I apologize for the double post. But I just realized that OP (
Original
Poster: coinomy.net) had two misconceptions unanswered by others, plus what seems to be an implicit question; and I want to be sure that OP sees this:
From the way you stated your question, I infer that you are wondering about fee estimates reported by your wallet software. These are usually stated in terms of the fee which is
estimated for confirmation within 1 block, 3 blocks, 5 blocks, 10 blocks, 25 blocks, etc. For reasons I will clear up below, this is only an estimate of the amount of time which will be required for confirmation of a certain fee. It is
not any sort of directly calculated fee or agreement that “this transaction will be confirmed in
x blocks”.
Now, Misconception #1:
Miners calculate for the transfer but who of this miner gets it?
Note: The correct word in common use is
transaction, not
transfer.
Miners do not calculate any fee. Rather, the creator of each transaction sets the fee for that transaction. A block has limited available capacity; and if the block is full (= too many transactions to fit), then miners will pick the most profitable combination of transactions which they can fit into a block.
At that point, people are competing with each other to get their transactions into the next block. Due to competition, fees rise. They do not rise because of any authority setting a fee: Rather, users compete with fees to make their transactions more attractive to miners. Almost always, of course, miners will select the transactions with the highest ratio of fees paid to block capacity used (ratio of BTC/WU).
The fee estimator in your wallet is only estimating how attractive your transaction will look to miners, so that they will choose it. If you set a high fee, then many miners will want to grab that transaction for inclusion in the next block. If you set a low fee, then many miners may ignore your transaction—pushing it aside in favour of transactions which offer higher fees.
Unless the fee is ridiculously low, then eventually, as the level of network activity fluctuates, some miners will probably decide that the transaction is worthwhile to include in a block. Most likely, you don’t just need one miner to choose your transaction: You need enough miners to choose your transaction to make it probable that one of them will get lucky, and solve the block.
When your wallet software tells you, “fee x for confirmation in y blocks”, that is somewhat misleading. What it really means is that based on a complicated calculation over recent network activity, your wallet
estimates that for a given fee, it will probably be about (say) 10 blocks before a miner who liked your transaction gets lucky.
Misconception #2—some pertinent concepts are explained above:
How long does it take to solve the puzzle of my transfer,
(This was almost-but-not-quite touched by bitperson.)
There is no “puzzle” for any individual transaction. As aforesaid, miners choose transactions and fit them together into a block. There is a “puzzle” of sorts for that block. The correct term for the “puzzle” is
proof-of-work (POW).
The difficulty of work required for the proof-of-work is set based on a
target such that
on average, it should take 10 minutes for somebody, somewhere to find the correct solution to the proof-of-work. The actual time will be shorter or longer than that: Sometimes a miner may get lucky, and find a solution sooner; other times
all miners may be unlucky, and it will take longer for anybody to find a solution. Moreover, global hashrate (the collective power of all miners) fluctuates; this also affects the actual time for solving a block.
If a miner chooses your transaction as one of many transactions in a block, and then finds the correct solution, then the time required for your transaction is the same as for everybody else’s transaction in that block. The solution (technically, a nonce) is found
for the whole block, as a unit which includes your transaction and many others.
