Topic: Can someone post examples of unusual transactions to help me understand.. (Read 177 times)

That graph is incorrect.

The most recent stale block* was at height 599587, on 16th October this year. It had the hash 00000000000000000012e55a5c5b0464ceab40b29af6cacb4b02ef6cc009a4b4, and was replaced by another block with hash 00000000000000000008fc27ad8c46149aa12bb12866fde846c93365591314f7 mined 15 seconds later. A stale block occurs every couple of months on average (but like the bitcoin block time, the actual timing varies significantly from the "average".)

---

*Orphaned block is a misnomer. An orphaned block, as the name suggests, would have no parent. These were blocks which a node received but was unable to verify because it had not yet received the previous (parent) blocks. These haven't existed for years since Bitcoin Core was changed to download block headers first. A stale block is what we are actually talking about here - a block whose parent was known, and was once part of the main chain, but has since been replaced by another block at the same height, and is therefore now inactive or "stale".

In addition to the above explanation, you can see transaction defined as double spend for a legit transaction as well. When you use Replace-By-Fee method for having a faster transaction (increasing the fee), explorer will mark that transaction as double spend since two transaction has been created from the same input, and same output has been used as well.
Edit- pooya87 has already covered it before me 

there are many ways you can have a transaction not-confirm for a long time or even ever!
for example paying anything between 0 and 1 satoshi/byte means your tx won't be relayed by majority of the network and as a result it will not be confirmed.
other cases would be creating a long chain of unconfirmed transactions. in simple terms you send coins to address A then from A send to B before previous tx confirms and then from B to C, from C to D and so forth. it has happened with some services like exchanges and sometimes lack of propagation of children means long confirmation times.

an orphaned block doesn't necessarily mean "transaction reversal", when a block is orphaned it means there is another different block that most probably contains the same exact transactions in it.
"reversing" can only happen if there is an attack where the attacker first spends the transaction and have it mined in block X then with his higher hashrate creates the same block and instead of mining the same tx, he mines another tx that spends the same output but creates different outputs.
the cost of this attack (51% attack) is so high that it requires that tx to be huge (worth millions) and people who accept such transactions amounts won't accept them with only 1 confirmation. and since reversing a tx that has more than 1 confirmation costs a lot more and it increases with each confirmation, this attack becomes nearly impossible.

This happened during the 2017 bull run when there's a high fee requirement because of too many users sending bitcoins and the alleged transaction spam.
But you won't be able to see those transactions in any blockexplorers anymore since they usually drop it to their mempool and database after a while.

These "conflicting transactions" are more common than you think, it's just because most blockexplorers instantly remove the conflicting tx from their database as soon as they receive the replacement tx. Others are keeping and labeling the transactions "double-spend attempt" (for weeks/months) like blockcypher.

Most of these are legitimate RBF (Replace-by-fee) transaction bumps but others are successful/unsuccessful double-spend attempts.
Or as long as the same input was spent, it will be marked by blockexplorers as "conflicting".
But all of those aren't "uncommon transactions".

The fact that the transaction may not be confirmed is usually caused by setting too low commission for miners. In practice, the more transactions, the longer we wait for the first confirmation. To avoid long waiting, it is best to check how much we should pay to be sure that our transaction will be taken to confirm in a satisfactory time. Before we send transaction (in some wallets fee is suggested automatically) we can check it for example here: https://bitcoinfees.earn.com/

How to calculate exact fee, you can learn here: https://support.earn.com/en/articles/64966-how-do-i-calculate-my-transaction-fee

Normally a transaction not confirmed after 2 weeks is dropped from the mempool.
Since lately there was no huge congestion, afaik there are not so old transactions to show.
https://sochain.com/ tells that the oldest transaction in its mempool is 17h old.



Afaik "reversing" a transaction happens only if the block which included the tx got orphaned.
This means that soon enough nobody will have anymore that block, so there will not be any record to show you.
If it helps at all, this statistic shows that there were no orphaned blocks lately (1.5 years).


Most probably double spends. Scam or mistake.
Such a conflict occurs if at least one input is no longer available, afaik.

It seems all transactions so far I've encountered get confirmed as normal in the blockchain from next block onwards. Can experienced users post examples or explain atypical transactions, for example:

Unconfirmed transactions after many days of the payment (does this even happen)?

A transaction which showed confirmed till 4 or 5 blocks but then was 'reversed'? (if I read right, after 3 confirms this is rare but possible)

The only thing I could find myself is here https://tradeblock.com/bitcoin/explorer/ this has the feature 'Conflicting transactions'. Some are from very old blocks e.g. from 2015:
https://tradeblock.com/bitcoin/txout/e276d10865d7780f2ef6436cfe49874131a53ad4199df7c7ee47cd883c29ac5b/0
What are these? Double spend scams?

Thanks