit (total number of unique bitcoin addresses used for outputs of CoinBase transactions).  However, there isn't any way to know if a miner associated with any of those addresses is still running, and there isn't any way to know if a single peer is using multiple addresses.

[WiW]

Oh, well in that case there's very little to do about this...

Is there any way to get a hint of how much hashing power is distributed across how many nodes?

[grantbdev]

This thread should probably be moved to off-topic.

Anyway, How at risk am I with OpenJDKs installed for development (no web plug-in) on GNU/Linux?

Oops, not sure how this got posted here.

[DannyHamilton]

Oh, well in that case there's very little to do about this...

Is there any way to get a hint of how much hashing power is distributed across how many nodes?

No.  There is no way to know at any given point in time just how many people are attempting to calculate hashes.  There's know way to know how many unique people/entities have successfully found a block over any significant time frame.  You can get a upper limit on the number of entities who have found a block by looking at the total number of unique bitcoin addresses used in outputs for CoinBase transactions, but many of those addresses may be pool addresses, in which case you'd need the pool to report how many of their users submitted shares towards that block.  Also some people may use multiple addresses for their mining, so seeing 3 different addresses in CoinBase outputs doesn't necessarily mean 3 different entities.