Bitcoin: insane fees spent on one transaction

Yesterday, a very anomalous thing happened: a user paid as much as 19 BTC as fees for a Bitcoin transaction. 

The countervalue in fiat currency is about $510,000, so an absolutely disproportionate amount. 

However, it is necessary to put this anomaly in context in order to interpret it correctly.

High fees on a Bitcoin transaction 

There have been more than half a million unconfirmed transactions waiting in the Bitcoin mempool for a few days now. 

The fact is that individual blocks on the Bitcoin blockchain rarely contain more than 4,000 transactions, with many not reaching 3,000. 

On average, since a block is validated every 10 minutes or so, and since only transactions entered in a validated block are confirmed, the queue will take many hours, if not days, to clear. In fact, more continue to be added. 

Yesterday a total of about 560,000 transactions were confirmed, and on only four occasions since the end of July have more than 600,000 transactions per day been confirmed. 

In cases like these, to speed up confirmation transactions it is enough to pay more fees than others, given that the transactions to be placed in a block are chosen by the miners, and they obviously choose the ones with higher fees since the fees paid by those doing the transactions are collected by them. 

While the average fees per transaction on 3 September were just over $0.8, yesterday they skyrocketed above $2.1, or more than double. 

Thus, in this situation there are those who agree to pay more fees than others in order to get priority for confirming their transactions. However, this does not explain the absurd amount mentioned above. 

The record fees on the transaction in Bitcoin

The transaction with the record fees was confirmed yesterday, and consisted of five mailings of very small amounts to Binance and other recipients. 

In total the unknown user sent only 0.07 BTC ($1,800), but to get this transaction validated he spent 19.82 BTC in fees. 

It is possible this was an error, perhaps due to the fact that the user had to set higher fees in order to get priority. 

The address from which this transaction started still holds nearly 416 BTC, or more than $10 million in Bitcoin. Thus this is a whale, and it is an address that has already made even more than 61,000 transactions in its history. 

However, it has only been active for two months, so for it to have made so many transactions in such a short time it must be managed by software. 

At this point one imagines that perhaps its management software may have made a mistake, or that the people who use it may have made a mistake, also because there was no need to spend 19.82 BTC in fees in order to have priority. 

Surely 19.82 thousandths of a Bitcoin was enough, equivalent to $510, or even just 1.982 thousandths of a Bitcoin ($51) to be among the first ones. 

The other hypotheses

Truth be told, the error hypothesis does not convince many, so much so that other hypotheses have also been put forward. 

However, if it is not a mistake but a voluntary decision, did the user know to which miner those 19.82 BTC would go?

In fact, the fees of all transactions in an entire block are collected only by the individual miner who validates it. 

The block in which that transaction was included (807,057) contained as many as 2,652 transactions, and in addition to the 6.25 BTC reward it also collected more than 20 BTC in fees (almost all of which came from the single transaction with the record fees). 

However, that block was not mined by a single mining farm, but by the F2Pool pool. Pools are organizations that combine the computing power of many miners, and then divide the rewards in proportion to the percentage of hashpower made available to each miner. 

Thus, the 19.82 BTC was actually not collected by a single miner, but was distributed to all the F2Pool miners in proportion to how much hashrate they made available to mine that block. 

At this point, the hypothesis of a voluntary user decision loses strength. 

The other blocks

By moving on to take the previous block, and the one after it, as an example, one discovers even more of the exceptionality of this affair. 

The previous block contained as many as 4,400 transactions, yet of fees it collected only 0.17 BTC. It was mined by Binance, i.e., a single mining farm. 

The next block contained 4,700 transactions, and collected only 0.16 BTC in fees. It was mined by the AntPool pool. 

Incidentally, to date AntPool is the second largest pool in the world by hashrate, while F2Pool is third. Binance is fifth, while in first position right now is Foundry USA pool. 

At this point, the question would be: since the 19.82 BTC fee mentioned above went right to F2Pool, how did the user who sent them know that they would go to that very pool? 

The answer is that he almost certainly did not know, and so it is extremely unlikely that he put them in specifically to send those BTC to that pool. 

Moreover, block 807,057 was mined by F2Pool in about 11 minutes, so there is no anomaly. By contrast, for example, the next block, the one mined by AntPool, was mined in only 4 minutes. 

In the end, the error hypothesis seems to be the most likely one.