And come on chet, doesn't it strike you as a little odd that this figure seems to indicate that there aren't really any more users using this system today than there were in April?
Much of what I've put out here focuses on the fact that the early history indicates a small number of users mined up all of the coins, went out of their way to drown out new users when they came on from Slashdot, etc. And then based on the RSA paper, the vast majority of said coins from that era are sitting in holding accounts of a few users and you've seen a lot of effort to hide that fact and do some kind of strange things with the coins. The "top 100" list is a complete sham and this still seems to be controlled by a few users. That's what most of my investigation here has surrounded and I haven't focused a lot on current numbers.
I actually didn't even realize that's what it looked like until I just spun those off a few moments ago and I'm honestly floored by that figure. How can this increase 10x in value without there even being a growth in users? After it had gone up like 8X in the few months before this. That's ummm....a little bit strange.
It would be more interesting to compare BTC price to number of users IMO.
Why should a dramatic increase in hash power lead to more users? New technology led to increased power--as I said above, I bet the number of miners decreased. Your graph only goes through the end of October. I'd be interested to see the user data from November on when BTC price went up dramatically.
Will put that together an address vs price at some point. I want to figure out how to pretty up this enormous number of tabs. My guess off the top of my head based on what I know of this data:
1) Current data is going to look a lot like the "speculation index". You're going to see a run up in price without the run up in users.
2) The data will actually diverge from the "speculation index" back during 2011, the first big bubble. During that time you did see a large increase in addresses that correlated with the run up in price.
Will also add November and even December to the current graph. I stopped because it basically looks the same but will make your complaint regarding number of addresses in that one table look even worse. I stopped so you could actually tell that addresses was flat lining.
My main point with these graphs was to illustrate the "ArtForz" incident from the early days. But seeing the addresses flat line since April is a little surprising. Like I said, that's kind of strange and might deserve some further investigation with some other metrics without a doubt. And like I said, I'd like compare price to some metrics to find some correlation there.
And you're correct, a higher hash rate doesn't mean a higher number of users. Sometimes they correlate well, other times they really don't. During the times where the price grows relatively normally and there isn't gigantic leaps in mining technology, you see a very strong correlation between hash rate and number of users, transactions, price, etc. The price tends to demonstrate a pretty strong measure of stability. Then these technologies come along, radically change the cost of mining without yielding any increase in number of coins distributed overall, and the price starts to rapidly increase without an increase in number of users, transactions, any of that sort of thing . That's where I'm coming from in that hashing power (in the form of these huge technology leaps) seem to be the main driver during said times. If you look at this history, this even happened to some degree during the multi processor days. That's where it ran up 10 fold from .6 cents to 6 cents. First guy values it off of how much electricity it costs him to get coins, this is late 2009. And it stagnates for a long time and few people do any trading, because there really aren't many people using the network. Multi-processing is released, bitcoin forum itself kicks off - this is start of 2010. And there's still no users basically, but now the difficulty is going up because you have people throwing multiple processors (more hashing power) and more electricity to get the same exact amount of coins. Some people talked on the forum about whether it would be profitable to carve out Amazon EC2 instances to mine, to which Laszlo the pizza guy warned about increasing difficulty. Their forums are like 3 pages from that time, you can read most of the stuff from around there.
Sometimes the price lags because they can't drum up enough demand. Like from the July 2010 to October 2010 run up in hashing rate. The price flat lined there at .06 because they didn't get almost any long term users off of the Slashdot deal. You scared away all the fish, Art, what are we going to do? Well, we can release public GPU mining and pooling so they can at least get some coins, and now we can get some people to start mining and basing the value off of their current and far more dramatic costs. Afterall, mining guy isn't going to make his guest room unusable due to heat and noise for nothing:
http://dealbook.nytimes.com/2013/12/21/into-the-bitcoin-mines/?_r=1&