42%, up from 38% 4 days ago. Would be good to have a graph over time.

* what would be the implications of reaching 51%?

* what was the most recorded % of any pool? I thought BTC guild had 49% for quite a while.

* Why are people adding and not running away?

1. BTCGuild and GHash.IO
GHash.IO overtook BTCGuild this week - they now have a quarter of a percent more of the network than BTCGuild.

Last week I wrote:
"GHash.IO keep increasing their share of the network, most likely due to CEX.IO trading. I'm not sure how I feel about this - it's an increase in the proportion controlled by a single entity, but in effect any pool has the same control of hashes. It should make no difference that (as far as I'm aware) CEX.IO maintains the hash sources locally rather than the sources being distributed. Is there a downside I'm not understanding?"
I received a couple of good responses on bitcointalk.org:

From eleuthria, BTCGuild pool op:
"Only extra worry from cex.io/ghash.io that is different from past concerns is accountability.  If a public pool *attempted* to do something nefarious that pool just committed suicide whether they succeed or fail at the gamble.  When a pool that has ~1 PH/s (based on estimates looking at their speed fluctuations during known pool issues for public vs private) privately owned, there is no accountability left.  Pair that with using a 0% fee for what was already the 2nd largest mining entity before the public was allowed in.  If they attempt something, there is basically no downside.  If the public hashrate leaves due to an attempt, they aren't actually losing anything (no fee) other than the mining time on their private farm."

From gmaxwell, bitcoin developer:
"The excuse giving for years of why consolidations of ten percent, twenty percent, or even more, in the hands of pool operators didn't effectively disprove the Bitcoin security model was that pool operators were more obligated than a typical miner to behave with the public interest at heart because the hashrate controlling miners could vote with their feet.

I've never been too fond of the argument: evidence (e.g. miners voting with their feet very slowly even when a pool is clearly robbing them) suggests otherwise... But that argument doesn't even exist for GHash.io/CEX.io: the miners are captive and cannot leave. Worse, there is a moral hazard because an unknown portion of the hardware is paid for by other people (at top dollar rates too) and so if some stunt they perform debases its value... so what? Heck, perhaps it drops the market value down to nothing an cex can buy their obligations back for a song. This means that CEX.io can probably profit from an attack even if that attack ultimately fails."

Thanks for explaining that, guys. I feel a little foolish for not having taken market forces into account.

* what would be the implications of reaching 51%?

That's much worse than central bank. Central banks are a public companies... And those structures are regulated by law.

There is no law or public activity, so this "central bank" is able to do whatever the owner wants. And nobody can stop or regulate his activity. It's absolute monarchy as is.

Isn't this kind of centralized hashing power too dangerous, even way below 51%?