I know people tend to get obsessed about the confirmation time, but the important factors are the number of nodes and the speed at which any given transaction is seen by the network as a whole.
I believe this is under three seconds for a transaction to be seen by the entire bitcoin network.
Why is this measure important? Because this is the so-called "window of opportunity" for a double-spend. In fact as soon as the transaction hits the network, the race is on, so we are actually talking about milliseconds.
Satoshi provided an answer to this somewhere (I can't find it right now) which explains what the real issue is.
A large number of nodes means that any double spend attack, has to work harder to get the transaction seen first, and hence included in a block first. But again this is totally arbitrary because nobody can predict the winner of the block (who gets to choose which transactions to include).
So this "fastest confirmation" metric is a red herring.
1The number of confirmations considered to be 100% is arbitrary. Bitcoin could have chosen 1, or 100, instead of 6, trading off speed and security. If we turned Bitcoin's difficulty way down, we'd have a currency that got hundreds of confirmations per second, but you'd need millions of confirmations to get reasonable security. – Nate Eldredge – 2014-04-16T07:07:41.240
2If average block time is all you take into account to declare fastest, "42coin" would be fastest at (you guessed it) 42 seconds. But it's moot: you can design a crypto-coin with a block generation time of 3 seconds if you wish, but the massive amount of orphaned blocks would force you to wait for a lot of confirmations, thus negating the speed advantage. Ripple uses a totally different way to achieve consensus which results in faster confirmations, NXT could potentially be as fast or faster as everybody knows who will mine next block and so transactions can be sent directly to that node. – Joe Pineda – 2014-04-27T11:42:17.103
No, no, I was comparing "42coin" to "Fastcoin" - block generation time is a bit faster for the former. Yes, Ripple's method is ingenous and achieves much faster confirmations, plus a number of other niceties they baked-in altogether. Nxt's idea of deterministically chosing next miner seems interesting as well, but opens the door to a kind of DDOS attacks. – Joe Pineda – 2014-04-28T12:59:02.100
@Gracchus, What do you mean by "not the only way to skin the Fed"? – Pacerier – 2014-05-23T15:08:40.350
@Gracchus, Oh and I always wanted to ask... what do you mean by "asker oppression"? – Pacerier – 2014-05-23T15:38:31.137
1
@Gracchus, Might get left behind? I consider it already done: http://meta.stackoverflow.com/questions/251758/why-is-stack-overflow-so-negative-of-late
– Pacerier – 2014-05-23T15:59:29.847